From 4a70d15d3d5c3af9b296aec91e7b7a1b4be5b79d Mon Sep 17 00:00:00 2001
From: ageorge <anjith.george@idiap.ch>
Date: Tue, 11 Jul 2023 15:16:53 +0200
Subject: [PATCH] Initial commit
---
.gitignore | 20 +
COPYING | 674 ++++++++++++++++++
MANIFEST.in | 2 +
README.md | 92 ---
README.rst | 129 ++++
bob/__init__.py | 3 +
bob/paper/__init__.py | 0
bob/paper/ijcb2023_caim_hfr/__init__.py | 4 +
.../architectures/__init__.py | 2 +
.../ijcb2023_caim_hfr/architectures/common.py | 186 +++++
.../architectures/iresnet.py | 216 ++++++
.../architectures/light_iresnet_txl.py | 169 +++++
.../ijcb2023_caim_hfr/config/__init__.py | 4 +
.../config/database/__init__.py | 0
.../config/database/arlvtf.py | 6 +
.../config/database/mcxface.py | 9 +
.../config/database/scface_hfr.py | 9 +
.../config/lightning/x_iresnet_CAIM.py | 150 ++++
.../config/lightning/x_iresnet_PDT.py | 150 ++++
.../mcxfaceprotocols/VIS-THERMAL-split1.py | 6 +
.../ijcb2023_caim_hfr/database/__init__.py | 35 +
.../ijcb2023_caim_hfr/database/arlvtf.py | 212 ++++++
.../ijcb2023_caim_hfr/database/cuhk_cufs.py | 100 +++
.../ijcb2023_caim_hfr/database/mcxface.py | 229 ++++++
.../ijcb2023_caim_hfr/database/scface_hfr.py | 94 +++
.../ijcb2023_caim_hfr/database/tuftsface.py | 98 +++
.../ijcb2023_caim_hfr/dataset/__init__.py | 0
.../ijcb2023_caim_hfr/dataset/datasets.py | 55 ++
bob/paper/ijcb2023_caim_hfr/dataset/hface.py | 424 +++++++++++
.../ijcb2023_caim_hfr/losses/__init__.py | 2 +
bob/paper/ijcb2023_caim_hfr/losses/loss.py | 63 ++
.../ijcb2023_caim_hfr/scripts/__init__.py | 0
.../scripts/hface_metrics.py | 164 +++++
.../ijcb2023_caim_hfr/scripts/version.py | 13 +
bob/paper/ijcb2023_caim_hfr/test.py | 9 +
.../ijcb2023_caim_hfr/transformer/__init__.py | 12 +
.../transformer/lightning.py | 84 +++
.../transformer/xlightning.py | 189 +++++
buildout.cfg | 40 ++
conda/meta.yaml | 79 ++
doc/conf.py | 269 +++++++
doc/img/bob-128x128.png | Bin 0 -> 6547 bytes
doc/img/bob-favicon.ico | Bin 0 -> 4286 bytes
doc/img/bob-logo.png | Bin 0 -> 6266 bytes
doc/img/caim_framework.png | Bin 0 -> 180345 bytes
doc/index.rst | 42 ++
doc/links.rst | 13 +
doc/references.rst | 11 +
doc/running_caim.rst | 38 +
doc/running_pdt.rst | 38 +
pyproject.toml | 3 +
requirements.txt | 9 +
setup.py | 56 ++
spec-file.txt | 286 ++++++++
version.txt | 1 +
55 files changed, 4407 insertions(+), 92 deletions(-)
create mode 100644 .gitignore
create mode 100644 COPYING
create mode 100644 MANIFEST.in
delete mode 100644 README.md
create mode 100644 README.rst
create mode 100644 bob/__init__.py
create mode 100644 bob/paper/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/architectures/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/architectures/common.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/architectures/iresnet.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/architectures/light_iresnet_txl.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/database/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/database/arlvtf.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/database/mcxface.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/database/scface_hfr.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_CAIM.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_PDT.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/config/mcxfaceprotocols/VIS-THERMAL-split1.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/arlvtf.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/cuhk_cufs.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/mcxface.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/scface_hfr.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/database/tuftsface.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/dataset/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/dataset/datasets.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/dataset/hface.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/losses/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/losses/loss.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/scripts/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/scripts/version.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/test.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/transformer/__init__.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/transformer/lightning.py
create mode 100644 bob/paper/ijcb2023_caim_hfr/transformer/xlightning.py
create mode 100644 buildout.cfg
create mode 100644 conda/meta.yaml
create mode 100644 doc/conf.py
create mode 100644 doc/img/bob-128x128.png
create mode 100644 doc/img/bob-favicon.ico
create mode 100644 doc/img/bob-logo.png
create mode 100644 doc/img/caim_framework.png
create mode 100644 doc/index.rst
create mode 100644 doc/links.rst
create mode 100644 doc/references.rst
create mode 100644 doc/running_caim.rst
create mode 100644 doc/running_pdt.rst
create mode 100644 pyproject.toml
create mode 100644 requirements.txt
create mode 100644 setup.py
create mode 100644 spec-file.txt
create mode 100644 version.txt
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..f7d22a0
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,20 @@
+*~
+*.swp
+*.pyc
+bin
+eggs
+parts
+.installed.cfg
+.mr.developer.cfg
+*.egg-info
+src
+develop-eggs
+dist
+.nfs*
+.gdb_history
+build
+.coverage
+record.txt
+miniconda.sh
+miniconda/
+.ipynb_checkpoints/
diff --git a/COPYING b/COPYING
new file mode 100644
index 0000000..20d40b6
--- /dev/null
+++ b/COPYING
@@ -0,0 +1,674 @@
+ GNU GENERAL PUBLIC LICENSE
+ Version 3, 29 June 2007
+
+ Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+ Preamble
+
+ The GNU General Public License is a free, copyleft license for
+software and other kinds of works.
+
+ The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works. By contrast,
+the GNU General Public License is intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users. We, the Free Software Foundation, use the
+GNU General Public License for most of our software; it applies also to
+any other work released this way by its authors. You can apply it to
+your programs, too.
+
+ When we speak of free software, we are referring to freedom, not
+price. Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+ To protect your rights, we need to prevent others from denying you
+these rights or asking you to surrender the rights. Therefore, you have
+certain responsibilities if you distribute copies of the software, or if
+you modify it: responsibilities to respect the freedom of others.
+
+ For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must pass on to the recipients the same
+freedoms that you received. You must make sure that they, too, receive
+or can get the source code. And you must show them these terms so they
+know their rights.
+
+ Developers that use the GNU GPL protect your rights with two steps:
+(1) assert copyright on the software, and (2) offer you this License
+giving you legal permission to copy, distribute and/or modify it.
+
+ For the developers' and authors' protection, the GPL clearly explains
+that there is no warranty for this free software. For both users' and
+authors' sake, the GPL requires that modified versions be marked as
+changed, so that their problems will not be attributed erroneously to
+authors of previous versions.
+
+ Some devices are designed to deny users access to install or run
+modified versions of the software inside them, although the manufacturer
+can do so. This is fundamentally incompatible with the aim of
+protecting users' freedom to change the software. The systematic
+pattern of such abuse occurs in the area of products for individuals to
+use, which is precisely where it is most unacceptable. Therefore, we
+have designed this version of the GPL to prohibit the practice for those
+products. If such problems arise substantially in other domains, we
+stand ready to extend this provision to those domains in future versions
+of the GPL, as needed to protect the freedom of users.
+
+ Finally, every program is threatened constantly by software patents.
+States should not allow patents to restrict development and use of
+software on general-purpose computers, but in those that do, we wish to
+avoid the special danger that patents applied to a free program could
+make it effectively proprietary. To prevent this, the GPL assures that
+patents cannot be used to render the program non-free.
+
+ The precise terms and conditions for copying, distribution and
+modification follow.
+
+ TERMS AND CONDITIONS
+
+ 0. Definitions.
+
+ "This License" refers to version 3 of the GNU General Public License.
+
+ "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+ "The Program" refers to any copyrightable work licensed under this
+License. Each licensee is addressed as "you". "Licensees" and
+"recipients" may be individuals or organizations.
+
+ To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy. The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+ A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+ To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy. Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+ To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies. Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+ An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License. If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+ 1. Source Code.
+
+ The "source code" for a work means the preferred form of the work
+for making modifications to it. "Object code" means any non-source
+form of a work.
+
+ A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+ The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form. A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+ The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities. However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work. For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+ The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+ The Corresponding Source for a work in source code form is that
+same work.
+
+ 2. Basic Permissions.
+
+ All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met. This License explicitly affirms your unlimited
+permission to run the unmodified Program. The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work. This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+ You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force. You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright. Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+ Conveying under any other circumstances is permitted solely under
+the conditions stated below. Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+ 3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+ No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+ When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+ 4. Conveying Verbatim Copies.
+
+ You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+ You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+ 5. Conveying Modified Source Versions.
+
+ You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+ a) The work must carry prominent notices stating that you modified
+ it, and giving a relevant date.
+
+ b) The work must carry prominent notices stating that it is
+ released under this License and any conditions added under section
+ 7. This requirement modifies the requirement in section 4 to
+ "keep intact all notices".
+
+ c) You must license the entire work, as a whole, under this
+ License to anyone who comes into possession of a copy. This
+ License will therefore apply, along with any applicable section 7
+ additional terms, to the whole of the work, and all its parts,
+ regardless of how they are packaged. This License gives no
+ permission to license the work in any other way, but it does not
+ invalidate such permission if you have separately received it.
+
+ d) If the work has interactive user interfaces, each must display
+ Appropriate Legal Notices; however, if the Program has interactive
+ interfaces that do not display Appropriate Legal Notices, your
+ work need not make them do so.
+
+ A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit. Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+ 6. Conveying Non-Source Forms.
+
+ You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+ a) Convey the object code in, or embodied in, a physical product
+ (including a physical distribution medium), accompanied by the
+ Corresponding Source fixed on a durable physical medium
+ customarily used for software interchange.
+
+ b) Convey the object code in, or embodied in, a physical product
+ (including a physical distribution medium), accompanied by a
+ written offer, valid for at least three years and valid for as
+ long as you offer spare parts or customer support for that product
+ model, to give anyone who possesses the object code either (1) a
+ copy of the Corresponding Source for all the software in the
+ product that is covered by this License, on a durable physical
+ medium customarily used for software interchange, for a price no
+ more than your reasonable cost of physically performing this
+ conveying of source, or (2) access to copy the
+ Corresponding Source from a network server at no charge.
+
+ c) Convey individual copies of the object code with a copy of the
+ written offer to provide the Corresponding Source. This
+ alternative is allowed only occasionally and noncommercially, and
+ only if you received the object code with such an offer, in accord
+ with subsection 6b.
+
+ d) Convey the object code by offering access from a designated
+ place (gratis or for a charge), and offer equivalent access to the
+ Corresponding Source in the same way through the same place at no
+ further charge. You need not require recipients to copy the
+ Corresponding Source along with the object code. If the place to
+ copy the object code is a network server, the Corresponding Source
+ may be on a different server (operated by you or a third party)
+ that supports equivalent copying facilities, provided you maintain
+ clear directions next to the object code saying where to find the
+ Corresponding Source. Regardless of what server hosts the
+ Corresponding Source, you remain obligated to ensure that it is
+ available for as long as needed to satisfy these requirements.
+
+ e) Convey the object code using peer-to-peer transmission, provided
+ you inform other peers where the object code and Corresponding
+ Source of the work are being offered to the general public at no
+ charge under subsection 6d.
+
+ A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+ A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling. In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage. For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product. A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+ "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source. The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+ If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information. But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+ The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed. Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+ Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+ 7. Additional Terms.
+
+ "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law. If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+ When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it. (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.) You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+ Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+ a) Disclaiming warranty or limiting liability differently from the
+ terms of sections 15 and 16 of this License; or
+
+ b) Requiring preservation of specified reasonable legal notices or
+ author attributions in that material or in the Appropriate Legal
+ Notices displayed by works containing it; or
+
+ c) Prohibiting misrepresentation of the origin of that material, or
+ requiring that modified versions of such material be marked in
+ reasonable ways as different from the original version; or
+
+ d) Limiting the use for publicity purposes of names of licensors or
+ authors of the material; or
+
+ e) Declining to grant rights under trademark law for use of some
+ trade names, trademarks, or service marks; or
+
+ f) Requiring indemnification of licensors and authors of that
+ material by anyone who conveys the material (or modified versions of
+ it) with contractual assumptions of liability to the recipient, for
+ any liability that these contractual assumptions directly impose on
+ those licensors and authors.
+
+ All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10. If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term. If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+ If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+ Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+ 8. Termination.
+
+ You may not propagate or modify a covered work except as expressly
+provided under this License. Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+ However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+ Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+ Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License. If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+ 9. Acceptance Not Required for Having Copies.
+
+ You are not required to accept this License in order to receive or
+run a copy of the Program. Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance. However,
+nothing other than this License grants you permission to propagate or
+modify any covered work. These actions infringe copyright if you do
+not accept this License. Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+ 10. Automatic Licensing of Downstream Recipients.
+
+ Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License. You are not responsible
+for enforcing compliance by third parties with this License.
+
+ An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations. If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+ You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License. For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+ 11. Patents.
+
+ A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based. The
+work thus licensed is called the contributor's "contributor version".
+
+ A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version. For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+ Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+ In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement). To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+ If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients. "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+ If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+ A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License. You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+ Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+ 12. No Surrender of Others' Freedom.
+
+ If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License. If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all. For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+ 13. Use with the GNU Affero General Public License.
+
+ Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU Affero General Public License into a single
+combined work, and to convey the resulting work. The terms of this
+License will continue to apply to the part which is the covered work,
+but the special requirements of the GNU Affero General Public License,
+section 13, concerning interaction through a network will apply to the
+combination as such.
+
+ 14. Revised Versions of this License.
+
+ The Free Software Foundation may publish revised and/or new versions of
+the GNU General Public License from time to time. Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+ Each version is given a distinguishing version number. If the
+Program specifies that a certain numbered version of the GNU General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation. If the Program does not specify a version number of the
+GNU General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+ If the Program specifies that a proxy can decide which future
+versions of the GNU General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+ Later license versions may give you additional or different
+permissions. However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+ 15. Disclaimer of Warranty.
+
+ THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+ 16. Limitation of Liability.
+
+ IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+ 17. Interpretation of Sections 15 and 16.
+
+ If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+ END OF TERMS AND CONDITIONS
+
+ How to Apply These Terms to Your New Programs
+
+ If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+ To do so, attach the following notices to the program. It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+ <one line to give the program's name and a brief idea of what it does.>
+ Copyright (C) <year> <name of author>
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+ If the program does terminal interaction, make it output a short
+notice like this when it starts in an interactive mode:
+
+ <program> Copyright (C) <year> <name of author>
+ This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+ This is free software, and you are welcome to redistribute it
+ under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License. Of course, your program's commands
+might be different; for a GUI interface, you would use an "about box".
+
+ You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU GPL, see
+<http://www.gnu.org/licenses/>.
+
+ The GNU General Public License does not permit incorporating your program
+into proprietary programs. If your program is a subroutine library, you
+may consider it more useful to permit linking proprietary applications with
+the library. If this is what you want to do, use the GNU Lesser General
+Public License instead of this License. But first, please read
+<http://www.gnu.org/philosophy/why-not-lgpl.html>.
\ No newline at end of file
diff --git a/MANIFEST.in b/MANIFEST.in
new file mode 100644
index 0000000..9d10504
--- /dev/null
+++ b/MANIFEST.in
@@ -0,0 +1,2 @@
+include README.rst bootstrap-buildout.py buildout.cfg COPYING version.txt requirements.txt
+recursive-include doc *.py *.rst *.ico *.png
diff --git a/README.md b/README.md
deleted file mode 100644
index 9cb7ee2..0000000
--- a/README.md
+++ /dev/null
@@ -1,92 +0,0 @@
-# bob.paper.ijcb2023_caim_hfr
-
-
-
-## Getting started
-
-To make it easy for you to get started with GitLab, here's a list of recommended next steps.
-
-Already a pro? Just edit this README.md and make it your own. Want to make it easy? [Use the template at the bottom](#editing-this-readme)!
-
-## Add your files
-
-- [ ] [Create](https://docs.gitlab.com/ee/user/project/repository/web_editor.html#create-a-file) or [upload](https://docs.gitlab.com/ee/user/project/repository/web_editor.html#upload-a-file) files
-- [ ] [Add files using the command line](https://docs.gitlab.com/ee/gitlab-basics/add-file.html#add-a-file-using-the-command-line) or push an existing Git repository with the following command:
-
-```
-cd existing_repo
-git remote add origin https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr.git
-git branch -M master
-git push -uf origin master
-```
-
-## Integrate with your tools
-
-- [ ] [Set up project integrations](https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr/-/settings/integrations)
-
-## Collaborate with your team
-
-- [ ] [Invite team members and collaborators](https://docs.gitlab.com/ee/user/project/members/)
-- [ ] [Create a new merge request](https://docs.gitlab.com/ee/user/project/merge_requests/creating_merge_requests.html)
-- [ ] [Automatically close issues from merge requests](https://docs.gitlab.com/ee/user/project/issues/managing_issues.html#closing-issues-automatically)
-- [ ] [Enable merge request approvals](https://docs.gitlab.com/ee/user/project/merge_requests/approvals/)
-- [ ] [Automatically merge when pipeline succeeds](https://docs.gitlab.com/ee/user/project/merge_requests/merge_when_pipeline_succeeds.html)
-
-## Test and Deploy
-
-Use the built-in continuous integration in GitLab.
-
-- [ ] [Get started with GitLab CI/CD](https://docs.gitlab.com/ee/ci/quick_start/index.html)
-- [ ] [Analyze your code for known vulnerabilities with Static Application Security Testing(SAST)](https://docs.gitlab.com/ee/user/application_security/sast/)
-- [ ] [Deploy to Kubernetes, Amazon EC2, or Amazon ECS using Auto Deploy](https://docs.gitlab.com/ee/topics/autodevops/requirements.html)
-- [ ] [Use pull-based deployments for improved Kubernetes management](https://docs.gitlab.com/ee/user/clusters/agent/)
-- [ ] [Set up protected environments](https://docs.gitlab.com/ee/ci/environments/protected_environments.html)
-
-***
-
-# Editing this README
-
-When you're ready to make this README your own, just edit this file and use the handy template below (or feel free to structure it however you want - this is just a starting point!). Thank you to [makeareadme.com](https://www.makeareadme.com/) for this template.
-
-## Suggestions for a good README
-Every project is different, so consider which of these sections apply to yours. The sections used in the template are suggestions for most open source projects. Also keep in mind that while a README can be too long and detailed, too long is better than too short. If you think your README is too long, consider utilizing another form of documentation rather than cutting out information.
-
-## Name
-Choose a self-explaining name for your project.
-
-## Description
-Let people know what your project can do specifically. Provide context and add a link to any reference visitors might be unfamiliar with. A list of Features or a Background subsection can also be added here. If there are alternatives to your project, this is a good place to list differentiating factors.
-
-## Badges
-On some READMEs, you may see small images that convey metadata, such as whether or not all the tests are passing for the project. You can use Shields to add some to your README. Many services also have instructions for adding a badge.
-
-## Visuals
-Depending on what you are making, it can be a good idea to include screenshots or even a video (you'll frequently see GIFs rather than actual videos). Tools like ttygif can help, but check out Asciinema for a more sophisticated method.
-
-## Installation
-Within a particular ecosystem, there may be a common way of installing things, such as using Yarn, NuGet, or Homebrew. However, consider the possibility that whoever is reading your README is a novice and would like more guidance. Listing specific steps helps remove ambiguity and gets people to using your project as quickly as possible. If it only runs in a specific context like a particular programming language version or operating system or has dependencies that have to be installed manually, also add a Requirements subsection.
-
-## Usage
-Use examples liberally, and show the expected output if you can. It's helpful to have inline the smallest example of usage that you can demonstrate, while providing links to more sophisticated examples if they are too long to reasonably include in the README.
-
-## Support
-Tell people where they can go to for help. It can be any combination of an issue tracker, a chat room, an email address, etc.
-
-## Roadmap
-If you have ideas for releases in the future, it is a good idea to list them in the README.
-
-## Contributing
-State if you are open to contributions and what your requirements are for accepting them.
-
-For people who want to make changes to your project, it's helpful to have some documentation on how to get started. Perhaps there is a script that they should run or some environment variables that they need to set. Make these steps explicit. These instructions could also be useful to your future self.
-
-You can also document commands to lint the code or run tests. These steps help to ensure high code quality and reduce the likelihood that the changes inadvertently break something. Having instructions for running tests is especially helpful if it requires external setup, such as starting a Selenium server for testing in a browser.
-
-## Authors and acknowledgment
-Show your appreciation to those who have contributed to the project.
-
-## License
-For open source projects, say how it is licensed.
-
-## Project status
-If you have run out of energy or time for your project, put a note at the top of the README saying that development has slowed down or stopped completely. Someone may choose to fork your project or volunteer to step in as a maintainer or owner, allowing your project to keep going. You can also make an explicit request for maintainers.
diff --git a/README.rst b/README.rst
new file mode 100644
index 0000000..f86901b
--- /dev/null
+++ b/README.rst
@@ -0,0 +1,129 @@
+.. -*- coding: utf-8 -*-
+
+.. image:: https://img.shields.io/badge/docs-stable-yellow.svg
+ :target: https://www.idiap.ch/software/bob/docs/bob/bob.paper.ijcb2023_caim_hfr/master/index.html
+.. image:: https://img.shields.io/badge/docs-latest-orange.svg
+ :target: https://www.idiap.ch/software/bob/docs/bob/bob.paper.ijcb2023_caim_hfr/master/index.html
+.. image:: https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr/badges/master/build.svg
+ :target: https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr/commits/master
+.. image:: https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr/badges/master/coverage.svg
+ :target: https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr/commits/master
+.. image:: https://img.shields.io/badge/gitlab-project-0000c0.svg
+ :target: https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr
+.. image:: https://img.shields.io/pypi/v/bob.paper.ijcb2023_caim_hfr.svg
+ :target: https://pypi.python.org/pypi/bob.paper.ijcb2023_caim_hfr
+
+
+==============================================================================================
+Bridging the Gap: Heterogeneous Face Recognition with Conditional Adaptive Instance Modulation
+==============================================================================================
+
+This package is part of the signal-processing and machine learning toolbox Bob_.
+
+This package contains source code to replicate the experimental results published in the following paper::
+
+ @inproceedings{georgecaim2023,
+ author = {Anjith George and Sebastien Marcel},
+ title = {Bridging the Gap: Heterogeneous Face Recognition with Conditional Adaptive Instance Modulation},
+ year = {2023},
+ booktitle = {IEEE International Joint Conference on Biometrics (IJCB)},
+ }
+
+If you use this package and/or its results, please consider citing the paper.
+
+Installation
+------------
+
+The installation instructions are based on conda_ and works on **Linux systems
+only**. `Install conda`_ before continuing.
+
+Once you have installed conda_, download the source code of this paper and
+unpack it. Then, you can create a conda environment with the following
+command::
+
+ $ cd bob.paper.ijcb2023_caim_hfr
+ $ conda create --name bob.paper.ijcb2023_caim_hfr --file spec-file.txt
+ $ conda activate bob.paper.ijcb2023_caim_hfr # activate the environment
+ $ pip install pytorch-lightning==1.5.3
+ $ buildout
+ $ ./bin/bob bio pipelines vanilla-biometrics --help # test the installation
+
+This will install all the required software to reproduce this paper.
+
+Configuring the experiments
+---------------------------
+
+Once the environment is ready. The MCXFace database need to be downloaded.
+For downloading the MCXFace database, please check the following link `Download Instructions for MCXFace <https://www.idiap.ch/dataset/mcxface>`_.
+
+
+After downloading the database, you need to set the paths to
+those in the configuration files. Bob_ supports a configuration file (``~/.bobrc``) in your home directory to specify where the databases are located. This can be set with the following command:
+ $ bob config set "bob.db.mcxface.directory" "PATH_TO_MCXFACE_ROOT_FOLDER"
+
+Also set a directory to store the experimental results.
+
+ $ bob config set "ptemp" "PATH_TO_A_FOLDER"
+
+And it can be tested from the bobrc file.
+
+ $ cat ~/.bobrc
+ {
+ ""bob.db.mcxface.directory": "PATH_TO_MCXFACE_ROOT_FOLDER",
+ }
+
+After, the installation steps to run the experiments are given in the documentation of this package.
+
+Check the "sphinx/index.html" for details on how to run the experiments.
+
+
+Running the experiments
+-----------------------
+
+The database class provides a standard interface for training, development and evaluation.
+Now, using bob package, all the steps can be automated: training, and evaluation.
+
+To run the experiment, use the following command:
+
+.. code-block:: sh
+
+ ./bin/bob bio pipelines vanilla-biometrics -vvv \
+ bob.paper.ijcb2023_caim_hfr.config.mcxfaceprotocols.VIS-THERMAL-split1 \
+ bob.paper.ijcb2023_caim_hfr.config.lightning.x_iresnet_PDT -m
+
+.. note::
+ Note that the first config file specifies the database to be used and the protocol, one can change the config file to
+ run the experiment for different datasets and protocols. Please inspect the configuration files and pipelines to familiarize the experiments. Please run this job on a GPU machine for faster target.
+
+
+Running the command should launch the training, and once the training is finished, the evaluation process would
+start automatically. At the end of the experiment, the score files will be written to the folder corresponding to the
+specific protocol and dataset.
+
+Evaluating results
+------------------
+
+To evaluate the score files run the following command.
+
+.. code-block:: python
+
+ ./bin/python bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py -sf path_to_scores-dev.csv
+
+
+It should return AUC, EER, Rank-1 accuracy and Verification rates at different FAR points.
+
+
+Contact
+-------
+
+For questions or reporting issues to this software package, contact our
+development `mailing list`_.
+
+
+.. Place your references here:
+.. _bob: https://www.idiap.ch/software/bob
+.. _installation: https://www.idiap.ch/software/bob/install
+.. _mailing list: https://www.idiap.ch/software/bob/discuss
+.. _bob package development: https://www.idiap.ch/software/bob/docs/bob/bob.extension/master/
+.. _conda: https://conda.io
+.. _install conda: https://conda.io/docs/install/quick.html#linux-miniconda-install
diff --git a/bob/__init__.py b/bob/__init__.py
new file mode 100644
index 0000000..2ab1e28
--- /dev/null
+++ b/bob/__init__.py
@@ -0,0 +1,3 @@
+# see https://docs.python.org/3/library/pkgutil.html
+from pkgutil import extend_path
+__path__ = extend_path(__path__, __name__)
diff --git a/bob/paper/__init__.py b/bob/paper/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/bob/paper/ijcb2023_caim_hfr/__init__.py b/bob/paper/ijcb2023_caim_hfr/__init__.py
new file mode 100644
index 0000000..edbb409
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/__init__.py
@@ -0,0 +1,4 @@
+# see https://docs.python.org/3/library/pkgutil.html
+from pkgutil import extend_path
+
+__path__ = extend_path(__path__, __name__)
diff --git a/bob/paper/ijcb2023_caim_hfr/architectures/__init__.py b/bob/paper/ijcb2023_caim_hfr/architectures/__init__.py
new file mode 100644
index 0000000..7e9a070
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/architectures/__init__.py
@@ -0,0 +1,2 @@
+from .light_iresnet_txl import XIresnetTxl
+from .common import PDT
diff --git a/bob/paper/ijcb2023_caim_hfr/architectures/common.py b/bob/paper/ijcb2023_caim_hfr/architectures/common.py
new file mode 100644
index 0000000..9cafd3f
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/architectures/common.py
@@ -0,0 +1,186 @@
+"""
+ Common routines for models in PyTorch.
+"""
+
+from logging import raiseExceptions
+import math
+from inspect import isfunction
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn import Linear, Conv2d, BatchNorm1d, BatchNorm2d, PReLU, ReLU, Sigmoid, Dropout, MaxPool2d, \
+ AdaptiveAvgPool2d, Sequential, Module
+from collections import namedtuple
+from torch.nn.modules import module
+
+from typing import Union, Tuple, Optional
+
+import torch
+import torch.nn as nn
+
+
+from torch.nn import init
+
+class SEBlock(nn.Module):
+ """
+ Squeeze-and-Excitation block from 'Squeeze-and-Excitation Networks,' https://arxiv.org/abs/1709.01507.
+ """
+ def __init__(self, channels, reduction=16, round_mid=False, use_conv=True,
+ mid_activation=(lambda: nn.ReLU(inplace=True)),
+ out_activation=(lambda: nn.Sigmoid())):
+ super(SEBlock, self).__init__()
+ self.use_conv = use_conv
+ mid_channels = channels // reduction if not round_mid else round_channels(float(channels) / reduction)
+
+ self.pool = nn.AdaptiveAvgPool2d(output_size=1)
+ if use_conv:
+ self.conv1 = nn.Conv2d(in_channels=channels, out_channels=mid_channels, kernel_size=1,
+ stride=1, groups=1, bias=True)
+ else:
+ self.fc1 = nn.Linear(in_features=channels, out_features=mid_channels)
+ self.activ = nn.ReLU(inplace=True)
+ if use_conv:
+ self.conv2 = nn.Conv2d(in_channels=mid_channels, out_channels=channels, kernel_size=1,
+ stride=1, groups=1, bias=True)
+ else:
+ self.fc2 = nn.Linear(in_features=mid_channels, out_features=channels)
+ self.sigmoid = nn.Sigmoid()
+
+ def forward(self, x):
+ w = self.pool(x)
+ if not self.use_conv:
+ w = w.view(x.size(0), -1)
+ w = self.conv1(w) if self.use_conv else self.fc1(w)
+ w = self.activ(w)
+ w = self.conv2(w) if self.use_conv else self.fc2(w)
+ w = self.sigmoid(w)
+ if not self.use_conv:
+ w = w.unsqueeze(2).unsqueeze(3)
+ x = x * w
+ return x
+
+
+class ChannelAttention(nn.Module):
+ def __init__(self,channel,reduction=16):
+ super().__init__()
+ self.maxpool=nn.AdaptiveMaxPool2d(1)
+ self.avgpool=nn.AdaptiveAvgPool2d(1)
+ self.se=nn.Sequential(
+ nn.Conv2d(channel,channel//reduction,1,bias=False),
+ nn.ReLU(),
+ nn.Conv2d(channel//reduction,channel,1,bias=False)
+ )
+ self.sigmoid=nn.Sigmoid()
+
+ def forward(self, x) :
+ max_result=self.maxpool(x)
+ avg_result=self.avgpool(x)
+ max_out=self.se(max_result)
+ avg_out=self.se(avg_result)
+ output=self.sigmoid(max_out+avg_out)
+ return output
+
+class SpatialAttention(nn.Module):
+ def __init__(self,kernel_size=7):
+ super().__init__()
+ self.conv=nn.Conv2d(2,1,kernel_size=kernel_size,padding=kernel_size//2)
+ self.sigmoid=nn.Sigmoid()
+
+ def forward(self, x) :
+ max_result,_=torch.max(x,dim=1,keepdim=True)
+ avg_result=torch.mean(x,dim=1,keepdim=True)
+ result=torch.cat([max_result,avg_result],1)
+ output=self.conv(result)
+ output=self.sigmoid(output)
+ return output
+
+
+
+class CBAMBlock(nn.Module):
+
+ def __init__(self, channel=512,reduction=16,kernel_size=49):
+ super().__init__()
+ self.ca=ChannelAttention(channel=channel,reduction=reduction)
+ self.sa=SpatialAttention(kernel_size=kernel_size)
+
+
+ def init_weights(self):
+ for m in self.modules():
+ if isinstance(m, nn.Conv2d):
+ init.kaiming_normal_(m.weight, mode='fan_out')
+ if m.bias is not None:
+ init.constant_(m.bias, 0)
+ elif isinstance(m, nn.BatchNorm2d):
+ init.constant_(m.weight, 1)
+ init.constant_(m.bias, 0)
+ elif isinstance(m, nn.Linear):
+ init.normal_(m.weight, std=0.001)
+ if m.bias is not None:
+ init.constant_(m.bias, 0)
+
+ def forward(self, x):
+ b, c, _, _ = x.size()
+ residual=x
+ out=x*self.ca(x)
+ out=out*self.sa(out)
+ return out+residual
+
+
+
+class PDT(nn.Module): # better than the other InceptionTranslator; ~600 parameters opposed to the huge ones
+ """
+ Prepended Domain Transformer
+ """
+
+ def __init__(self, in_channels=3, pool_features=6, use_se=False, use_bias=False, use_cbam=True):
+ super(PDT, self).__init__()
+
+ self.use_se=use_se
+ self.use_bias=use_bias
+ self.use_cbam=use_cbam
+ self.relu=nn.ReLU()
+
+ self.pool_features=pool_features
+ self.branch1x1 = Conv2d(in_channels, self.pool_features, kernel_size=1, bias=self.use_bias)
+
+ self.branch5x5_1 = Conv2d(in_channels, self.pool_features//2, kernel_size=1, bias=self.use_bias)
+ self.branch5x5_2 = Conv2d(self.pool_features//2, self.pool_features, kernel_size=5, padding=2, bias=self.use_bias)
+
+ self.branch3x3dbl_1 = Conv2d(in_channels, self.pool_features//2, kernel_size=1, bias=self.use_bias)
+ self.branch3x3dbl_2 = Conv2d(self.pool_features//2, self.pool_features, kernel_size=3, padding=1, bias=self.use_bias)
+ self.branch3x3dbl_3 = Conv2d(self.pool_features, self.pool_features, kernel_size=3, padding=1, bias=self.use_bias)
+
+ self.branch_pool = Conv2d(in_channels, pool_features, kernel_size=1, bias=self.use_bias)
+
+ if self.use_cbam:
+ self.cbam = CBAMBlock(channel=self.pool_features*4,reduction=4,kernel_size=7)
+ if self.use_se:
+ self.se=SEBlock(self.pool_features*4,reduction=4)
+
+ self.dec = Conv2d(self.pool_features*4, 3, kernel_size=1, bias=self.use_bias)
+
+ def forward(self, x):
+ branch1x1 = self.relu(self.branch1x1(x))
+
+ branch5x5 = self.relu(self.branch5x5_1(x))
+ branch5x5 = self.relu(self.branch5x5_2(branch5x5))
+
+ branch3x3dbl = self.relu(self.branch3x3dbl_1(x))
+ branch3x3dbl = self.relu(self.branch3x3dbl_2(branch3x3dbl))
+ branch3x3dbl = self.relu(self.branch3x3dbl_3(branch3x3dbl))
+
+ branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)
+ branch_pool = self.relu(self.branch_pool(branch_pool))
+
+ outputs = [branch1x1, branch5x5, branch3x3dbl, branch_pool]
+
+ concat = torch.cat(outputs, 1)
+
+ if self.use_se:
+ concat=self.se(concat)
+ if self.use_cbam:
+ concat=self.cbam(concat)
+
+ img = self.dec(concat)
+
+ return img
diff --git a/bob/paper/ijcb2023_caim_hfr/architectures/iresnet.py b/bob/paper/ijcb2023_caim_hfr/architectures/iresnet.py
new file mode 100644
index 0000000..de4d68b
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/architectures/iresnet.py
@@ -0,0 +1,216 @@
+import torch
+from torch import nn
+
+# code referenced from https://raw.githubusercontent.com/nizhib/pytorch-insightface/master/insightface/iresnet.py
+# https://github.com/nizhib/pytorch-insightface
+# ported weights from pytorch
+
+def preprocess_img(x):
+ """Assuming images are between 0 and 255"""
+ x = (x - 127.5) / 128.0
+ return x
+
+
+def conv3x3(in_planes, out_planes, stride=1, groups=1, dilation=1):
+ """3x3 convolution with padding"""
+ return nn.Conv2d(
+ in_planes,
+ out_planes,
+ kernel_size=3,
+ stride=stride,
+ padding=dilation,
+ groups=groups,
+ bias=False,
+ dilation=dilation,
+ )
+
+
+def conv1x1(in_planes, out_planes, stride=1):
+ """1x1 convolution"""
+ return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
+
+
+class IBasicBlock(nn.Module):
+ expansion = 1
+
+ def __init__(
+ self,
+ inplanes,
+ planes,
+ stride=1,
+ downsample=None,
+ groups=1,
+ base_width=64,
+ dilation=1,
+ ):
+ super(IBasicBlock, self).__init__()
+ if groups != 1 or base_width != 64:
+ raise ValueError("BasicBlock only supports groups=1 and base_width=64")
+ if dilation > 1:
+ raise NotImplementedError("Dilation > 1 not supported in BasicBlock")
+ # Both self.conv1 and self.downsample layers downsample the input when stride != 1
+ self.bn1 = nn.BatchNorm2d(inplanes, eps=2e-05, momentum=0.9)
+ self.conv1 = conv3x3(inplanes, planes)
+ self.bn2 = nn.BatchNorm2d(planes, eps=2e-05, momentum=0.9)
+ self.prelu = nn.PReLU(planes)
+ self.conv2 = conv3x3(planes, planes, stride)
+ self.bn3 = nn.BatchNorm2d(planes, eps=2e-05, momentum=0.9)
+ self.downsample = downsample
+ self.stride = stride
+
+ def forward(self, x):
+ identity = x
+
+ out = self.bn1(x)
+ out = self.conv1(out)
+ out = self.bn2(out)
+ out = self.prelu(out)
+ out = self.conv2(out)
+ out = self.bn3(out)
+
+ if self.downsample is not None:
+ identity = self.downsample(x)
+
+ out += identity
+
+ return out
+
+
+class IResNet(nn.Module):
+ fc_scale = 7 * 7
+
+ def __init__(
+ self,
+ block,
+ layers,
+ num_features=512,
+ zero_init_residual=False,
+ groups=1,
+ width_per_group=64,
+ replace_stride_with_dilation=None,
+ ):
+ super(IResNet, self).__init__()
+
+ self.inplanes = 64
+ self.dilation = 1
+ if replace_stride_with_dilation is None:
+ # each element in the tuple indicates if we should replace
+ # the 2x2 stride with a dilated convolution instead
+ replace_stride_with_dilation = [False, False, False]
+ if len(replace_stride_with_dilation) != 3:
+ raise ValueError(
+ "replace_stride_with_dilation should be None "
+ "or a 3-element tuple, got {}".format(replace_stride_with_dilation)
+ )
+ self.groups = groups
+ self.base_width = width_per_group
+ self.conv1 = nn.Conv2d(
+ 3, self.inplanes, kernel_size=3, stride=1, padding=1, bias=False
+ )
+ self.bn1 = nn.BatchNorm2d(self.inplanes, eps=2e-05, momentum=0.9)
+ self.prelu = nn.PReLU(self.inplanes)
+ self.layer1 = self._make_layer(block, 64, layers[0], stride=2)
+ self.layer2 = self._make_layer(
+ block, 128, layers[1], stride=2, dilate=replace_stride_with_dilation[0]
+ )
+ self.layer3 = self._make_layer(
+ block, 256, layers[2], stride=2, dilate=replace_stride_with_dilation[1]
+ )
+ self.layer4 = self._make_layer(
+ block, 512, layers[3], stride=2, dilate=replace_stride_with_dilation[2]
+ )
+
+ self.bn2 = nn.BatchNorm2d(512 * block.expansion, eps=2e-05, momentum=0.9)
+ self.dropout = nn.Dropout2d(p=0.4, inplace=True)
+ self.flatten = nn.Flatten()
+ self.fc = nn.Linear(512 * block.expansion * self.fc_scale, num_features)
+ self.features = nn.BatchNorm1d(num_features, eps=2e-05, momentum=0.9)
+
+ for m in self.modules():
+ if isinstance(m, nn.Conv2d):
+ nn.init.kaiming_normal_(m.weight, mode="fan_out", nonlinearity="relu")
+ elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
+ nn.init.constant_(m.weight, 1)
+ nn.init.constant_(m.bias, 0)
+
+ if zero_init_residual:
+ for m in self.modules():
+ if isinstance(m, IBasicBlock):
+ nn.init.constant_(m.bn2.weight, 0)
+
+ def _make_layer(self, block, planes, blocks, stride=1, dilate=False):
+ downsample = None
+ previous_dilation = self.dilation
+ if dilate:
+ self.dilation *= stride
+ stride = 1
+ if stride != 1 or self.inplanes != planes * block.expansion:
+ downsample = nn.Sequential(
+ conv1x1(self.inplanes, planes * block.expansion, stride),
+ nn.BatchNorm2d(planes * block.expansion, eps=2e-05, momentum=0.9),
+ )
+
+ layers = []
+ layers.append(
+ block(
+ self.inplanes,
+ planes,
+ stride,
+ downsample,
+ self.groups,
+ self.base_width,
+ previous_dilation,
+ )
+ )
+ self.inplanes = planes * block.expansion
+ for _ in range(1, blocks):
+ layers.append(
+ block(
+ self.inplanes,
+ planes,
+ groups=self.groups,
+ base_width=self.base_width,
+ dilation=self.dilation,
+ )
+ )
+
+ return nn.Sequential(*layers)
+
+ def forward(self, x):
+ x = self.conv1(x)
+ x = self.bn1(x)
+ x = self.prelu(x)
+
+ x = self.layer1(x)
+ x = self.layer2(x)
+ x = self.layer3(x)
+ x = self.layer4(x)
+
+ x = self.bn2(x)
+ x = self.dropout(x)
+ x = self.flatten(x)
+ x = self.fc(x)
+ x = self.features(x)
+
+ return x
+
+
+def _iresnet(arch, block, layers, checkpoint_path):
+ model = IResNet(block, layers)
+ if checkpoint_path is not None:
+ model.load_state_dict(torch.load(checkpoint_path))
+ model.eval()
+ return model
+
+
+def iresnet34(checkpoint_path=None):
+ return _iresnet("iresnet34", IBasicBlock, [3, 4, 6, 3], checkpoint_path)
+
+
+def iresnet50(checkpoint_path=None):
+ return _iresnet("iresnet50", IBasicBlock, [3, 4, 14, 3], checkpoint_path)
+
+
+def iresnet100(checkpoint_path=None):
+ return _iresnet("iresnet100", IBasicBlock, [3, 13, 30, 3], checkpoint_path)
+
diff --git a/bob/paper/ijcb2023_caim_hfr/architectures/light_iresnet_txl.py b/bob/paper/ijcb2023_caim_hfr/architectures/light_iresnet_txl.py
new file mode 100644
index 0000000..fd43e00
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/architectures/light_iresnet_txl.py
@@ -0,0 +1,169 @@
+import copy
+
+from collections import OrderedDict
+
+import torch
+
+from torch import nn
+
+from bob.bio.face.embeddings.pytorch import IResnet34
+from bob.bio.face.embeddings.pytorch import IResnet50
+from bob.bio.face.embeddings.pytorch import IResnet100
+
+from .iresnet import iresnet34
+from .iresnet import iresnet50
+from .iresnet import iresnet100
+
+ARCHS = {
+ "iresnet34": iresnet34,
+ "iresnet50": iresnet50,
+ "iresnet100": iresnet100,
+}
+CHECKPOINTS = {
+ "iresnet34": IResnet34().checkpoint_path,
+ "iresnet50": IResnet50().checkpoint_path,
+ "iresnet100": IResnet100().checkpoint_path,
+}
+
+
+class XIresnetTxl(nn.Module):
+ """Takes an iresnet architecture and a translator module.
+ Then, provides a multi-modal pytorch model."""
+
+ def __init__(self, arch, translator, modality, **kwargs):
+ super().__init__(**kwargs)
+ self.arch = ARCHS[arch]
+ self.checkpoint = CHECKPOINTS[arch]
+ self.modality = modality
+ self.translator = translator
+ self.vis_model = self.arch()
+ self.vis_model.load_state_dict(torch.load(self.checkpoint))
+ self.vis_model.requires_grad_(False)
+
+ right_modules = []
+ right_modules.append(self.translator)
+ right_modules.append(self.vis_model)
+
+ self.right_model = nn.Sequential(*right_modules)
+
+ self.model_per_modality = {"VIS": self.vis_model, modality: self.right_model}
+
+ def forward(self, inputs):
+ input_vis, input_mod = inputs["VIS"], inputs[self.modality]
+ embedding_left = self.vis_model(input_vis)
+ embedding_left = torch.nn.functional.normalize(embedding_left, dim=-1)
+ embedding_right = self.right_model(input_mod)
+ embedding_right = torch.nn.functional.normalize(embedding_right, dim=-1)
+ outputs = torch.pairwise_distance(embedding_left, embedding_right)
+ return outputs
+
+ def train(self, mode=True):
+ """
+ Override the default train() to freeze the BN parameters
+ """
+ super().train(mode)
+ for name, module in self.named_modules():
+ if isinstance(module, torch.nn.modules.BatchNorm1d) or isinstance(module, torch.nn.modules.BatchNorm2d):
+ if 'vis' in name:
+ module.eval()
+ module.weight.requires_grad = False
+ module.bias.requires_grad = False
+
+
+
+class XIresnetTxlKLDIV(nn.Module):
+ """Takes an iresnet architecture and a translator module.
+ Then, provides a multi-modal pytorch model."""
+
+ def __init__(self, arch, translator, modality, **kwargs):
+ super().__init__(**kwargs)
+ self.arch = ARCHS[arch]
+ self.checkpoint = CHECKPOINTS[arch]
+ self.modality = modality
+ self.translator = translator
+ self.vis_model = self.arch()
+ self.vis_model.load_state_dict(torch.load(self.checkpoint))
+ self.vis_model.requires_grad_(False)
+
+ right_modules = []
+ right_modules.append(self.translator)
+ right_modules.append(self.vis_model)
+
+ self.right_model = nn.Sequential(*right_modules)
+
+ self.model_per_modality = {"VIS": self.vis_model, modality: self.right_model}
+
+ def forward(self, inputs):
+ input_vis, input_mod = inputs["VIS"], inputs[self.modality]
+ embedding_left = self.vis_model(input_vis)
+ embedding_left = torch.nn.functional.normalize(embedding_left, dim=-1)
+ embedding_right = self.right_model(input_mod)
+ embedding_right = torch.nn.functional.normalize(embedding_right, dim=-1)
+ outputs = (embedding_left, embedding_right)
+ return outputs
+
+ def train(self, mode=True):
+ """
+ Override the default train() to freeze the BN parameters
+ """
+ super().train(mode)
+ for name, module in self.named_modules():
+ if isinstance(module, torch.nn.modules.BatchNorm1d) or isinstance(module, torch.nn.modules.BatchNorm2d):
+ if 'vis' in name:
+ module.eval()
+ module.weight.requires_grad = False
+ module.bias.requires_grad = False
+
+
+
+
+class XIresnetTxlMMD2(nn.Module):
+ """Takes an iresnet architecture and a translator module.
+ Then, provides a multi-modal pytorch model."""
+
+ def __init__(self, arch, translator, modality, **kwargs):
+ super().__init__(**kwargs)
+ self.arch = ARCHS[arch]
+ self.checkpoint = CHECKPOINTS[arch]
+ self.modality = modality
+ self.translator = translator
+ self.vis_model = self.arch()
+ self.vis_model.load_state_dict(torch.load(self.checkpoint))
+ self.vis_model.requires_grad_(False)
+
+ right_modules = []
+ right_modules.append(self.translator)
+ right_modules.append(self.vis_model)
+
+ self.right_model = nn.Sequential(*right_modules)
+
+ self.model_per_modality = {"VIS": self.vis_model, modality: self.right_model}
+
+ def forward(self, inputs):
+ input_vis, input_mod = inputs["VIS"], inputs[self.modality]
+ embedding_left = self.vis_model(input_vis)
+ embedding_left = torch.nn.functional.normalize(embedding_left, dim=-1)
+ translated=self.translator(input_mod)
+ embedding_right = self.vis_model(translated)
+ embedding_right = torch.nn.functional.normalize(embedding_right, dim=-1)
+ outputs = (embedding_left, embedding_right)
+
+
+ batch=int(input_vis.size()[0])
+
+ vis_feat=input_vis.view(batch, -1)
+ trans_feat=translated.view(batch, -1)
+
+ return outputs, (vis_feat,trans_feat)
+
+ def train(self, mode=True):
+ """
+ Override the default train() to freeze the BN parameters
+ """
+ super().train(mode)
+ for name, module in self.named_modules():
+ if isinstance(module, torch.nn.modules.BatchNorm1d) or isinstance(module, torch.nn.modules.BatchNorm2d):
+ if 'vis' in name:
+ module.eval()
+ module.weight.requires_grad = False
+ module.bias.requires_grad = False
diff --git a/bob/paper/ijcb2023_caim_hfr/config/__init__.py b/bob/paper/ijcb2023_caim_hfr/config/__init__.py
new file mode 100644
index 0000000..edbb409
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/__init__.py
@@ -0,0 +1,4 @@
+# see https://docs.python.org/3/library/pkgutil.html
+from pkgutil import extend_path
+
+__path__ = extend_path(__path__, __name__)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/database/__init__.py b/bob/paper/ijcb2023_caim_hfr/config/database/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/bob/paper/ijcb2023_caim_hfr/config/database/arlvtf.py b/bob/paper/ijcb2023_caim_hfr/config/database/arlvtf.py
new file mode 100644
index 0000000..dbf18df
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/database/arlvtf.py
@@ -0,0 +1,6 @@
+from bob.paper.ijcb2023_caim_hfr.database import ARLVTFDatabase
+
+PROTOCOL = "VIS-THERMAL-G_VB0--P_TB--split4"
+
+
+database = ARLVTFDatabase(protocol=PROTOCOL)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/database/mcxface.py b/bob/paper/ijcb2023_caim_hfr/config/database/mcxface.py
new file mode 100644
index 0000000..7820843
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/database/mcxface.py
@@ -0,0 +1,9 @@
+from bob.paper.ijcb2023_caim_hfr.database import MCXFaceDatabase
+
+# In case protocol is comming from chain loading
+# https://www.idiap.ch/software/bob/docs/bob/bob.extension/stable/py_api.html#bob.extension.config.load
+if "protocol" not in locals():
+ protocol = "VIS-THERMAL-split1"
+
+
+database = MCXFaceDatabase(protocol=protocol)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/database/scface_hfr.py b/bob/paper/ijcb2023_caim_hfr/config/database/scface_hfr.py
new file mode 100644
index 0000000..a2be432
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/database/scface_hfr.py
@@ -0,0 +1,9 @@
+from bob.paper.ijcb2023_caim_hfr.database import SCFaceHFRDatabase
+
+# In case protocol is comming from chain loading
+# https://www.idiap.ch/software/bob/docs/bob/bob.extension/stable/py_api.html#bob.extension.config.load
+if "protocol" not in locals():
+ protocol = "far"
+
+
+database = SCFaceHFRDatabase(protocol=protocol)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_CAIM.py b/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_CAIM.py
new file mode 100644
index 0000000..f1191a5
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_CAIM.py
@@ -0,0 +1,150 @@
+import os
+
+from functools import partial
+
+import numpy as np
+import pytorch_lightning as pl
+import torch
+import torchvision.transforms as transforms
+
+from pytorch_lightning.callbacks import EarlyStopping
+from pytorch_lightning.callbacks import LearningRateMonitor
+from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.loggers import TensorBoardLogger
+from sklearn.pipeline import make_pipeline
+from sklearn.preprocessing import FunctionTransformer
+
+
+from bob.bio.face.embeddings.pytorch import iresnet_template
+from bob.extension import rc
+from bob.pipelines import CheckpointWrapper
+from bob.pipelines import wrap
+
+
+
+from bob.paper.ijcb2023_caim_hfr.transformer.xlightning import XLitDSUModel
+from bob.paper.ijcb2023_caim_hfr.architectures.iresnet import preprocess_img
+from bob.paper.ijcb2023_caim_hfr.dataset.hface import AGLitDSUDataModule, AGLitDSUDataModuleNP # testing
+from bob.paper.ijcb2023_caim_hfr.losses import ContrastiveLoss
+from bob.paper.ijcb2023_caim_hfr.transformer.lightning import LightningTransformer
+from bob.paper.ijcb2023_caim_hfr.architectures.light_iresnet_txl import XIresnetTxl
+
+database = globals()["database"]
+PROTOCOL = globals()["PROTOCOL"]
+
+POOL ='6'
+SE = 'False'
+CBAM = 'True'
+BIAS = 'False'
+ARCHITECTURE = 'iresnet100'
+
+def str2bool(v):
+ return v.lower() in ("yes", "true", "t", "1")
+
+# print("globals()",globals())
+
+OUTPUT = (
+ output
+) = f"{rc['ptemp']}/{database.name}/{database.protocol}/"
+BATCH_SIZE = 90
+EPOCHS = 20
+
+
+pipeline = iresnet_template(
+ "passthrough", database.annotation_type, database.fixed_positions
+)
+cropper = pipeline.transformer.steps[0][1]
+preprocessor = CheckpointWrapper(
+ cropper,
+ features_dir=f"{rc['ptemp']}/{database.name}/{database.protocol}/"
+ + f"preprocessed_{cropper.estimator.cropped_image_size[0]}",
+)
+
+test_transform = transforms.Compose(
+ [
+ # lambda transfomer based preprocess_img
+ transforms.Lambda(preprocess_img),
+ # convert to tensor
+ transforms.Lambda(partial(torch.as_tensor, dtype=torch.float32)),
+ ]
+)
+
+train_transform = transforms.Compose(
+ [
+ test_transform,
+ # random horizontal flipping
+ transforms.RandomHorizontalFlip(0.5),
+ ]
+)
+
+
+def trainer_fn():
+ # we need tensorboard, best checkpoint, early stopping, and learning rate monitor
+
+ model_dir = f"{OUTPUT}/model"
+
+ callbacks = [
+ ModelCheckpoint(
+ dirpath=model_dir,
+ filename="weights-epoch{epoch:02d}-val_loss{val/loss:.2f}",
+ auto_insert_metric_name=False,
+ monitor="val/loss",
+ mode="min",
+ save_last=True,
+ ),
+ LearningRateMonitor(),
+ EarlyStopping(monitor="val/loss", patience=10, mode="min"),
+ ]
+ logger = TensorBoardLogger(f"{model_dir}/logs", default_hp_metric=False)
+
+ # resume from the last checkpoint if it exists
+ resume_from_checkpoint = f"{model_dir}/last.ckpt"
+ if not os.path.exists(resume_from_checkpoint):
+ resume_from_checkpoint = None
+
+ trainer = pl.Trainer(
+ gpus=-1 if torch.cuda.is_available() else None,
+ logger=logger,
+ callbacks=callbacks,
+ max_epochs=EPOCHS,
+ resume_from_checkpoint=resume_from_checkpoint,
+ # debug flags
+ # limit_train_batches=1,
+ # limit_val_batches=1,
+ )
+ return trainer
+
+from bob.paper.ijcb2023_caim_hfr.architectures.common import PDT
+
+translator=PDT(pool_features=int(POOL),use_se=str2bool(SE), use_bias=str2bool(BIAS), use_cbam=str2bool(CBAM))
+
+
+def model_fn():
+ modality = set(s.modality for s in database.background_model_samples())
+ modality.discard("VIS")
+ modality= modality if len(modality)>0 else {'VIS'} # hack to handle VIS-VIS protocol
+ modality = list(modality)[0]
+ print(f"Modality of database was found to be {modality}")
+ model = XIresnetTxl(ARCHITECTURE, translator, modality=modality)
+ loss_fn = ContrastiveLoss(margin=2.0)
+ model = XLitDSUModel(model, loss_fn=loss_fn)
+ return model
+
+
+def datamodule_fn(train_samples):
+ datamodule = AGLitDSUDataModuleNP(
+ batch_size=BATCH_SIZE,
+ train_samples=train_samples,
+ train_transform=train_transform,
+ test_transform=test_transform,
+ )
+ return datamodule
+
+
+embedding = LightningTransformer(
+ trainer_fn=trainer_fn,
+ model_fn=model_fn,
+ datamodule_fn=datamodule_fn,
+ preprocessor=test_transform,
+)
+pipeline.transformer = make_pipeline(preprocessor, embedding)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_PDT.py b/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_PDT.py
new file mode 100644
index 0000000..f1191a5
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/lightning/x_iresnet_PDT.py
@@ -0,0 +1,150 @@
+import os
+
+from functools import partial
+
+import numpy as np
+import pytorch_lightning as pl
+import torch
+import torchvision.transforms as transforms
+
+from pytorch_lightning.callbacks import EarlyStopping
+from pytorch_lightning.callbacks import LearningRateMonitor
+from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.loggers import TensorBoardLogger
+from sklearn.pipeline import make_pipeline
+from sklearn.preprocessing import FunctionTransformer
+
+
+from bob.bio.face.embeddings.pytorch import iresnet_template
+from bob.extension import rc
+from bob.pipelines import CheckpointWrapper
+from bob.pipelines import wrap
+
+
+
+from bob.paper.ijcb2023_caim_hfr.transformer.xlightning import XLitDSUModel
+from bob.paper.ijcb2023_caim_hfr.architectures.iresnet import preprocess_img
+from bob.paper.ijcb2023_caim_hfr.dataset.hface import AGLitDSUDataModule, AGLitDSUDataModuleNP # testing
+from bob.paper.ijcb2023_caim_hfr.losses import ContrastiveLoss
+from bob.paper.ijcb2023_caim_hfr.transformer.lightning import LightningTransformer
+from bob.paper.ijcb2023_caim_hfr.architectures.light_iresnet_txl import XIresnetTxl
+
+database = globals()["database"]
+PROTOCOL = globals()["PROTOCOL"]
+
+POOL ='6'
+SE = 'False'
+CBAM = 'True'
+BIAS = 'False'
+ARCHITECTURE = 'iresnet100'
+
+def str2bool(v):
+ return v.lower() in ("yes", "true", "t", "1")
+
+# print("globals()",globals())
+
+OUTPUT = (
+ output
+) = f"{rc['ptemp']}/{database.name}/{database.protocol}/"
+BATCH_SIZE = 90
+EPOCHS = 20
+
+
+pipeline = iresnet_template(
+ "passthrough", database.annotation_type, database.fixed_positions
+)
+cropper = pipeline.transformer.steps[0][1]
+preprocessor = CheckpointWrapper(
+ cropper,
+ features_dir=f"{rc['ptemp']}/{database.name}/{database.protocol}/"
+ + f"preprocessed_{cropper.estimator.cropped_image_size[0]}",
+)
+
+test_transform = transforms.Compose(
+ [
+ # lambda transfomer based preprocess_img
+ transforms.Lambda(preprocess_img),
+ # convert to tensor
+ transforms.Lambda(partial(torch.as_tensor, dtype=torch.float32)),
+ ]
+)
+
+train_transform = transforms.Compose(
+ [
+ test_transform,
+ # random horizontal flipping
+ transforms.RandomHorizontalFlip(0.5),
+ ]
+)
+
+
+def trainer_fn():
+ # we need tensorboard, best checkpoint, early stopping, and learning rate monitor
+
+ model_dir = f"{OUTPUT}/model"
+
+ callbacks = [
+ ModelCheckpoint(
+ dirpath=model_dir,
+ filename="weights-epoch{epoch:02d}-val_loss{val/loss:.2f}",
+ auto_insert_metric_name=False,
+ monitor="val/loss",
+ mode="min",
+ save_last=True,
+ ),
+ LearningRateMonitor(),
+ EarlyStopping(monitor="val/loss", patience=10, mode="min"),
+ ]
+ logger = TensorBoardLogger(f"{model_dir}/logs", default_hp_metric=False)
+
+ # resume from the last checkpoint if it exists
+ resume_from_checkpoint = f"{model_dir}/last.ckpt"
+ if not os.path.exists(resume_from_checkpoint):
+ resume_from_checkpoint = None
+
+ trainer = pl.Trainer(
+ gpus=-1 if torch.cuda.is_available() else None,
+ logger=logger,
+ callbacks=callbacks,
+ max_epochs=EPOCHS,
+ resume_from_checkpoint=resume_from_checkpoint,
+ # debug flags
+ # limit_train_batches=1,
+ # limit_val_batches=1,
+ )
+ return trainer
+
+from bob.paper.ijcb2023_caim_hfr.architectures.common import PDT
+
+translator=PDT(pool_features=int(POOL),use_se=str2bool(SE), use_bias=str2bool(BIAS), use_cbam=str2bool(CBAM))
+
+
+def model_fn():
+ modality = set(s.modality for s in database.background_model_samples())
+ modality.discard("VIS")
+ modality= modality if len(modality)>0 else {'VIS'} # hack to handle VIS-VIS protocol
+ modality = list(modality)[0]
+ print(f"Modality of database was found to be {modality}")
+ model = XIresnetTxl(ARCHITECTURE, translator, modality=modality)
+ loss_fn = ContrastiveLoss(margin=2.0)
+ model = XLitDSUModel(model, loss_fn=loss_fn)
+ return model
+
+
+def datamodule_fn(train_samples):
+ datamodule = AGLitDSUDataModuleNP(
+ batch_size=BATCH_SIZE,
+ train_samples=train_samples,
+ train_transform=train_transform,
+ test_transform=test_transform,
+ )
+ return datamodule
+
+
+embedding = LightningTransformer(
+ trainer_fn=trainer_fn,
+ model_fn=model_fn,
+ datamodule_fn=datamodule_fn,
+ preprocessor=test_transform,
+)
+pipeline.transformer = make_pipeline(preprocessor, embedding)
diff --git a/bob/paper/ijcb2023_caim_hfr/config/mcxfaceprotocols/VIS-THERMAL-split1.py b/bob/paper/ijcb2023_caim_hfr/config/mcxfaceprotocols/VIS-THERMAL-split1.py
new file mode 100644
index 0000000..7d6906d
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/config/mcxfaceprotocols/VIS-THERMAL-split1.py
@@ -0,0 +1,6 @@
+from bob.paper.ijcb2023_caim_hfr.database import MCXFaceDatabase
+
+PROTOCOL = "VIS-THERMAL-split1"
+
+
+database = MCXFaceDatabase(protocol=PROTOCOL)
\ No newline at end of file
diff --git a/bob/paper/ijcb2023_caim_hfr/database/__init__.py b/bob/paper/ijcb2023_caim_hfr/database/__init__.py
new file mode 100644
index 0000000..64020de
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/__init__.py
@@ -0,0 +1,35 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+
+
+from .mcxface import MCXFaceDatabase
+from .tuftsface import TuftsFaceDatabase
+from .arlvtf import ARLVTFDatabase
+from .cuhk_cufs import CUHK_CUFSDatabase
+from .scface_hfr import SCFaceHFRDatabase
+# gets sphinx autodoc done right - don't remove it
+
+
+def __appropriate__(*args):
+ """Says object was actually declared here, and not in the import module.
+ Fixing sphinx warnings of not being able to find classes, when path is shortened.
+ Parameters:
+
+ *args: An iterable of objects to modify
+
+ Resolves `Sphinx referencing issues
+ <https://github.com/sphinx-doc/sphinx/issues/3048>`
+ """
+
+ for obj in args:
+ obj.__module__ = __name__
+
+
+__appropriate__(
+ MCXFaceDatabase,
+ TuftsFaceDatabase,
+ ARLVTFDatabase,
+ CUHK_CUFSDatabase,
+ SCFaceHFRDatabase,
+)
+__all__ = [_ for _ in dir() if not _.startswith("_")]
diff --git a/bob/paper/ijcb2023_caim_hfr/database/arlvtf.py b/bob/paper/ijcb2023_caim_hfr/database/arlvtf.py
new file mode 100644
index 0000000..84dddf7
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/arlvtf.py
@@ -0,0 +1,212 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+# Anjith George <anjith.george@idiap.ch>
+
+"""
+ ARL-VTF database implementation # iTERATION 1, COULD BE BUGGY #
+ TODO: Scaling for thermal to 8 bit , take care of it here or in the preprocessor:
+ UPDATE 1: For now, converted to 0-255 as float64
+ Handling unpside down: changed the default to true in allow_upside_down_normalized_faces=True bob.bio.face.preprocessor.faceCrop.py
+ #Update 2 : Test subject name to subj0
+ #Update 3 : MinMax scaling in the loader for thermal
+ #Update 4: changed all the protocols now train= train+dev, can filer by the column
+"""
+
+from bob.bio.base.database import CSVDataset
+from bob.bio.base.database import CSVToSampleLoaderBiometrics
+from bob.bio.face.database.sample_loaders import EyesAnnotations
+from bob.extension import rc
+from bob.extension.download import get_file
+import bob.io.base
+from sklearn.pipeline import make_pipeline
+
+import numpy as np
+
+class ARLVTFDatabase(CSVDataset):
+ """
+ Only VIS-THERMAL protocols are implemented
+
+
+ The protocols are as follows: SOURCE_TARGET_split, where the SOURCE
+ is the modality used for enrollment, and TARGET is the modality used
+ as probes. We implement several combinations as the protocols. In addition,
+ we have normal recognition protocols where both source and target are the same
+ modalities. This is implemented using the protocols provided by the authors
+
+ .. warning::
+ Use the command below to set the path of the real data::
+
+ $ bob config set bob.db.arlvtf.directory [PATH-TO-ARL-VTF base directory]
+
+
+
+ Parameters
+ ----------
+
+ protocol: str
+ One of the database protocols.
+ """
+
+ def __init__(self, protocol, annotation_type="eyes-center", fixed_positions=None):
+
+ # Downloading model if not exists
+ urls = ARLVTFDatabase.urls()
+ filename = get_file(
+ "arlvtf.tar.gz", urls, file_hash="c927ad20ec1721d5ab91486837260e4c",
+ )
+
+ self.annotation_type = "eyes-center"
+ self.fixed_positions = None
+
+ directory = (
+ rc["bob.db.arlvtf.directory"]
+ if rc["bob.db.arlvtf.directory"]
+ else ""
+ )
+
+ def load(path):
+ """
+ Images in this dataset are stored as 8-bit jpg
+ """
+ img=bob.io.base.load(path)
+
+ if img.dtype=='uint16': # convert to 'float64' in range 0-255.0
+ #img=img/255.0
+ face_c = np.ma.array(img).compressed()
+ face_c=face_c[face_c>1.0]
+
+ #min-max
+
+ t_min = np.min(face_c)
+ t_max = np.max(face_c)
+
+ data_n = ((img-t_min)/(t_max-t_min))*255.0
+
+
+ #mad
+
+ # n_sigma=4
+
+ # med = np.median(face_c)
+ # mad = np.median(np.abs(face_c - med))
+
+ # data_n = ((img-med+n_sigma*mad)/(2.0*n_sigma*mad))*255.0
+
+ # std
+
+ # n_sigma=4
+
+ # mu = np.mean(face_c)
+ # std = np.std(face_c)
+
+ # data_n=((img-mu+n_sigma*std)/(2.0*n_sigma*std))*255.0
+
+
+
+
+ # Clamping to 0-255
+ data_n=np.maximum(data_n,0)
+ data_n=np.minimum(data_n,255)
+
+ img=data_n.copy()
+
+ # img=255.0-img # invert temp
+
+
+ return img.astype('uint8')
+
+ super().__init__(
+ name="arlvtf",
+ protocol=protocol,
+ dataset_protocol_path=filename,
+ csv_to_sample_loader=make_pipeline(
+ CSVToSampleLoaderBiometrics(
+ data_loader=load,
+ dataset_original_directory=directory,
+ extension=".png",
+ ),
+ EyesAnnotations(),
+ ),
+ annotation_type=annotation_type,
+ fixed_positions=fixed_positions,
+ )
+
+ @staticmethod
+ def protocols():
+ return ['VIS-THERMAL-G_VB0--P_TB0-split0',
+ 'VIS-THERMAL-G_VB0--P_TB--split0',
+ 'VIS-THERMAL-G_VB0--P_TB+-split0',
+ 'VIS-THERMAL-G_VB0--P_TE0-split0',
+ 'VIS-THERMAL-G_VB0--P_TE--split0',
+ 'VIS-THERMAL-G_VB0--P_TP0-split0',
+ 'VIS-THERMAL-G_VB0--P_TP--split0',
+ 'VIS-THERMAL-G_VB0+-P_TB0-split0',
+ 'VIS-THERMAL-G_VB0+-P_TB--split0',
+ 'VIS-THERMAL-G_VB0+-P_TB+-split0',
+ 'VIS-THERMAL-G_VB0+-P_TE0-split0',
+ 'VIS-THERMAL-G_VB0+-P_TE--split0',
+ 'VIS-THERMAL-G_VB0+-P_TP0-split0',
+ 'VIS-THERMAL-G_VB0+-P_TP--split0',
+ 'VIS-THERMAL-G_VB0--P_TB0-split1',
+ 'VIS-THERMAL-G_VB0--P_TB--split1',
+ 'VIS-THERMAL-G_VB0--P_TB+-split1',
+ 'VIS-THERMAL-G_VB0--P_TE0-split1',
+ 'VIS-THERMAL-G_VB0--P_TE--split1',
+ 'VIS-THERMAL-G_VB0--P_TP0-split1',
+ 'VIS-THERMAL-G_VB0--P_TP--split1',
+ 'VIS-THERMAL-G_VB0+-P_TB0-split1',
+ 'VIS-THERMAL-G_VB0+-P_TB--split1',
+ 'VIS-THERMAL-G_VB0+-P_TB+-split1',
+ 'VIS-THERMAL-G_VB0+-P_TE0-split1',
+ 'VIS-THERMAL-G_VB0+-P_TE--split1',
+ 'VIS-THERMAL-G_VB0+-P_TP0-split1',
+ 'VIS-THERMAL-G_VB0+-P_TP--split1',
+ 'VIS-THERMAL-G_VB0--P_TB0-split2',
+ 'VIS-THERMAL-G_VB0--P_TB--split2',
+ 'VIS-THERMAL-G_VB0--P_TB+-split2',
+ 'VIS-THERMAL-G_VB0--P_TE0-split2',
+ 'VIS-THERMAL-G_VB0--P_TE--split2',
+ 'VIS-THERMAL-G_VB0--P_TP0-split2',
+ 'VIS-THERMAL-G_VB0--P_TP--split2',
+ 'VIS-THERMAL-G_VB0+-P_TB0-split2',
+ 'VIS-THERMAL-G_VB0+-P_TB--split2',
+ 'VIS-THERMAL-G_VB0+-P_TB+-split2',
+ 'VIS-THERMAL-G_VB0+-P_TE0-split2',
+ 'VIS-THERMAL-G_VB0+-P_TE--split2',
+ 'VIS-THERMAL-G_VB0+-P_TP0-split2',
+ 'VIS-THERMAL-G_VB0+-P_TP--split2',
+ 'VIS-THERMAL-G_VB0--P_TB0-split3',
+ 'VIS-THERMAL-G_VB0--P_TB--split3',
+ 'VIS-THERMAL-G_VB0--P_TB+-split3',
+ 'VIS-THERMAL-G_VB0--P_TE0-split3',
+ 'VIS-THERMAL-G_VB0--P_TE--split3',
+ 'VIS-THERMAL-G_VB0--P_TP0-split3',
+ 'VIS-THERMAL-G_VB0--P_TP--split3',
+ 'VIS-THERMAL-G_VB0+-P_TB0-split3',
+ 'VIS-THERMAL-G_VB0+-P_TB--split3',
+ 'VIS-THERMAL-G_VB0+-P_TB+-split3',
+ 'VIS-THERMAL-G_VB0+-P_TE0-split3',
+ 'VIS-THERMAL-G_VB0+-P_TE--split3',
+ 'VIS-THERMAL-G_VB0+-P_TP0-split3',
+ 'VIS-THERMAL-G_VB0+-P_TP--split3',
+ 'VIS-THERMAL-G_VB0--P_TB0-split4',
+ 'VIS-THERMAL-G_VB0--P_TB--split4',
+ 'VIS-THERMAL-G_VB0--P_TB+-split4',
+ 'VIS-THERMAL-G_VB0--P_TE0-split4',
+ 'VIS-THERMAL-G_VB0--P_TE--split4',
+ 'VIS-THERMAL-G_VB0--P_TP0-split4',
+ 'VIS-THERMAL-G_VB0--P_TP--split4',
+ 'VIS-THERMAL-G_VB0+-P_TB0-split4',
+ 'VIS-THERMAL-G_VB0+-P_TB--split4',
+ 'VIS-THERMAL-G_VB0+-P_TB+-split4',
+ 'VIS-THERMAL-G_VB0+-P_TE0-split4',
+ 'VIS-THERMAL-G_VB0+-P_TE--split4',
+ 'VIS-THERMAL-G_VB0+-P_TP0-split4',
+ 'VIS-THERMAL-G_VB0+-P_TP--split4']
+
+ @staticmethod
+ def urls():
+ return [
+ "https://www.idiap.ch/~ageorge/arlvtf.tar.gz",
+ "https://www.idiap.ch/~ageorge/arlvtf.tar.gz",
+ ]
diff --git a/bob/paper/ijcb2023_caim_hfr/database/cuhk_cufs.py b/bob/paper/ijcb2023_caim_hfr/database/cuhk_cufs.py
new file mode 100644
index 0000000..3d97f4b
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/cuhk_cufs.py
@@ -0,0 +1,100 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+# Anjith George <anjith.george@idiap.ch>
+
+from bob.bio.base.database import CSVDataset
+from bob.bio.base.database import CSVToSampleLoaderBiometrics
+from bob.bio.face.database.sample_loaders import EyesAnnotations
+from bob.extension import rc
+from bob.extension.download import get_file
+import bob.io.base
+from sklearn.pipeline import make_pipeline
+
+import numpy as np
+
+class CUHK_CUFSDatabase(CSVDataset):
+ """
+ CUHK CUFS database ported
+
+ .. warning::
+ Use the command below to set the path of the real data::
+
+ $ bob config set bob.db.cuhk_cufs.directory "/idiap/resource/database/CUHK-CUFS/"
+
+
+
+ Parameters
+ ----------
+
+ protocol: str
+ One of the database protocols.
+ """
+
+ def __init__(self, protocol, annotation_type="eyes-center", fixed_positions=None):
+
+ # Downloading model if not exists
+ urls = CUHK_CUFSDatabase.urls()
+ filename = get_file(
+ "cuhk_cufs.tar.gz", urls, file_hash="3e1587a95c3438cee3b92882ac43d8bd",
+ )
+
+ self.annotation_type = "eyes-center"
+ self.fixed_positions = None
+ directory="/idiap/resource/database/"
+
+ # directory = (
+ # rc["bob.db.cuhk_cufs.directory"]
+ # if rc["bob.db.cuhk_cufs.directory"]
+ # else ""
+ # )
+
+ def load(path):
+ """
+ Images in this dataset are stored as 8-bit jpg
+ """
+ img=bob.io.base.load(path)
+ return img
+
+ super().__init__(
+ name="cuhk_cufs",
+ protocol=protocol,
+ dataset_protocol_path=filename,
+ csv_to_sample_loader=make_pipeline(
+ CSVToSampleLoaderBiometrics(
+ data_loader=load,
+ dataset_original_directory=directory,
+ extension="",
+ ),
+ EyesAnnotations(),
+ ),
+ annotation_type=annotation_type,
+ fixed_positions=fixed_positions,
+ )
+
+ @staticmethod
+ def protocols():
+ return ['all-mixed_p2s',
+ 'all-mixed_s2p',
+ 'arface_p2s',
+ 'arface_s2p',
+ 'cuhk_p2s',
+ 'cuhk_s2p',
+ 'search_split1_p2s',
+ 'search_split1_s2p',
+ 'search_split2_p2s',
+ 'search_split2_s2p',
+ 'search_split3_p2s',
+ 'search_split3_s2p',
+ 'search_split4_p2s',
+ 'search_split4_s2p',
+ 'search_split5_p2s',
+ 'search_split5_s2p',
+ 'xm2vts_p2s',
+ 'xm2vts_s2p']
+
+ @staticmethod
+ def urls():
+ return [
+ "https://www.idiap.ch/~ageorge/cuhk_cufs.tar.gz",
+ "https://www.idiap.ch/~ageorge/cuhk_cufs.tar.gz",
+ ]
diff --git a/bob/paper/ijcb2023_caim_hfr/database/mcxface.py b/bob/paper/ijcb2023_caim_hfr/database/mcxface.py
new file mode 100644
index 0000000..cbc02d6
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/mcxface.py
@@ -0,0 +1,229 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+# Anjith George <anjith.george@idiap.ch>
+
+"""
+ MCXFace database implementation
+"""
+
+from bob.bio.base.database import CSVDataset
+from bob.bio.base.database import CSVToSampleLoaderBiometrics
+from bob.bio.face.database.sample_loaders import EyesAnnotations
+from bob.extension import rc
+from bob.extension.download import get_file
+import bob.io.base
+from sklearn.pipeline import make_pipeline
+import numpy as np
+from scipy import ndimage
+import bob.io.image
+
+class MCXFaceDatabase(CSVDataset):
+ """
+ Collected at Idiap for the BATL ODIN project, the MCXFace is derived from the HQ-WMCA dataset.
+ The database implements several Face Recognition
+ protocols, ie between the same modalities and heterogeneous face recognition as well.
+ The database has only the train and dev splits due to the limited number of subjects.
+ A total of 51 subjects are present in the dataset collected across several sessions. 30 subjects are in
+ train fold and 20 subjects are in dev fold.
+ The dataset contains the following channels:
+
+ VIS: From Basler Camera
+ DEPTH: From Intel D415
+ THERMAL: From Xenics Gobi
+ NIR: Is again from Basler NIR camera, we use wavelength 850nm
+ SWIR: From Xenics Bobcat we use only 1300nm wavelength.
+
+ All the channels are spatially and temporally registered meaning, one
+ can share the annotations provided between channels. The left and right
+ eye centers are provided as annotations.
+
+ The protocols are as follows: SOURCE_TARGET_split, where the SOURCE
+ is the modality used for enrollment, and TARGET is the modality used
+ as probes. We implement several combinations as the protocols. In addition,
+ we have normal recognition protocols where both source and target are the same
+ modalities. For each of these, we have also prepared 5 different splits by
+ randomly splitting the clients between train and dev sets. Subjects who have
+ only one session are always assigned to the training fold.
+
+ .. warning::
+ Use the command below to set the path of the real data::
+
+ $ bob config set bob.db.mcxface.directory [PATH-TO-MCXFACE-DATA]
+
+
+
+ Parameters
+ ----------
+
+ protocol: str
+ One of the database protocols.
+ """
+
+
+
+
+
+ def get_depth_normal(self,depth, sigma=2, inv=False, abs=True):
+ """
+ Normalizes the depth image after computing normals
+ returns a 3d image with normals with is uint8 format
+ NB: increased sigma say upto 10 seems good may be 2-3 optimum.
+ Inversion doesnt seem to have much effect
+ Left right not invariant, so flip variant this is an issue.
+
+
+ """
+
+ assert(len(depth.shape)==2)
+
+ depth=depth.astype('float')
+
+ if sigma!=0:
+ depth=ndimage.gaussian_filter(depth, sigma)
+
+ sx = ndimage.sobel(depth,axis=0,mode='constant')
+ sy = ndimage.sobel(depth,axis=1,mode='constant')
+
+
+
+ sqrt=np.sqrt(np.power(sx,2)+np.power(sy,2)+1)
+
+ three_channel=np.dstack([sx,sy,np.ones(sx.shape)])
+ three_channel_sqrt=np.dstack([sqrt,sqrt,sqrt])
+
+ tnorm=np.divide(three_channel,three_channel_sqrt)
+
+ if abs:
+ tnorm=(np.absolute(tnorm))*255.0
+ else:
+ tnorm=((tnorm+1.0)/2.0)*255.0
+
+ tnorm=np.maximum(tnorm,0)
+ tnorm=np.minimum(tnorm,255)
+
+ tnorm=tnorm.astype('uint8')
+
+ if inv:
+ tnorm=255-tnorm
+
+
+
+ return bob.io.image.to_bob(tnorm)
+
+ def __init__(self, protocol, annotation_type="eyes-center", fixed_positions=None, sigma=1,abs=False, inv=False):
+
+ # Downloading model if not exists
+ urls = MCXFaceDatabase.urls()
+ filename = get_file(
+ "mcxface.tar.gz", urls, file_hash="3624815522a3d42e87da619101f92258",
+ )
+
+ self.annotation_type = "eyes-center"
+ self.fixed_positions = None
+ self.sigma=sigma
+ self.abs=abs
+ self.inv=inv
+
+
+ print("MCXFACE sigma, inv, abs:",self.sigma, self.inv, self.abs )
+
+ directory = (
+ rc["bob.db.mcxface.directory"]
+ if rc["bob.db.mcxface.directory"]
+ else ""
+ )
+
+ def load(path):
+ """
+ Images in this dataset are stored as 8-bit jpg
+ """
+ # if 'DEPTH' in path:
+ # data= bob.io.base.load(path)
+ # return self.get_depth_normal(data, sigma=self.sigma, inv=self.inv, abs=self.abs)
+ # else:
+
+ return bob.io.base.load(path)
+
+ super().__init__(
+ name="mcxface",
+ protocol=protocol,
+ dataset_protocol_path=filename,
+ csv_to_sample_loader=make_pipeline(
+ CSVToSampleLoaderBiometrics(
+ data_loader=load,
+ dataset_original_directory=directory,
+ extension=".jpg",
+ ),
+ EyesAnnotations(),
+ ),
+ annotation_type=annotation_type,
+ fixed_positions=fixed_positions,
+ )
+
+ @staticmethod
+ def protocols():
+ return ['VIS-VIS2-split1',
+ 'VIS-VIS-split1',
+ 'VIS-STEREO-split1',
+ 'VIS-STEREO-split2',
+ 'VIS-STEREO-split3',
+ 'VIS-STEREO-split4',
+ 'VIS-STEREO-split5',
+ 'DEPTH-DEPTH-split1',
+ '3DDFA-3DDFA-split1',
+ 'THERMAL-THERMAL-split1',
+ 'SWIR-SWIR-split1',
+ 'NIR-NIR-split1',
+ 'VIS-DEPTH-split1',
+ 'VIS-3DDFA-split1',
+ 'VIS-THERMAL-split1',
+ 'VIS-SWIR-split1',
+ 'VIS-NIR-split1',
+ 'VIS-VIS-split2',
+ 'DEPTH-DEPTH-split2',
+ 'THERMAL-THERMAL-split2',
+ 'SWIR-SWIR-split2',
+ 'NIR-NIR-split2',
+ 'VIS-DEPTH-split2',
+ 'VIS-3DDFA-split2',
+ 'VIS-THERMAL-split2',
+ 'VIS-SWIR-split2',
+ 'VIS-NIR-split2',
+ 'VIS-VIS-split3',
+ 'DEPTH-DEPTH-split3',
+ 'THERMAL-THERMAL-split3',
+ 'SWIR-SWIR-split3',
+ 'NIR-NIR-split3',
+ 'VIS-DEPTH-split3',
+ 'VIS-3DDFA-split3',
+ 'VIS-THERMAL-split3',
+ 'VIS-SWIR-split3',
+ 'VIS-NIR-split3',
+ 'VIS-VIS-split4',
+ 'DEPTH-DEPTH-split4',
+ 'THERMAL-THERMAL-split4',
+ 'SWIR-SWIR-split4',
+ 'NIR-NIR-split4',
+ 'VIS-DEPTH-split4',
+ 'VIS-3DDFA-split4',
+ 'VIS-THERMAL-split4',
+ 'VIS-SWIR-split4',
+ 'VIS-NIR-split4',
+ 'VIS-VIS-split5',
+ 'DEPTH-DEPTH-split5',
+ 'THERMAL-THERMAL-split5',
+ 'SWIR-SWIR-split5',
+ 'NIR-NIR-split5',
+ 'VIS-DEPTH-split5',
+ 'VIS-3DDFA-split5',
+ 'VIS-THERMAL-split5',
+ 'VIS-SWIR-split5',
+ 'VIS-NIR-split5']
+
+
+ @staticmethod
+ def urls():
+ return [
+ "https://www.idiap.ch/~ageorge/mcxface.tar.gz",
+ "https://www.idiap.ch/~ageorge/mcxface.tar.gz",
+ ]
diff --git a/bob/paper/ijcb2023_caim_hfr/database/scface_hfr.py b/bob/paper/ijcb2023_caim_hfr/database/scface_hfr.py
new file mode 100644
index 0000000..e0d3b9f
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/scface_hfr.py
@@ -0,0 +1,94 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+# Anjith George <anjith.george@idiap.ch>
+
+"""
+ MCXFace database implementation
+"""
+
+from bob.bio.base.database import CSVDataset
+from bob.bio.base.database import CSVToSampleLoaderBiometrics
+from bob.bio.face.database.sample_loaders import EyesAnnotations
+from bob.extension import rc
+from bob.extension.download import get_file
+import bob.io.base
+from sklearn.pipeline import make_pipeline
+import numpy as np
+from scipy import ndimage
+import bob.io.image
+import os
+class SCFaceHFRDatabase(CSVDataset):
+ """
+ SCFACE one more column with modality is added which is VIS for enrollment RGB mugshot and SURVEILLANCE if its not a mugshot
+
+ .. warning::
+ Use the command below to set the path of the real data::
+
+ $ bob config set bob.db.scface_hfr.directory /idiap/resource/database/scface/SCface_database/
+
+
+
+ Parameters
+ ----------
+
+ protocol: str
+ One of the database protocols.
+ """
+ def __init__(self, protocol, annotation_type="eyes-center", fixed_positions=None):
+
+ # Downloading model if not exists
+ urls = SCFaceHFRDatabase.urls()
+ filename = get_file(
+ "scface_hfr.tar.gz", urls, file_hash="bf8d33cfbb3e5b4ac4ba844b12b24ba8",
+ )
+
+ self.annotation_type = "eyes-center"
+ self.fixed_positions = None
+
+
+ def load(filename):
+ extensions = [".jpg", ".JPG"]
+ for e in extensions:
+ f = os.path.splitext(filename)[0]
+ new_filename = f + e
+ if os.path.exists(new_filename):
+ return bob.io.base.load(new_filename)
+ else:
+ raise ValueError("File `{0}` not found".format(str(new_filename)))
+
+
+ super().__init__(
+ name="scface_hfr",
+ protocol=protocol,
+ dataset_protocol_path=filename,
+ csv_to_sample_loader=make_pipeline(
+ CSVToSampleLoaderBiometrics(
+ data_loader=load,
+ dataset_original_directory=rc.get(
+ "bob.db.scface_hfr.directory", ""
+ ),
+ extension=".jpg",
+ ),
+ EyesAnnotations(),
+ ),
+ annotation_type=annotation_type,
+ fixed_positions=fixed_positions,
+ allow_scoring_with_all_biometric_references=True,
+ )
+
+ @staticmethod
+ def protocols():
+ return ["close", "medium", "far", "combined"]
+
+ @staticmethod
+ def urls():
+ return [
+ "https://www.idiap.ch/~ageorge/scface_hfr.tar.gz",
+ "https://www.idiap.ch/~ageorge/scface_hfr.tar.gz",
+ ]
+
+
+# Trying to debug, changed the extensions from filepath and now trying again see: did not work
+# tryiong with ,drop_last=True on the dataloaders to see, its not the case too
+# issue is probably only one sample for reference modality in training
+# tryiong >1 limit in the subj_sample loader class; THIS WAS THGE ISSUE ALL ALONG
\ No newline at end of file
diff --git a/bob/paper/ijcb2023_caim_hfr/database/tuftsface.py b/bob/paper/ijcb2023_caim_hfr/database/tuftsface.py
new file mode 100644
index 0000000..e932aa2
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/database/tuftsface.py
@@ -0,0 +1,98 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+# Anjith George <anjith.george@idiap.ch>
+
+"""
+ MCXFace database implementation
+"""
+
+from bob.bio.base.database import CSVDataset
+from bob.bio.base.database import CSVToSampleLoaderBiometrics
+from bob.bio.face.database.sample_loaders import EyesAnnotations
+from bob.extension import rc
+from bob.extension.download import get_file
+import bob.io.base
+from sklearn.pipeline import make_pipeline
+
+
+class TuftsFaceDatabase(CSVDataset):
+ """
+ Only VIS-THERMAL protocols are implemented
+
+
+ The protocols are as follows: SOURCE_TARGET_split, where the SOURCE
+ is the modality used for enrollment, and TARGET is the modality used
+ as probes. We implement several combinations as the protocols. In addition,
+ we have normal recognition protocols where both source and target are the same
+ modalities. For each of these, we have also prepared 5 different splits by
+ randomly splitting the clients between train and dev sets. Subjects who have
+ only one session are always assigned to the training fold.
+
+ .. warning::
+ Use the command below to set the path of the real data::
+
+ $ bob config set bob.db.tufts_face.directory [PATH-TO-TUFTS FACE base directory]
+
+
+
+ Parameters
+ ----------
+
+ protocol: str
+ One of the database protocols.
+ """
+
+ def __init__(self, protocol, annotation_type="eyes-center", fixed_positions=None):
+
+ # Downloading model if not exists
+ urls = TuftsFaceDatabase.urls()
+ filename = get_file(
+ "tufts_thermal.tar.gz", urls, file_hash="b4ca445f72b436f14be23984234d9e1e",
+ )
+
+ self.annotation_type = "eyes-center"
+ self.fixed_positions = None
+
+ directory = (
+ rc["bob.db.tufts_face.directory"]
+ if rc["bob.db.tufts_face.directory"]
+ else ""
+ )
+
+ def load(path):
+ """
+ Images in this dataset are stored as 8-bit jpg
+ """
+ return bob.io.base.load(path)
+
+ super().__init__(
+ name="tuftsface",
+ protocol=protocol,
+ dataset_protocol_path=filename,
+ csv_to_sample_loader=make_pipeline(
+ CSVToSampleLoaderBiometrics(
+ data_loader=load,
+ dataset_original_directory=directory,
+ extension=".jpg",
+ ),
+ EyesAnnotations(),
+ ),
+ annotation_type=annotation_type,
+ fixed_positions=fixed_positions,
+ )
+
+ @staticmethod
+ def protocols():
+ return ['VIS-THERMAL-split1',
+ 'VIS-THERMAL-split2',
+ 'VIS-THERMAL-split3',
+ 'VIS-THERMAL-split4',
+ 'VIS-THERMAL-split5']
+
+
+ @staticmethod
+ def urls():
+ return [
+ "https://www.idiap.ch/~ageorge/tufts_thermal.tar.gz",
+ "https://www.idiap.ch/~ageorge/tufts_thermal.tar.gz",
+ ]
diff --git a/bob/paper/ijcb2023_caim_hfr/dataset/__init__.py b/bob/paper/ijcb2023_caim_hfr/dataset/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/bob/paper/ijcb2023_caim_hfr/dataset/datasets.py b/bob/paper/ijcb2023_caim_hfr/dataset/datasets.py
new file mode 100644
index 0000000..e30008f
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/dataset/datasets.py
@@ -0,0 +1,55 @@
+import itertools
+
+from collections import defaultdict
+import numpy as np
+
+np.random.seed(2021)
+
+
+def subj_samples_to_heterogeneous_pairs(samples, required_pairs=10000):
+ """Takes a list of samples and converts them to all possible gen and zei pairs.
+ In each pair exists a VIS and another modality sample.
+ """
+ # we want to create all possible genuine and zei pairs
+ # label for genuine is 1 and label for zei is 0
+
+ # generate all possible gen and zei pairs
+
+
+
+ samples_per_modality = defaultdict(list)
+ # make sure VIS is alwaus left
+ samples_per_modality["VIS"] = []
+ for s in samples:
+ samples_per_modality[s.modality].append(s)
+
+ unique_reference_ids=np.unique([s.reference_id for s in samples_per_modality['VIS']])
+
+ target_modality=list(set(samples_per_modality.keys())-{'VIS'})[0]
+
+ ntimes=required_pairs//(2*len(unique_reference_ids))
+
+ print('subj-unique_reference_ids:::', len(unique_reference_ids))
+
+
+ gen_pairs, zei_pairs = [], []
+ # label 1 for positive (genuine) as expected in:
+ # https://github.com/tensorflow/addons/blob/d26e2ed5f68092aed57016a7005ce534b1be3dce/tensorflow_addons/losses/contrastive.py#L67
+
+
+ for reference_id in unique_reference_ids:
+
+ src_match=[s for s in samples_per_modality['VIS'] if s.reference_id == reference_id]
+
+ target_match=[s for s in samples_per_modality[target_modality] if s.reference_id == reference_id]
+
+ target_non_match=[s for s in samples_per_modality[target_modality] if s.reference_id != reference_id]
+
+ #if len(src_match)>1 and len(target_match)>1 and len(target_non_match)>1: changed for scface
+ if len(src_match)>0 and len(target_match)>0 and len(target_non_match)>0:
+ for i in range(ntimes):
+ gen_pairs.append(((np.random.choice(src_match, 1, replace=False)[0], np.random.choice(target_match, 1, replace=False)[0]), 1)) # genuine
+ zei_pairs.append(((np.random.choice(src_match, 1, replace=False)[0],np.random.choice(target_non_match, 1, replace=False)[0]), 0)) # negative pair
+
+ return gen_pairs, zei_pairs, list(samples_per_modality.keys())
+
diff --git a/bob/paper/ijcb2023_caim_hfr/dataset/hface.py b/bob/paper/ijcb2023_caim_hfr/dataset/hface.py
new file mode 100644
index 0000000..d51e01d
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/dataset/hface.py
@@ -0,0 +1,424 @@
+import torch
+from bob.paper.ijcb2023_caim_hfr.dataset.datasets import subj_samples_to_heterogeneous_pairs
+import random
+import pytorch_lightning as pl
+from sklearn.model_selection import train_test_split
+import numpy as np
+from torch.utils.data import Dataset, DataLoader
+from multiprocessing import cpu_count
+import random
+
+random.seed(2021)
+
+
+class NormalDownSamplingHeteroPairsDatasetMMD(Dataset):
+ """A dataset that returns a heterogeneous set of pairs of images.
+ It randomly returns zei pairs.
+ """
+
+ def __init__(self, samples, transform=None, num_pairs=100000):
+ self.num_pairs=num_pairs
+ self.samples = samples
+ print('samples::',len(self.samples))
+ self.transform = transform
+ (
+ self.gen_pairs,
+ self.zei_pairs,
+ self.modalities,
+ ) = subj_samples_to_heterogeneous_pairs(samples=samples, required_pairs=self.num_pairs)
+ self.len_zei_pairs, self.len_gen_pairs = len(self.zei_pairs), len(
+ self.gen_pairs
+ )
+
+
+ self.pairs=self.zei_pairs
+
+ print('pairs::::', len(self.pairs))
+
+ random.shuffle(self.pairs)
+
+
+
+
+ def __len__(self):
+ return len(self.pairs)
+
+ def __getitem__(self, idx):
+ # rind=np.random.randint(2)
+
+ # if rind==0:
+ # samples, label = self.gen_pairs[idx%self.len_gen_pairs]
+ # else:
+ # samples, label = self.zei_pairs[idx%self.len_gen_pairs]
+
+ samples, label = self.pairs[idx]
+
+
+ #data is a dictionary data['VIS'] and data['THERMAL']
+ data = {}
+ for mod, sample in zip(self.modalities, samples):
+ d = sample.data
+ if self.transform:
+ d = self.transform(d)
+ data[mod] = d
+
+ return data, label
+
+class NormalDownSamplingHeteroPairsDataset(Dataset):
+ """A dataset that returns a heterogeneous set of pairs of images.
+ It randomly returns zei pairs.
+ """
+
+ def __init__(self, samples, transform=None, num_pairs=100000):
+ self.num_pairs=num_pairs
+ self.samples = samples
+ print('samples::',len(self.samples))
+ self.transform = transform
+ (
+ self.gen_pairs,
+ self.zei_pairs,
+ self.modalities,
+ ) = subj_samples_to_heterogeneous_pairs(samples=samples, required_pairs=self.num_pairs)
+ self.len_zei_pairs, self.len_gen_pairs = len(self.zei_pairs), len(
+ self.gen_pairs
+ )
+
+
+ self.pairs=self.gen_pairs+self.zei_pairs
+
+ print('pairs::::', len(self.pairs))
+
+ random.shuffle(self.pairs)
+
+
+
+
+ def __len__(self):
+ return len(self.pairs)
+
+ def __getitem__(self, idx):
+ # rind=np.random.randint(2)
+
+ # if rind==0:
+ # samples, label = self.gen_pairs[idx%self.len_gen_pairs]
+ # else:
+ # samples, label = self.zei_pairs[idx%self.len_gen_pairs]
+
+ samples, label = self.pairs[idx]
+
+
+ #data is a dictionary data['VIS'] and data['THERMAL']
+ data = {}
+ for mod, sample in zip(self.modalities, samples):
+ d = sample.data
+ if self.transform:
+ d = self.transform(d)
+ data[mod] = d
+
+ return data, label
+
+
+# class ARLLitDSUDataModule(pl.LightningDataModule):
+# def __init__(
+# self,
+# batch_size,
+# train_samples,
+# val_samples,
+# train_transform,
+# test_transform,
+# test_size=0.2,
+# **kwargs,
+# ):
+# super().__init__(**kwargs)
+# self.batch_size = batch_size
+# self.train_transform = train_transform
+# self.test_transform = test_transform
+# self.test_size = test_size
+
+# self.train_samples = train_samples
+# self.val_samples = val_samples
+# # self.train_ids, self.val_ids = train_IDS, val_IDS
+
+# def train_dataloader(self):
+# dataset = DownSamplingHeteroPairsDataset(
+# samples=self.train_samples, transform=self.train_transform
+# )
+# data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+# return data_loader
+
+# def val_dataloader(self):
+# dataset = DownSamplingHeteroPairsDataset(
+# samples=self.val_samples, transform=self.test_transform
+# )
+# data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=False, num_workers=cpu_count())
+# return data_loader
+
+
+class ARLLitDSUDataModule(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ test_size=0.2,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.test_size = test_size
+
+
+
+ self.train_samples = [s for s in train_samples if s.group=='train']
+ self.val_samples = [s for s in train_samples if s.group=='dev']
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.train_samples, transform=self.train_transform, num_pairs=100000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.val_samples, transform=self.test_transform, num_pairs=20000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+
+class ARLLitDSUDataModuleMMD(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ test_size=0.2,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.test_size = test_size
+
+
+
+ self.train_samples = [s for s in train_samples if s.group=='train']
+ self.val_samples = [s for s in train_samples if s.group=='dev']
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDatasetMMD(
+ samples=self.train_samples, transform=self.train_transform, num_pairs=100000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDatasetMMD(
+ samples=self.val_samples, transform=self.test_transform, num_pairs=20000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+
+class ARLLitDSUDataModuleMMDCONTRASTIVE(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ test_size=0.2,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.test_size = test_size
+
+
+
+ self.train_samples = [s for s in train_samples if s.group=='train']
+ self.val_samples = [s for s in train_samples if s.group=='dev']
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.train_samples, transform=self.train_transform, num_pairs=100000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.val_samples, transform=self.test_transform, num_pairs=20000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+class ARLLitDSUDataModule_ablation(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ train_subj_percentage=100.0,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.train_subj_percentage = train_subj_percentage/100.0
+
+ self.val_train_subj_percentage=self.train_subj_percentage
+
+ self.train_samples_all = [s for s in train_samples if s.group=='train']
+ self.val_samples_all = [s for s in train_samples if s.group=='dev']
+
+ train_label_unique_all = list(np.unique([f.subject_id for f in self.train_samples_all]))
+ val_label_unique_all = list(np.unique([f.subject_id for f in self.val_samples_all]))
+
+
+ # subselects the subjects in trainset and valset to what is defined by train_subj_percentage [0,1]
+
+
+ train_IDS, _ = train_test_split(train_label_unique_all, test_size=1.0-self.train_subj_percentage, random_state=1)
+
+
+ if self.train_subj_percentage<0.05:
+ self.val_train_subj_percentage=0.05
+
+
+ val_IDS, _ = train_test_split(val_label_unique_all, test_size=1.0-self.val_train_subj_percentage, random_state=1)
+
+
+ self.train_samples = [s for s in self.train_samples_all if s.subject_id in train_IDS]
+ self.val_samples = [s for s in self.val_samples_all if s.subject_id in val_IDS]
+
+ self.train_ids, self.val_ids = train_IDS, val_IDS
+
+
+ print('self.train_subj_percentage :', self.train_subj_percentage)
+ print('self.train_samples : ', len(self.train_samples), len(self.train_ids))
+
+ print('val_samples : ', len(self.val_samples), len(self.val_ids))
+
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.train_samples, transform=self.train_transform, num_pairs=100000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.val_samples, transform=self.test_transform, num_pairs=20000
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+
+class AGLitDSUDataModule(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ test_size=0.2,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.test_size = test_size
+ org_label_unique = list(np.unique([f.subject_id for f in train_samples]))
+ print('UNIQUE----org_label_unique:::::::::', len(org_label_unique))
+
+ train_IDS, val_IDS = train_test_split(
+ org_label_unique, test_size=self.test_size, random_state=1
+ )
+ self.train_samples = [s for s in train_samples if s.subject_id in train_IDS]
+ self.val_samples = [s for s in train_samples if s.subject_id in val_IDS]
+ self.train_ids, self.val_ids = train_IDS, val_IDS
+
+
+ print("train_IDS:",len(train_IDS), len(self.train_samples),self.train_samples[0])
+ print("val_IDS:",len(val_IDS), len(self.val_samples), self.val_samples[0])
+
+
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.train_samples, transform=self.train_transform
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count())
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.val_samples, transform=self.test_transform
+ )
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=False, num_workers=cpu_count())
+ return data_loader
+
+
+
+class AGLitDSUDataModuleNP(pl.LightningDataModule):
+ def __init__(
+ self,
+ batch_size,
+ train_samples,
+ train_transform,
+ test_transform,
+ test_size=0.2,
+ num_pairs=10000,
+ **kwargs,
+ ):
+ super().__init__(**kwargs)
+ self.batch_size = batch_size
+ self.train_transform = train_transform
+ self.test_transform = test_transform
+ self.test_size = test_size
+ self.num_pairs=num_pairs
+ org_label_unique = list(np.unique([f.subject_id for f in train_samples]))
+ print('UNIQUE----org_label_unique:::::::::', len(org_label_unique))
+
+ train_IDS, val_IDS = train_test_split(
+ org_label_unique, test_size=self.test_size, random_state=1
+ )
+ self.train_samples = [s for s in train_samples if s.subject_id in train_IDS]
+ self.val_samples = [s for s in train_samples if s.subject_id in val_IDS]
+ self.train_ids, self.val_ids = train_IDS, val_IDS
+
+
+ print("train_IDS:",len(train_IDS), len(self.train_samples),self.train_samples[0])
+ print("val_IDS:",len(val_IDS), len(self.val_samples), self.val_samples[0])
+
+
+
+ def train_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.train_samples, transform=self.train_transform, num_pairs=self.num_pairs
+ )
+ print("dataset", dir(dataset))
+
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=True, num_workers=cpu_count(),drop_last=False)
+ return data_loader
+
+ def val_dataloader(self):
+ dataset = NormalDownSamplingHeteroPairsDataset(
+ samples=self.val_samples, transform=self.test_transform, num_pairs=self.num_pairs
+ )
+ print("dataset", dir(dataset))
+ data_loader = DataLoader(dataset, batch_size=self.batch_size, shuffle=False, num_workers=cpu_count(),drop_last=False)
+ return data_loader
diff --git a/bob/paper/ijcb2023_caim_hfr/losses/__init__.py b/bob/paper/ijcb2023_caim_hfr/losses/__init__.py
new file mode 100644
index 0000000..83f802d
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/losses/__init__.py
@@ -0,0 +1,2 @@
+from .loss import ContrastiveLoss
+from .loss import MMD_loss
\ No newline at end of file
diff --git a/bob/paper/ijcb2023_caim_hfr/losses/loss.py b/bob/paper/ijcb2023_caim_hfr/losses/loss.py
new file mode 100644
index 0000000..7aa5ec1
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/losses/loss.py
@@ -0,0 +1,63 @@
+import torch
+import torch.nn.functional as F
+import torch.nn as nn
+
+from torch.nn.modules.loss import _Loss
+from torch import Tensor
+
+class ContrastiveLoss(_Loss):
+ """
+ Contrastive loss function.
+ Based on: http://yann.lecun.com/exdb/publis/pdf/hadsell-chopra-lecun-06.pdf
+ """
+
+ def __init__(self, margin: float = 2.0, reduction: str = "mean"):
+ super().__init__(reduction=reduction)
+ assert self.reduction in ("none", "mean", "sum"), self.reduction
+ self.margin = margin
+
+ def forward(self, dist, label):
+
+ loss = label * torch.pow(dist, 2) + (1 - label) * torch.pow(
+ torch.clamp(self.margin - dist, min=0.0), 2
+ )
+
+ if self.reduction == "none":
+ return loss
+ return getattr(torch, self.reduction)(loss)
+
+
+class MMD_loss(nn.Module):
+ """
+ MMD loss function. referenced from: https://github.com/ZongxianLee/MMD_Loss.Pytorch"""
+ def __init__(self, kernel_mul = 2.0, kernel_num = 5):
+ super(MMD_loss, self).__init__()
+ self.kernel_num = kernel_num
+ self.kernel_mul = kernel_mul
+ self.fix_sigma = None
+ return
+ def guassian_kernel(self, source, target, kernel_mul=2.0, kernel_num=5, fix_sigma=None):
+ n_samples = int(source.size()[0])+int(target.size()[0])
+ total = torch.cat([source, target], dim=0)
+
+ total0 = total.unsqueeze(0).expand(int(total.size(0)), int(total.size(0)), int(total.size(1)))
+ total1 = total.unsqueeze(1).expand(int(total.size(0)), int(total.size(0)), int(total.size(1)))
+ L2_distance = ((total0-total1)**2).sum(2)
+ if fix_sigma:
+ bandwidth = fix_sigma
+ else:
+ bandwidth = torch.sum(L2_distance.data) / (n_samples**2-n_samples)
+ bandwidth /= kernel_mul ** (kernel_num // 2)
+ bandwidth_list = [bandwidth * (kernel_mul**i) for i in range(kernel_num)]
+ kernel_val = [torch.exp(-L2_distance / bandwidth_temp) for bandwidth_temp in bandwidth_list]
+ return sum(kernel_val)
+
+ def forward(self, source, target):
+ batch_size = int(source.size()[0])
+ kernels = self.guassian_kernel(source, target, kernel_mul=self.kernel_mul, kernel_num=self.kernel_num, fix_sigma=self.fix_sigma)
+ XX = kernels[:batch_size, :batch_size]
+ YY = kernels[batch_size:, batch_size:]
+ XY = kernels[:batch_size, batch_size:]
+ YX = kernels[batch_size:, :batch_size]
+ loss = torch.mean(XX + YY - XY -YX)
+ return loss
\ No newline at end of file
diff --git a/bob/paper/ijcb2023_caim_hfr/scripts/__init__.py b/bob/paper/ijcb2023_caim_hfr/scripts/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py b/bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py
new file mode 100644
index 0000000..bd2f709
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py
@@ -0,0 +1,164 @@
+
+import numpy as np
+
+import os
+
+import bob.core
+import pandas as pd
+
+import argparse
+
+import yaml
+
+from datetime import datetime
+
+# datetime object containing current date and time
+import bob.measure
+
+
+import bob.bio.base
+
+
+def read_as_dict(filename):
+ data=[]
+ with open(filename, 'r') as stream:
+ try:
+ data = yaml.safe_load(stream)
+ except yaml.YAMLError as exc:
+ print(exc)
+ return data
+
+
+def parse_arguments(cmd_params=None):
+ """
+ Parse the command line arguments.
+
+ **Returns:**
+
+ ``file_name``: py:class:`string`
+ An absolute path of the score file.
+
+ ``legends``: py:class:`string`
+ Legends of the score files.
+
+ ``save_path``: py:class:`string`
+ An absolute base path for saving the performance table and ROCs.
+
+
+ """
+
+ parser = argparse.ArgumentParser(description=__doc__)
+
+ parser.add_argument("-sf", "--score-files", nargs="+", help="A list of score files in Bob 4-column format",
+ default=[""])
+ parser.add_argument("-s", "--save-path", type=str, help="An absolute base path for saving the performance tables.",
+ default = "performance.yaml")
+ parser.add_argument("-u", "--uid", type=str, help="UID for the experiment",
+ default = "")
+
+
+ if cmd_params is not None:
+ args = parser.parse_args(cmd_params)
+ else:
+ args = parser.parse_args()
+
+ score_dev = args.score_files[0]
+ save_path = args.save_path
+ uid = args.uid
+
+ return score_dev, save_path, uid
+
+
+
+
+def main(cmd_params=None):
+
+ """
+ ./bin/get_metrics.py --h for help.
+
+ ./bin/python ..scripts/hface_metrics.py -sf ../scores-dev.csv
+
+
+ Sample usage: . """
+
+ fname,save_path, uid = parse_arguments( cmd_params=cmd_params)
+
+ all_metrics={}
+
+
+ negatives, positives=bob.bio.base.score.load.split(fname)
+
+ rr_scores=bob.bio.base.score.load.cmc(fname)
+
+ R1 = bob.measure.recognition_rate(rr_scores, rank=1) *100
+
+ EER=bob.measure.eer(negatives, positives, is_sorted=False, also_farfrr=False)*100
+
+ AUC=bob.measure.roc_auc_score(negatives, positives, npoints=2000, min_far=- 8, log_scale=False)*100
+
+ # Verification rates
+
+ thresh_1=bob.measure.far_threshold(negatives,positives,0.01)
+
+ far_1,frr_1=bob.measure.fprfnr(negatives,positives,thresh_1)
+
+ VR_1=100-frr_1*100
+
+
+ thresh_5=bob.measure.far_threshold(negatives,positives,0.05)
+
+ far_5,frr_5=bob.measure.fprfnr(negatives,positives,thresh_5)
+
+ VR_5=100-frr_5*100
+
+ thresh_0_1=bob.measure.far_threshold(negatives,positives,0.001)
+
+ far_0_1,frr_0_1=bob.measure.fprfnr(negatives,positives,thresh_0_1)
+
+ VR_0_1=100-frr_0_1*100
+
+ thresh_0_0_1=bob.measure.far_threshold(negatives,positives,0.0001)
+
+ far_0_0_1,frr_0_0_1=bob.measure.fprfnr(negatives,positives,thresh_0_0_1)
+
+ VR_0_0_1=100-frr_0_0_1*100
+
+
+ metrics={}
+ metrics['AUC']=float("{:.2f}".format(AUC))
+ metrics['EER']=float("{:.2f}".format(EER))
+ metrics['VR_5']=float("{:.2f}".format(VR_5))
+ metrics['VR_1']=float("{:.2f}".format(VR_1))
+ metrics['VR_0_1']=float("{:.2f}".format(VR_0_1))
+ metrics['VR_0_0_1']=float("{:.2f}".format(VR_0_0_1))
+ metrics['R1']=float("{:.2f}".format(R1))
+
+ print('Metrics:', metrics)
+
+ now = datetime.now()
+
+ dt_string = now.strftime("%d/%m/%Y %H:%M:%S")
+
+ metrics['time'] = dt_string
+
+ all_metrics[uid]=metrics
+
+ with open(save_path, 'a') as file:
+ documents = yaml.dump(all_metrics, file)
+
+
+ # Dump the result as CSV: can do the same with tex as well.
+
+ results = read_as_dict(save_path)
+
+ csv_save_path = save_path.replace('yaml','csv')
+
+ df = pd.DataFrame.from_dict(results).T
+
+ df.to_csv(csv_save_path)
+
+
+if __name__ == "__main__":
+ main(cmd_params=None)
+
+
diff --git a/bob/paper/ijcb2023_caim_hfr/scripts/version.py b/bob/paper/ijcb2023_caim_hfr/scripts/version.py
new file mode 100644
index 0000000..6d1f941
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/scripts/version.py
@@ -0,0 +1,13 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+
+"""
+Print the text
+"""
+
+def main():
+ """Print the text"""
+
+ print ("Print test text")
+ return 0
+
diff --git a/bob/paper/ijcb2023_caim_hfr/test.py b/bob/paper/ijcb2023_caim_hfr/test.py
new file mode 100644
index 0000000..50daa58
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/test.py
@@ -0,0 +1,9 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+
+"""Test Units
+"""
+
+def test_version():
+ from .scripts import version
+ assert version.main() == 0
diff --git a/bob/paper/ijcb2023_caim_hfr/transformer/__init__.py b/bob/paper/ijcb2023_caim_hfr/transformer/__init__.py
new file mode 100644
index 0000000..2d27515
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/transformer/__init__.py
@@ -0,0 +1,12 @@
+from sklearn.preprocessing import FunctionTransformer
+
+def per_modality_transform(samples, estimators):
+ new_samples = []
+ for s in samples:
+ estimator = estimators[s.modality]
+ new_samples += estimator.transform([s])
+ return new_samples
+
+
+def PerModalityTransformer(estimators, **kwargs):
+ return FunctionTransformer(per_modality_transform, kw_args=dict(estimators=estimators), **kwargs)
diff --git a/bob/paper/ijcb2023_caim_hfr/transformer/lightning.py b/bob/paper/ijcb2023_caim_hfr/transformer/lightning.py
new file mode 100644
index 0000000..feb017f
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/transformer/lightning.py
@@ -0,0 +1,84 @@
+from sklearn.utils.validation import check_array
+import torch
+from sklearn.base import BaseEstimator, TransformerMixin
+from bob.pipelines import SampleBatch, Sample, SampleSet
+from collections import defaultdict
+import pytorch_lightning as pl
+from pytorch_lightning.callbacks import ModelCheckpoint
+
+
+def data_per_modality(data, modality):
+ # takes a list of data and a list modality and returns a dict of data per modality
+ samples_per_modality = defaultdict(list)
+ for i, (d, mod) in enumerate(zip(data, modality)):
+ samples_per_modality[mod].append(d)
+ for k, v in samples_per_modality.items():
+ # convert a list of pytorch tensors to a pytorch tensor
+ samples_per_modality[k] = torch.stack(v)
+ return samples_per_modality
+
+
+def data_to_samples(data, samples):
+ # first detach each element of data
+ for k, v in data.items():
+ data[k] = v.detach().cpu().numpy()
+
+ new_samples = []
+ counters = defaultdict(lambda: 0)
+ for s in samples:
+ mod = s.modality
+ d = data[mod][counters[mod]]
+ new_samples.append(Sample(d, parent=s))
+ counters[mod] += 1
+ return new_samples
+
+
+class LightningTransformer(BaseEstimator, TransformerMixin):
+ def __init__(self, trainer_fn, model_fn, datamodule_fn, preprocessor, **kwargs):
+ super().__init__(**kwargs)
+ self.trainer_fn = trainer_fn
+ self.model_fn = model_fn
+ self.datamodule_fn = datamodule_fn
+ self.preprocessor = preprocessor
+ self.trainer = None
+ self.model = None
+ self.datamodule = None
+
+ def fit(self, samples, y=None):
+ datamodule: pl.LightningDataModule = self.datamodule_fn(samples)
+ self.model: pl.LightningModule = self.model_fn()
+ self.trainer: pl.Trainer = self.trainer_fn()
+ self.trainer.fit(self.model, datamodule)
+
+ # restore to best checkpoint which may not be the last one
+ cb = [cb for cb in self.trainer.callbacks if isinstance(cb, ModelCheckpoint)]
+ if cb and cb[0].best_model_path:
+ path = cb[0].best_model_path
+ print(f"Restoring to best checkpoint: {path}")
+ ckpt = torch.load(path)
+ self.model.load_state_dict(ckpt["state_dict"])
+
+ return self
+
+ def transform(self, samples):
+ if isinstance(samples[0], SampleSet):
+ return [
+ SampleSet(
+ self.transform(sset.samples),
+ parent=sset,
+ )
+ for sset in samples
+ ]
+
+ data = SampleBatch(samples, sample_attribute="data")
+ modality = SampleBatch(samples, sample_attribute="modality")
+
+ with torch.no_grad():
+ self.model.eval()
+ data = check_array(data, allow_nd=True)
+ data = self.preprocessor(data)
+ data = data_per_modality(data, modality)
+ data = self.model(data)
+ samples = data_to_samples(data, samples)
+ return samples
+
diff --git a/bob/paper/ijcb2023_caim_hfr/transformer/xlightning.py b/bob/paper/ijcb2023_caim_hfr/transformer/xlightning.py
new file mode 100644
index 0000000..9d30ced
--- /dev/null
+++ b/bob/paper/ijcb2023_caim_hfr/transformer/xlightning.py
@@ -0,0 +1,189 @@
+import pytorch_lightning as pl
+import torch
+
+
+def get_stair_case_learning_rates(epochs):
+ # Let's do 75% with 0.1 - 15% with 0.01 and 10% with 0.001
+
+ iterations = epochs
+
+ run_01 = int(iterations * 0.75)
+ run_001 = run_01 + int(iterations * 0.15)
+
+ return [run_01, run_001]
+
+
+class XLitDSUModel(pl.LightningModule):
+ def __init__(self, model, loss_fn, **kwargs):
+ super().__init__(**kwargs)
+ self.model = model
+ self.loss_fn = loss_fn
+ for name, param in self.model.named_parameters():
+ param.requires_grad = False
+ if 'translator' in name:
+ param.requires_grad = True
+
+
+
+ def forward(self, inputs):
+ # in lightning, forward defines the prediction/inference actions
+ outputs = {}
+ for modality, data in inputs.items():
+ outputs[modality] = self.model.model_per_modality[modality](data)
+ outputs[modality] = torch.nn.functional.normalize(outputs[modality], dim=-1)
+
+ return outputs
+
+ def training_step(self, batch, batch_idx):
+ data, label = batch
+ pred = self.model(data)
+ loss = self.loss_fn(pred, label)
+ self.log("train/loss", loss)
+ return loss
+
+ def validation_step(self, batch, batch_idx):
+ data, label = batch
+ pred = self.model(data)
+ loss = self.loss_fn(pred, label)
+ self.log("val/loss", loss)
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(list(filter(lambda p: p.requires_grad, self.model.parameters())), lr=0.001)
+ return optimizer
+
+
+class XLitDSUModelMMD(pl.LightningModule):
+ def __init__(self, model, loss_fn, **kwargs):
+ super().__init__(**kwargs)
+ self.model = model
+ self.loss_fn = loss_fn
+ for name, param in self.model.named_parameters(): # DOmain Invariant Units
+ param.requires_grad = False
+ if 'translator' in name: # currently only translator is trainable, modify to dapt other models as well
+ param.requires_grad = True
+
+
+
+ def forward(self, inputs):
+ # in lightning, forward defines the prediction/inference actions
+ outputs = {}
+ for modality, data in inputs.items():
+ outputs[modality] = self.model.model_per_modality[modality](data)
+ outputs[modality] = torch.nn.functional.normalize(outputs[modality], dim=-1)
+
+ return outputs
+
+ def training_step(self, batch, batch_idx):
+ data, label = batch
+ pred = self.model(data)
+ loss = self.loss_fn(pred[0], pred[1])
+ self.log("train/loss", loss)
+ return loss
+
+ def validation_step(self, batch, batch_idx):
+ data, label = batch
+ pred = self.model(data)
+ loss = self.loss_fn(pred[0], pred[1])
+ self.log("val/loss", loss)
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(list(filter(lambda p: p.requires_grad, self.model.parameters())), lr=0.001)
+ return optimizer
+
+
+
+
+
+class XLitDSUModelMMD2(pl.LightningModule):
+ def __init__(self, model, loss_fn_op, loss_fn_ip, **kwargs):
+ super().__init__(**kwargs)
+ self.model = model
+ self.loss_fn_op = loss_fn_op
+ self.loss_fn_ip=loss_fn_ip
+ for name, param in self.model.named_parameters(): # DOmain Invariant Units
+ param.requires_grad = False
+ if 'translator' in name: # currently only translator is trainable, modify to dapt other models as well
+ param.requires_grad = True
+
+
+
+ def forward(self, inputs):
+ # in lightning, forward defines the prediction/inference actions no neeed of translated here
+ outputs = {}
+ for modality, data in inputs.items():
+ outputs[modality] = self.model.model_per_modality[modality](data)
+ outputs[modality]= torch.nn.functional.normalize(outputs[modality], dim=-1)
+
+ return outputs
+
+ def training_step(self, batch, batch_idx):
+ data, label = batch
+ pred, translated = self.model(data)
+ loss_op = self.loss_fn_op(pred[0], pred[1])
+ loss_ip = self.loss_fn_ip(translated[0], translated[1])
+ loss=loss_ip+loss_op
+ self.log("train/loss", loss)
+ return loss
+
+ def validation_step(self, batch, batch_idx):
+ data, label = batch
+ pred, translated = self.model(data)
+ loss_op = self.loss_fn_op(pred[0], pred[1])
+ loss_ip = self.loss_fn_ip(translated[0], translated[1])
+ loss=loss_ip+loss_op
+ self.log("val/loss", loss)
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(list(filter(lambda p: p.requires_grad, self.model.parameters())), lr=0.001)
+ return optimizer
+ # boundaries = get_stair_case_learning_rates(10)
+ # scheduler = torch.optim.lr_scheduler.MultiStepLR(
+ # optimizer=optimizer, milestones=boundaries, gamma=0.1
+ # )
+ # return {"optimizer": optimizer, "lr_scheduler": scheduler}
+
+
+
+class XLitDSUModelMMD3(pl.LightningModule):
+ def __init__(self, model, loss_fn_op, loss_fn_ip, **kwargs):
+ super().__init__(**kwargs)
+ self.model = model
+ self.loss_fn_op = loss_fn_op
+ self.loss_fn_ip=loss_fn_ip
+ for name, param in self.model.named_parameters(): # DOmain Invariant Units
+ param.requires_grad = False
+ if 'translator' in name: # currently only translator is trainable, modify to dapt other models as well
+ param.requires_grad = True
+
+
+
+ def forward(self, inputs):
+ # in lightning, forward defines the prediction/inference actions no neeed of translated here
+ outputs = {}
+ for modality, data in inputs.items():
+ outputs[modality] = self.model.model_per_modality[modality](data)
+ outputs[modality]= torch.nn.functional.normalize(outputs[modality], dim=-1)
+
+ return outputs
+
+ def training_step(self, batch, batch_idx):
+ data, label = batch
+ pred, translated = self.model(data)
+ loss_op = self.loss_fn_op(pred[0], pred[1])
+ loss_ip = self.loss_fn_ip(translated[0], translated[1])
+ loss=loss_ip#+loss_op
+ self.log("train/loss", loss)
+ return loss
+
+ def validation_step(self, batch, batch_idx):
+ data, label = batch
+ pred, translated = self.model(data)
+ loss_op = self.loss_fn_op(pred[0], pred[1])
+ loss_ip = self.loss_fn_ip(translated[0], translated[1])
+ loss=loss_ip#+loss_op
+ self.log("val/loss", loss)
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(list(filter(lambda p: p.requires_grad, self.model.parameters())), lr=0.001)
+ return optimizer
+
diff --git a/buildout.cfg b/buildout.cfg
new file mode 100644
index 0000000..242ae58
--- /dev/null
+++ b/buildout.cfg
@@ -0,0 +1,40 @@
+[buildout]
+parts = scripts
+versions = versions
+
+
+extensions = bob.buildout
+ mr.developer
+
+
+auto-checkout = *
+debug = false
+newest = false
+verbose = false
+
+
+
+eggs = bob.pipelines
+ bob.bio.face
+ bob.paper.bob.paper.ijcb2023_caim_hfr
+
+develop =src/bob.pipelines
+ src/bob.bio.face
+ .
+
+
+
+[versions]
+
+
+[sources]
+bob.pipelines = git https://gitlab.idiap.ch/bob/bob.pipelines.git rev=a44cba9267c1fa58d4dd9d02dd0e3841ff48c6ea
+bob.bio.face = git https://gitlab.idiap.ch/bob/bob.bio.face.git rev=caf882c2ac73b9d2052ae4c19a0a09dceec129f8
+
+[scripts]
+recipe = bob.buildout:scripts
+dependent-scripts = true
+
+
+
+
diff --git a/conda/meta.yaml b/conda/meta.yaml
new file mode 100644
index 0000000..fce4b26
--- /dev/null
+++ b/conda/meta.yaml
@@ -0,0 +1,79 @@
+{% set name = 'bob.paper.ijcb2023_caim_hfr' %}
+{% set project_dir = environ.get('RECIPE_DIR') + '/..' %}
+
+package:
+ name: {{ name }}
+ version: {{ environ.get('BOB_PACKAGE_VERSION', '0.0.1') }}
+
+build:
+ entry_points:
+ number: {{ environ.get('BOB_BUILD_NUMBER', 0) }}
+ run_exports:
+ - {{ pin_subpackage(name) }}
+ script:
+ - cd {{ project_dir }}
+ {% if environ.get('BUILD_EGG') %}
+ - "{{ PYTHON }} setup.py sdist --formats=zip"
+ {% endif %}
+ - "{{ PYTHON }} -m pip install . -vv"
+
+requirements:
+ host:
+ - python {{ python }}
+ - setuptools {{ setuptools }}
+ - pip {{ pip }}
+ - bob.extension
+ - bob.io.base
+ - bob.measure
+ - bob.bio.base
+ - bob.pipelines
+ - matplotlib {{ matplotlib }}
+ - scikit-image {{ scikit_image }}
+ - tensorflow {{ tensorflow }} # [linux]
+ - imageio-ffmpeg {{ imageio_ffmpeg }}
+ - opencv {{opencv}}
+ - pytorch {{ pytorch }}
+ - pytorch-lightning {{ pytorch_lightning }}
+ - torchvision {{ torchvision }}
+ run:
+ - python
+ - setuptools
+ - {{ pin_compatible('matplotlib', max_pin="x.x") }}
+ - {{ pin_compatible('scikit-image') }}
+ - {{ pin_compatible('imageio-ffmpeg') }}
+ - {{ pin_compatible('opencv') }}
+ - {{ pin_compatible('pytorch', max_pin="x.x") }}
+ - {{ pin_compatible('torchvision', max_pin="x.x") }}
+ - {{ pin_compatible('pytorch-lightning', max_pin="x.x") }}
+
+ run_constrained:
+ - {{ pin_compatible('tensorflow') }} # [linux]
+
+
+test:
+ imports:
+ - {{ name }}
+ commands:
+ # runs tests for package only, report only what is in the package
+ # creates html and xml reports and place them in specific directories
+ - pytest --verbose --cov {{ name }} --cov-report term-missing --cov-report html:{{ project_dir }}/sphinx/coverage --cov-report xml:{{ project_dir }}/coverage.xml --pyargs {{ name }}
+ - sphinx-build -aEW {{ project_dir }}/doc {{ project_dir }}/sphinx # [linux]
+ - sphinx-build -aEb doctest {{ project_dir }}/doc sphinx # [linux]
+ - conda inspect linkages -p $PREFIX {{ name }} # [not win]
+ - conda inspect objects -p $PREFIX {{ name }} # [osx]
+ requires:
+ - pytest {{ pytest }}
+ - pytest-cov {{ pytest_cov }}
+ - coverage {{ coverage }}
+ - sphinx {{ sphinx }}
+ - sphinx_rtd_theme {{ sphinx_rtd_theme }}
+ - tensorflow # [linux]
+ - imageio-ffmpeg {{ imageio_ffmpeg }}
+ - pydeprecate==0.3.1
+
+about:
+ summary: Source code package for CAIM paper
+ home: https://www.idiap.ch/software/bob/
+ license: GNU General Public License v3 (GPLv3)
+ license_family: GPL
+ license_file: ../COPYING
\ No newline at end of file
diff --git a/doc/conf.py b/doc/conf.py
new file mode 100644
index 0000000..93f3d9e
--- /dev/null
+++ b/doc/conf.py
@@ -0,0 +1,269 @@
+#!/usr/bin/env python
+# vim: set fileencoding=utf-8 :
+
+import os
+import sys
+import glob
+import pkg_resources
+
+
+# -- General configuration -----------------------------------------------------
+
+# If your documentation needs a minimal Sphinx version, state it here.
+needs_sphinx = '1.3'
+
+# Add any Sphinx extension module names here, as strings. They can be extensions
+# coming with Sphinx (named 'sphinx.ext.*') or your custom ones.
+extensions = [
+ 'sphinx.ext.todo',
+ 'sphinx.ext.coverage',
+ 'sphinx.ext.ifconfig',
+ 'sphinx.ext.autodoc',
+ 'sphinx.ext.autosummary',
+ 'sphinx.ext.doctest',
+ 'sphinx.ext.graphviz',
+ 'sphinx.ext.intersphinx',
+ 'sphinx.ext.napoleon',
+ 'sphinx.ext.viewcode',
+ #'matplotlib.sphinxext.plot_directive'
+ ]
+
+import sphinx
+if sphinx.__version__ >= "1.4.1":
+ extensions.append('sphinx.ext.imgmath')
+ imgmath_image_format = 'svg'
+else:
+ extensions.append('sphinx.ext.pngmath')
+
+# Be picky about warnings
+nitpicky = True
+
+# Ignores stuff we can't easily resolve on other project's sphinx manuals
+nitpick_ignore = []
+
+# Allows the user to override warnings from a separate file
+if os.path.exists('nitpick-exceptions.txt'):
+ for line in open('nitpick-exceptions.txt'):
+ if line.strip() == "" or line.startswith("#"):
+ continue
+ dtype, target = line.split(None, 1)
+ target = target.strip()
+ try: # python 2.x
+ target = unicode(target)
+ except NameError:
+ pass
+ nitpick_ignore.append((dtype, target))
+
+# Always includes todos
+todo_include_todos = True
+
+# Generates auto-summary automatically
+autosummary_generate = True
+
+# Create numbers on figures with captions
+numfig = True
+
+# If we are on OSX, the 'dvipng' path maybe different
+dvipng_osx = '/opt/local/libexec/texlive/binaries/dvipng'
+if os.path.exists(dvipng_osx): pngmath_dvipng = dvipng_osx
+
+# Add any paths that contain templates here, relative to this directory.
+templates_path = ['_templates']
+
+# The suffix of source filenames.
+source_suffix = '.rst'
+
+# The encoding of source files.
+#source_encoding = 'utf-8-sig'
+
+# The master toctree document.
+master_doc = 'index'
+
+# General information about the project.
+project = u'bob.paper.ijcb2023_caim_hfr'
+import time
+copyright = u'%s, Idiap Research Institute' % time.strftime('%Y')
+
+# Grab the setup entry
+distribution = pkg_resources.require(project)[0]
+
+# The version info for the project you're documenting, acts as replacement for
+# |version| and |release|, also used in various other places throughout the
+# built documents.
+#
+# The short X.Y version.
+version = distribution.version
+# The full version, including alpha/beta/rc tags.
+release = distribution.version
+
+# The language for content autogenerated by Sphinx. Refer to documentation
+# for a list of supported languages.
+#language = None
+
+# There are two options for replacing |today|: either, you set today to some
+# non-false value, then it is used:
+#today = ''
+# Else, today_fmt is used as the format for a strftime call.
+#today_fmt = '%B %d, %Y'
+
+# List of patterns, relative to source directory, that match files and
+# directories to ignore when looking for source files.
+exclude_patterns = ['links.rst']
+
+# The reST default role (used for this markup: `text`) to use for all documents.
+#default_role = None
+
+# If true, '()' will be appended to :func: etc. cross-reference text.
+#add_function_parentheses = True
+
+# If true, the current module name will be prepended to all description
+# unit titles (such as .. function::).
+#add_module_names = True
+
+# If true, sectionauthor and moduleauthor directives will be shown in the
+# output. They are ignored by default.
+#show_authors = False
+
+# The name of the Pygments (syntax highlighting) style to use.
+pygments_style = 'sphinx'
+
+# A list of ignored prefixes for module index sorting.
+#modindex_common_prefix = []
+
+# Some variables which are useful for generated material
+project_variable = project.replace('.', '_')
+short_description = u'Example for using Bob inside a buildout project'
+owner = [u'Idiap Research Institute']
+
+
+# -- Options for HTML output ---------------------------------------------------
+
+# The theme to use for HTML and HTML Help pages. See the documentation for
+# a list of builtin themes.
+import sphinx_rtd_theme
+html_theme = 'sphinx_rtd_theme'
+
+# Theme options are theme-specific and customize the look and feel of a theme
+# further. For a list of options available for each theme, see the
+# documentation.
+#html_theme_options = {}
+
+# Add any paths that contain custom themes here, relative to this directory.
+html_theme_path = [sphinx_rtd_theme.get_html_theme_path()]
+
+# The name for this set of Sphinx documents. If None, it defaults to
+# "<project> v<release> documentation".
+#html_title = None
+
+# A shorter title for the navigation bar. Default is the same as html_title.
+#html_short_title = project_variable
+
+# The name of an image file (relative to this directory) to place at the top
+# of the sidebar.
+html_logo = 'img/bob-logo.png'
+
+# The name of an image file (within the static path) to use as favicon of the
+# docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32
+# pixels large.
+html_favicon = 'img/bob-favicon.ico'
+
+# Add any paths that contain custom static files (such as style sheets) here,
+# relative to this directory. They are copied after the builtin static files,
+# so a file named "default.css" will overwrite the builtin "default.css".
+#html_static_path = ['_static']
+
+# If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
+# using the given strftime format.
+#html_last_updated_fmt = '%b %d, %Y'
+
+# If true, SmartyPants will be used to convert quotes and dashes to
+# typographically correct entities.
+#html_use_smartypants = True
+
+# Custom sidebar templates, maps document names to template names.
+#html_sidebars = {}
+
+# Additional templates that should be rendered to pages, maps page names to
+# template names.
+#html_additional_pages = {}
+
+# If false, no module index is generated.
+#html_domain_indices = True
+
+# If false, no index is generated.
+#html_use_index = True
+
+# If true, the index is split into individual pages for each letter.
+#html_split_index = False
+
+# If true, links to the reST sources are added to the pages.
+#html_show_sourcelink = True
+
+# If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
+#html_show_sphinx = True
+
+# If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
+#html_show_copyright = True
+
+# If true, an OpenSearch description file will be output, and all pages will
+# contain a <link> tag referring to it. The value of this option must be the
+# base URL from which the finished HTML is served.
+#html_use_opensearch = ''
+
+# This is the file name suffix for HTML files (e.g. ".xhtml").
+#html_file_suffix = None
+
+# Output file base name for HTML help builder.
+htmlhelp_basename = project_variable + u'_doc'
+
+
+# -- Post configuration --------------------------------------------------------
+
+# Included after all input documents
+rst_epilog = """
+.. |project| replace:: Bob
+.. |version| replace:: %s
+.. |current-year| date:: %%Y
+""" % (version,)
+
+# Default processing flags for sphinx
+autoclass_content = 'class'
+autodoc_member_order = 'bysource'
+autodoc_default_flags = [
+ 'members',
+ 'undoc-members',
+ 'inherited-members',
+ 'show-inheritance',
+ ]
+
+# For inter-documentation mapping:
+from bob.extension.utils import link_documentation, load_requirements
+sphinx_requirements = "extra-intersphinx.txt"
+if os.path.exists(sphinx_requirements):
+ intersphinx_mapping = link_documentation(
+ additional_packages=['python','numpy'] + \
+ load_requirements(sphinx_requirements)
+ )
+else:
+ intersphinx_mapping = link_documentation()
+
+
+# We want to remove all private (i.e. _. or __.__) members
+# that are not in the list of accepted functions
+accepted_private_functions = ['__array__']
+
+def member_function_test(app, what, name, obj, skip, options):
+ # test if we have a private function
+ if len(name) > 1 and name[0] == '_':
+ # test if this private function should be allowed
+ if name not in accepted_private_functions:
+ # omit privat functions that are not in the list of accepted private functions
+ return skip
+ else:
+ # test if the method is documented
+ if not hasattr(obj, '__doc__') or not obj.__doc__:
+ return skip
+ return False
+
+def setup(app):
+ app.connect('autodoc-skip-member', member_function_test)
diff --git a/doc/img/bob-128x128.png b/doc/img/bob-128x128.png
new file mode 100644
index 0000000000000000000000000000000000000000..787d77b3ec4f143f8dd66162227b5f385f5183f1
GIT binary patch
literal 6547
zcmV;E8Eoc>P)<h;3K|Lk000e1NJLTq004jh004jp1^@s6!#-il00004b3#c}2nYxW
zd<bNS00009a7bBm000Cy000Cy0p^vXPyhe`8FWQhbW?9;ba!ELWdL_~cP?peYja~^
zaAhuUa%Y?FJQ@H184yWCK~#90?VWjm9YvkTKfla`WWwZ_E9Zm|?tm<U0=j|-awrm5
zK|n+ikVEkRQBd$$z*RsXT!Nr-A1lbQAdV1ZSd<kHL<BVuAS5K2Gnq*y0|_K~zx|_M
z_p9!%KHhshlQ;2wlbP3DU9bDsQ=j_Pud2JNMOCTlUstu8+JLH{zO8*Ca5He;;GqXB
z9L{5p8qR}N6;^#)`&{5T;7H*9`nL8l!+HFw2iQcaZ)>jyZUy!S8uQ6J0k;Fk4woHO
z53mVT-`0K*@K@mD6bc`l3VBd{Tl-1FWs6jUz~NJUTYDYw6X2sjQ%u%n!2W}W9`Nc=
z+f)T$m{i}^z7KE{upQ|<?~A~mgNGjQ_E6eYjRA&O^=<820M`LW04+s)bq;VF@X4XD
z%T+Z77(&&zwZ8{=9ypQW3Ll)xd2oGO`-vHw<x{l@7y{L|wbugkfMb9uW$tb%PCW2h
z*@jtE1t4S9Kk~%8fV*+JbfXb>i}*7JRueR2@jx{O$VlTK`pqcdVi8UNrg|Il1s$qp
zHUPx3!C9SyGg5eb2B!LVU-baxYy5+c?*RN9*crh-k|&lD4<-qyndsl%51ci4*oSW~
z$JUSuRRBuY`1|Jz@Lk|EU^);{vm^X{n!Ntir3fEdGxxwnMVcC^DgY&F!o82pLih#n
zE}+i0QGzaCthGI)G+qK;KxEG~M;!Q0Q3gk-+6^p06Mp{a*U9PAY}7s3sCDF&;EyEz
zY%DB$-`g18Pv}RPLr!nb>RhTGpa@O4yKM%@y})k3mOykgA05y|uiHFDK3V;F`hYXn
z9Ch%W=`4>{6@ZkQc*n2K0KS39R@Mwg!Kt^NqU)EGFB+1iz<vWqAM$#f-ASqfkf4dT
zJu(@%2Y4?~UsgaD$d?+F=sYg~dk!4?k&@$q>Nr4*Cf@SP6M+kWS-uUiZa*QPt(O<`
zJ&oG#)Y~~Y7nnX1VNs&09>AhWKYO?lxCh|_;jO-Qx)vC%#-^COzGVM;tOr;JXjC4B
zjR>a<eCp79EJsRI)dSd+)5*a10ZTPicXeOj%Rr0gp?!5JpAf9gzJ2Cb#8L^JJRQK9
zh^!!3QN{PMD&dSGRrLTRXwr=jZGrF{uuEuT=|P>RyxU(7T)+CHxo4GUZM>=;pae-y
zF9&we=(agIEx6H)4V>f!z&RP18>=b+MXKSt2X_IE0!CXq1QITBD=0xS{CT<%KDX-B
zBi5H|Z<4A26rqM|9}t|obf$ZWDI$M6DB+|$C>wx>R(<|Io-NVLqErQ-2uV)QQ`p&`
zXv>%)&9a`41s4KmmS|$pssfNwjq~nn1x^zQU-<E_sO4f%biP2I9^mx;FC96gH-eg~
z0+3cYy#tsPKb&if*hkiQ0}?+aIfP&Lf90sBGqO09ssN-=(+}@G5!k~^D00LvD$NIp
zun^@7B^g+Ts^b6&YP#f}2H?-Ywq9b)xPNFY$hRChKF|vs*Y~xfpD4%HQdIQ-Nph~e
zZyQRF@B+3jelSk}k{s|v-#Nz&d*MSBfLJws@4r8Su#a_Px}eUw9)RRE;MDRgEnQUr
zBGi20-DAbMx4${*rVn)sKXSy+T?s+565;HX=Y4W88>@>~6@Vz^^cr9XAUdE6j^f7j
z5G;@=@Y|K&KJJlh3@%Mo04!?$&YimhbAcK!vH0O!Uz!DyMZihrSz5lT00h)>{_Qm)
zax><`9S}Ek>sK)HS1Tm_z&BT1aQy1>tSw(v0D{Wtg}|Jm`|%NZ>=%M~X2nIHzOO_J
zGgTD;LoMgtHXAqrbWB!b{FV{H*PUZjhN=J<5@)z~W+5S*M|}f2NnO8C<i85Ip!ebv
zR%T#sCaMCE*W|NrJ_F?6e2Hl;?Z!e9iv0i5d+BF0eYUS$RRI7dpLtUY@HK?-p^c%S
zE@r!%WQSxiL-2HOxvByHNKS7>WSh{9mk=JlwcPQOBJv*qF73JEr0$GN&P+uCF!huh
zCJ=m{FzP`Go6=^D9a-@w+poKeHE+Z|4gubuls9>>C+e<WK=@nFl_%d2vG+}?)K>$i
zp8Vs(f$M=>crQgj;C&}eo@4;*={CPySPT@@7?8=4`wgXut$+jTzJ&D%OM$l)-Uiko
z@;0y<=mP$X@H=2(&y^<+`=!}b08Bga`YnOK06W_EQak|GL6l<l8~6Qmk69;Oq78r&
z<5hx@f86%)10D*jP*?->A@U0FYv8G#t4@ByVs)lU2fzu}-HP-0k>UPJESzHU#sClt
zP<9hC7-@C@tdNushyvXRUBCkhH}+ipxgLvUWh)VYX~+L)FNB8(4vi+2a#;|hD3rKu
z5dheSZqs^;gt#ICl0-n_;1F*Y9f&N$xg+_G?rT1m+10~X0Hz;zZ7t4B*1OsZAzdk_
za`(R7l?p&n!C!tooG}$Bar?2>@Dy+)dxj0b8^CP{mvmoyO4+l7!~<~bHJ9StqMUx>
zBWu32*2I4y5+vLA%FP?sZrf}r6_7YMqP@mDI9C&|K;)+GADvQi@E8fe^rNrd7I+Gn
z5x<8b1HfU7{^NluBB)EsBMHY%3rC#iu!Av(@EY*d?(0u|k`fdNz>K4=`U9|c@*ZLW
zHEaGfR+SiFrU~f0^rS)y<cS2JOhDpnPUAJ4kL3#Ux^Fmb9!1Iuz>JSy`DNf@+<9=q
zUyIpeF-MO|5B$N>?muln7tEU?sEfY#p9Davw%vY=NC2$I>_g;1;IytAPYaLLL?{Hn
zjKk-(0M7!2xo|=!pHRw0xA_qXse?Z^%}GoICX2{F36MlM62K`9jCH{85f1PA$?0#B
zLLmTJ54++q;9#p!$%#M_H8zx|Q~-(y=)nKSjs0GUft|OW2Qn=HX}m@h7=?62coFzm
z*Uvt`j0E`rXg&1uBXQ=OnUwU!pfS;T((K2cIks@3zLy9<M9mAP1-(hp9wOZ~4FOOD
z81YcNg2-N7H+?>K_7Lx4Q0qr7+Y-2pw1pouGK>6C6)q8gf~=D&d1-=Iq_F_x^93tn
z5D?Y43qa9!O5xKtJB*Rf4u)Fx>rXcr+796^;A6zd^8hms{^4CXZ>LB-d#GmJ1suxg
zKiFp5@&)q526l1V<Ge?js6VyGi1HR*`YPb-owuEF6A=<sW#)mG?gjh?^ET*6l~_nA
z^2Gub=>;M-ILTx1-P)bhrWX<u0dYRw1~vkMFc#BiM0o&-*nhELM8z0xdLeBw!oHoi
zf6*Ebh^o^1{;zCL)PT>Z(Uxx&ZX=N3c6WVLtu=qa$k0v`w%fNwLL;1p^IEpCg|sxD
zBFO`I1-J&`YT$`{ViO{>fL4TQICBe51{x8!Sq-)+#xc@>Q8E<rf55(-cYZN^`^4t>
zbnwiBe%Oq2oqcOJ%?e74t4+)!{At-EUyB_mn{oKOQNX(qJ`QpOur=mwXGy_b{&=xX
zNjfOr1TO5j`%9MvZQML>3CrovaQ@hN31Yj0*V*JLan`SV;D_^Gzt1L_vHan)b2t~%
zo&&x!WA41U2;aha<g6h%xC2CeoVa#)71st%2nyX2;W(Ti-+zjmn+Is!>(mb-qwi-_
zZ8P9BNmA=HRij2bO_SLqr48)xyvTasr^_BY=ZhuVc-R$(1J~r642#>(|B4Ro!uHub
zg}?de0ru;-=PNInwro}a%-H$h8k|?M?uyjp3*3oGL~08Fj2aDME=K@jhvOw+_oa`Y
zyRIBtTMxaw6_JM>TMbW~FX&96C~W&_$1l!2&9r0lY63L(ZQLh%RUx+lSPK%EZ{J%q
zAV~+v3P#nQv$XwNWfeZavbL`;YyIfu?*o2?Z~$)iJAc=)S+-6q;iQKnes_ByuQ-1L
z|9&Ji_wi=q0jBSCz*LYwV9x9vsKMb7z!qT#0I3~kdXGn!J$vg%v$VVQkjus)JOeln
zrP!i=Ynt;bveyub2LzmPfw{{cILn)eYV&zzjOtBjZkBV?at=_(3#fwTKy#?(P}RZ@
zx%`i_Zszqo$U6h*JALT7?#236$4Ob6z3j1b-jbXiCOKV<V!Rqg6J8(V<P|kuL!*IL
z(3p9h{Y@gLUp0@vSpYC?$Ndgga(mkg(kM_j-Gw9;kX!)})i+VSVfjDjcMXO8OXr`v
zlpNocoDK%xEmRg1sz?rh9@Saqu}3@r({}jaXjHF6wJEZud$ry97cBU~x)R|*7OLh#
zU8qd8VeOr4T))I$&{$tM<w0QrVS*CpC31{24?E%kKyw$Nx$Qy)=@zc9Frxs8Z3;S3
zJ-Oqh_S837l%l2o`=6?Vi!af^C4;)Q2VK_(8`i?c^+sW1g~IA3Y(O|^b*+b8;SJx8
z=w86o?f$(L_!P(%PKr4=BF={z>4lA<yk<UQ=w&15qn!(%dVVPE4@w7@+=bM9liG2+
zh}5FBbx!bE<9teuyFS)uyO@_0^QLRChfo;A+5%xJ&P%68HUOyJ0&EpknAWbJH|k~s
zBm~DIkS`C3?G`Qn=X{+p=b)EZKd_4eypy-5W^4h3+B&C;T3hFIPiw~	I5paez=
z4Is8ena)TDV9GXo&jt1nvx`(@Z4U$DgQFkiJDsomF4e77SxM1*Q4SXvov*r1Qy^>6
z^=tCx=DVjgb?B&CsL4Z6Q<nya3F^H1?vF4+0hls-uQ9--2#x-6j9^dS+2%oFt?l`=
zFdTpFdiD1uy@;tSY3^~LTai}BOa}t-Y!}cCuzny6gxhSy10f>nJ>xkd0f5{kz;@xm
z?*I&LxobqB6A8x>;KU3KPOXlY+WV*Pxc_QITm3*5E)GQkfxtWcb`N1T88u@kyT==G
z08HL`kJ-RcPDAJ_UzqN1w{edH#J&C=hGQ-0ZQYChxHN0q(@WQ37BF}PUlb6ohsehe
zQd5`r5Ov=4bR!OceBd|JR@>beK0H3E0XT3YV#099s|Xhjf#s>Cs)cI}W;vG%gc(CP
z*FXB)-H#{$TDIEl7;(qUeOJK}YCDpTg6;VXIJ!Vj?_Tum#-Xr3wKVsZH>R&nAqLGZ
z83<L-4c;teBL;w$nY-43T%_KO6!||%Ty6J0>|VjmkcT6O@>I{_=bjmwGbl#d=N5c^
zoj4%;-B~k1Y=5SlDE;o&M+^X{&O_9hw?C#ZfkJIBta$lCO+f*UMZoEq+g*ZGHGdDG
zQ}E8|L<6Bd*j?{n$nPE2Dl)PFXl~tkd*CqCyp_iouM3{J50D5r{1B}c^anj}{B4uZ
z-3x%`>a7JF1w^F#H#Qf%!vOHId#Dis0IJ^tT0`x+^MS3lOQ7Hja9BOZ%RNh9xFuVo
z%aM+?7BB>AygEZxp?Cl<umA4^q6~Jt{A90tq>%!^dw}pb5BQldRy4BULovcS&}%ZY
zH3O;YSSWb;xGtgZ^u6Xk$~%AtUNl|t05nbCaTlEX>#eod$Ap**W^-UJxHZB9V>xht
zh9;LU)v<^>Mqzk0TL&jH9x%NFUK&RG4<i5olv9AoK3DJC>`Fp~?c*Wxi@IRN@&#qz
z<Cm4D%zp0#V2tUpym1Z#2;U<Nz8YWnTA`=lgW>Py@hTqxbU)`KYI{`ZYnHLCCx9Ce
zT~)4ySxQwu05pqPvr+#!#CnI=reXlmht1<vGysiLx1WhbPfwSR>oGq<`7EBp^AAb~
z@MI>&W+lxXjmTK1**9vu_sfK#0Eu}`%-U2qF!ZAQ-aK4I0|5G2l&SVUzhyYU2S*YZ
z3h>939WM{*oj%1&cMaJrPUp@oc%OGLo4sThy1HI_#`jA4iUvTHxrpzX7fYrHIFiBW
z1+K`#&aBk3)ox>P-M=}a7j*ymWM{P~7=9BxX2k-~FlAc-CWpE@o+ST20*+)bI{P|b
zc`*Z1vy<xkfEJ&Jcej-f!%9^`_cv550HBkBz_aIpLbbw?8jSfF7@4thp8?s@?Dv-r
z1BEX*X`S7#|7k(+h!qQfs?G$q1Tl);JUdhAz$nba%$vZkGBmerHMj0kE3l8*?E@$s
z1~X<LhlFcCR4@QI)6`AKpNd~b^yljdB!FSGhkd}`vNALisUE39Yhm5>V#UHhf&M~k
zuoUGw+hHpg0M(tbJ+}}DYr`YK0Wj=P4EA@w)*sxmX{BeP#zPy%xRwM8)X5&s=bdi2
z8~9SslIN`FTd@Erg&P;mM!@=BABYCS4n;*3e)II5W&v9lO3&Yhsp00+#sNmYRahp5
za8<DYIQ@Tj(eC0C!H9-pNS^U4Rf?VuGQHsTAsz-2UB7XzulBz2{1Z{fu3P}nap5E|
zDX@W&5{e;t`$j2hn!e*`VE+Q7g@#)~rL0C3H!gVlnh2j)=tKBs-0>?G09Exy=p>?H
zD`YMjii!asf3L&VzJVWiQ@$$u;~5ta%h*OR27raVZ@d^;`=NpX0A1&uy4k@sB=+m0
z`J@08ejqxsv$J$HPTPJW$fw193dlPRKMd-o#d-xViwnX^;OrzDDi;7{wLOZ50vP^+
zkM$Hr^P%XP><leK%H@b|;}tMQZFlXJLT2Z8lvfCW@a&3Z|46*SwPFEKv=fZ>%~yqk
zH9p!?i2uU5yRcI>R+g^DsoS*zhZH6-R3mtk00>T+-z(&S@rC_f^RO8BbQ(J<768zt
zU<5@*!G@uT?C)E=gn&Dd)TY(-_#SZacl`YzK-eop292vhE?TjC!HP6SR4xF@5_iFq
zcdk)Tr1268FfBVni&w*xZFdLu0`TtgO#q?r+YW-)fmEJb*|Ff3w6;_%0E#XF)(Gy@
zs(~RdS9#YL92blSLLy#;GxTbhvdvhK+khzm-jwrkAb8C~A_$8?K2?l86$=2W&w})*
z3xJq)!B`5m`)Pm>-DCiu+=}SVc|(kJ!|eyUgTe4V0)#GvGgfxIRMa3>#m*5lxT<Rd
zXoquliaYm@H-(S!&p%X#mltE3v*b+@HEu#<t7RJWq1B|xvrkm?Ae5Q{xWT5%owvxJ
z=3jKr#?t`k4J$id{&lf7RV)AisLl`nk^Ddq0K%8wTQz|Ig%(9WF^p%zq{*{q0~Y{|
z1rQ-U2$u75_On6Z$(5Z8FD%8biUvU87pinS<rl8iUM_+#1c5NKDyY_ho;6(OF{x$i
z(Wu@RoTINO5RCNz2rq&jTdHjp4ZxeLyI%(FHC<kTgCrQDyn~l@*7%0$vxnn66ukl1
z)md>*z_&e6{uU7OQ`(u|0$l(c(%1RQ+ojr8(EtFTEbv9cX8rcc+VA4cFMC0f;TiHv
zQ-E&|wfPg9XZ;s&grfeKBLsq`0>Nk&76Kpc>w2Y^GF0)CMUR`%vbDf(ftgN<_=4wx
z0{w%)Rgq9W=i^m@o!9htuD4_?TN9hN`U!9h$VC64Ep~<8Rw98@@NIt$_)vfMYl-g#
zOj5;rfVT#E7c2U-FN8B(yxlv*c?Mt1Fhk+ItZkmqG_z)6^H%o(#{m<A!JARn4YMq5
zzj<1Ph1Y=tvnYHR0UiKg-1z1xz@HHJb+)0=0PLP2l~?EjIdX7SXC^a)OlX=p2KXhg
zuLxtq$CJQEYHyH3SO9#izkAV=ENq1l-~j;MTC<`Xm|uwH0<|3oc?LHv;1w*v<5bYw
z>nBYOJc3rdCN#C~4*E2(KQPwwLOzw`4F(k+McFgU!iN#!0RYC0Z<+)=56lYHszA+7
z;}tCJ>+8TtgR8opD%Q3MjWc!txdzw`-?NPVgNvnR`#-GSpa=Nzs-D*`DAukt8ZiI>
z)QxZYgg95xS_-LRS(Lz*)F5;N{{wt)a8=jxBzwj;PHzRd8rU0{;XA55c#DR>Gt;+y
z32^GF-o=k6*;1-T6aWBq^^NC)d;?)Z;7@B=ECmDsSdJ5E&j)z{Sb|V1usg61$Tq-4
zpb4Qh?0u3T2&RQ>-zdo6fx}kyF3u#_i_(Y#0HChE@mAoIAY;O1S_%lU!Qw_>Ab+`T
zl5KC!YPk@2>uYcx;yiA3?~;-Sc+zU50x$~a*}naOQEq|8gJ6LmkS)R+g!d^?5Co49
zeGITd&;fGo>J>{aVVE=`0RX70Zx{zMAJ_w^b5esKL>Ew!HyCCJdVptvFRWhi#v2Ty
zMl1jTb@dISKrRE0$GyP<DBN7(B?K0WXcDY0JP1MbF`T7H2>OA4i14M=E0%7OXB9^@
z0Iurl8$JwjBd~2DRRZHFLk&Suo?sAUA#m=%%B8XQ?GLp!3jhG>>KmFt9su5rc^g@5
zEp--#4uVi%xDHr^@PmPs%kD}(tD)Cs0{}o>eS_e<ru}?i8=xr}f(&8;3FQp#zX#|Q
zxCP{@fxcy%b~bWDn-u`2>gpTc2Yerx12S!xA;{s3C3GYF7UY&Seart&1!}Vez*Sv+
z<95JBAag`$!I^b*bjc7zHw9}EIs}%3JP7=JO@Bv4J%w!3W)Fa=y75g7;*1>}1?&jY
z0@Ndn199f(^>SMQ7!(*pcn9Qd;H~^mFUTDTZG-)tZw=GwWTJm50D>AfzIhz5BhZM@
z0x|=bE-)2n$oD6^f!9G^!FktSrwE;ctGb8vb#g<c{|8XDqFsBy#!dhL002ovPDHLk
FV1j@GAe{gJ
literal 0
HcmV?d00001
diff --git a/doc/img/bob-favicon.ico b/doc/img/bob-favicon.ico
new file mode 100644
index 0000000000000000000000000000000000000000..4cc3264302627d40868261add69eb755856611b6
GIT binary patch
literal 4286
zcmb`Lc~BE~6vr2J29j)cv)MD7gm4qFctEW<D4>FtQQLY@tG0My4=wf>9c>xM*w%WY
zsNjhxh!&729^h4lzuJzr(@xu&I@3R-9Xrm{>L{IBm=RvzZ<Am^La-*xm)-sSe((GF
zz4tbd4Mm0FPgE2|ep59jijpXba-wjE%0iKG?S*qpm`9!giUNSA7RKsKI^7D6)|Z)?
zh(?Ll|0J@yW`(886i&ZP<&5KS&EWsz9V~{`>1J_;@E^^z4tS#u1fw2AMh_Be02!s2
zXs{4*#@njdxDxjf7R+;?^PEAq2DzO)vIS&|kJ%3tR{t%s+X-9cXt0`%U_;q?=5nk|
zKQ~ZDpml;?pTg4-4|;GJ%$cDZSeE5ToSkK3Ro1x5YNmg;@eE-@q`-iKXYWQ!?5by}
zg<v#X)#{A3CUy4ewk%VI-g~wQbY>*a+;a#lM2Rd!izbMXOb{n>rM=e<<l2NU<ULfh
z{u;dfC3QBQ2^5dqIng5fZ!~fVJ67f(PUgDeEnHEcm_Xwy%TBgnPqn^s#_rx)1!ira
z7&0CsavQx|h*!*zU@=3Y#oXbv30eKc^`5iy^m6aL<msueucC2D@3sEd(UK&`E8L%6
zE;v;lhN6a9#jo&Q805dUzw;=8nM-=<>!)8&*Y_7k<`OM@3E?JLk!uqm*)G5cy9gsA
z<)y)5ee(%6ofUi7q4n``FG)XZd}{(7lOoN-l5IlQaE&R#D2E8CQ4%~KE!GE$>1!PC
zn;%FIeSP%S^j>nL=e^fgjBhN(A>7iK$bBJNf-x~NjE#}+rmCvmw<g%pshTTD9km`}
zEt9N*R?lH}etEFiv{<<%jWA<nNRN|YT$}<K@d{)*tz(15`sR~ua`JGy@Hgq*h)B{u
zPc2g%;!Vs$@U83Xh`bZyk(ZI6KvseUCL~&5qSFG|PTS1B)(09(jgmKx#u|>s8v5!;
z@=A+U_6CX}<H);Cc$1te<P23|{wOOHjkm$(iRz;M;(E`eCkn>&c<Dwu){oRs%gK{h
zN0w9R^F2K^IWneTwA$v%MBXG+jt#a<vBK6Dtx%k+j_S<_c0D0pou}0gYdFc9Q#R(5
zlVrV?m-poVTx8X2yj&}6n~Kb7sFzgO{<8X8Fjk=Xq73_)O%uKOWP5X(j5$rVYJXc@
zE^1KmbZc7iH1$I9bd3c&UQuDkOchFISzu?L1$O0G8i@av99zMrY}>US%E4<C?z@Mu
zX4kgn;#jWr2Fay6AHwkD)P|hzOJ=BlmCQocY%gmL@?NvR?zyP>3Y5-QZtk9|ItkyE
zMGe7yeogk-L&VSHNoob#$^K_}rRCxA_c3<9iVPoTp5o!Wu0YuvsC)(XEJQ7mVb5ae
zk39vJRR7xU%u~l<O&*fE5Fej3Kd)!^{)smVbAVFxFZKHSGxjc)Veb+d_7$L(%CP@U
z)LT-=fp?@+70cx@{%qfQ*#c!7dbDX;fJYSW`LPzsZMgq7^ln3MP3Yf-{+%9wWV|f_
zGT^{62@Wok?v%ePUoS6|+~q5z&4&u55k6L*NBo?2cwtn;)9PTMyrX=D2#4N7y)Qz=
z2dEE4II>c#?URE&59N4up_w_dn*XJ8wE&fCP#+0U^)aeQ=sLQV&krRlkk>I6KeO7!
zcj6I_t>@wRCuTVQDe5zGt9yfB2^1H~@tTcXb?s&|)NbLRb}I*UMB6y1-)^c3B`1)V
zdzaaEqQrE|y~_md-6l9$iYn98>E36a9Vjjo<L>=zO2a-=e#3sY=JWv;P9H>-vvB4R
zs)B*D70j<^kC^f`9cJP~!SwfNZl;Eut7JEvt7QIZJj#HkY9llqLmfB5`7cm4cwT4x
zrmE`6&+GnP&vM>aOJ`iDr@y;cPs2sGN0(00aOo5cmmBCimrfZ*K8ycAV(aQC-IYc|
z3920hUUgk*)E5jSK2&S2o;OUn(q#CV@V{&{l!waqX?&WRC<gcOd2<~_2aIdS$K(6d
zI%uUbp(ToimNhK69)*LeYcMJdT;Ng5Iu~flbU`bTgM#~P7_>*=HwyR_13vG&C@Q>#
eq8u&VkKC?xfI+BC09Uw2tt<i^0FQ92EAv18#ibAc
literal 0
HcmV?d00001
diff --git a/doc/img/bob-logo.png b/doc/img/bob-logo.png
new file mode 100644
index 0000000000000000000000000000000000000000..b60858a7068bf45c1ed8e3da12fe244ccdcfe85d
GIT binary patch
literal 6266
zcmV-=7=`DFP)<h;3K|Lk000e1NJLTq006rH003kN1ONa4bn%HB00002VoOIv0RM-N
z%)bBt010qNS#tmY0+;{*0+<0d(?{O`000McNliru;0Ox~6aXI@MOFX+7z;^6K~#9!
z?VWj?Tve6Ff9KU+UDa!PNvAstS%44(41^`fW|R>aSyTcdvdo780Z~DQ5oDr##sPKY
z1A-zThL3ziHU*JQ1OkFWSdu_UA|Wfiq?7J+*S^1d=8vlCdiCnwSFgJ|fu2vQK1o%*
z<(~Jx_uO;OJ?Gxn#Mt+m$QRY26$6~IK1$mMYC@s6ed(HWZpD$+j&trqHDSp8o%>yV
zN9WNweQTKdc+)=Ydjopw85bXRMPoPM!pWy@{3rdIcUv62&$dqha&P;Luin|Y*zGHB
z*}CPn@8fZ`h2Mw0cIZ=1JMZxCG|ez>6Z*=P4_(i%>@zZZ2DzuH>#94uj`R3Uo{Hnp
zL$_W?{rurR>=O^Y<>a#$UE4eh5CAB@t+3DMx3cj2jLdkEyMxot`+4^Xo&Ze9Y4n(F
zTW-G*f0SAHeb~nbz4@fm7v0c2+pxui93Tk&t(A{RE&N_KGE`?dbW3t>5+HZ_IxoJh
zYl$zYmqC|+cPuz`>qRmPzgMx-TpVDrTVMq~;n;~V?a0p_^Zn-8L;}{j4iG&4lTO{Z
zh3@B%J~kt>bpG_Gb#J-6(l4gS+WSx8>?a$0oBNA*w|~*oOiJWb(E@g?`}sBemyb!?
ziN3u&Fp~|a%PSzd`_&si>Ck~oF4DWQujd2+uj@FMN2TqA@JFt9H2{Qq&smO-txLW1
z`YAU#97Wt_sk;8pJD#|jKk@xLQSlnDe5(0WGx8GlMC6sPzVY<O9+UaPwE5p8hbZ9Y
z3T@xD?uj4pByg&V*LcCX9iJ6y%PxAKTKT+HIeQ8F%-^~@^fKdOWE>xP{Tm#fJnrKW
z^+)|k^XgD}?8BQMyvo{O?iK78T+({1D1&W8FZ7;=HoP&Jy!KDs>z}38^;C5Abym^G
zy@dVP=0m;@w_f4B6hC)=@Z|L@CJ*$pmvo$>OOR8+4DI;K8&6d$V=rKz_Z!c2VS--z
zKN))G3Jy&k=(9Xa?sNu3g@;~zg||L_d3BF$k7HkYM(gK<$(EO(v%cf;b$^>o24?Z?
zzInRDz~sMW;_Gfm58CjE_N@DGVBImR>*lh=PtX4feemGXJTiHppSrN~1y7^Oab)g)
z@(*Worg}MhIEDS`U-;$<l-~O2TR<QYdH-9J2|8)b_c<GjSFcyxRcQP3m)U{7hmFkX
zi(AiBxs41KJD+;*g~?=KzWm?b1wuJ4&BQm|6dABtfqTx#w7O1SRkuLajTSneO^+SP
z!;=U4@w2DD?rBx|>cGxt{&W_n?Ul6$wf*>?c<1X<4+a1-ku6tDCg?!>1FlxJuE)0h
z<KOH;H({q1cqi&251)3x`9dr2LW}4L%sXCPwU(w$9_Xi>+q|&okyS?5v&j$c8cb0G
za!9d0CL=Sy?tuIC1~ysX%HS1OfRz1i*e@`<2)VScxIZ23|A!l_ReCsPDb=@DGm~U{
ztJRS$hxpoLzB++Hi0xee@UD%!2a&`u29ZDx8HCjpV8H&h-!`Am2B<8-&eC7TGYZ<Q
zs4O1U7nY7#QPJ^Qgr&ynhuGRj((NVPW@#pg0O+Y~GMC6kGm*^juJs#M^JZ*AQa%(M
z#eU}E)@R)HT!yb47E8=U@n0oq;}N9jR9T9T$AG@tF*K&?An7?|aTH0bfxsfMJKD45
z#m(<vC5Aboz)sHdSJfSC)>>H=l-dEAs>67OwZ6;41Xk#4*Y#@EP#rlb0|I3D$Dq;g
zdN$m*>A8(BU?uv?qfQ0vXa8^0cSvK2d6B`ksswqI(cDI?@Ur4_ycIw;_mkpyk|0-s
zFRug!LC=JH);u)u2W&7-+#m+9WdE+eyIbVuMP3E@Uvkvu@2u!_<voNN`9zN}$O65r
z>$Icm@+wG%H$3^la=cSAHpBqVeZB4|nRy{e$69GzWV&zeX6<-IW9wLXf=rBcXKIDq
zc&51$%vF$0Z2$W^cj0-CkNWLrT-S8CQ0lfEwi>lveYfgZ<-qrbw#&YeE%Tz*bpqh<
z&pg2qd^3tdFvNLUQ~gzN3MKC1cVW|!tZWI@EdJQEo%aH8Ks61caw|#no@6b<)p_Xb
zi)agP0RU&q-MS|;q}Fx0_b%T`KCKQ)d#@)Eher#J#i!9zoQ65`#N;Q7(X^gKN(+$7
zE`9=#OHc};0&>0PCZ*ZYBb3#G<R+!wm;chdb58DGMfY$NAUF|(J`s&}R@0m}U2Sq>
z!VvndU%YqYocgwf&ZDWDrqO(@wYpCe?|*k$+v8b>aeu?y;M0zVeESm>Z9rs3mc{OC
z>hK<;Euao<Djh_P`S`Y0fS`>c(V!K6MK-Vg4qh%(cs}%u#(Oo7%BE+-o6iZYoHr|Q
zw04lA15+Ifw1CXKlvlvehp+uRzYI7zhY5;!pZK7&O-OAz1Bie4-ZG9Lf?oUy0FIgO
zx=lMC?o$6yipulo|9WfPv3(D7)F@}p6UhekRixJR-NA7TVr~BS%z5r*bP6@X!kB2_
zPXD6Bv&I>ib6vh0*rQC|apbQ(FK`YcSe4&!+#>f4+7iuAl1eI7KzczT6-!UDg?
zoL2X!=2dm4+341#J<poYIpPv=H64&o7%+6H^-ga7=rT4$<!4^*zD;dF^x=UevDN0S
zbN_{Ir+L;kKDM^+>%6A)d-gAV-yk@kL4yPNZ;sJ3Qrq7>Dp3(zKj|XvhgvsPbiQQJ
z`Tmuwmh$!q$}V*{Tt6YNG20@H85oYd)GJA7ymfEli0C^Bq>)8#RO7kq?d<S;#t_4A
zaQp{c#{FfvD?R>5d`+dC51+#c;-3ujyVSZEo-T)BS!35cZD|f!q>w-i5ri>}-Pny`
zgb_nDIq=hjB%$%ra{gkjPanN+z3l4pzXQ{8_=F|(X8qW8K&22J86*+IVDEZGHhn`3
zUlLnciacM6R@ZPWjjCncue2HEQjBM0HV<4cpIf$Z2Vbo35|U~yN^&#>4cJ}xQFHlz
zhCmi+BoM{Ws%^iKb9iwLU!9AYksH5zTbdV65cUP#t_yT37M$0SFh&a65ls*NEYq(#
z_k->DC9@29@w?W9!pfk|Bk5Ebw)|U<MgCKrnceg{{#dEwMv3m}#p<1Ykz<ir4NBj$
zyM`{~q)N_;NBFPCK1w%L0NkwyOc3_q{o0fqQiv5`hY&#m>4M=%uHJR8)!E`UeVtjy
zjX<al?EpaJ{3Qj|rXsZ>{Iq54<_58|@;rl_LiM)z9XG;lhF$zFa*2mx?_0?U|5qfE
zK*j{Tr1PFO(k1}=f-k2Rk6;8L#E?L`P|KqYe`x43PFS74DTg<$%nbR^4FH|jX@{7t
zdFhO7@6Z+2S^F)+{Ju=fi`yuG4gm1qNwehbPCy{@;lMI$9&6r@ZH{0V!w4aUByuJn
zjRL5zIhWe(Y55=Bqly{|J^4`Ly{ggjKC7`PdC>$+Y4LrN9V)*|q!YJid#cLWUMia^
zM#SVpD>io=>^X}^Q+GfqLk0kPBy?@{Ic)q99|mNRMH+wuPB`I&tKbQdAAy>&H{R-)
zE0iJI%!dR2Reb`rSY;LFVh3`7`#037E-1Tt<SrY@ke*5z2K)|W0O!5bP#T(1Ie1RL
z6nWkTl#h&8nnNxhmnooULb^um)8_fkVP%9o7q#l-3<y?oJinMs23BOSvwewc2`gie
zVt4rRVr<V+&T9t6m9!Z2O})~!P)PNh0pxmzuC%k_gtr#YGm6V0hctk60+AZAd+yVk
zSsC6+uLwVDr<CnGx<^o_3WkX^1|7@cR|drN)T602cI1e+5v6)r<q-gp@G)J>&dU7-
zgsH@B**-gR_*DqCvyzE(ZQ4HLeAgniu4k%tat0vKl&KDpY`uaGWMytCw`t@Cn=3(b
z>QV?IR)3o|olWMYly1+g8M)iu-Biw(F>1XYu5o&jc5U9#k_EFwCVp?W*IqeYPWV|l
zMS~O0b68WQIXE7_$?h~KZfBn=M8Mz-eL<B_A?U-QZ`m<q0KjvgsjXSIw<hd!rqLXA
zBt5$$bdTMVxV4Bwrtw+>rmB@W*OPeCww!LbSb-IAbWmf-<d~1kZcMMTEoZjHK0Vb_
z6ZYJ<H6JSy7mVyW9<Y6xGvQ%n3WMsQbR5B%mpNuz&H_~^i~#{3OH4v>4nX2z`zypn
zw>oVu(wTKNVdoQBnWPZ$$UXK;5@(X8av(*_YKE8?iH$|hF`Cj*6(bZhbf4zP=FY@p
zV~}&OCbjASkR8shtO5J1Mw-XUG+nkg^U4@1aG}M%gw~=qG5R)MKL$Amvogm7qK1lr
zTFG_XfIhoYrW4QOiVk@c`zG|&fStdNpxnFwK<4eStViN`q!=mzAUISWJ?p7w?a#7p
zYd|+kObi2n;5au+Oi%tZ8Gpe3s#$R`+Sr79bY5I0_Iu_ULSZ}_w?J8p17PU!$g}oK
zYW+NQ7|(~HL_IDx%eEI{B%$OA`*QS4(4Z<y4~gBf?ZxCt=B+WvxuTR2Fqf7cia!I?
zpzWfaph`Y-O6(j%QQys9h0Mp$cWSJj9VSxf8H1vn%N0YQQTL$F=~EdPK*X`b_PCu;
zF`h1Ier{9VSdClTMMt@Y8+Aob*|&skK^)FyGZl=1R`Dh`lxtoY$T8q(UO@9r<AKs!
z*PlYLT*u4qE{dpa|L$y>QQ0o@<%mq;*Lf`(f8e*p*oevniG+O|`VXl|p0)r)fBbvJ
zj!yO#5!iPge+X}-gks06-^&5{H2@3|8pCqs%i*Y;Ia&x#uZumA-#-2>mCTNqK_8=t
z*efOA^L$02$>3b(f!wxcP1)IkN9=ZV!7k^CT;abIt~;5&#g|}*x?9g}w@YQpG0uf7
zl|m4a=(WW;5aajT!*b*dKu_4!4%e;JrJCy~HpXuO%!480G5Vn0i`DKy6HBZ%0OVFD
zAM~Y><T{m?le29ywOdhV>_3bnfy|1;s8c<D>?7*w|KufawP!j5K(O8YWi!fBtR;OD
zhXDDLHws5J+0^wTwc}JnAu*i#HYb3lIHsT?G(Z5LX)QK&eW~LxV@$JbIMyG!sx;0W
zzikgy&VC<760irvaSx`Ldxp`t;v1<a0DwlIu(<%WkPWpVC*4tu=2w)%`qSBW0DvQA
z2eR~Cn`!HAeE30ex|Nv?XVc5GhUbNi-*4Y3N6rAqi>~S=)L*DAfx>$b+tZik2SU6<
z?z3~gdRwc@sar;esVYw2khmsq5UCfC`!aymLN-<~Hs0;nUq%A}a&LuxTxN{Mt?iE@
zs}e$9%&B&|f426cV(P+tvtnoFYnjc3QeTx<41<9>98^`#V!DEzGXI1jwx!Mzh5mt8
z%aH>N)LTB$FAx2E?X04kVWy(54}Yb6LgDzaza_TIkTV$6>^_!!mJL7PdRXg(^hS9k
zb5-)S(L)~*yHy@FwP;6?<sCR%wdH7%kAcNN`n=2sqj&IoPO-wsINE7ZmCBvtxP|KE
zW>;*Bo^GBeG;ZuWnBJ8k2LN(=Pr$(z&>MpHX@`~HTCoHX&)$?WMEk-KS`VqW%|O8Q
zuh`E*PK*0d%#{ZO#BlcV^gj%8-seq_R1N@8ljjVwFlWwA?_*jgOBWl7fyB~e#Uu1<
znJ2RxX;dK3e>$hqiPWZm(Hyv6I}@68sFljymHb7y`^kUId*?FiOf&FVQgq*1yZ3QA
zR&Jg+np>8B!gRv(az$V;x#nXD1?6(wGlS3MXYiOQ0Q&CarLi|93~HJ2+x|R0hl(g<
zA@sW`atoehlV?qSY7X4*IKh~>Fx+55Jdn8BJeet=;cXnI7%~a{{q)Uv&K&QgAhmiQ
zp-;;)YXlC9pQUaw&jAPqu$tzWqmYGfxy+l%pWsFFD1sVXGwL33>|cKLsMkM~y)L|5
zHl(W!`(l2Uf~tiunN87~@f=b}aRMMeT=09&@LaAPRO)(+=y_25DDf*<Ah+&f+yjSv
z$ONIU&D@MvkVcYI0QuoYz;%J+LYgCwt#hBaD*dD^R}}mKmZ?@RaW?x#@~3zcX(TyK
z1V7xU_h0E=s&&g3m5D8x3!|^e$I&&;Z*RK~7ZqmiNI>ScWu8vI&AYG#UQBf#=s8cD
zUwoKE#S)rL*$Y#DS2}^u;bhgyCg@_5{-pjc2C)r6bT|)poK15|9<(W@xHk8d><5Za
zg@;z)=t|^;Hj>rXW&WI9k6q|N6FPj0U1vJ>gG0KsOxIV&&q*ueqqSn!_!&CcV#TT8
zs2CO@tpN=bG!7=pK6U-|^tm~Uc?Nzz{)UC>MP(2i=G}-<BN`|Gmm-6d;lE@q7DE;>
zqIO=P9VpZG!e>r|d5BWfNDUMuhx|B>i-$s2vqfC0=G7c+el`K&F@A~=Ka{wRs|3${
zU09JH4X9^_NpG-p+%EoA_Q%%KV1dA)cnZ_h%Ourn_D0Q7Z|<hZZ>_Ul6R?P;wTU0b
zNk!W`Hs~VGEA^|>>+EFZz5C(!SR_ZymcRjoc!juD-(n|X88`y>;oxe(<=}acpGjPw
z?ynlKn3%OoEjQC;XsopHmg`SbVvYX2)a&-Mp<XBM;ibhz2DYy}hBw4_^>y|qTm=01
z9ZrLDY~T?RucW`5+H9w4Cv3@B%V%&s79%M0Y9xz;N3c=<S@P*>O}_Hw`4X<=BjK}!
zo-u^=yi9+?e$|e4;+N7z_;hh`S{3kuck#Q~?<LmSt=b8FL|fyhv>R{~brz>@O5h6~
z;0^pz_IH_qv7;D`(jQOP@IrJz{q&_A`mr9*h{t$n{7;3iJBu#EA=C;|YIp{{yq5oz
zd@480V_4lg;ZMf%J6(&l)3HCMaR5Ge;o=nHh$74*7#1(4p4EpZ;)%WvADx6#Fc<9z
zz|S5ykVFCzgfN0({8_w=@I*YC+|%IvJT0L)+(rTT;2{kO#1W+shB2r=8h?u;V=r}&
zJ&=$HZa7_jM<5dtKpM$C^L%#F;D(dx;NuYxGGjMeHsv_rg3}!!PdWrb%1#<*uVW`C
z8qkC$c;JF7KaCLS!aI(^+Ai84$Au;|p#dJa;LJZhpdclKC<g6L^sXw$gC;bg9&We_
zqk?QfN*ZAdjrA|+&>nbP$P`RRW0i*%MKOpWj@Hzg+R=^rDqU(ALm09<L0X~ZQ_+RG
zDso0Jh#~vRG@;lD=s-98wogc72YOi_2l8p?g2(m=N%UgJ*o_=H(2Y*GY@ZNA4|eh1
z#ZE5FwVPBnyt3%Qz*vxw7yBbHW;rQ*gu$`gE!E+GaXgMRj_uXgYfTt-3PKWovW=t{
zyX{>GHDf<G#wjO(9=lUwsU6eDosEc~$9mTN1YxHD=GuDbq7{-Y)jL3Hw8|#gM6#`V
z2T1Lx^+CA9_(<(2PY8B$V~)j)8yyM6kO0984}!`zp<{a$9r0tfMdjuYM;r+_;DrbE
z$}7rXyZV^{06-0xVQCiQh$9Inyzrnw*(Orhrc$H{z)l+WvoJ3)^cXES5gO5|M6y%;
zH(bbNX<kOKgUhd~87+$A8U5-%I7D8|vZ&l$JY?#2r=UrJUPrIoi<$`REoiYY9etJE
zB@w15JKrdF${QzjSh$N346w`-X;*f>BN$R-e!8Ka^)Sps(osr>(tjJ0ts|@%J2}v;
zY-VZnt9%q{M6=Qe?5K1q$!#&^DUO}29!1nE-AaiEMLtYb$_cC5A`#k@zJ09HsjMA)
z6M_~EGG*BftHJLy1EDDv<AHvwPLdiez*o4G)S)z(j6RD_5@=C^kD6V|8W+D4;8FgK
z(;O^EZ5$&gfM{r<VKzrzRC3bRo(YGD6+m2QK-gq1G^HHNdCUyj4OaO1KFXFm3$;FB
z!wCV^#l)(%q!^qE-EtyWwt~WC@ds?HFD;0`p}12I)s=O6ZVNe8LJSNFp;=*Y{ggAd
z{MxbSEQbE5dZA&Z1Lh*1!OgRhQ`OsNrZDOi-ie2DhRDA|DTk}a?O9MrY0&(nmrAG=
zyJmsIMpT`dVoICsqTJ|(k_8SM-4p8+Q~D<!((`7D@}^NWs<83m9L1HfA}{GhGgT{g
zQ5z0`bCgnEg_ClHQq-o*X3B)K<WPECF3LZ9gDl|M4cMijDZPGI(E`+pT~}U?eHSTf
zajXU%ARt>yn;He;w2&hT*HWV#2Bp=5I?@WW>1%Ny$?{JD*pDg_S=ymVCdayzd@(c$
k{u1*4!<V)X!N>9XKYrHF6~+H2b^rhX07*qoM6N<$g7`X0ivR!s
literal 0
HcmV?d00001
diff --git a/doc/img/caim_framework.png b/doc/img/caim_framework.png
new file mode 100644
index 0000000000000000000000000000000000000000..0293859f4a4538406434d9618dbbea6c9b6a58ee
GIT binary patch
literal 180345
zcmeFZby$?^+BXa%NGd8KN(+i~NH>UdGt|%sLx<8egeWK=B}0eQP(ycvC`fmwbb|~b
z{od$ut-aT?_q%-m`;Oy#JPu%-xvx6={GI1DZ<Q6LuHlm4qM@N(laYR^iiU<GjfMu2
z$G!@@L!Cc;hK7c&V<9fCEF&&XrEG6!YGDmUL!%3~2@{h8soZ`YYD%&u`s!&A$bzQn
zK?dB)LB%&nEU0~oSJC*xLCn3kA0$Ll)Js_{$uY|5u*W5_#l%YUZEke7m{i*#RCBSS
zPhRkD+w=Okh?*kjt|;fly4xBFSC{L*xvKwF3PbGGjVL0&jED4ccRaVJ&4e1856>Pn
za4}`&g_oo)eVv9f9@x8_kIbN6J{(du_wM~fgoyW%dJt8v(Bc-&n+m3zgLTJsv?c_x
z81N=7>iU0=a#g19csOVFtY|dDN>4X$5MF>U2-C{TTd*ML3x4vEy7&#X?`_lYR?dA?
zgLlg}vqDUN@|Dq7{*UF4<7r=~fZ(n;Ty$^59c>l+9=P>1oUEJgzc9?laRrm-J;&93
zhyGDVDhTKH>&G+SW04&qdG1WfMd9+5WR_@WSmkH+xEQoP2Wd7HB;`}W=*LMuR_lbh
zEzkUJ%<H`ZsI39jk}*|KKw|)2W22!5SfF75uh4-X65t07?Mj3{8W!;P7Vz`*1L)VM
zIMN@k{CW+NzkE;(EG{Dh`~@4^L!q_~=5~&@G0qLZR3jE@T8>%@^8ChjHY`RacF&+J
zt~SpvhoA|%@&hkzppHgVt~S=T4*ae{)IXl!2VP&k%}P!6;}J(IA!;oJWh!wydngq*
z3mXd?wJ<If6_ud9i7CJ8Q^`Mu1OF4EHg|M<&d<u~;^M;M!ogx^Z^p{b$H&LY_L%kY
zV`kt9W(PN0M<Z8eTL+q-ll+?JDb&H(-r~8Vg`F+c<-A7E?3^5hsHra(`t#q<eL`I=
z{#wb_;g4eh2V}kcgq59zjrGsDfuVwz@A4~KxI(QVPc3W!@&IcHbMf*D{&@a>eDc?d
z|1whRuaS>Ac)0&{=)ZjQ_o3<zP<wGZ8(>XG;lEh($KZec@Q;Cltd~ds7gGGB^N)7{
zLJQ*xvi`|5VcesSZx_+fMA2lPimADxZ%pBMPPo7>zEv=FL~IL$BzANsmR7|tt-wna
zv_IUI<q1&<u_?T`8yr$p^rGaNmQf+PpAA`>o&`PmqE-r~?5dcYT>O${1b2Y(atL{=
znRQz+boJt_t;m|Mn7J*2xud|2vf<dpvu9Z|>ix?6_Q)MJnFkm*MA6WH?-$iO41NKD
zp^=fyU``T3ak1YW2R!!?13L_NeirnbD*+?=Wqi=G9pe;QUZR1pseFFxmk|g<`t^(F
zzQ6VGk4aY_i2A7UchMUE{$YIC&_U@Rc`JXLY-qlYx2QxJ5>4M1{pLA<t%=5<p{p8^
z>Ha^A;K#W^1y?L-)B1nls60Sqw(Bha4<ix*PReGEnK1n>Q?RLS&;mrhi$(nVBf=B`
zPG$^PKfd#Ol6fhK*n|L)gZ=P+e}wcSfRp_e-1qQ)Tf(qk0YoNhX~O*d5$W#%POfss
z2V?%e@c*y6{I9zFue$uUCjYOx{C`wks5mwPf8QX}LS~0F4Tiop2l&-hE=K8?R7^ao
zlHwGBV`7KJwU<cE-25${>&toj`$0uRp7%+GH(m&`qy$`5Rb{~|>(8kIvM%CL=`Z#B
z{Q!ZGMbu~T)%yXe-#7U3k$}M`e5fGs`~HsYE8y=Gc=@hT|6ZU2%-r}C@I+kyzdlSD
z-&f}Te_{++eYR2GV!QiGT@RZTVeS4@c=8M4KRvNwKM>Buak`U8{$GRH7)y}vBkzYj
z6<dL$oA^@sx){<DrC(`($pl;MrC`M#m%jUb!IB{Y!j<e=wUFN*@}3@$MU%Aw?qBQv
zv#>l+2LinGT~)r{A0m$c#P@x+(!}qJZ<rboNJX^EE&l!xfMkHgBZO^e{`nAp#K5Ae
zfE|!!V1@twkVX?+6GhYcf9A^zYz&a9lF}X~`E`2ZFXK2u`fX}A)!R$u4(?Mj9($@9
z7^q1=ImU_;{M^pGvMXG6+>s?nrZ!V`dv-U4hBSp(M33{pq3;5RK1}8Mxb1i+rwzda
ztJ>kpG==W_#uVGS`*B_;X`2=I3b?j>2hEN*;sYIsnyr|cJK@G(bdS1RJu>HX-_3Jz
zhtFs>&&ajxACtAlw4T6-20UhW4kNzR3cC@Ixq6U%TQ3}m={UtN^l~4ou4$f`p;kAn
zTSigvo=sn&sk=ixNxD10{t;|@{~xF-x`tD`TRdquLE_Y4Uw`aVx{pvk-`ZSXu;U}F
zrr7+nj}TF5F{{7b-hY@)&~}145g$oQ95oyMax~iA#1tw-eJC}}<to$y=?+bfzQeNG
z0|iA+t~0jMv47`XW)-2SJM>E-tv{@l5UNSKt{3V&rv)n`I}|2szg4*`vUs$X`DDa6
zabbh4Tv5}z{8+ncrqT#WowCg5a#pc7KJjE^dc+{3o<r(#RbD%<_zi<oNxGnm5=!u4
zX#M2|AEcQhbQ@q$whj)Z_Oobj-XD88YlzZ7;w{?SA-T~l`Y=nO7ppVv&dXBb&=cVu
zMYrS==!$zOVQasW{b_-P2bIu{h`sk$yQ}#vr*Wvko7w55A_mR(s_eAcO=ErW$CA%o
z#%#S$8Z+F_pT~PTojd_HV`2mId)(me0&K=WBzJ2Y#^oq9{+78$d5m?Cz|?KUt4hRU
zd3OJ<8DE1imAdY!KL5;v!+qk0))qQ5q3)xV23b1~7do?^34O+yuX-u#eToSqwTcIJ
zo~)4%cgk45I2_J;$gPOIdLP_rUFDoe;aTjGIk`P4{Ns2$6cPDV!jZWr$LhN-33!)e
zJ#H2Ow>;X$g|42gU@e>tUHIoq2^96{KUhv<#W!`^dt~puQRg>eoUkZZE1sj@o7^!U
z<<Y#sE{iYReco#ZJ(+L7o?3w&Zmz+OTrRg$JZz)BG$n!y#qnIW+Vz}I{>dnd$6{5q
zT5Ylc1np-5n)}I8-!2kZrx<M-9!8=*Ka|j(WsBA!ZX>TIJ^JLfE?{!8F?$$(K`ETN
zl{`(zhbpYOmr=6lzPs2|qjWC2MR3i_n?U%DXB6tg(NVGW&RKJvS5%uhXf9QH%=?^u
z?EKjdtmAOmJ3Wb*GBh>)EkT`U8rh6DXC%)~-S!vMV$<P3uis5Y`y>I{MWQ_sAt&9f
zkv#JI%NZ38=c^(aJhcz^8a*cN>?iWxy_99Q`OV^s^$Kwa$l!aVpa|Q(M7XtEM!VOR
z^liT85+I%1Y9jgv^O((Om>1%(6zO<uA`ipY>X3ji=6{0AuckoLPHHO}G7L-nHpf#>
z3wZ|>O6>bZC(1>cuIU7^V49HjZ#U_`vZUNM`Jl%*(2G*R)APLB>0Pv$0(*l%aw!#+
z^Js0mw)=;2Xl&2!;eVsp7#KBcs=B8-+&MY=nq{dl@U_Qh-U0Di<D}gC1wH3mpN5^a
z8Ulhv;+hDUNm075>nK0<Ef>~H8ceUDdcERv&HYK*IZCZN9iVX}+QMoUk&j+PD?F{a
zGJVCsc-28GR;#voxqVh*%+kC3!^r2yMfKZQLDwR=a-}3LYz@aLmdJGq0_r56F$p?$
zvKwJ?5THU9J({&%HwYWAyEba4JZ}^zPd)q067|lhQv)o1LcaOr?UhaZ$g*pCQ2W`h
z*akRZB<TMC<T}z5M8Rp31$u-I6@nC@p2RG#+WDS+7cFVE)*R+Yw~i{{QR~rq<VDEk
zU?e>zKM<;=8eId15qC^<4to-h(7esTEf@<aaFLz}sb9H*^rS5?Gly6S<lj}zq$gcj
zbt$f?d3NkGV^bcE8=reum}`sOfy1%0s1v?#gzBxb*P4KQ4x6-fG0#UnEAqTCWn3N<
ze8Oc~9!)yF-KkY>#CcZRJCxn{nfdr}r|z|}H};)<E(9L$pb(uYiC9n5b#a$bE2iV^
zT)CxoQWzxHLr+g{*T7mJjia31!~LT{a?!gnB#6hP6GGe|m3ki)?-QPavgRqM&;prF
zPR+X#aL#Co8zhI^EuSa_aRnSqd=1$YdcJa~?(cRHytu#0wXGvQLqOJN%Wg@VK2jSM
z&<$@UZH~%EcrIh2U`{V|MmqD0hGDqgn6Cbv@BwR0OM?8T(3gH`(sfJApL(kT7r9I^
zH}Dg4F9{bh;c0N~w6?hY^De~-SH00$h%>3qKuAxg`e;w?-Q}Fzl7Uw&hu%D6T_;9N
zGn!K^bQ@fEntO8fOE9g0Q~HbOXoW+rIKY#W?VRV{<&P9~$}l-7>Hkw^7K7#+jG5bF
z{W;m=#YsXCpAA<Klrfgwl#fx|(2z0TcM`51i0Cpr=->5KJu^VzrLkJc?p?D6o58ya
zKdW)GA#ElS+=`u?_YRGb7d`pcdsK8Oz!R|YCN_W3D7gc=CG!tr-eP@?7CX6l?kUz)
zM@EN|wORNxUW{yMkJr4gaB-V+_LrAF4{q0QY!vR^!;6Z?31%_;V!*vo#_N&)IIh)8
zPv_dzrQug)xpNU&t=2?)6C7A0mTjGZdDbc~TqjU>;`BZv>l#}WecSoI#xgLZ+6oxH
zl=Mf4zEBy2=Gnu<Phw*4fg#>1ZM_Jh7#@RrD*60`F<d-IWP_*Vo=v>hb%pYTgq4oC
zm$0RMvZy6ibxAYz(Y$aL{|2n_&cj&d@ofu~(A_9p<9NoI&$B|gA)?4_-_KAZ1+1AX
zHEH-7aYC#}q^xlaBiP=Y7u-m+?qT@7C-;hQP4|h7c8zN+P85l!-L>u6&$+QBAMJXA
zzb+5$Y@3yA>JEymBiYTCUzH^m89uM6E5y3V0&3x5x_1#=zA}A<E0Z5-zOn7=hwx=0
z47;;wHBm}Jhwm+deO^m)>>yBUW~AvG+dOl{OmdU_OWXT-GgG`TEGUktJxJ^l>@Y;+
zTCPa{Yls_*|IUsy*K>kD1L^-5KKoiw-}rt^ym&OMcT+0<>le6{nYGT01k$`T{+-#v
zw6hzXRLh#x5Nh8;d?ev#lo!SVa+z|9-l{s;ntkrPJ)LJwcvtF4<E`hD_3FMY$5bs^
z-sfwaxhDIu>Qz_8o7kTI$A$m%w)*9%Jjts$Ny(c30qv!#!?2DJ%Jr5WF4r*aczEki
z@6fmcM6K_(n!*2|0IE_SQdRkCNEZ%Ipf>l!pZvMWyE;D12QA#xSbt9WkKw<(3J_c)
zRzCIF{KKOE0^<JeGTHZMQ8HMEGBB?(Aqsj{1DJ!AYMR~NpD?}j<=U!_&sOgIU-t_{
zlcNbNWA;HbE}dV4Jw?J@;(OZ8q<Ao_N@Fj~ko6icg9e^(W@X}EvxGTc;=5CkkE#Bz
zXYx(QsXg>-=f=YMlc9!d*e3Fc`ThUIfqx-}C>1wwif1Tlo?pxVLvReRME*W{`Dpx;
z_Y#`FhdBWIi|Gcz4$tM$8w`%zsv@Q+h%nJrRoKJ-+WMbD@ZV4jTAjw^c;byGR3Zl4
z@yt0}g`f4+<W2q(p8q!P*I9kD=^AlW!7qkaIDY%21Ikos*<!o<ZNo{-&=}?e)2@49
zU$kuYw_pB&NR1M{t==j1tzLIxo4<ygIT*Y<4X*hs|HHXicyTmw+Z-7^F&K5J`O9tW
z>kmaO<*6V4LrJ5T^#*UH(zL-j6?MQqIyt)PO^eF6zqCX?N3<oqTyg%dh5qrC=q*ew
zF^0F5YX6X&i$1{&l!c>)ux{^QEqrqyG|QLyYwk2W0BoryQ2GaHkS0gZ1dqGF#|hi*
z7#WM@KTCb|YYtIdJb<6LO4s$o|4|VBWazg`j7L~R_V+X1e_y6wY<8jT{qB)xPIk6T
zz*n}b|3%#l?J?s>hR&8q{b0{AK*f2LPlS?B#upCW4*xo^?`>j3%<S9ARMi2O>fz*T
zaK0NIC*tkJCc0*0=Mi7OpJ%r^T4W4{Y3)(|sjdHh(ik#y@8r|2Oh2g)i@}4_+<a#r
z)PHTr5ErNL<>2i&rs}}U`Jw2HbnX0JJARZH0V;ZWdGAPhEZ2wg#_M4<NUOi>@h`FX
z4<hyU4uG%pr4C+hm+A%_4ITV@%4nEyY2pdP=6*2l?aep0*~%;iXl=%dUlSp2iOVzp
zV(5QeyD^d&yY_;nxN6G2Dxr8R4G$K`G7_0-g!W@(HZ@z#myBDQ2+Q~>`@>_YjbYKo
z!7BuUH#&`Nnxq;B-{=bdlb)kvAYS`yACF+PuD16?TbF0G$or`MvJ}#hjhjKc`{Qmb
z`^P~}A2KvO;@wLdsvd3mWasHrSWcAN6oay)|M3vO>F|lEL@g_^M}9m@NIm-{Oe31_
zWk94{<KU~T0qp-`V%VH8j@>=>{jez2taNnx`KOH5gtav@GvuNf1Rg!2-z3W|(5x6C
zwWhgu?YpNi`DlZIdWmUvZLQ!vg{;L>w?tky)ehDgc?bj|W6_-B*q16%y^4sB&=4W|
zp#<^<m)>xDlEWU59I6Zq>=@bX)ejv1<vfk)c$0@`+v~|97e_d&t=nFQHR%<o?*-n8
zd)yIIjU_$8@f);z^OP67ci+JG{Ar8Om73X)VAC=iO@lW4cEiswQzsRz)<PnZXnY#A
zKoFPnBFB2Mx_hST*8{*10wH0In>VSrXm8&8p|P~$RI%vbi3+=sS~nD}u~zAyX6x_r
zc_TyA!NDPtuPsUu#LUdR%9-%SnD2X{QoF;}(@$oTWO{}J(G>5QJNVpblc;?}$MG-}
z$-y6r^u0Lx@uh6l<{yeOFwj2!0rJR~i>6*dSXx?^zL=M^nW|RDkXJa^=okKP?5S?_
zpSVLmU_mJaCR6f0oV`zfLs>zl{Csig)5Wt*Uk3`<@wfS=B=y1D7GM2iU+mfMQP)sp
zzh8c$#5N%AUFx{k`9k>2Z<`W7iixp%IP~1cBcUu<ccW-oY29JQ(_#T%@Mm!K?|Sr}
zhHcs>+r4DD#z}Ce@k$f=4K^-5lyvTY=G%=lQRjSJJI@LbLd#C60K|o7j5M?~by+fL
z|0?V06IsR)+MvYH?rbh;J14?#(@B9G_}Ur9h+{LlVS0lWVi3W8HFqn4sb*3LW*<{P
zlrc7(_X@H5Oanufjt+jG?KLKd)_i|8Gg-sHAUVl<tEyKz$#V{eucSmBFg2KXcV2Bd
zyc8`3`H6(5F^dWRG*PO`RQJ%q@!U53QOr78MhuwC@htxlWn;6!#Knbw3%y`ym?=o=
z?zZS?Z2VzG`@xN-C9CHr+anweXEvi~7st(%QJp2*-WMlEhgG6f+JI!I!l@}QIoB9@
zOIodQu#vJDNrXW9&|AsX+f0;OOxJ#5EfdV}S!Mb!?EU4mWo}fW{3r)GdISk4K~CoT
zq&(KW<7HL>({p6LSpbY07#A4)gZ5yEk56{CpvNBhe=*uWjgetBwr>5i*ZEb)Yv)wi
zzGCr1X~M3{iB~i|C8$g<qbu69ANEj`ANvLy{CdfkyT5$czfBwCias%E3DtW6Tkt0j
z`Ep3;R*|N6&g!M@D=G@96UMZ>lXbYcEn|W=_!<pl!Ju5~`p(|8ll+CFf=7~uote%`
zRjze8E^)r{dBW#iC#S5l?~InP#m{mY`#40i7`_;|lNK*ilH3l)de`3#rQLcXppZ)~
zY8j^}*!%<E$0okC<_*7WqLkDR6Wok<ttzLWAlj+$;mnQOiZ?|K$MH_*h#PvjMs(9^
zYilj+(En!ZMM-cbk4>R*#{ChQa#I8|q?W?Q`(?M#Rqqoh-jK%hT~6{)%+<_SQw%(-
z84bQxpfYwGFZsYSSL6nMSn(LS_Q9iz52-|&uNU(kYmx^zn_qZ)e!|ELK8=;py3cas
z4VcQBh+ZQ!5FGwY9N-m6PSYRoJuxQICz}fzGJAh9XT$)Rf(|aDHEbs|YaBe-oZxcV
z`qGfn^x)FgtWxhR@iah$goLz>qQvXRnF2hUukn}z9uz}9qXUEE`Y*TpFAm!G4d`&j
zyFqwQ(>p_8`rGxhRmzJO?U=HO=ZjNog6hP>F5Y$(lbKc-_%%_rD8i=MtSwaL;NHeX
z_WY;&i(T3u+DT{fUp`#tS|8_G*&7hSVeG*KwcWU@B@sB37Lg)ww)Q~O^0<qS<_9a@
zh*R~UgHg*pBQ5Vp{d_?f@}Jr7-xfpjrKd5(l-``I;&Aoc&oxlj(71yR2AdwN(ESIA
z7p(<tZf+j8j5Oq*)F0MG;ze2(Oy!n0*U((gd;V1xjvPxBwyV%T3T=2#^Z@!WH~8e#
zg89VTU~p1rI@3WQnJrhLObY22kNrTFsxmNgEMpe7?O~9blb&S31e97hlRW)@lxJ=<
zG`e0vtgWq?kYCZ+y~##&p{slg@WN3*Q~|cCUt#{Adkf14HMpI=fbDb(ynkFPoXo0b
zmoA#!+k9J)KBk6gG`A#lh8Q=5WW{vdm)LP<f-<_IOssCJ+joNS*r^MynN#7?>+*y&
zw7AI4lsH_<`&$s5L#AYTJ|6Mj39(lJp;^fKm|4=$Vs6xWm6=yT*NF9CEQi^mlZSdq
zA;07HP7@6SGqZ+;8)Rj=o7lzb^V;GXSHT4h)_SS?W?*kgi5cm_mZMcu>QRb^)b){^
z!b)Y1_73qw=grYHC}Y&sisyoRvUwT>I@u^@1^45s*8S$>$%2prMP}Vvs>p_e9TlS6
zOBRw65-Zb5iL6#_yRRdd#S69svdY#q1CqEBtyFb%YVI)W->r7plTL&zGE2eX9O8!u
z8{?T0a?2gzvDq*W{l03(kylAvJc*hfI+@(I<LUBzdo2^T?*~@p-@Q$ctShvs)hLhZ
zh>W#}W*)dX^aUp@Fz~JXPmdAI5*F-}t*}I(;WGImz_$KqsgHD-9$*7t-2JiSt8`iP
zoX4QnZ9K1+BlG_rJ+nbTNi4~e%u~C%D{u}TFyCuPJ9C^irS{?e_o1=zyzVH9F>>z3
z1%XJNwFn^&<6^CWay`J9r})aZQDhO(?I&(~J{s*DYO}QtJ|?b1fp-(gt3T_DI*FVS
zlzi4t+_hpMh}tSGp9l)mJ`f`Py7g&l_vDz?aeD3^B-zbI0X}HujOKc9*s~&M0E}M>
z$&D0f&KRk4@f)*X7e{ehgC!;XbFREL<wYN}F7K^#BOPfoJu@4at-Cf7d%EN=(@~*u
zwMVMJ*ya^~{<WbZ$Bxb*r%99ggF_m_<M5lqp<FlDD%}f1)=hPh#?P6W*R!T8$M;;n
zYvtn(6|KbW%B}4j?zhbg7bNzO_a)Wn5AGi2cA2?vW0PQ_e>dpIbU#0!d}+grLZvxB
z5@lE&cyi^^BMlOpTU!sV4rG>@zB>~4-+x5j9|NMfDxbLC=Y=Ak7&omDtfRR=cQbsQ
zy|rZvN^cpYy#hq?M7|@|bVIq11o@|DHK=Y>jSQbm_c9+{1tId;JLn9&&-R=0WKg@i
z)Wl0H|3i)B<<MPfG_jHdGeS&z9?rAUbq#;K#-Xko>(N%CR^7LZ$~|~}HIaYRUZ$d0
zH#W}0Ep{o7p=RM8&n+4@yJ^8Zg9c%Cc&s*s`$|e-b;43KXlAd4XN^^O2OS<iWbf>K
zKyK9Y==h_GGWGi1INYvw6%C(AAyyIBnjVi;pvZ9g-pOks-WF^y^Ch`JUZVl&9ME%O
za_}&&oYm2Bh0)-m*c#KeZadjFRQZ9bp=s&i)tn%1?{rUk-GC#i-c|dRah>5W?tDqv
zLV!g*lBzzeIgFbfUeVwnxl3;FND(N_zAZsK1Z=TkJMP9zg8|R9tMLxO=CHwy{F*I(
zz~L+8Y8KEMGPFa;?Hbi0yTYyk-c%amn=#~e+?Y3PM?#cS$M`|WX=+?^h7jC1s9C>D
zbdwEU43=;GyZeo~?Tfcu=e`=AX8Pj2kF42Z?gu6*zK4qY&ihxdAV}86OP7m`wH?4^
zfu;!4u*m%&6V@~2G?HXd#Y?6fm$6eN%I2l0wY`h$kn%*%cr?l2z*MFLDFf|izA4hP
z(N+<P(+wJxd21W1_hqkVb{eyu4v9>=1D2N{_w37bp(W_Nezt5`Ee|4LnPJ~<7BA@W
zaaW5tzj(}axftIqPVs8PHcMOB0%`Kmr$;6ovR&}6YwPQN4X3dtFJGx&Xj0aLH75lP
z9PY3vY9$U7QJ0KHEiti-lvm^pJrYK}fSs2Z%%TQv!|nqekg0^?6~O6z)oz<xTK1=f
z3X=$)Z*8B4_+-D3@0UggD<!TCvsz8GsBlOCVVUnBF<ocbYBkYVQ<{ijyB!EoJAXZ<
z`l~<bOGf#{CB%}5Gvd6nSNz`{rf(L;)`^`P<#*>#$l-tkh=?2i9EUBf)CXIav{Tcg
z5TcQuf#9`sxSfAF=i%<IA=*%%-g4PiS@x0L)oW4{eoZPI#J5+AnM(qkMO=lw%OEel
zV7Mi^@7FHn)4@3cLS0$KWQ$10Oh0E;)a~b`5T&UiJs!_+;_mHBZVbPEAHdZtNUvS5
zVwQa3rbQv8pP~U<h%c_$-D~$geK&r63)`dn^We?OldWy@sdsn^&m*}e(w|~Sb}{s4
z$9B2DXvG_UzS#0qO@=%ToA-RV;X#R$MWcnrMtf)Erg-w*hb|X1UhW|>Pk*2VzTp_D
zI58yIOoz>wfYL54jNydgW`4y5+@G(gk?mD<@PJ{`J?ei`T>!-OEjQhdMSCet-y9x?
z&KrIl$8tDMGL;jfd7GWLoy6EH7EFW+8q^wzD;Brqo5%?6#Aj}>)%YYlo_H^=!PMXR
zVl8I8$4!#V>eI)`gxkKXlTIU%Q>M9RJ1mfib)te=Nr&=!$fM-bnGY#lNf~Q8(}egT
zyL#q9mDPjkh6CoEpgJKTey?)HT)Q}82MBI3OQC18;MD%>p7lo}?yh^mg5(X2%>)d+
zS_QV?_xsDN#OSK?JQ+VvymIM6x?<SH)z#JeG({DE@;<n6$0xg-rQx9X<L!cL|1J-J
zTHd9-!R9e=R4ZJC+qrf#Eq&u9(-91PKaP2gjOB!m@6mF2!GxK{Yy05phhS%wi>PNp
zTVHa|<rRA@oF1Up81IcNNH3~>^>UF*3z<BDFvGT&IhV^1b5G-V+8q%3m_ZtCNmC87
zI%hz}^@%7;)d+vj$kBI3=rEzXEQvj6YcMz|Jd+*IgpZ@l;zV|HtV9Yh9U)bMm=TNs
zCfK&0;SB7UiW#$gmrPqLA+NcMqAYkfy2xR<yN%x;A+=sJI7HezZXaNY%K<H4eE2i7
zVapNiD<nc-$B@8-0)9H)KQq}?mM|kt$=I^B00UZKY2UdyN6K$+=23_D(B6}E_xUA}
zLC$*4&d*P2G<anQkh6J6_LeX%wyVkB5$-Z2t5mZnM@g>WC|FO<fyYJGuG0j(!mv!2
z@kEnnUD#W;)%)1M((5QTYwbSR64%kO++x$IL$$=gqL<OkS1(bI+=4A)j}!#-)$(Kp
zV_zD@>c;Y_Blp8yUDU|-lc9=}(%fv#lX#C7{Fq{X=Y?7USFPoTOFYl+M4PVf_A`MH
zOgDa@NTO<>a1i1kh6tgbzunhFNRep%9P7fDR~0%nH8o!4V7BwUU9+A#x+8;#QDY@1
zH^2WH&*TRHNMYcK-n>CQ_Tq095Go0bcE$zwXBDSj4g3OWc+S${Ndthe#F+S22h`K9
zzoAW>_(lo|7`spzgR1X3mF13*rDwA~-<f4(fPB1lwJo7-M(`V{9&D(r&H82a{0#ua
zy3QH+gX(*DTHcP@h8n)bAtJgl<PMFIWygctdGjXTFbw%`?LTnhH&TtbG2_M50<JOF
zj!p8|%9gWmV)#u#h>&=n?3TVO2<MH_`|DJO3>2{TKUw()XW5QGj}DSY#>UHU3iMl*
z7s~BBrVORcTvR%LvenLg9eyL`HhP0nlEWj~3=0y4kj5&9W%a_s2YhgFE^Z)-(`Cr>
zfvPg#({3l+u6_k%O!DgJAOvEqv(N5Q^y)BPqJtR~8YB4t$flqQ%K0~W0^p`E^!Au`
z@9v03u69N`nV5|F6H}s&!^Jm`KiaXLM{O5QI@c&1nXcDBIqdgG;u+P$iD#q=@;w7|
zZI|CLJ|p0?dN*X`Z2E@F?Ihi1W{R>X=HAPM5cw|p_I?4)o?b50F7?tp(6f~u+E~4B
zwa92Ln~a_@D$TDg;FNj*opN5^q#x#}CY({SnEdHI-#h_wW4f>}KW-p|Ir^Gl%1^n@
zzZ)-x;_2)@gk81xm+a{>&83oWT!$yIPv{?AY2R~ZR94HWSGQ^|mXF5sF!kr2s%Oh?
zPPBL#ou5Y^IO?BSp`@TEEaY9Q=RLeIJfg3e=0@8~H(a{0mZKH)DRRaZ&pha!pU4yX
z<hM*#``^_@Hu^HXVDjEo7E#lVkQc8KCZ7No3vTk-S8Av;Dm=HKM-Oj{8F)FT+oi9P
z7@U*s@NYO~A6_sy+fc@xiX-)?oBifGyLpT*S>x9GW@TbO>kEdQ0J9_qzTW#cL+&Hm
zj#tXnaW>-es@lQ;mUmq5Ww#Rd{u3@Tw>{h`ILbL;t6$1Xr67Qye3g}_`!4f(x?l$<
zquJT~yAmmzn$Md9iJ=rDS4?X@x?V{??iAVf-VUc3FWOFc%<;GsDh{`;@tV%sR2&I#
zbGY**N&LC2F3K(l_OP;YTxR|WhiPQT*PZll*pFPyVRc?^liwA$n4a1uqGf+!nUuJs
zJy2$TdQ~3Qx>xauiMckNQ-@Kuq}1Oh8z)9A?h?1?7Y7))7b7M4w;LgB3Yeh&Z?I*j
z{CuHG3B9%r!677lbXoPFsAZOo=BC`W>fLvE@Wo<+NyX8T{Sx~!ca8dk#Pz_M?s9*Y
zp-sz(5e$Y!;u_A2K>Nkp{Nno%%1V)oUZWvoQZ}T6y1<;Movc+xMiFs7TdWE@8Uyp&
zX?$krcz0&JK;U_PPv~fFRe8O^eXH{NaM@bgHiW#oRi>da)dhAHxND~Hp)gCi9+@O#
zB{D^N5C^B2=;lV8uw^cMe|FKv{V+f#^9PuOc-g3h4i1Q`8RT!l1^3i=?4N$#oDBf{
z77wF*_)q!0D+e6E=#}f=TMR|lum?A7RnN90q5NIfub3Le&a?8YP8MCyb5kfQMv*mz
z9LqlsugKLzFJU$+R+VuEuxi#%>M*k)&6+M;HI2h!RHCKeFaiPI)OPM7>q;=_HOO=`
z#Z=ngKc2IsiF<NIOr>?g<Jt8ve&^#SY`m6di$vAy;-rpC2|5>hU5V8f;mz(dI;`Nn
z&sA>_lnX%Czks_jStN(Y6P*Rrpdvtt@1xUlBv9{FdEg;RK1zYD%TgAX9@qB)E&wk*
z)y?^dA!q~V%gX-p=5O#!<14%b_cF+5xSMtvEdTKaccRc_zIcO3-c;!Ddf-TpW)zNY
z+#9ut$viiSu@_uUbD|~4mHe&>*6UZD^mw<2&NG$P8o`euWBHVz4ygGYp5Z3}U3ZbC
zg`bG5RLI#FB_ONK+u9ir4*Qp_2DMIA=WF>h+2`pvS~?me4vDvcZ7yIVbs9A4Um_ON
z6!X^1HwTNvJ)9=;D$>HAqQ?(p_;4r<BjU@ED_-Y{oyph{yyia-6eA46A&6<`(%FUM
z+qSjhf(Mq%K1p9jso$`J@3N`Sr1=4PRNu;VmQ~8^cR-f4Ia);H>bb~h5Z<xa-wtOi
zb3>W_VWNWU9UN$g7AgK+p8-nY6Rx0+hbA_`iyuGsM!BoY1unFA3{D$Rj!YYnkBsYB
z`om>6pi0SmQ;DWTxpzXRYwaa<;VuR%>TR8;y@8rXP({6O)g*O$DECe8-Sm7{n|H+l
z))mB>G-<bYu|V$MN_9jyb%?r|p6B#J&#XMHvr9TeW7LEnR9b3C=<iHdX(;V5UWfyc
z_l?mykI*~1kc3jA^n}>490u(V#itBVf!&49&hVa@G`yrcWiH7hN8Kfso3a~s9JvW;
z-)}Q$NV#;R(-7ZYw%8g!dOscNxGmaq=zWvynU`0+H8-QY<;GK=!BNSDAHb!wif<p_
z-?<J~cC>VMV;{nkD?)skj{q0c7mLsVT#%?G-qY(bYr`G;b+m>#9T@UXp(Jekb3v@_
z{Qg(v^_+&(kNczR(}ROYFT@^~m93Efot|$A3(cT(Gy@1c1qyW*mh0-ej~SIZxHVTB
zIWiS|`sm?t>#hTYPUOeKa?8|rw%5L8PYpCIPSws?pEW@r-bB4~;^?r@oTyT48%N)n
zGpoHGt(R+;lo?GSoHIo`YPH`rWUS6H=fL=8Jy9g*-S^4m!`!i{-1>SSoM&HpPk1^5
zW9*Re=!7QFhhPahzRo~>jotkt@k3rfsoIFOs-C?*Ju+Ra#(2aiZDI7;k~v??3Z&s*
zdoRX0p<N@tM43Z9?cq{&{wlQ2ed!^pe1uY$FDd`_V&wWRM6<(o-1B%Rv<0mGq}{!y
z&QfqBEUY@~K@m_OdpP`l{D+aDRiRRCRMV?bs)=(i69JMeRYJBhAe^FoA!$K^j-e_x
z8!koB7sq9#rmhahor=8$TLC1zFAA>Vf;Tk0IFgj6(Wn-x&&4R8XubE8h;9F<$6L7a
z<6z`n&cwg>b{Ia!kUv$BQLk4VDbE=pR_<DRyr`>dVS&JjSpy~@_v%Bc*_ShJO^<aA
zyFiGq!2C5;)n@x!_#km+<Fz(!Z|zG`b()^J3PDl3L#{{#Rw32Qs%W^Cq{2m%`<UDH
zys9cmfi7oXpx!~L&xBYHZwZUG7>Vd7>*6x5^b$+C12_Gw#u^I%#A`UNg0|4u*)LCN
z8F|+N<4Dz2l*u*nYL;JCR|`a1^r#vs7L6&CiadpS9mx_cs;yLw2Q=@tIuyBY-C5)d
zddaBv87%x2Dqp)I+H3xC9r5wSWuccv821NMm0s*a#8y_w#^GA)W>kK<B)+EHIOtaL
zQB5vddx}O1!9$>96zG$!mSl76j|OVH^}RPhIDLFvn_0V*uJYRCaGs8zP7{Dq)_pf@
zR12Y3OCMGwWK&<}LI+RJMtP=T%1g}P^8N54KmEiHO^HbY$ra@E7fmfYBFYsq;Cv})
z^m4}Cf&EmnoB7{0cPsPx4Ti%zIR~t7u5jpJQg~-T%m^y{LFq_tlx%`_4J=9DYc1V#
zSXKE&I#Z3~_9osO-93+R0pWv4GUDj)HEu?0$ZF6<oz#r$EL<7du|Μa$S+*o6Z`
zipa)nuo+iPC%jiOw<GE`rWBVm*+@OO0qnAE=#!&xPg}K01J4YJQ?3`GlM?u`%F<dS
zytiK78QN`~8OXr+sD{*Hrv7XQR_s<+3^j{i#Oorc=F$w|GX-F%2h;!z1t@wo_EY>A
zOAbKL01pYEBrF4+yg<x`0~)rgArtoGb)9tR5o$WC#fb@|tmb7E*U>b^3I!l`1NR;4
zQ@ARC_x=mD0F!LNFx%NSc#H4>m$#n8Cn5nP6(IYwgU{3TV#N==<q9>8&Ai!^CG<Th
z8@lUj)k;_x;U`ae9tw3Nhv&Vs(>8yZISxd`(I(Qp?^x_ayJ(fI6n&ddPn4|Tca{Xy
zgmT%mzHD#L*s^wb8Va2z!N1PgdKPhs+5)*wOk+PZXtBdeRtpv)AC2Qlg0Gk@R3U}c
zvsIaX?gMF`ogIr_*mPPihG=3Yv(oACJp)Q+4C>_^GswAR&IzC6!Bra`o!!q3yXuJS
zM?8!M`pJ)&p07uQgrFp)tTIE$$8#ra>tG37DxRFa!AhRWzAf%m3W&>l4+2UdQ9|FD
z>XUnoHRS&2CQvn?0(_H+M$BU=^ViuJ81j1hQ5j{6*SlC}Tdmg`a98ewus_kTY5c5G
zI4pI#<|IWy0W2YDR3}{zYXXsh?)Sr=|B%C<zYJ?Zv5GYz^@CYCpCycPr6K!E3uYGQ
z<GbIu*)5NHgq*aeOgnPyrnEUM>-#9Q#boj_YX+yL(;~*)jKpM}2`9NGtmmUU_ZUlX
zM>LZJ@{(&|*|G|=5T;cw%^o0fg$qs#Z}JTaYH$OUU7$n6d22M|b-bLKZd`Pj1_NF$
zJoYR0uv1EndO-IUN!j##S3oC&YVKm`OYP#_&T^0IM_1}n9+$}7u@W1@U91l5<X9i5
zJzpUB?pVE_bv?mNtX{YvCy>}uV#u__GLa+ZK5LB@kKRG|Y{?9n=kpISOK|29P)UYC
z2u=uzv{1cmqNNZS{^%~&oKB+Mhy1Swc$fLq=IPgpDn)ePQLo|*(STLZPtVhKOIo&3
z&7wR%-+rxS?5$o7C*<m25a8DV600^wjxgED{fCaBmRsKgl)SZhraVgkq@7b^SS1O|
z@pKU}tHMho#X)>0*Iq1T1L8-SAF9d=rQk}g{eXHTV3HTi<=}g-x&PVX9g__Ci@D7&
zUkNv>x6R7j_bOV|gla?yU~W3K>MGaKF<R&gZO^oiaBo$-RD(7d8m_!u@-u~J_FV}<
z`lJ)Dd8G)KB<M!-oDxfFc2#LO$VNRSGxBHV_|7x}NnN(|w$w;Ny?&X71Pup*uQaJ%
z8POcH=Fqw5s!6YBDp?!yB;-_25|L`iJ%g}itwqScn11f!K4`ydo{&KwB@fgeBeLbD
zE|nY?Pw?9(LhUk3xUAgCp`p^(d`0b%TSRO_TV3}>E&ENw{cnm=^?@o3G!dF19_lXD
z15B?c8+4C%xB+?epBCwx9oF<7M-Xu-=0Rp7%PepGXy2j|!c2l4K)RAe3;;-Il}}U>
zC{O;(UxD)bD*oc>tOcs6(7?7r2UhhKf^PCSpv-t#Q3z;XG4G1wsz!AS%;QED$1X;5
zH$dNx$rdO8AfVG1<SWO8&m|1W;14+uB^VwuF17voE`aREitM(eJojj$?Gp1GT3J>$
z!ZpZe>_GJgTy41Db9A+yfe^R|wZ`ZI)(=uid?{EId^ti`J(zq!czWaYFAhw4`+-DP
z_qI9jz=w~u-z#!Rug5MkVNb2C2vhP?Zl*UQIk^O(@k8mzp)}-ewhMBh3QZdz9Wdcl
z4;PVBm%!_6cfZHb#;j@N9NS$^K7gv1DmCqPYJz7?3K5rX50;n`)}I%3(d*s94cR^N
zaIAb$p<P1~x%p+FICj130EXG_LS$v}nPgK0b@FxU?aBCRr|NX@y%*7$hP#*d$CH@3
zE<GJR^UnEUgFyfR<?!xe2M}v%t{$%!i~tY|&=W&f!VR<xsN^i&1W3*H5Q9NB@%a7b
znfKWKL6pTzk$#=-jF4$#Bp0tsg+b`1IeWlSg)}NO+<~FE%+a=7Lqnh}&L^9m{lf9m
za=+1-9ayTkNcP;70lcJiJdfQl;3l2s-|^gHBVwjH`eoXEi!dH{wR&`D(ixUJtgmm7
zF}gfSswgjA0lEbgrS-xYvW4s)xOyCJO}QV|%#h<o21dG+dH}HlkU$bp5O=%16q9_{
zDln)uL#vrDPC8Sm>>Ld%WuQHaoeEEBmw4UOync|~WbH9x4ywN{>>?Xr(Am3dwVah|
z`aC{MX8ZZ8>j}ck!!Qw%)MR0mg5iTL!w{$`7gEF<1+6w}wuXkVV>Wo5z8IoVIA3Ed
zAmhEU2D4hvXd+1oNT3Qc1;NgR2~FZF#UK+4*V3JPGG^rXXm8N-Sh2maq$((zdk-HM
z>e@Y?jI(-~c6BS!n$Ntvf#V{^lReQwh&o^@;~u8NOBeeDH^ihHmrGK35Ap#tqrP6B
zOOKaGu%gWtlst^plXp?4bs<;g$bGdy$4OBEF?p?H`dRta&c`}CdWdlZY;9{3cagH^
z?Ua$G*Yt-r*R@pZXP5W3kdMkPK}E1^qun{E&imqg_FI>o7udN*uOo_?=CZ~?#%#FO
zaQKw?J|z@va@i!Os@S+%xEd${X=+~6WkLrJ8`LXty1r~RX1&D*w|_4e0C3G00~;Sn
zYksEMB6K)u{L<LQ=w}B`XDkO{?=AJeFfRF5jeXVpXJ<1PXHPkatc|IZbBc=}Z*eCc
zU566$!)?kl*S~!qoZrXffsxQ~KiWEoP+?=7$S33QPXJ|Bj|=Xe$SWtf?B(S~^o5$Q
zapJ|T@~|ciA5seC%xnwMpuYJ5YOTCo@AUS3Jpjvb0Kr#n;o!(9C1kEa7{B7UyU_O5
zHsat$45u%o#H?*uw)2jM2315^&VXpKum@R^X)HWBs<Zal_s!mYxmK?X{{uq)KKoG&
zp)w3jSN?%J6b~#gA^YD2Qyz;?;7^g+M=W=0Ml}=FI^@rr(D7PQ$rTjnyxVN$%5_yK
zyui$lc!eqLRO-C;{VU9Ufb`t(17AJj4REcG{_#!6D}XuxJqq>p^>5I@TwGk)y>*e>
z&AmR^CkJB-Tpd{NFR8jp4QQZatVPRZ3)81ZqK#^vwvm7M72m<zg?SCKV=!;hDu>}U
z#q?L$qwk=$!Hm!MV$o7*ON0{3)>PNW@S3O+U-nxwkEvy}V@ebYV2OMREq~ubu{{X$
zXubY`w{LA|<C$6MP)J-sU5cEU<;0MQM#I|X0X?GtxFkB-YPdhuuyHW!HI3i~*M-Je
zbSPQS&Vy5p+(+lLxfd{`e)4$>UU7M^M~Bs&^>-thRzq_UhF50JR-1PGD*<RshG?|P
zIQot)i^_}D@0xtLF}(x&$(kXZfWi5gaI55<-i55f6~8^HDqVPHCU`h}orzOtlUc>L
zqjs4usEgYbvk9Jbb#48E|M<i$gq|Y|sh(-8|1ICQqgT^vKg&1L3Y6n$G;8Vk(^q$$
zep7$4nvT7bS(;v`dRd4LqX3CtLHLK5Yl175l3$psSBbwi#k~1P?_x71HKzO~5QJ^M
zdnPBW5kr#S#&tQ!3uuJMm`Qj{{WI7^z#IK@O*mT9Gg?aQ?R0KhPal8iDPOYOKR>A>
z98RpL-OqDBTQ^AVJ=k(RnE-$&UzOp&G?!_&Os~^<N-M*YI%WTxibxS-VTTJbh?uH|
zXde=q5m%P}v}zt-du?k!_v@1$?7ZvK-R-5>pGpO$OE22MmPjkG{}iHs>?lWD7unBu
zWcvF3cdP3$3c0EG{Dz)nj|H;PgtX}9pBuimEm0lGZc!xX8_xfX!i8@J!@BaV@~b3?
z2n*+|VX!*q`aRVf&Vsq{&mtU=PYvxchVu*<4Vp%Mllu<32PW>}hB36W)IAVgQ|q|z
z<knm<oy&y8$)Dz|eMeNn9m~AULg^*2WED-rgY4`z%qqkpW=&|`m1-y46tcq1CM;uJ
zFW0dPEyq5hl{P#T$#|5N#`)uFcbKj(4-4}!W0(FAm&XDsw0^fQg|<T2ECe4#Tx^M&
z;OUS{&{kg>Bq3yjD>bUwKMEYb1X<`A3^C>NICOR)0dOSq5zuQRL=3cjD7nOWgnOR@
z9cvPg5^XR27BS{@v3b%#g;P#>mHITc+@{VLz)oiUTCL0eu*>|5GOXw!B`f`}l)+H6
z@%RKadq=YTi+3W$mxP@HD~*C-BF*4hbydM@EV`xA8innmhH6euPJ%;Ol?+Popg?tD
zR@TZc2VFQmqvQNX5JW0{(D;t7eN5#C-nqBb>cq9n?GEOB$FXr1ptNG$A}nHw9NrE!
z)IbI6+daDPZa(`Y#qmDKZ|OYNv6)aV3HP07C-;Sl*8NU7v%RD3jlpZ0-5SxEJcyi`
z{d|{IOTq{6jlK0&3Or0{cy#om8oo6tDkw$>S-$#^Y{{7}NF%KH>QKL%Bql9eyDapm
zM9upgK6})7QEO^dd$lBSsyA|(fx`WsU;(|(%{yd6Qv(^?hkRvD$(p$}G7@8p+8R~o
z5%%G>TDjWS(($J9`0Z`C8=;>H5|ftR9L`rkloN8t7<1<7Wn&)zt=4c8!yGBV10Y_C
z#vb8~oUl|x^>u=ySwGY%n7$5ON@Bl)K=S5PqmpL7p_*|q<@KD|4K0EE7Ho99N^==5
z(1V(<AgUD_2ZagQ_8+5z-#@10ylgCf0G@@(nK2GTrP!8a940N{jSuyw)Gbxex8la|
zJFbKw=aW3PaJ(v^nDRE$wGf?h>-&Zbv4oPB!R3#Tukiu)q$jGY8u)sG+jfSLL(urQ
zs@h&%Oo{Gg4=cI!8c%f&cP=sf%JMN%ZKa?ZrHiUG71o3e<@najlj-88$=+*`=YTQD
z=kkTjE8vZ2W~Jndmm#h-K<Z?h{Z*<ezaZ(j>Yb6B`4s%UIh2&<D#p>8b_PN;--E1%
zStT0G^s2~ST$O_`O2T$8^7^D^O0*li&Lh3d%cndR#^!h9na{U87T#!g$?8_v74*$j
zy`G1r3WBhM#SL5VQLwF@;57f8)JHm2>Uz@+#YPfhWl?cd2tPBIS7s#6&^j{$CAb7v
zicUcSOw4In|GJ*eqY~V!$rGLJu4x(FTL-<yE=mF2r4%@N1PQm7HL{orWQAb*`<Pd+
zgkmMNo|N^}x_C^!MZK$_jI2=uYC&=em6sYRZ{@37g1@uWj+|vr^2tut!;i_i2h>r4
z;_-3D=h;i;Q^C|VeVx5;>O{X@cnckzX8-NRCBh#8VyV$1HM@;TV1@SWs{9n}a0+N}
zZPP?c{4-^gyz{*gOPFo9`EF9}X*{8&)rzV=WQcyiZMF3SjF!z@(CYCexpu;h`KPw}
z5>ht|eIkeI6blTJT{mKc%bd2}<SIxA1cM7oS9(!qFUQF9JLNPo8DhEWXX$e1E<B55
z2<kl*)iGJ15G`7Wn9AX%?ciJIf^)ZtgxNaRi)f2A_pTv_mr`kCWpe?cL#r1_yoigV
z2EV3xL>FprDAYEv!oB^zPa`iTZV>Kb2Z_geuX()1VS=V{D?6k~MMIm{*lA}yrGz#1
zZqE*#QjL&YkSVhNbhpFL&HtQr26DSvaKN3@(KhjR3{fUS&{Dt_G{)c)iTqiUf^!Xx
zV{V6s*tt&+(^i<pF;@>$6c@NF@mk7qcc*n;a)ln`dPzi-OpT%-v>e?UPY4LiM%L9Q
z##+9qiH$wL(LH|T_kinZ*$?RkgAHrZ!BSng;9;@G!NCM?_sbg-@=>DwAcSRtw$M!W
zTy2WCZxT|a|FTnutqwDShz)AfR$(*2UG#8f&kM%x{)fSf@dD{ru-+w82RsdcDpfxy
zq`OJ2G-S~GmOE6~0+ezmx@s?kGH&KJ_iDpTsVL=lx5T8Z?Yk~MYR!GzP-5P)EhjGf
zr3dF}x{P1erLf{ktE!b<VnwqAyhbTs?U-rq6_D+p2`vv=X?O;@5M_VuCR&Ql64@#5
zc-e|~B5yd%8k&<)=i$no4`6kmKa=}}56lP1{B6FY;Cvl5T1^e_8<|88)b+}B^Ab&i
z1eMa<l*)5^)e)&=DE;@V`0s{spY0wW&k*w0ku!8}%xvxM<|yO{T@<@_mPAa(YICF7
z3+=_%clP(Oq!&#Vtsp~J%i=e8yJOVzLJ}s@9ul4De9SMXf@m~ql#E?xW372Ql+hvH
zq#{&jiq}|Pt(@HA5%HtU;n$=)pBK6tjtNRWjab;HYlYjSrl#R}E1oJ1eSyvY=!z1J
zyX>o_C2V86jLzt3f#~dZJX`gTy#GFs_YW$TJe=uZ2O8cyU(<sSb20coYsQyW3#{C@
zynf+Rn_5jp-=uP<?PdvysxM?{)gSYSe{7mm)jqBV%G23B(p<nDu=UwvjzfoosJl}^
z_Icz`=6s(1kv!U<O@|!I>fALHp3|Y+HTS9_o%NhftZp1>v+gPdk241_Br2~5_Pnl`
znGJ|<<l}zTSItlAXv0>=#skWnw>01j>Q~phJHM#9HEqn^%r5dTcPD4lS#@(~!JtgG
z7D}I+JV(opHLVd^zYBooo8PWJi4!zk%uG!vHXO!WWb`wLXHDqqYdOpLQgB_kqx1Pn
zF(f))mzYgweN*@-5@OT-G?bRkDaN~**V(+lwr2kTV@R4S9K$nr1k(;{7uD5s)>olR
zMyKR6=!0SK1O1%4br(oEL`yXHmso>^y}$3qUbzOTfEZbd`W5`ph$8IE+CX22lYX6j
zs?L69&BErpQ&cMwQ9yVNa6O|zxaVoB34zwkiSc=GBpXo&XV~Rei%91HeW5ao8JAZS
zF3Xg6Lw!b&j(wYNb^Xc~mBfq|%qA6bL~X^Qii<gh3^iUb#7;H5lu0E`h+S5es!+=0
z_l#CMJDDeEXs4K*UX8$4nWoGBqLP-~v0cEagDH9FYZs}GRRS?C?5~%s;O=a{mPBy6
zoqc>TyBb{kUHHk?H_W|vlp7;y4MN!g7iY9pWX^gQnFIuCl&9|N`&`F!0o<>q=7BqY
zv>JwoTSD4}DZD9+(h!=Jt<8Rv6r~FE8>ov>9~t3U5|?~rXTBSwp2}a#oFm{Zs&QkW
z*D>`-y~I6n>}A2TGURNI?*fmQ{5zZ0((R3Hr_GN%G3q(?U6yKRzG@mtSIi7JeWN8(
zsOZ=A%iHFicvfAcD2C@^P30Q;JD@y_nn$NXDb+fiQ65;N25NF<&nlW3>AD9d%aByi
zF7cd<?_Sjz<8&(ofUyAz(b!Lc&W3@(+`29y_s_U7k0&;dkV*)e2OtFXhv#?znMJ7>
ze9wL8GlyA=qs&;?yzdd1&lG*blF!R&>tDlUAzCbcXrM+YMtk|>TTmN*%;jw&1_p+R
zQIDVHH$bx-Z()0$0N;DCp8rO?{rN(B8Oj8>k0D>|b7Z!xS`s)g*34`l&pf#sq*pxF
ztjVBVn6K+Bb?V(ZBgkVgGj22z;Qc~gvwM}`qx+h9jdP;Ydf^fDt4lulMKV1eBg*ku
zIycFC<xL=i{o-@&Gx7}w2`ByMud`jbP<r>Y^<oRK4!N^4I?O;s?&@aeo9D^i9+FCN
zMQX%mR*GVtzWT{02ev}rSB$z6!gH6(ZMzt@2^zIsAg}W)Yvp?Yuhf%_93ih@8g7^P
z{O!80Jz%31ec_B6Uy^ywiMFYoXg-KO_YklOs=?QZ!cl8yZ;AlYBBD6{>WR)k`)TRV
zPYy$pCUTDvnvquJYaNfF9`2KCd*$)kCz`1qr$gHTyVn!6d?7s)M{hT1E8F3*<demA
z9}<0z9kYxycb>;4_i+<`+30l4&dmhZ3HOK+;hX{KT+TOL@h9u)#h<W&u<ovu<&WDT
zcg27<mFk?wjIPpdYmXbmZsEu9d$Sw1Ya1~v`Orsez+(xJ9pvpUucVSfd$0Y7xv}Lr
ze4hg4?Y!r)L<sv%Trdy@ltkZZ{y%Je1yq#l+V&Pl0Z~AZl9mpgL8M^-L69yf>28n)
zDFLNBhn5<oySt?YhDJcTJBRw;VV`}zbN=sJ>shW5$CsHWuKT*LJDzucc|Rk3@6dvX
zqLZos3boWx3EHu3_;M3}A3k7FpMnZ-R)!K=Ua(XqjYSq1xFlAWIcVNfjJ-%eo{#e&
z<4t5YOp1@uiQKy_B}}plkAjbPIj%MMpV)B4ZTZny0i<fb4t)Hhm8@OK$$GU0h<PBP
zO-6ChlMadNE8Sb-RDp3qOJmtcskB<X!F~G7I;h$_;p4FB&nv7R*9H?Xp>M7Fq$)96
zzBbL8)${V16IY#?b$e_6IoHgJJK|0(zdSZ99OOgGBSDYA^UogJk!<jsbrbHp%=R9h
zp_7HC@ZFHz2!+=*j1S*Z;9JSu2o=9_8Mv%<?1(@jr8QQGnbd*f_Iz^0Cd+-P$p+2o
znr6F*eYY-`iZ<e;h>P);3!s*}324MA6*1^w3y^Mn{VO-puqGTuXd|9^|GB4>La%FH
zT#4di5(jWa6l{5WdINNe+-5z_@oIB1&fR{%?(7=yjP5DPs=yY%0dXgctS_jncx(!O
z#w{S#uzo<WYDTo>!y{>Ct$z#o`nV-GkDc4ov?)Xa&&#y=z;av6@uoH+@v=_Gj9?Ze
zsb0;GYEeci?*oem)3a56zj0nSQXFLIC;U9$P$_o#oIlUvByR#F+0SBUE@Kon^_jUz
zM{@4k+^b~FWUl%ZE4TUjU#p2@rG!H)h{)i3>LO_})kk&cx5_}45`3lv|Mi@V&S&Y@
zt+-j|4`fq#^QTrlf@8?;nN(ezE#~#2g7ehgLMHxbDQmtHoaMY~Zvy|yHI4cCw-fos
zyFrgdIuxwi3r|#xkBcv9`!GTu5`-eFZ=TlB>gIJ@_Bit+UtWZ9)c|=-K!Vj9R{_Xl
z0RQE@fjratB(SNfl>cd!)4HjBeFM!^mt~8483B{7`O)eCDu;+gI(tpZPvKuhL_&#}
z$4+MtRQ2r=3F7f$KhAnE4^|}=rddw&S9=9Fde>13%%<mrRc-~(ENOU~eNRo-xpenD
zf!P|?ZkL&0TbFaakt<lUK9cHF>1Dg+Y}Pk++9&?8WGWTD`#RsiLvKbicDZ9!#Tn^y
zXh)XE*hli?)w$znp5$0R6UOiSB4&iW8!)|~HcYf=suG7O{fomy`t+5JVZ;B32z`F+
z@+kCmRhp%esjdS73UYPLABBK*nZIYQ3A8Sm_zx6quX6dBO9K0eIqBsEe6`P88|5(L
z(=WWfEvfa4WOsT0U{hAK1<_!6eNg4TV#2UW31N8b>xbHMZwN}z;G6@Dk0{1*>o1Yd
z<XRBU&6mG_JyKEua^2|WtE^@|C@4h2?H=F3{hdNcRV+WXckNdGQ9($cvGLIdx>r;T
zD_zeG{8^(7+y=L@rf!O=XNXu+R(cWr-#R}!WM3S8bPuL6J5vvDI^Q-+vECXFxngbf
zln9)2wB>6oCvr!&v71zEIrzt^h*tY_s<oT*jODh2Ie+w3m^fj7@WV7_PK?Z1nHY#-
z?`p~RuBz*Ac62}FVyIZN%-LR@A1uy&g}CJU@B@LEFsVAeJRLkj&R?%Bv#%Ynem%LJ
za>L)}Ek2w(t29)v5}S!yqxNN=93*RzuzVdE$d)c-)`u%OV{evZCn?92BcF;b<nk!_
zZF;qt$ma#l$JN>LpC45}dmg|f@jekW)9NJFZOe~S#vv7FdEwbSbkmnsRc#r1I;U}a
z!Y^g<d7^&qL?dlk@3yju4r?o;k=^yB%0&9|?b|Rf(DjO(W-;!_VfM~T@4{$ISW5!U
z{_>Jw@R7=32+5QC-%?FpWsH2RwHA|MZ91GCnf?&Cy?UM|OIJp1(+-ku0ZD(azzETX
zjHO_}>kPi==KX=`=9RuMXAwIGFOIZr|Au|Rl%_(n7%vDMA0Jn%G*^~R-S)zM>~ze4
zFKy*g%cng%_q4I`;<i3w>)5d!kyM;3aijF9M0wT*jb0ax{u4c>8pM%)!_5jGa_??<
zD?I&C6eh8ED^Dh2-M(#RLm8W|m~_=<;I_E(#4Qs|0TH#aaeBIE8XiBf7uqD^ztpCD
z<lndw<}1FKiDMZ!=<B~ofMsD|qG5Bj$KcfIBg)7j@RaEK;67N~?X}aN?Y#4}Vr_`!
zZ7u#sC+Dt2%7$k~J^8!F{e_{M8w9pgwVj`D!>U{kZ)eC<e>c(gviRTjsih%@b`CnK
zs?_*BKEC;&+HoID7BIN%mxn_q4UO8?TiOl#AyspIn!lElkvm%>JEm>BQGi%>yRGGz
z^PJ&v9jK<W;6JmIy4oa{Itr^F1xYU)<GOU_7lnv8HVK<Im^xQEcit2xb@eq~hxvRF
zd`j|wUykE$KDe(zUDA+~?E_NyD_(k7UB6H6sMmMxVc-!J_~(GeNlAIL?{9!S-mtw`
ziSMqQ&?e>j?GXJyDTh0-!6!HVv;OZ@--F{g(U5wBPpt?hO}ELvmsJQu4Wdki{s)3u
z<iCgRR}c~z)=`ynCT3vAH*8e>isF+7!f$LwU;G6I?`@)M?!GCvR0Cp`z^cdUL$k-p
z)%l8)@k1`Q-xmXGOE->g1{zY%pGolznjH@Mm*K;ITn23}N0crz8X8ztny!6dQMG|)
z^L<vOyMk36b2~e1XS09iTPpPAOjzZtYmhTD;X(1xi!?sPNCbM*1U`6YFKBBpfr3(#
z0}N*<#wC;Bbr!Z_O;jaHyBQhMw{Az&?8Y8FteHwJxGloDh3wQ{<PBr>Jmhu;cCM}(
z2&i!)IN|VUPc&Fc$Ny1~j8|E1jN{?+QjYW<R+cQOIoR-S3qmMBU6PQKVnXTA?t{r3
zOqkF;d{b-du(5pAoC~00*ri!_vu>*>c%osIgHSqUaa*w!@t}w0B5L+lvPl$jlH=8U
z{fu3}^6p{stRPkxb6m7tGHQ4KJwC+<i66DTe>E2Lt<MA-SDdYGJ;QIZGMDEge&pbB
zZM)l(xznol4YXotKIhB)9O)pV9xH|0+FG8~F`3Yz*(&bZSTjjpPC}i@?rr-`^nlF$
znVK3-I9;_$C4KQv(juWv`4<#3^HcaXYKbL0aYgEC__sIPH<O704LAG9&I2TtAGj_G
z9IJD^9D0x%UKDn_HRMTV0{#(>HIzs(iq|^E4yFhos;=t#v%KQrb<k?D7-aM$ad8CL
zV<A8;u)kY)lizU&tIlF@oXo4>>5t7#CAi^JXt{6s3A+C^hx%y1^bw~pBm!{o2~E8f
z(G5K<^=$nhCDjLFV6WJbf;eskB^>zPP778@RK332+KkL-zL@aT+4<*k<ffrAGB0f-
z<dzS!XnHPH(_B{*-gs?D=}1}=!k!UnyXX~Y*1ooGIOsyP3)LS?=5aNy19G_j*}8|g
zrgN13(Bqk>J@agEGFe)NHE{S`JrvgcN#9YKqTXtwr6|l)?ajXL#%c;V(_`&Mu35yC
z+XpI+>ecgEzvZ#dJ*4}KUX8m2Y9CbDG~*d6(_UBde>`=dTT6l&S!ZpU?}hr33$WGN
zoo|stqXMY$yzHY?x0pCo$P1g-f_MyGgthG@cbOfW49z}EyP3Bz*eRDepbWQ`cxQ1^
z8GFzhY}{RMBPeLC$#ykC#%vs)$pE?bVaYIevTL8FGTWP`6t{7+zg<UkeN*t7qf@9~
z_u)niKYyvg+2tL28+&uL!DW}dj-hFHKzHE2W$=#3xfIl3)?TH?Yhqpcl`nmf|Ayh>
z>vR>D#ot?60V5<{cT-C$KeOq@QSoRX8shTJJ$x7hL<{S+uE{RsYP!j+s<D(XV%2&>
z7d11N@(%*E;Sqyd5#oZ#VeYl$%wmCrl247c|AyvheC|ndf$*E}^YHBmn|t>j7H8(m
zSvyUi9^b6<A>!w1-Y_Y!?YPwnIMEkR<maV5A3K=^2_QReA5VOcrLebs7C8XTf0gRx
zu3F%>umX#o{ivK05Qc4OA!prU9=QDChoSrTbBnC{;ah8+H*VAKwU$%3*lSldhXhMr
zkX!8VKSNKfUn;mTz_*`!fK_n>R9J@xgX<WJ+&ftwF}F>rHokhU<2du3jl*Ux;!zfT
ztZGM@&Sg2wvnTIf)z6uaX&(IJ^GE5Tw6V8DKsxcKOe)MbgQg+HURxV#$IxJxuB9(<
z*`lBsYquSjnyMiMyb>Yvn7;%79TH^+YcQ7ar3-)0kZzrxbojx2OS5xXu<4pzW5FkM
z*GZ?2ITwz>Ibt4-x5NY8Js!Z>?|iqAx6-4$)-$$0WVschb%~w&=6pA`>3kaJjGiZ>
zDvv*o^+#ARq3TD7qoO$lus{&+McVzJG~pc~B<%WAWk#K~-t4$1YM5~A=4&UctihHG
zSBZZw#t2o(Ro#OF1Wt9=c&Looc##Xo?@lh(=bvbzyTm@)`667h-o?KWbvsOKOc|nl
z-=S4_EfpgpYfY`i1~xYiE&QHqIdXpeP*clpR7Gp717F<MG+f|R&HZq<erYqmIPCVO
zzL`{2+tuh+o$#&EWwRI0baf*)RZP>|?)vP_TTyjaO_w!sCEE{YO;)96>9_cWU_nw2
zthkPbq<IoYq9$fSO=odS;En_Ie&<XV49!&j@U@q<5Z>wf%sON91CTZ-l0p(xmy|k9
zQ7#Cn$H+)=p+-jk4s>3fE5!t81x(2AFP~i1p`@tQWoOTf>tB(?dAiI?UCO8#VEl;2
zrK3Q=5`PrFKvHUP`$w&OL-*<`+c#I&{7eF*WR>D?uya%>5$5faTaMR?sBO~AGo{2p
zw?>_hZj4#Cy~b7>ey~~zl}dryQ?A)hBFN|5Sa83afaOQ<x~Pfy0xR?2^oA&#_54Wy
z2miqtX)3Aj%(c|!onY3$)gAI-+XY>_rIE<x)gQgbv2I40AKo->B_$!_&3Aa8?FjfU
z#y3X>NW9)?Tdv9xbQ|!I9@@`u2qQ_K$XE#-uB&GGjD6aL*koe!@N|^Ur}!;ie+F%0
zZz{94dH#%YF&cWh(8pGo!C!bprw+YY4PEOvHtW7bqD$2rb}WT=SM-+F>OGd;nkbet
z6F{zK3cq@j{-F=xe;2hJu5uMO4N?ZH1Y8||maFPQ&#s1qXy56}t~KqHZhNGKb=q7N
zP980XN6o-L9wI-P{J~eVo$6JqzhNjn-W28Ed}9o1I^hxw61+V(>F!S>4}Ix9e9gpL
zRXUk;w=`cclWkGp4ZYdlWv58GkK%6L2lNnUW7h_2W95cjW#t}4cKtQBRs{0TBY4!*
zHJ7wHgk!GT_!eapAw{4nn|X@)o73vi{<nB{#TK?l16-GWxq83SRJ183`cqSl;nc8?
z(HQWX<DSsrDK{75R7>-JOe*k<h;X5q)FPOo($m<U!guUM<5{kkqXXWMyvJih7^5Mg
zj1H&$-qoBO^jmMlIx#77LKKw{(HO;b8g3cZvdRcf6*x)cGgryZpEjTM&uZOZQCHW;
z+r8eG`XZArMiNaonl5*-4qdW8FviX@JDw=3W^+jry#7<y*{ArJkMi8SnOjm5+tP(K
zwRSIjalu<_cxI}|I{w}Jh1DdRKz|BP#&LlPrdMt!Q7>kN|9rdp<viZWH|J@B^kVo%
zq_y$x&4yz$ZuCd&j+h<-HnYPRYO7ypxzu`Pk$2PEB1%WqiXa8UDCRwiTr>3dn`Bx*
z^o-uFfx70}FJPZNpsk4U!O4_{M2C_$3x2sBKYemr-`CEgHfm6COAkn?de%;gRjaO|
zhST9;<~vB`>!;F<TR!DW!K7jHVxJz09FE+r!wWQ^s#?51Re%q7=+LCKK)qaSxjP!Q
z-F34-*k`Amev}Nw_hv?(_o=)Lxsb{^tDi3Ov~j4U2uV1ySQ<$pe0LG^Cpe@?u%7eh
zQcBfL0~6=9r88nEH;hHwP36|kP1`*9rQA;7fW=t{L3J`mhUzx;Y~5`kq0q62fKu$T
zSULRq(qJf6fNf2>^ONQ8qrI!8_2~qU&bdgJ1D#z7?b0)SM5?{5`Pksr>0#EjpckWD
zaeYu^OWM^zTGvF=P3-yHb^cOUwHAp|e%CHxrT#M82rTKWvhzr8JD~H4)RJR$lU|lJ
zzYPrZ*>kBmhn?BN!k=}!Qpg+@VRZ)vsCl>>__1FeDB+(hH`<^8hB#RW+WJ9fepuD<
zF|?27@3&-&E=-xEg?e32c!f4!);CL8n=wgsma(W7Rp{o$U{ATRd80Axw|VVIb}Cdp
z8O5_5ZXu0~D@>#uVpzI#uIMZe-<8(UQVp2lDJiE(y(~@v2!&n0K()0{09O1^h;K?`
zsl}+b+7YFO{t8$dcO|n%#{A0>=3LEoS>NJl+C_hj{Q0^#6c2EKubYON7kp~Ais~1u
zW~`+If9f1{Odz`lMYA)kjU&o)p|w1Zyg#@OC<rh7@zt8rn~;k5tGcssE5*61uzHrT
z`I@hpM0Ng?s{@63aKe#!rk%MF+2jNs`}3{L6(KhSLN-*QC)B-gAcf7t*y>hos5Nz6
zMw{oOmd4op#MB(Y%1U&G-8AB+X}m*R*~$oc);Yc!)*-HSt7fy9xu(0AY5hICtfPCm
zXQkA>ujz+PL1=rAOn+sd%*Aq<=~UXeDe?Le@GknL%*;$aJfgt;Ef+GYp6Su3iMTih
zPmR-RI_PG2S7eO;XSa;{d@l3N{7Zwa&G$_>c$A}g@RPI3N{5aZrfB^6*n{XjXk^3N
zqePeAjy9qB#kFuyTGYsN)C0wA$JDvdf58>Xo&cc53DDFOPK5SP)JYlS_U1}0hnzZ5
zhL~Vih)OvReLtXr^~!kF87pZ<UgfN?M7ygDwBQujS^ehxo3UD>1sHN!&%%tc;6azH
z<c*lmu)N#-k^0?7==xm^i8YIkCJb%S-jPEv9deer2HMoC-POs9Ht(8GowJN-yuD`&
zPc-$VKZ1*G9)0~wwRxrCn7~^-M=777``rXM0O^-|ux-F+FoE^5RCg+*AaIL@7UDLT
z40x6^;@p?ZqfNKhd)+vcE>wzo2%Bq_+0(j-Cl`Lv(QXt;!vMy6R`54le<`Q&5ZAf>
zUAgmZnJ`x>(|o+G!+7T}(&tK~MD%YH%GRgrIbPHf^yQvo0_UEg=_1soJB*x9f64Dj
zX|b6NLY2Pl2mDFtr`C5p`GRTTwI{)Nwo0xJl#^swYg)eq$EU}hzf5^qT4Jl<HD3fz
zVaqOiZLBi5y*Gn6ecNBCK^)!3AU<&;6Bv$l**#sSu_R<-&zf9W(`<8glcJ!&MptDy
z_k8|*q&X*I{wj`dv4NmZU-BYmoqOM%HMzNBzG+8SgKT{=?VLNm*+%t{`Df5!km(Q0
z^}U(dCLUwep}^ee9I+X16MV`(y!2<6niM;Y<im(kFySDye%^D)yK&=ew}2${+5=qK
z{0wUq9MIJ@WqZjmzoWWIFkJoRf#N#{HSDqfu;?wIM+v~e|K`pcHh}uEnD_$uT7PI~
zU`_7cdmaAYBP4QfDm#>VY(Az+LOS#Et9K;XPk1<#OD)B;fL#Lr@jwaJ7Uj@*kI+M4
zy5?hf%C<lYmh)KmY>%tC6wG8oG_AtMVdwMr{X980Alna|UwGDJ!&iTk56{^_Hgi;r
z{2yh^4JUVS)FV;<S~ImK6TU_{PqKqvGuDlk%bN7I;0SrZ)%fSh1BQ@Ilh}atFfyoT
zyRw&;qS*{mvtPD%qpvtDCET5h?->EdiS3mCuFO|Ij?#lv9I+$za2#Ui%o@!8G(o|i
zlGnw-^~wU>eK(~`TCO_a2^+4rl((n9_qwjT$7`--Y_P4FWkyT6rcN#?&B;jWxf*U{
z%q-qOU+-!P41k!S&d#FtO2;y?hH^?#8m+rGZA>&JLYpxGmj6KLzpxaD4K7i`>dp&#
zQ7(pN4den{Zt8u;A=w>w^&ZA$O!`1bBvUM5&l;$Q1$U};e&~8Ku4P6H_>ik__$N@y
z@F-fys56UuGU=8xut0Ycu*O?nbd)~1q&5v~s4gHd4ht1M>wblVFPP70y$_khZ|%QK
zh^Da06!IV=P1U`d5{q602CJ&3uquCu>&`knTt4~y&H4U#&tmEi15_ubV5hB@L)WY>
zjdLMV#i8%Gngpt#=b;YZ=qL@hLC+<dti>N9Rs*g!@yydQe4kL%oM!*meOtq`Rc@H5
zwu*w4_lJX&?=*h{!#~>92pvSlC$|`itDvDwvGx*R_kbGk8@-E<9iXg_0PrE`WaD{{
zng(tw(CR?qbJ8Xk&jRfU5}cx)0RPn8uuyOnpodM%wqPZ+0D0bp@E33uRngJYDp<1v
z^7~(jRh6B~y*&zi2AkhGP3*pVS}R+6oeCYFkV|xAUVKx0+4Drr-wmg`-*wF7^@n1C
zdl(LT^W;PGw}EQN`Q5O0&7V~J>rMxTD-G-sDPHO^a``ftmN(;T4;#JB(Wnjy<!HQ{
z`Yp(gZXNhTTU~N97#2vgf^IMx@6hjGUlRBxvS^tE(uJ-~9OL|y!(iA)7zUh>vQf``
zqb*4IR-}~MjK}<rq+_E2Dqx|McF&pRp?OnWYN{|-^qybt3ZS06k^_F4=T~>k;pHoP
zdWX(1Az|DfNh*ZZI$F`jXy!ZH!DDjj5M^BdB`Qh(mSrY|(&N}xEh~>6i?YQ6dr1=;
zn#rD!2bHCl->yqtcM#lbWy^$h$#TD{0UA;zP%ikKXgmeN>9f*hB)QZ!Jz1B|G=9CX
zQnQYHz-Bbv68=f!%&8NJUM;46ktHO(Vu(gV_vuHz$9tu-+pBb!KP;5N>cur2=5atn
zVz5@k;MQFA%c`rsd9hI?L6Rs+Qefl3gs7hqpXXRmzgty9ai;$m^v}G7A1Ec^bg|Dw
zslMj0RkyxDjJtA=^T%VrQ_FRk1*y*fsRU4CsUzedm8IG2sk0M!fbVkq3#5c&tleYM
zib&v*D!wDUG<5Q+9G*V88rHRR&q+zepC_&*2p!;=nuY|~Tv0t0KY8bN`KtrFvwliR
z-1E<veP(?jm$ti|^Y$D9uQGHr<MmTw$LcdNH+yFXvh^@C$5`9BAf6m2CqGu>R|SDb
zIviXmkDMPl1nxMl9aB)|qQgET3p>4!o>`g89=hFTK`45$oLNl3FOvRw>IK98B3Co+
za(b3P_$}n80w)igV=!EV{N-?xAp<^l3aAWh4ISJ1AZMbfozIP_Mx~s*w)z90um3YU
z-<L6ysbJ=kqViyhkP|}b*sCcwUjRC?rT%I~bdFk#^uM|cJMQdS4Ra~WJpuG)o?5*;
zXZgSC^m=`Ag`c|KJ?l|wi>^;DC0+5dn+vFT8OYyK|3@M!(al>3pTC#bQiID_%oyCK
zQ_r1xyD>P~T+Z&)`02Y>idm(S{>?Y`B_9I&Om|5=tYAW&X;{<4Pu=#HnPwUm+X4<B
z2<6!v?GQtGVMqAXYCHXH?(cL3KdDMa3(cpmPaoNt?GDrfm=BW&u|_1yckc|K7W3Z&
zSRB}dJjK!Tyv38^Mdk~VQW_fA=W#A+g1uH=I{->BWvB?G2Y!SGM@yIKn!gpqo&$=J
zT0nH40BNynce19I4*S%251Si1XvUt?ax!8Sh`NDY2b|OaY)B5bSQ6JAN%CaNd)4&f
z7R_ko;_^3N5vaG9zVloDYK{fjE|iSiN<blH_PMgQ`eeorO~xZ|aw<k0ygQOcJi|c8
zIfj8tPbNS2Cx`tEYFX$X-}9tkj+N<k_0M~#nb?90tQv=?JSTuz|331-3r1EnpcxqA
z3XnV>V!<D6>5|=%s}LW+dS-rE%SC~|{0q<_CY(YO1!z*%u;=OjV-Nju@m$3o>64nI
z;0r^Qa#>t|PYd3P&@v~&W}_RzJO^jF9d-sxm=k?8s@=_95i%(%pKPBnP-)XBv**{G
zL6Wbh$N#m5WYJOjExJ*$-ha4*z%PgFg1QtT%Bso~QiLkq?}AyD%IIQC9rWHE&aB@&
z$+i^~1>YGe9?(O4X@SUwsJ+~4tAp*EKaEi~)-uYhNeV=&{RP3QoQTfF=U(eEzOyoS
zD8`M30pG2G(+_|qrl;Zqvi-m?B+xZ((yKk@fzku#er(ek(8r|ABPGi1G-am?$ESqv
zPmVryf`YtT`i+VMihx>;R|rRq(~g_4-5<HhWW^utj#aAOV%bIAh8O(`Gt^w_nWJ*-
z;Cdg*vK6>6i<TSssdp~(KlT%#sS2TGK9#O?q6p&{s-)*x%*PRZ4>{=~;IKU@%f*j?
zSKq16--LB5d~&sRQiMG}Z526)*jwb`w|*!D$c=47SQ~2g{6uvdKuW0YL#W)w;mJwv
zEN;ucC0t#!OzKbL8iJ6-jU~=AIj@U&gPlYU+uho=uKL@nzNM-#@}xV&qSEz7Bz)iH
z_^&hF0+QwPj(!BMa_ZW>?y<mpD1j$Ekfg5Xji3sdwL-0viBHS=110i6-Mn_6Ty|R*
zZ6FHL0_q9?c8>vQ#9I1021dp9Mo3WaAzOIG-hL!v4NE5F;7Ci$VAtz9bqUG`IPONy
zhd~`|vi@~A*%UZuoi*F{)=~ugM>#&uV0T56o@SqvUX|@mT^no+Hi^9#I?FM!*&#`1
zkz@B}qmJ+(?Cb6A>c~lv)<h$GFEd+yHuzuPK=&>FPZrp;17y6}0g@{44LFy+N2{o}
zd<Pr{AV45}DA_cgIw$P>MGb{R2QGIw6olIT?`);9M?hkty1%_FNroHJWzD5T#hg7%
zM&$>@4XeGx-6Fc#L(eF9dEJ&$ykMHuJ0GGz^yF^erWB@hC9H>&04QN3K(i+V36}Rp
zO<2n&D)MA$sZ(*#N^|}#mH+!3GCHDN`{WWR_o$e3gu&Ynio>hAohi(iPYT_<?harX
zD*iB)dD4Z#Tiu_it$uB)CC*<~%ExJK9q8M5A^322Yb*E(DQDJ2u+EkK;U6h=R*foF
zmBNn%o^<mO;Vpjj5N0$iqE}>N?sBi70KlxU*~~R8Vl{}q*d=;T-g9eW5NJ6%!$^;o
zV4^T&!uD-Xtvg(3Pe4Z|Wnz3B{x&zgz|b2Z&p6ffuLFzvD5y`<_EZSsIKz?;VCE-=
zy(hqSih1oFjR_y$u?aR2|J05+B&oD@713&7?g`GziG2;-yJO^%K0-yhdD-P)mEx_(
zz`GfIa_96g;BjCbioZQ?+SD*T%f;;Fxb%UcwKNEm>-Er+>Qzeny%NlZ>m8f$sdAI<
z8~DNHN_J6Ek&E6aT&>E%OgfbxN|~}dS<4P2HW05Rdz_Y#JT`i(16-tw+{iIdc)^&T
zpHJEwqc>AgS=|k^0#!^mz2I(I-JGS9((iFyd%_2mBPK<QcBmpz=$&+_BElyYxYvkg
zcbvMdmw`u|Cb+`DEVFk9u>bo0f6v0-e||T`0E9+t@{Qn}B)-KS_&9Nx?xwhgbGBQm
zh7vgje9m?Es-~o=&Xk*59tA?MpP<eB-o-Er^O48?fTA{DtV3z|8xl(({oAhZ<?ydx
zkjR6$?YVIGGmYVsPQju2=CYto7;<^M!C;q+#GBN*M$Zsm0En3APJ$dnhMr<&`3(%>
zclm!w>pb|wtI+rnxM56g?()Cb>_6}igbhx&r|`p}NK+sLRveg^n21@v_s817&Tgrl
zF0ds~`gyMI=14*%;rk~<O*?l-MhGRA3&3U}C%Q<<ra`sl+MO&e=@oqG14$@|0^q?v
zsreq?nHO(i%orKo&*C`MQ9KDk)VMz%WzjCMe61mCPhgNg*x%$HnWBgWf?clh3?n%+
za}K@VStEGqUXcEex&61@1zrf*;!jptY0TPold~E1#?&lFN%`QI+th8)W5H!@3mAx`
zftcB{MIdmHm0S)=?m$+E4POWXl%4u9y$&RWKg57@()r}JUkHi<=S0dT{_Pvby~Z;!
z^~32G*09APJ<+}^^=(pk(8b25?YkrQ8AL1eX9&V@XDm$7(yq$`ekWzJW4>-|9YxFa
zfxbdF`ENt@kGqr;D<mf;$A7ULZFbW})btAs`wVo#O!1W8B8AI(lHGrxgoU($A+JU?
z785+sWua4<sNbiHMkh>FtZQy1&VT;FgujfsR&TR29Sh#u*SD0LlPdX6iSAx^(%*=|
z=!w>C9nm{Ct#p-ih#i#B@)rrRbgdkAwlq1!dT*Y}(fGp6@opLZS?qr!4YvWP%qMrq
z>*fdm9A9*t)KXZGnI0y54lvc{um_5r4h~^}A5G=olHxo^0yI&w1#0X7ZE5bgR2*wh
z=_h*#8o+0btQvTFvX1?74SaI_0rubr2)fvmP%?lk{_7JyL#q>Jz6xr5%Y)Yq7<7mq
zK&hUKz?_W!?>Xy~7V^oR6>vZ4OtJkvOL7pD;K*iW=ek=m^Y`4mll`Jv3O!%&J&?OS
zmfMgeeEIR2s*XqH10@>n34D;!4&ZyF@_ufA=961FR2|lmp92%Z_Exvg+-zN2e1G_;
z;B;_sJ!w1q(hj564EQ*2(eU11Jj{KHc3i>6fHERov-RY%Z_sW?M@9K;_wIbv|NngJ
zw&APEtKt`1CQ}Fs{-OYLREy3tdGejlYfn=wGZ`6`8X9|O@@Bp&GW}7hYu+X(@eo*6
z`8-gS_3!6~Sa30~wi6DcjUB~0z0H)gX_DLXh16lbD*r$kstZ2#H*er*xiqmmYzrqM
z_({^j+n#`YbFklz(~VlIbtzFxsrXl+6HTX;fO(&!oR<PO#ynEmsM)RClC7!T17c=!
zHn4h>pxDb&EWLJpBzS9IQBfgkWo6Z{V`Z~-_ReT$y0X-GAhBYvanED#3mccTE;p$6
zIH>WDbXg(l#5}_U2w2eacQT+K!+##We_n?Wc;_1r1l-yVPvzFHrp~!!<Z_JDlC&aL
zMIo`*PX@LV1$xl=<rs>iXl|E|^jbd<hHar%PreOg_dqF_2ppDQRL9(?xA`kwgs$PJ
zbpGSOn|Tk&>4%+R&?Enr`Ky8X4hF%C9#boF7mZn<0B}j}W?61g)$2rcRF~3qITBSS
z5HSvJzB*KE{Oq{CO2K=(fOA9Tak8;@HMn<-l?6Qrl<R=YB#rPW<;aF|H64VSrGV$J
zcb%8m(=Mc^J%`-p@Qni2lcjY*pkp%ZQ=7J@HOJ5Ji3z;7%UzaXUu-^L3%I_;2jYWo
zDVw?ePG$M$RR6sX=^?|I(rJoMxrQv>v$LWS41$*Za#^k<n*wiKTv1Vwcaq<M5a2W~
z4$kxIK5#P`0Oxe=z5@ORqhYCPuf2lmGy$g~1l?P-g2K&NoNPC^tZ0y1QQ@gLe26SO
zcH5p91U<=>%U)!J$Hd3yRKEpU!7(HzB^lh_TpzUJctTKtt-g3L<wERC+HT|7-1z`7
zardqnEZooADZ`Si>zIHEaC*%ip3qW)Da}h?)EA`ax6^@-+RfjE02eI+IbPjFu;(^^
ziB#?tx%lJEoHTnd>-370Ny+}we{}G~Hjl%9=~8-nbvIb;@WjYq8LTJAe;~@LR`T7D
zU<oMp#V#UU0ms)imv)?FGe}*ehAOcQoH*0RoY!9VR)^1W%T*SSuF!B5%{p-o0!?>O
z$c+pSwDj5~tNt%;c4srY{7@cW8XcE(A!X|mr8U6ytEY9PO9S|$J@ix`1s6HKEn!|Y
zjHxa~e4aH`G@fNm?Bk8Uo)7H!V!@w{-o(w->`$bIb=v1I=4FrBUehVtZpA9w?($>7
zJAP(khG2R()@K~W1p_T8@jn1We)!01UXZ?M$IuYuv^z_3Hft}(>pIuqdT>(gCD3@W
zvC!(9Yq|-(Jp^Y~S93uRBf~qIulkykOMd1EUT?T;Y0N(_Hy+^CW7R=_F;>6V#8;!-
z+XL!~`T`B+vkMwh?Y5J-1PIX6pPrtDYe$Bs6zsrXh^ZnL|GZsi^FSWM*iB@a8-%}l
zpNo=|J_&T<)a$0)K2}-`u9F-~Xy<+rr%4@rW&3g&xcMRTg+jxUDw}C?RMR%sU$*T2
zA>Yv`k4rt>+}zCNQknxIh}+TGaFaHlXKrt71TrbXv!E-%o>#iX9mJ(AVg}m^7V=TV
z-LL9xu5im&^>b_o+8F{pfS$~7%{S;?hNoWy5-t*@dfe6Y#Z$_j;a~bO0QEy3jvxR<
zlBzAWvGpSggc7$ZN%WVrG7R0r)ZjL@a!ozn{;56m<;v_yB~yY(^J*EkRoS4%(eW%K
ze9}}Y%9TPbUQ<VBem=Z#@n;+&Hem=!TPPFg^%|2DsVihmKHYD?nzT$aBd4{KmzQ_5
zMkEN>m}|CPhMR!!((NJ4MLn6$SgncwZN0n?=pY{fKYH2G-LNrg$S3ze0PMyE*f;M|
zD*4DO=Z#7tIVtv=&p70PY0%%~f*&WI;s6G}G6z?>u_2%R=WPYj^iB5zr7<AbmiRzP
zy!1RD;1!rp7vj%rnSwgNM>?)6v1Tblwm0YDx7^w;f4&KEJN|Z@$iA-C0jSZU2qjY9
z{i082ko<Xt$Gpc`Cy<?2q*mv!@tu<b;{6BjSB{sv_S&7(l`S?-ntA+BFMx#_X4;`a
zTsB6D57gA?1{j1tk@pgIGGzsq@L=*eX`GIj47dV6{lDTBh>Vl1GU<iOf1_3K`^pla
z`V&KSS)>iyJyJcml7q<0a5}_H3;BbcGOd8R72PR>#d^^I8VKDk%@*AjZu8@m@LZ>%
zA5RA-t*eH_s0(8n917=h!{julpqo=3CKHUYb*DHwr$g%05iho;zU!DIzjRbN))Vm~
zN+{3Kf&SQf-1(97IGRVPLx46V+L8wg??ADeGVsWR;yPw?C@y5!ICV8I-0>iNvcw>y
zQN9<d#d`5pb5h<I->E-RfU_6UqCLdwc+rLKRF<A*d&M{`)&RsFqW0f)g+;keqo2}V
zjdb(>!w>$=M<Cwpatq5oxmA<1vr$)<ue_24dYVsIVZB3LKtm_Mc>Dx0)5`^Iy-J8P
zUL>p41xTL^AX8jakOLv>7Suf$igiF7MOj7)>U=Rh(Zu;3JhL$VqtWfGtGfAm#|6MS
zz8?XS_Qr$B2~B0;Lr=W+&OE_Dv++GM6gDJ&(Jr0fniSj=YxE(9wLmZpQ#-d!hT4zy
z?eX8Ez%pw6UOHMrj0QZVhv+NTXN*$eXKn6Cvp%9WbUdrE(J%uYLn);j!+f^s=iTBe
z9m37XALUJWj#GJZwj*6$Ras#_zn;)8QqgZe3trXr^(tGGO`+34bj6)Wl!>fIZYiBk
z>V+4Mcl&n{#9tYo!J(^)Ikcse7t(x^`@yG!<C82QwE~hanAlpwL`nJZJQlM(UD^{p
z5!VpWSrJNu#a!28g*BdKa8}3~bSp8Zw%}zmf3zXyC9zUe+BW|p6%)l~M;OkwVLk_k
ztVD7MdzZV>YM-oQSIIX9I{U!Ls7l7z<K(B{p;y`ZCNIAl#?1DLjc`NN{G*_o=TSJT
z5$mb~JNrcci%NR?Q0caQ5GM2WMJ=TFsN}*2pInZfCrU3FT<PY4q{fGj;MO%+yw(R#
zlyXj38)3j&P2x_L8A)vh3<((nkzv-H2If7WOR}V(^98D=-(@*cT$a}h2(xUMa0s5a
z;wa8U2Nuh2{`94f@c#m(>GymH>+aXO+Y&J#piBRO2o5z`LLCUpe9QkucE_XVqg!2u
z?KrZc*da^eg73}pu{ka>(rEQ?(H?ifq~3l{uiea~Bz<kr+aYF5uD$b4%PA!#WkZLo
zWbCG-uj{M8izkxX@5?5pF^u<_>Jf|=Z_`#wi8|<xBWYix>sU_4<5@sdnfSygE$|ua
zOWVO{6jA5$@!!AM+Bw;T#Ov#aZzmnCI@$yE_)MhS8~hG_`!uZ~JiJ7kS1SZMyjYl7
z3qOh9Qm!iLGiXST>uC)J9)yL3Rm6M>bS$$dec0q6du}3tlt|2Rns(jB+F}dOVvnPg
z6re1NOhGaQzO&s@AFV;21ieKDip&dk47lrO4Z=1n2q=&CWypa03v*w(Pft%Pl5T!>
zIuP!PyEtZ-<*N^O7`?f(pQuUu_mGc3NeMp<Y~@iu15!3vVb4_?oPeOYJP35&@U>%&
z)?GVqXx9%9Nua9v6m{jw642&cksq*=2Si>6L^;W)Q$wDdDn*(kg@uI&rQsf|yERK;
zN`s}|n2*1Gv5t_0{vZviwUXAI#oSLQjS>VCV#edHUtZltU$xFeYbGnTKM@o0vGWlM
zQpb4ylNOWcP{8Gw8to5`+v~Q6cwYO_xSYB)I(UK*KOH>IkRymxK_oV<4sJd~{Afd5
z=!M(rk3rbY$nf0UoFIBqLCUGHgmH)?qZy0w6_c5VKDG60j{<QmsF7SCWwvDsbeyJ?
z{)nWT%d;ZRJ+C9~tCfoT+qBbQot<w<(fcpq=in7u=yrmMD646-?n6E!B(bkERO2A(
z$(m5y0;ZQ?$6J1McS;DlSI65fus+kOnbvAVNlsXbe!uEa(GYjjhsEfV<Nk$sHUTMw
zM2N+cPVr{<(VXiZ6(@edspgKQXcH`3s;zP_WQmF3qL>E18ak;voBvJ0ON;1d_c{Bi
z4f-^9_&ARLk0X}x0*E1;xF}f^NE{;X)40*iXR~iE_WDiN=o@QEC%In{4Fx?4@kQkm
zd*jhVexO{(D?<^J0&;*xse%i`w!kvhusZ#Z$$We3FHh(~erce>Rj<(iM=sEx(v$m|
z-(VCjQ|f+kc#rS=9B2XFAi4L_##L&L6pqW|aeFx*zB5zZt#f;#lVig|vOtwnYC4>8
z$H9X#opz=Pyl&6D(hUB<F+)QBO6!X}zWDvK!6gpmJA?>&w_^=S67eW%B$+?*E7RBZ
zekOSS6GIS}mr){%iu(QwbW2*?R#}57Je_ml^A@sALnPh8uqN`2sy<d>xUSCy?dl&9
z&O~tyj}VsJ@F>r5DyD~H8z#xDY$}|)1tq@3vT+r0mZnlv87|8rRUy#<Qu+*C9Ws_B
z!xY5vPZr<kcT?zLy(KJdGht#l9ndbn9Cr;0KR;lVc>gYU4c$8Vp4ID{nv?h2v|ro8
zq)Ar|SD7B##@S#7y^$O8-ac*=OD_0EnF4m~6q%Joko3_{nWcp&bc3r7<-Eo@C&E)n
zkgnJqvMFjS{?`ls(9@2OW!W-{l-G$lY+dTXU(+ll1m-D}XuR0|m-h|iw_(7o6DA^z
zIfrLQN%5qw`GE``1Yr3zUf-U6j`Etfo^_@gxZQU@40h;g<`_y+rxbsn<Os;C<371p
z>$7)>pe8XbK=NQyDe|1<f)&P_DZbeW@M=2@&CeUz*8*pAZZ_^WM`?RDf`A0EQ7G@O
zeTed*O7+_PZrGH~qj8*{N?aq0Z$E!#i9yP#yof*oca4yca#{0df8R9p<TplBM>Fb%
z7B9IvzV?1g1$f6&8awhB6;gtJ4UZ@(%C>#|O70!JtqSMVNB@eACkByCivr;!iu1Af
zeGlC$P(#zWNY~*!&l)7oc#LPiB4?3xzq<)H7`cK^jh=xueNOWyylMO6@sbXm<LVDN
zHpn3FrW^y=7Cu&}LK#D#r6~)vGvlH|L`7d<N7<v6uBo)$#%b>``nUb=tfFbTAc!9i
z8px#Cp?>+5&M6@G$B|~lAMpeM_EmG}bx6BvgX7>7bbOrqs@5^k(dC-bV6rT3oaWVd
z1bDN5f`T|F4ChKA{!<|DX%LCV)z5gOnNs|arG(&346Z{dc1S=9YZW>+*?R+#XI)!o
zE~z~oU7i)M97`uYo-?RiEmCUOw)si3hM37s7wg*tLc+gc_3sz*9|6-OO)d>mx(uI#
zaA>xEFgmp9rNe+nC1n>$mpxF*J}GeF0{&OFh%TfRffG#%r+D7=L<xwN4YL19feDr3
zyP%v@luy!dz8GwKdwrIZCU`vpzE%XXxae<AZEiWQ`+Ws&7Qr`*-!(jDX66WWE_zk<
zPd3MVrw!51ufD48G~S#P@7-F`^X=AQhJ5`39WCcKr#<F3X~0~+Alq}?Zsx|2mUJKu
zL{ibyef-8KS0NZJ=h3$<@j&saRq}0#kQgu!U(W9-fb^e&u<^uohhQM=3OZOsjc?Tz
zL{-8gvHgA#bi5nf_UI~NT`5p*n<O)>?kmwGW0yl)DW}KeYgM)p<0nzbjcAvzyhphH
zxHww@(W?+y#qDV>Cq#XH{b=p^gm#Hhs(TN*Rp&#C&d67vwz1eE4aIXg7$wd#fkdfS
zogGaJzU!FQ5Tj;@afmb}9cRpE&rW|Ro{3aOA$$7Eu5)S&@@=(X+ktFE!ePTYk32uG
zTgQ8T`*Zp6gx5MIgRs*MUc)ZKd_^englAw%UDVZ&LZ0*dqM|V14vd_|96vulPhsL>
zB1<zf37vd#%>sWR$;)L&w__zq5F_=o8;-CiTNXMULX@nf?Zv0>1p0Ii>QKqN|2-JB
z&_-G8ftZ+|?xyac&7T`mIG!j5e9-Dgc%KW(i_O0L<zfP?_&d3U7@yoVjm+$&a}&&3
zQcqV^lL$H(uZ6X;=8Ed!ioP8nD(jMA?z{e6g96t{F^<j7lFhz&Hwc${RZ9y67?E+k
z><{}R1kfhdcgWBWxJrCA@jO=NBK90bd!O0Qz*?9v_tnXnE^8N=7^Z+AOk~S}Tury_
z)B0DEXwhC5<J@I#_73#J@yZ{%5e(I-c9gCU6#EzbDFx4u+i@b_s{CrP=*AU`pc{{n
z_o?cj#Grg5%k=c?i*~^(JZD{B%IRa83M!*_c+Y7lwkscAcgW~)Iw4~dd<8sLluj#Z
z|M*woRh?(hQmoRL_7u^%#hyoc6;ccKFM%auCZo$QO(@)c@h{@i{MPKy#IskP;DK;$
zktUs<o+47&?3ftTqVyh$9HbYSpgr7QBzvzu{`s^c0?%zVZ;-4sE4<*yuX=p3zcb0x
zj&qoPd^U@FL)CKX<iiB7`C1nP_Iaz{38Ms}G~O%KEY<3(!O6)<t8!8QT7t;zyB><3
zetaF59)@|oNwzM{Z8d*RR?}cikExnK2~Q>tv0M*x>?+y+|0L_8;rj#eiA^COFH=8&
zOPU(?`SYIln_Cf|ZBXk65g=%^eET^0Z;9|xFpyM;ppkR#4tHorFoBdb@JgTYr0H&s
zv%oufj%2TAR=&OK7)D8K_feWXHLAH!_vJj$L#y&L9?t<uzs5Xr{FvY~z#bWX_X4?y
z<iH)es_w|qG})bF9$6<v>v@zg=Ca_8t`D>>+c0^FoTcaMcEl%U_Lk^ba{!7fnzNIo
zzvE);laxkzH3fd{i^StrNQrb=q!yX{E=sN^QWfqW4z=vTpfioOyVW%&bHk)vCgO7+
zDUQ(+xgl24^4yo`^~5)Q<?%e;CbmC4j9ctrR6K){@-VZe^5H3BibaAgJ2V;m5l3S@
zN@um)<~3*CABj-{5$JpnjYU$S#Vgw-ozwG?R6`weIya+R!_sV7L6V1y(ZB0g6{3!3
z<zF0nCw+D4XVA+x`NCVpYcP!{_sMk_DFsg&4-wwt;p=5^{8*Ao;IyA2rMsS7l(L!{
zDei|wt&SMWY;VxmTRgKpVs0d1H_lp`JcxJ3%u);XSW3E}#f0w*Qx-g9xJk$-F75W?
zgih8k8bSjVJt_!jvVpp|f3(YgzYx+D8+@W<KnMH$&aNG?LMVMfjS7f+`_O*T-9$>k
zvlISu10)1bAj5f3?Xs=M2G+|0`>2v~o*$-B4B)!`OMpE4I?L0`e+3Q7f@J3}2AxNS
z-g4T^C3_B*x*RMXA4P9F%^id}gLrh1nS#hnsx^4p^|nc^mRVA@4~j**cF&rjwbOkc
zS}sdt!=CW}JG3H?`ItFk(LdF5pE`)@=>AjN`<S-(Nk}mnXrBvPQuB3SOD49%JZsED
z6qlGQp*!w@o>P#)9S+(+K{elyl=1e|I#J?}QrXrmk&RzT&kemoEP$YFiJqrZvA%Gc
zd#TNIkenwhKg)3)U!+UTx74x?eA%PLH)AYdS(aP|mCPN~cGRA-SJApch~bKg)#6CG
z*u(ex4o9}mxs^a7>ZFzl`SwOnkg({J)|zo$asiY5ZrQ5O)|39o4TD!C_wX>i&eP@W
zONj1wib%AD)pY^f=xF9q9Y@flD*;7y*dOra<1((R{$w8Zm9F!_6u!|$s;~M(0r|M<
zrGgz^=e|VH(x>24bb!d=hj%ES8<AW7<gIp~`9C}uU9`JU|8)T%KU10p$0uq;c`dC<
z^N#mBw<DsKW8LF{SP&{1hy5-Y2Niq*BEXXbfFwkM^F$nSvM}esx507&)DJuf)ZHWm
z<b*!~<n6f_h=G7SL8Rs}omUUfWOT!n-f6PZ!2Y=zeajl@8W0XK(;n+fXj#XMp-UT%
zJik~pMEfT6Whm@&6o|bv?*w5!j&2d^94PJKH#7_+{Xr8Wq$x~Eau3w@5IJVd+xGCa
zL0PtJMku)J;M=}8TL=CWEj9Wno`Y%AHY}b%-tgVCf^<=GO4>rH{<2}>n-8izWZMBE
zZ=bqTEq@O!5q0ZOIe5V<%=)&iSYMZ7wgSqw73(=^20aS!cAYGv$0S{a#92tCo#w#G
zF5i<&e*tgvBz00)?^Cz?1G;#I+)@nfpmZC>^f_=>d-vzV+$a+vlTPj<Sf--I2z?Kw
z<CMHPI7l2?N;5ouB$`X>HXezyqiU6^V6?iCr@DDnb?>HRrIv<G2b2Djsj2Dut@^L=
z@i^`Iyv8Msw-C0z*AXr9Zflnb0;5Cx<Aa{yz+nAP|C=A_qFrEws6Z;&T<Iv)i<d<p
zjyuT-2fOxbYZ`;7+8SUy5CW2Dv_29{FeJpxL5x~I+)EA-A?8Bs0W)`l8~t(M8C)0N
zOzw0fViq)HNaWDU9Sna*V34Wulg+4~`jewQ`1&M(^<0n97RTx<7VY*^c}SQ*V3GlC
z-g~|%a6fgA>|2}x;Sld9F9HjG`9i1UjmtrbF~{uTjw`_@hH@!crJBvtIH$Co`B6}P
z)i{_)L)lh|-<AC0eK)*p_r+jQ4bLB=V7IBoXwOOxxh>;MrJU}jpYe-5VX^vJ(w$oU
zrb?<Rq!0B6Nq)&mq=AGhdU)@<?|Y=1FxAI>6`1^0yq4^^FRU+Z@nhigF<YElT+n?D
zft<;85<_eWx89<5Vy6&lwJ0z5GIS(ycky_RY}!?sT-Uy^;JiVe{UxCf{c9({_0Ih+
z&AI<;2yL9!de!{pxO|d;>0U^5LVhIyj?@2P&*&i^&<v<yn^5--wA^&G>J$_@KKlz5
zsrQVUg;ZMqC|>THg*cv+svchW<)U#fpke5ZTtab`_pS@k0y?Ns7}`RN&wX0EMC?|{
zhfmDoz07Vv=cUbx_exxH0p$HG43{<nm&K6s^)v>>`v7u7ani{9n53dl`bTR(&+!Da
z{D4_%mJDQD*2#E;JjnbCe!XRmoxM5Ple1Z-+M{Xtv!hpYjKqMlKu%xQ$>RbZUOR1J
z?%uC`qpUBIl68R?^ZzMY+T+JF$^2@1(v;%5<L9hp?=-|;b5kO@L+MhPuP@4;hTTs-
zpx=FvEDoiM{POq2bHeavJdrjW`s%s*0(+UBes$|HEq4^9WqF{}czmi}XCHThOQ87D
zx$cu`$;n~yaehgC`Kt=oCloih+_w17xx|LM{8ZO%MV?(|$&hV_;*okj48XEFeN^so
zdD;o&SZ4q=oh!c(h?w2r$!GVW1m%la;kH*6;JhvA5~G*@Tk!;{5P-TH47gZ*)fy`I
z2FODpL96x2T@E!o7|r1uRXs$3R5*~*o^YSFLzyK?3iR`CJ>~U5=lx+0v=^3xJ*d<R
z$xtD{vCaU2FloO&#dv}H>OE~M@mpMEj8Px4$Jp3du7EwR&_c9y6nOJ0V(cS=B%0iw
zZd0g4n3Qx;FHBSzgMoIkg<)$rjuMHh`D+}<5m};OSr9BJX&GD6-MlJU)9Nx#v}9Nq
z{Ttn*jLv#HX5B>dF>X?cVA;6wbpbbMp0+Et67UGH!P5<0#R&Yzu?~dh?LwLh80?me
zbc5cxE_%ME0qhI41|d)ZmUAPhM?oKVap3YP?Zux<-;looH4<IME2<Vpki2}zjnQ%`
zenY%c`R<(4yeJNd!<?u+a*x|}Hy;ykmvxhHJn=tR!_VwViVF*6JoIyF0;4<FCCNIz
z!X60|vq8|)3Enb^Dgv=VyJq>vH~V_zn+3i)<Awsr3a(B4Q1VOT)NwiA|3k?H7J~-u
z0;CkaXA>AMaC1ytsEhK?fbmUCSMt!Uv>{BccsW{k)8B$`Dg8-GDJPJ%Sh;i#hOc%t
z8{JWOq#z|9c&;RaccR}4pI8u0<}zcgxl!Q2ixy6M>qXJ`Dl}E_s^h%SWC(aFzMs(T
zn(dXP?Q3r|s_#P`nAx`%tY2!#-rC~*WQovc2_V81ppsum7kNxAjJNopYn`D_?7CkZ
zAtZ?QG>{Aye(ue~>>8jRoW*PPl_W9v<sz25BEwQN?LBmRn4F-kA^z#1jG$xGX?ERd
zTa9}Um~&HAfMrqlu^!fzlK8f;guX{-QO$;lfRQDM2n0>z_0fBCrFgGR$q;-7iwKy%
z4&L<?$LJvD{yUS@mF%PM;P~J{;kUHZi_)&x5{Otcc7OZvi+y)?9Vo7<g1CH#J+mI3
zZA9Qb${?xiE8Em5b|aZ7d4>)3oU{Uu;i1Hi{Jm`Ta^q~bbewu568VPyRJ?%SZE_La
zn}+kS@8f5<ZtzQk0$)KSO1&i9HzMN%TJy?yEa@~k&=+9gA9NBKK%!r=MVRuT@^b&b
z(Xv4WaJF&RP+Pe%#N@=32XEZ*bJW^FRzXciDSI8io#)B(RSJWWY?^qAns1Pp2f-CI
z4|sqmQu?xTfvka)E^wjh!C0}Uqh-pfi$;mA@T}`v3Qyy^5q<=|zqZ;8rKS4L@l74e
z-9w+FM3HDd_wDl4;qyaHlq1Oa2JL>VXm>_5w>P}wd7YL7+C$`1Y55mnR0@)!!c5d(
zyOc5wobvioIZtUtpPpy;55@{w{$Mfie=o8WQ+s4O8m@CHu3?ei&(Ia||M>a}s3^Pc
z`&U3hff+(Vx<N!3knT=FL}?hhLnNdIq?@6X25C^bM5Gx8P#PQtEJ||da2V?U_{R7B
z>i@0fTC>1m(0iYI&ffd%b1sY9(8$Xad<cp6dcU}@J+j_WmVjlXmo{524S7pYj1nP~
zEI2pC&Bu(KK}NsVn&)vxu~ICnXkgN1(+4z2(cL&-OY|Y;sT!x3O<ucJuysY#c!6!}
z*y+ZI@y%|a&XY?CwZrTcjsodN5^Z2)PRRj@L9`*)Ipl=>WFo%OspjVvJB(zdwwjpX
z@gi5KiYUF~=)<TQoS1A_8(sx}GTd^lLqjsZl$x5FPSLe^y6tO{BGi!}F57D0v!&;j
zrvD5||I;<`modu9i2xflFvuKqcl0+Ho`yc^0K5u-^nANZ;-%H5{45h)F;8w%Mh{TD
zUgBUm9nz=(Ie(A8#UF3PDq6}k|7T}Uw4OrY8Cht#V^$9qyx{G~1@o(E)uy7m7q<K+
zXG8~}FqLeR5zI0OhFGSdqo1$n#qr-zCfP~1ki!DAD0C;_(AD)AAy<f6+Up3(SiYen
z>xj<PQsgl0{DGb<!#f<`OHmeS%7&D69zqf+p9FGSVi}FbUh-6;prP-1s`;UL4pf)f
zF`Xxs`@vu7IK<*KoNOy3=3>x`ne8Xx?p53;S~hidW11v{%?(8}^pD~ty{!hDIL&Z$
zjs;^GxoH|U{5yfnZIW@`tA3mfYa^$YY0y(!0Te>q7R%001LPr-6YS6PTQLO|k$S}W
z%|srS({FeE)wVV_r`lFWveU<Y9{bz5amKUz3CqPrl*&?qS~qQ>iIaj2RA&I`k<duA
z9%SB2c$(0DGAMD@TX5IV#1awS`G3)%zt2+79`Ro2E5Oaaadj2F=DTFPcBA|f^Y+R9
zew|05#2+07f65(UB+6Eg4kVd13n_H7l7fEQm8imdysm3Hd?3EX`E=pfs=4V(A+1ue
zj~K}{G1R8{_IwX<?^g`(iB=*Tmeu+#q$W-S@0zOE<SRW)*0_uoNlN3EmT?$wM5F+$
zxg{B8FOOvt8N(&A$yr+<m|``-#a7CU@gQrWY{;o%2A<x0tUNiRL<(|9MtEOXt*>yY
zC?i&-f$!d&5DUX7hoNws1YHHfl60z<@}iH~w^d5~gq#SS09uQBL1KnncVv9k>|0yk
zuU5=6rnxp?YKLj)1?2CbRbM~9^5uAvLah<Dd-)7pUHi)1G{U~;`b#lu8b5NetQ1WJ
zR5iVMstGizwzjrXg<c*hOYfiLA=F6`@#}H#tFJL=d6m?x8C;+n<<8taJ@r%#WoAc5
zN1Yi8iR<bI$wQ7(oy0{&%YLYTbsP9~3zCYiI4Oov5RpLtFL?Hsf(ZC^c~bbZmacv_
zRyCbz%se^0T5i;7JU$q9NdLLm{$~(W*Z{9z$?Txlyvv$Ic1w*(7`x@~oOVbw*4TVY
zU1sLTOaFe#K;d2k5uFr4td!8c%lr7;GAu*oMl365S`US9+%2%0-SyjvTSVM=$r$;{
ztP%*0Yw2O(`GT_M67j$bg$U8?k^uA)yjpHER)BLM#WQK`RqVq(0qKE{rUwyZ?8!hZ
zZw7@?+aiW!YBcjtH>bCy5LL1wx~~@0_NnkMO5K>C2f5_j&}Qg~QhvtqcrfKYq~O+!
z(^ZvXT!=M%{sXozB?Pjp(@-=@r)fTro3^G9&fCU*Fxgag*auHTsXj)uxwS-V4HQ3e
zHE85Mtoj5M(?#{cQ3J^%BATm`s-&VpCzvDA=ds0fv)hkGrY&PrB@?Om-C=rmv(`V*
zLgliH5so)0xVV(~FtJJg6wcQCC`ye*qh)lxoZrX9bi}69mQ!xPf7*wBo}SPern9;U
zDsEtkv}pYT6T?!P_m<7)gEws(2hLvKxPYdLjjLQLmduJmh5kP65{KTOC6@kM%U7NU
zR}1Q)+0P%3F<)pN$(~zT(x#_@*10csVh)qv|LKsvk^kE+f~)c^Qs2S?F_T+3tl(XT
z$hXUx6=AvyjMOc72Oi%8wbu$L`D9LGE5?S%G|Xpp%|eXxGKZPVy|QzIOi)eHaZ?vv
zdg!r%aOQ}y*kYmAv9@vXd$e%0LIkHgdJ_I3!t8b1HlY`FF?k~jK~A9|*1#<mt1$#H
z^=wIEYmFmFGSfw)bz%sr2wsYQJXHa~Jp1(;a-Q+bCiG)ol!~Mf4S&)!YED>}DGc6b
z@Qj8N3D$t<vLx=4IgY00ql%J11@v(%V2X{7IQ;H86_&&O8m|FfeM3}%W15$&+06Uv
zH}q#Z0uHBSfNMjJvQ=55peUY@a|qn%I<Tu*dexYlW+!SkdAa2?dYZu*SV{)1)dtVt
zRNgsp*oY!^WoK{nX?6csU<T1lvYgux=W5!<^Z#L;{;zs09evn`P<Q=~v(<U*V$PUx
zXDvT7tem6Z%Y4Ckw^>N2#W&bKlp%Ne{Jp1|POhm#@8giiJF*W5xQ)bBK4X5CKx^)v
zUTs0Z8cYF86_XdgLbJsSej9FG!X8_S?EvC0Z!q4)jTb9|B)G4&dnJC6Wjd^9lKZ8a
z7kW;7o^wGA>UccFrJM&TR^e$ex_Bf5;V%@!ZxWCuYRiQEp=?8fp*Q-7YT%jVDZ1fq
zlftyMF2;nsuMR`sfy9Wpx&p|IWdTD}*xWE%BQ~O}QS8fSa#NpLDSaS;wnBW*Dn{bE
z+4%xo?%W<GLQ8QY1pFdpqa^96*1H{6(#2IK1P8I<UXr}O#*2Q&0iABj3r<WESI{G<
zpNGriKBxr#n8mo-N%qp+88}%mZ$p@5pEza5!XW>G1jl9`{$bS?slB#lA>~@6XRt8Q
zRpM4ascv_HyjykzlW&V?GRx}b*~5JRf*>A{k&$h9e;m~Zt{zP~JzT5c8n0Yk%<P23
zIHt}n+W#7F8l>|($j^T(hvFG#`MQVCv1%~8?Sx6_OrtM(5C>h|$57qGyW2&yKB(sP
zC=-M#-T9|e;Wy0sAAf{|@eUmg5g$C?=!r1<VRpA^buR8x=*Y_B_0GJ_G#YRX?w=VV
zM>PG7I$r-lVkZw@KJ0b-@JGL|#-F(DF5cj0vB7dS-4qSwK5z(jQZH9(`J+xr^4_)m
zNQ~38-V#wouoH7J2^X$W-n_{ct?`&mk;sNp*xMNHJ6x^U$6|96Lox7>2U2UiJ{x?Z
zS`2U5tMUiuvMOZD!2;X4`^Z)Js>3EU&w4GrIcX*fC|;~{GYt%zqS;>b>Zz-%OF`dZ
zdS0AnR4aZ&#ih35IO~g`TW{_y#w}^*zeLmK*PMJjkqA8*q0=0{E;b{HPm)r<j9Jvo
z51FUjAL<^%3&oB<uo*W$oivpKp(aH5r3Z!>^SSofp0PXM8)s=~AVSRw=zksZxjbE~
z5Lef1E|shflO1OmXLDUN3SxhS>00Lz(Yrt-%2Pr%CCigNQHX*y<4cp0vNA4<1mgw{
zKgw}Q0G~cG`C*z1!QH}l|Gy~f-^VPhf$euI<hX-J7SQ74HioqPd*v4rDF9*M{?=i3
zemY5>Gv3@<%6NIcjsn4*%au@-``ZcqKjR9@ZM1YQm*kzCCJq3#;1w|~C$(aA>MiB<
zqE9z&gN3Xz^R=1Ggj*Xc<c*G4iwJBvN!TAR;d@1)P~z9PV6lSW^b{x8yw;N<x`ydM
zw0;HLOE}10Aeens0jmh~0O*Rqh;z__VmUFOlPLLB?T6v=aZ2m`%+doIEFl@r>3c<y
zdi&E7wwvia+A>zJW#hAhUp^bQY|0jy%C2O&<V0HO<Vs<N0_7w`BBc|jL3+Ux<(yS^
z=8i=n^H{4y$%9W4zQJ4By@un2%ZL&zucm2)*p@Kul}v8zFMm15k*rC<@BPu$YcG!d
zX?|*-p_){)VXMY^t~PHks_YUu2{T5F9FkG2hFy-l0b@UNPa6$bE`O$H4UPEzr<eu>
zVoG^V8bdkC4;e9NIV?WIP|lgve)P57xaF(3PSS4G+=supQKG;#m~PB$SKc^i(2pxf
zB#Z*AA(!qa92SDNHH%37mK(8^;y>VJxo^<DKlzbF3|GX%M&K|r97}gU`DvI6pw^Tj
zom}O-AM5dOLY=hf6j{3JIh+&Q&-Dk@@W|u%p3OQic2kB60Uider{R?<;xdc0(2>;o
zh+Owd?=v!!CEm^0hgt9;{7_;!jgi6R_j4+Y05&q&pVD(+8i*47+)z?ruRc=E5k@V>
ziPJER2+%reh^D-}uLmo8<N|#k6qy;RTOSUxfAtRGT;B(JcX#`Hr*z9_x(n=-Hykt;
z8G<{i-q4}vp_$M4sxmNdx4PG{V0O4)<~62r)+5;EliVCYNI?c+(VsDK%zyd)84B=9
zdscXJITYX*JODBasu%>VRz~VJRO!h%+3Ho>%_B-%aeISZ)clMAs$5$K7ZIVz=2n}2
zTJN}<KQIY$NN-zxoqPY=wfZlXgUswXkIpkXu)l5qIG<O(pbE~fEH3HV+s~F7l180d
z0U4U(PZ=5@uL^9Gyn1psA1`@J0X>z=1m}ze`y9Mwe<Z(k7XFa7r}%CR@OyjVvlB+%
zBxdw~NmG}jJn~{a+{jsTq9@qxfQx|gpci-oQ?mL^Gmg@qZO4VF#A&E3k*>1Uvr-U2
zAhbVX#r9&z`Z$rA)%{)TnNZ+lyv_sMrX%^%;=q0E(`JF-l2|gXK2S@x9+EaL1Z5|%
z^iZZ+kKGXYvSE;L6`lTgkjG0nGwWx>`{rzQ7Kp7jOPtraZU*NGh9$awHoHd=j|pjo
zFP_D=w0O%`hQFEKtnI1weNoYY)zRE(fMA~2Cy(W}RHpMfr#CF>tlL_k`ha`r$4zDA
zISa0%2Hu@RqR9?dKxXIrrm{)U;QfJ(rm9mcyDsGxBoiU>z@`gTQM>D9lOcN;m;V09
zT0J*Ipo-rCkUkmF>7UCv7vU}R5oB9?I8B6^9j>_?Wi%}HpD>(b^6$zVzN?h&v4O?@
zy2f<_TEJ**Atq4!p;YS4or};Hjypb`&n77<^KQSK?SiD~aV$OcJfI&qrKR~s!w58z
zC|FYtg>cV4#~k+dCC=zSPyr&FH~Y?yAZFPx0HLYeAm#xJJ>=nVQK&C6ORG$$kY86U
zO%{lQkf=~|+uS6e#J;3czT}F04ULoIT1~ZwgKQ85$8mZmh}2v5U?O)nQME!9t&x6l
zC3?#6I|N+BJ^nEIaB7uDiTt_*ca|aqw_h3?t9wl%-`SMdF~hK|9^cyg>4N7t4ow)?
za&YU8OsC}t5Xtv9{P%DTsj}N~Db0?uR8sg@-b2P7Qbkao_}D8k$KunB`oUL@s*Z_J
zx@q<N(!<YLxMpVe&G@J$gt+ShEgN-QOFhRuAhB#m(Y+tm8Rs77^24&}3m2*kOCid=
za8Bv;{o<0j_tmO=8zwudWnrASy$3W@xE7Yp#Np&LSpOfX)FkCt+eFNNNahdQ2*5hu
z0=!{4anIC&>YIM=V@LL^O-3A<!Bxh2_}WaP1{7$Ke!R%s#=(CVZ`x7q6#9cDJ?(03
z*Tqw?8vVO7RY9LhBJDI$w-BG^IGQnd68P%e1hC!OssaLZ1_602Ew+ef67ME^B`bux
z0pA@*VPz%AMtBj+=;W$XOr&+8@Y!4|*kSFHIM)S2iBCRW{g87+AtNkQ>yen@KrG8>
zNro}u58_utU++YMqon%iRrsql8L>#3R8qMXF<;?0FJX7B?IbDJG%(VMv6t5U+*Vm<
zW?vxaRD;K{tFkZ+O$~BiUGik-JgMs3JOxup*ojn28bmtfHO$wSkOQ|z#QOX1NP3RN
zQF?*6<OW!R@>y&t%{4~`=dAS1_Py>^_tN#r06xKrwLiMihcXA|&pynBU%!c4M(9F>
z;3_QMM>10}WH79OO&#XJzLU=3y3yT{;X$Br75w3Rh+N)oPDq!Duv%Nw#sYI+a9gK~
zm%PQ|zx6TycNvYEIjn(N=Luky)*h=#cmfo}%7KmsS6Kh!MW!~6zY-I#CLS&?b(Cp1
zxWG6q_qaShT;YUUM9vn-ER$dfMb;>dYt}@dOoSvcjgz)SFG!7miI^5Fkz!Y8wLYk;
z#O_9+rs2j04UBEFH)Ji61giBT5?!}4H;HA9)x$x^I4TNK5@@xyTOylfZ^4&{%5;+$
zh7wMtML)Jx4&%xB5w3>XlUV4fQJiNZGOOQw6-=wiSk)wF!xSlwAP!y9M_o2>K?)w2
zlp|gvauE8Paj-j0p_`m-z2<A``Hqp^KgGB`td&@~%rHPD+UCprd`_2IqwMJe(sPIL
z{M}5<uWR1Mh2EbE@Ll&9A%&1JGaxkf3fkN)P&}XTaCOyQPS0|^JU_h{=Ra(rS*yF4
zt`ngpdYQ;acyed|RRPQB=U}bV(VriK8{M)#eb+$LLgwd=O%lFL4q@Q3{rX#W7kuQw
z690#S`Oi*J6aB6SNPYM#XI#2H``v;(;I*3k*Nn2)SOC?IA3FLBgcf`c1Bx&Tco8>B
z!Q!kvi$pB!MDvtkz<l)%MWE6|(}sK%aK2Zn%e_HVIX}X~KM5z>N@-L|#DHv2rAa+X
zU0Le#;px@D6eYcILAN2|;+1qUcKTJeM~bgiG}TI#h9$a8VmbjU<|&8e5}U0u=8j2>
zSX^2_2Ko(%;UxTzXfFi2kZPhb3Dp2UG}igqMJ!P(LT%U_SM^D~{b?L5O#G~blfU74
zR-fV;*VtENq(uDTEfx$|xR>TchClC?H+Xcz&7XXIFnTR4aYB8Z!4_LOlMRUx;hNN2
zWjyYfL^YLV?R$+$t{vTM2c{yN{R)>~vVVZG3dI<aX5~I;ZUWpzAeOmDhuQogIoi3U
zt>+D5aqs8kKJ(FYpa2XAjQG^uo)d8=aKaD)9hkUOxKDWZ|5T3uS$E+3)Ldcay)!4}
z%KQDIObvk+j^{_&WB%4=K?hgir|-nFtTsObm}`xp{J`7WlZ0yBDS-oe9*hg!xYao<
zI4vl{ANb>_mGXQ2I|pFuZ9B1@yv4}(uiovbUsBvD+N@`@{g}gS*R@u_8QVhw5PGH8
zFZ>n5iI1nbIbyBy5??502b5xDkWL)ItPq}D&e)WP%C15nYf3I#V%|M?7XCGo3q;sZ
zq$K4_tB5u}27EvCr%Ys4#6!Ysq+H9@aT?LW((q%=qc~85(|A0Yw6HB_aQrRzIV$|K
zN9H3%y4_ky)9McHr}l|*tIqB<6{3w>`hq`8W&4T5>cim@ldC30rND1ntMzhoo`ydp
z36;UObED%BDHmBK1Q@|rTom-%B^MibEty3t_-QRdoiRQ$4Kk1fzG?l+rc8k1)_B?K
z27S@v)nNKeGJ8LrwK_SilLzI=vJEM**|I<9?bhDZp1!|q@~W3_>jmjkd(&0iOQnU)
zWqcV+e*fR+bp1Ex_ppP2=Dtp{v}diJ$Ok&04OLKBT&FC&>eJih`GgA9CuQ<V{NF43
zQw!4+g3Nqir%e|Q_L)#x>#)c2AR_>qsp+zw&E;Kx#Id%X4SaBvCTpF-4Yr6#G8~&m
zAp)FHYR-Ub;D89a3Km73TPa&yHk2jnZBnMQQVzFL7^l<^M0=5FIyI4#=#w03G?pcW
zHlB0HO!A+!X%aUis0b6=QnPp)n{|QV6R+e^`uQ+Od8fl{nV#P~t+yuP9ydbWLAFLS
zE{RS^m+{E;yS^U2CzgT;X)OKa6|5n1PSzP-|EL@UVRUcvYRwY~=a`pKrX1+SE2u9k
zTvKv{r^Gm_l#LN~k2oeH!I$fjlZP)namMLcaoTa%A)xi>M(2ELOI%c%zh?6=Cu%?)
z1>bycFU7WY6H-sK&vB1ypi<7wE+%Py#OP!4v^wdLexoP4_c|MW(2JY4Ge7mp&R20W
zrvC}l0&`h_o-Cr@?nEZQok*!;5$Iu2y%f!Nvs`P9ZD02<^ouN1ICx)>Cg=M2OBqQX
z2>^^H3K<IHsCa8c4}osq&6F&HeLNT^g3@k&vp+e|(5bgSj>(o-x>6eqaJ~t_FoFw8
z8GWtSk+I=Mx%xKabX?)osM66SeJmR}>Q$0Wv0dI5*I0EMddmj2T*<_(6Xr;R#0K&>
zG^bk8Sn#*Nei?hW{eeJqYKuDVWj(D3AW7l|w8b*2qmn3n&mRfv7VDyNsw<3}TeVL4
zK2vhraOwsYNJ^6=ALVTX^ct>(GsQu&(vI#PFnTT1ElJ_Pmk;|g1F8jzsBP^$a{CCc
zVk5J(b<mWc3xw?Rov%l3CYfs~Yl<ZDh)Qp+d#fqFtuC&~xX<MXZt-EBGGq%)>yfL@
z@?6Mgz9vl!n&oTb60iS40;KW0U*))3aP~it2yVi#$s6U&-_VRJ!QoX~0M{!AD)4dp
ztO1#S9gBYU9!DQ7^<<db8=&CvyYl4sTWJFGX*^Vt%>}z@jIdfqtB?}Qf~8X`wZBFR
z5-?L<2ql9^1S>`K*l(a%MRdCfsAF}hX_uQRvL@64Tl^DwF83wn6vbi?+YPP@P6A$H
zqu2-*Y;3I$;J-iG0O6?!%yV#EY7$#psq)DsPIEg(86G8M(=Bm)QaLMS=}v4QI*ILx
z^xrFK<NaxHF)uCGY;<0!8_LIh!owU$==1%leV*7jPa+}?l3y`)UH+&)sU7P@-YjWW
zz=GjB&V$61zE5o8&OWqd9<M!L$V>#uh9>eYCwKuofs2nBx35=RATX_V5GCvSXIQbH
z6VmgmPPZ^PjE!nvVkI24?pn2j?YlJ++SV%`WgSO=kn|kOdRq{+!v9q8Kydzc)%W}<
z4%%w`IldB7$gn&A*C+#jg2uvM_VUt<1+-?lJ!py{xflTMV+`vnY`BD+H@GO0nrwyV
z!{hms=aE1-chIb44Z|xV08?VD)a7l6;@Ip2+3>m8K?q*EIOawWB&3o3mK&#&wLl|*
zVSLyzr$iA#e2MaWqC5i|nML;rsg%!IL<oqc0S6^8h}x^32<-k_Y`txsN|l$J@lAbF
z&9hem%RJigF=hZLq7FJ?q^=0AsyTk_((zc^KW(mC_-p0cNS5rih%OA{W**tejOOi%
zS_pE#m(K0w#`LMJUf-*=1Vj#{#(tj7ThaY6iFk$)@czcx^};;S!}?jd#<g)keZQ9x
zT%@PDHDM?#t=qlocorx7!Dwzs&Q%-gk=BVhf+@&nuib8-i}Rh)OKkR)3;r=|v~%0z
zFYChpnOb#p0Gs=FZ*Pknw))PEhcbIjz;u}zUIGJJqk#vRPX=4MtX*v69)EB7tDKR+
zbe2d>4a**9NcwZO#Em3{@o!KK<(!<3Q`gGCa03AB@bPM{vtGkrB8w#Dd^y9#tO!<m
z)uXvYP5=T#qIDkauYlPXW!<I1PRTi`*GKr{D>8yuE0>#QzB)p{*Tbv0uyu7BNz_M;
za%37Pg3NL3Jk*JmsxNs^e(y&$pCZu(f-{Utta*P}=5q_JmfOlrh7+DJ(hDK>#Rz{*
zQ@E6Ov-~_KZs<*d!Mwf=6wP45I9IS)zoN5%^j4KOpW*S7Vrg^SYR*>fBieEz=PKrQ
zlK^SqI3t%%CnX!iy!jdyPXX7I26PTE8N1e#d(#avL($99xpW3L<|eR=BMOnT;g(D$
zvBq0XZdCz|={57WD;_(LvQ&Q}b8GJ<V>XG#P6ZEWFywx{u-W<#6w3<m?gQ!a70{7w
zH+2Gwvx*0rIdo*d(F?kH>RlGzmtOYoZ2BMwV}b(n8$4CM;@(7K-@yfCJQi~P){u0-
zc({(_?M;y$l15+8CIu3^LIA^<yk3;5#^ZTvevv`N3o_clNA~md5*HH@Fi?gAq(@Fg
z9nC7oW`Ds(Y7LkHBO@tI;bPt<f<b3gArw5$0vK?oQ<sgkEQwGv%@XPPvcY8oha8H1
zT95nQw*}>Hgg{3;JOstrP#}DPrYKw*5->3v$%1=QUZB&^xyDtySB|@~3tP2gXIJWJ
z_9oM<cg?<wwe}L)_9z%%V|=YZnWLzrJe5e8^Q7^%aEJxlQK4#XY4(F0vGfcAVotV5
zPWO-B8AuBjWbNh}Wt#1it^z{q<bD;rwP(<?kR9Lh^+7d_tp)k``gRuvbzzyM{i*@?
z4A)^IzoVNmz8r5~ikHfA)%QJYl`G%f$~pA~;>+gV6fd8ujhjJ2=|arEej2n*IgVTL
z_o|ihyPUAXfAkZ|YIW@1;pb2PLA%PP!dGyRes4Lxu47#ry8ncqrzX{FPFr16?s0uO
zbiP`!wS?){e$iJu{_ci?r#49|k+nAhy$W1JYLEG%gM+q-CGe78a4h_PUwRaXb*gUg
z=#Z&p&C?MZ;L~l13|}jUp(&mcFkVydp%}6t=<Ae>yOO1*w3Nn6R;Xp2h+(}BVY_!X
zgH~RHEJgc1k+ZU3JS0w0g=p28gPqTrG&~i#N2G~o%^arDD{(RvF{30yVKNleHN=2@
zk%ypj4<3%e?A4s_C6KFb^<^R-FDv6+yAky4HBvBehTQ`alM?Yvk@lQ!x(%fW#Z9qs
zx6UN;F*ktEt2psxZP7yabf}e$A(@~0aD1_-8(mhF(<ME4j8(p!zN2)r$25M}G`-c*
zw{-i)p0dv-zANd?TDFc*rH2cg4s{<W&{{LYHqjthJg4P5WTm|}sft&9nSlHZqMSSV
z;I(X)6D&SLZ(DiJi;Okii=D>%ocv?5T$IE?Q4258>HhpbtMd2a{I8d=3sozjzL^Wb
z`|~(~(X*|oW2WDW2)r*WPnoq#aQ2(`59u|UjhnYkiB*2}^(U<qz5L)9q(Q-ZzAZR<
z`?=Y330|I*dv|xgv7lW$<Oulg`Y%RQE09epmPTQ(Lp%oUZfKC9*_$U5eAyxtGQezj
zOt)6SN6sR;55k<3n9<IRvU6z>&!UDRoyesr<nOE1h`w<kor=~lNwnpGZc?cS3=0zQ
z#Q^{hip(HdKH?j3mSLtMM5K}LJBw6^iFq-$&<_e0OK9>8hQ5H`*dpXlEpDq_AiUQ;
zcEq**)+sRuqWP7ZoSorwqO<3e8vLR-z-1J;%1fzRzL!v60ETlJ^uz87bRO5+hw2Zy
zZ!9W%-!5^`DGtCjiweRv78AC5_}qvOt?IUVgBjtjDun?_pE76v;{*Ep;br8-gIEKX
z`hEHMqYbNI(cbyb11+>d5xc>&FWbx3Ed>4j&J|`n`HNcL+1dtFedBrGBL-hq0zM6r
z6gt*B4}Zy#II<p1+-kV?h?s!LCDqq|B>P-nc|w8ZD#dU;LE_p5vG6T#;zRVFO`<&W
z^5F2JL@IokZiTbkipkQq4fW+o?&x?8`n<c5`N30$GCtrT!<9(SgV!h){EdfHv{zWA
znARny?L^JkvKA4bdwF6&_XMG$S7Bsbfkh}mvnCzLgU;DL@@=eOL3>u)GqJVObdS(M
zeXlx|p=T7ZxhjXdG#2WM<9R}kiAqr0&Woo4HZ1GfF*)!3;RtUp=H~2l*%oA)8ZbX3
zGwG;`5Ak`!K1-O}VwB_^x&>7X(}DWUP$m@(?!F1xrOp-Idd|Q^0LA#OF1g+*{C#%+
z6$XUBc(R^JX&dVT!wQ>nJOAI<@YfH>5+i=ykzNMPGqM4}i>*i2UNabEY72bgS%g9%
zRjEo~Ec6`h#LicLEjwapA_zZTg)NO@g({YBUD*$uN+PR4i57_=XAi%Axb9jY=|i0^
z8LXU~2Au<&agrXf(X=!-b9Sxhb*msEG$K8JR++l?6Ga7A4QCo7GFg21@7$q;HV>2h
z$S{tPJQpfw?_Q_ORZoEcWbzzL6*Lz{a!4#i>BAel`Z*K2vFb9~{+=+lRsahjr%6nr
z7vZjoV)E`Y6g}xnhp349O)*p_H&h5nyZnrR&c~krs%_YorP<;^F2*cX{V?Mhdm?op
zol^S@#r&8Q`I5Gq!R3>ns8lRU>+U>`N#E_b>UjROL>&Ir;=h+nGUAUPBVE5vX@cH6
zz-zutIf#BStM;EEmM(8cuj<@uUap{>(;VCzh`oRw6p%NGhkE$VKUE1SfSq0#U?Mj-
zrft9f?<f+HpkU??hc<C^l+Su({xb`JUOxPhtY73Uc<}Wr6~^8?*=CNXR>YAPmO^g^
zRog=KQ%=J#!%lr&fq+A|18mxw4~u7;?b6nBAo_e&d<{~JFWc`Uic?%6M9#nm>C`8T
z!WLfKv%3+GOLUg8e<@j|26GvrSl<{h66^pArF>tW;v)B6A=-!&^Cp%XijqX`4bW{H
z$3T24KkG6^vS3RVh{DNQbbb2@50s$+l}K85m&h!aCQFj#bxkAK#v5++rBr-&*r~N;
zUmmB2>)lxwRHds{E_+^2RF{Uq(PSB+*)3~!XpSQz)A6#Hw>i}+*zG4grwcP-OR8Tx
zwJjT+<U<*;r`+nra+TIrgC60FcG>t<`?C=J(M;;dc7$46Y$kC4GE;%t1$EZ}wAuzz
zAtiS$$-e>gbE)N@5V1f^DogOH*7IH6)tfOf&*h0Pi5M45vw*Mr$$Huh8dFu~_j$n9
zZ82{TvxlFOT<hWLBK-%E$L7Dc`0s-gq`I(wtHy7!!f5Bo?!!A@PfsOJ+KpZ^e+2J*
zp%_wPt!w<KHX<SCdiNWjO#PMpbO@pAW4*>SM{%CQs>8>(X#dOu<TQ8zi(TfXfjf@n
zXTZ)U|AE&Idf>D}|3LzeD7@zd0UrgWiY}Osot#^kz?-;~h~gz36+NC({OZJ(=W_8U
zi4_W6{UvH%j13+|g)-avf_yT(lbA+I@kZ$SVvN|zO%()azHwTjlqV@QMne!i$YWJh
zm+tYWlgK-1IfVmu@VZ<tH~irzkoSU$PL2M1p%~`nX1)e_Vy<;|pcC~w^^vzl4<%_}
z9c>=(LSL|niAlJKztFv}Ho<{s;MF$`%1c{4S*t;3kK|EB0`j=d$&RC8{31g$Z$$hc
zq01o7<!$l>xm}&aN;EmFwEK2Go!(K)tn=3MtYhIZI-EJfMHfG_cPu77`Rd=#eHdFp
z&x2q0(KWwN^rdB;!*7xvoXi#8pUWMqY5ZkjO7UF;GupTIPb&R8qyCdv`(t~cmhK$N
zUS7Wu3+pU+;zANtc9!+NRSgOUo&r)R0S;Ws%AxS^wSxKUJ-|L1W-1uDpTM#AG9!bN
zS6Gmxc^{u1xQ|FiHd){@OF1)~yrGh7T`_0D%Ws_(z#i&8^~%%?>{Nj%kgSM~NV)cP
za6}H!N&d5<rbb{J+K@e4C3Y}yn>Xdn-ayGYZ{lJmBi*cNgRRc;5u0q0GJ`BLg9c+T
z`%0wXPfZpE)C;+;tbRF~fr;Cb!@iTIQ6!@NRTHxzQ!NCvwGgRBD&0O(H{YbdcfRiS
z%1Pv^CvWJ}HDcrr%?*|ez^lu_Gv&-fpk?eM8JB@cbGKZA8}bwVg03p1Z^c^blP*M<
z6P|4`MaW^<(cWY#Vs<R9iCZt8y3<XI>RV!;|G7ec!I7f4o^<{%n6>AgU+Pu@wvKX%
zw%mLle1PrbFS}*P-?Q!B^S1vBx%~I#{?}Jlp%e?u`K7;P<&yxE$*4=P1Y3We0OC1#
zPtV`47LOX_j=)|EP*kBTSAAwxfaj&iNqAf(N~l=X+<c$y&c;B7WU3aOL8{&mRqPPW
zNCGk7=JT=7<c(qRGT%roMA}Lt_u$OMIet3E#UTB-Je4t4S`{&DSrU_bEK4k^(tPro
z-Ufsv-DU>3v{c5e=V4b>HkI{t3nVXPjou2?&o`&$wZ0ZrCi1REW-GiFK%^wir{l1Q
zpyNr>H(1)@mmPJsoYgv<8PQZZXA@hVuN#>*+&TEvc90=pOYEwmcg9-#)2P8VbVT16
z#LjeD@k(&@7Sz^DXLT4qSAxZ-3{LKT{*<YwP!6GhJJ)4Jmnd6X25_Zc_Wv6qMpNT4
zExIE^FG7O4C5_<xjhZi2Tbru-#Xh&93&uu*xajYO1iXWtAU)2`gWlwEay{(u`wMjc
z>trOe6Z8!}zXwF1r)?AqO-EegF`(oA&d9w17(f5JG5RLZa~kGx;8#8ahlQ_$f*jk{
z)|QAF9{|xm4_@9OlUWp@z!FZqLKv?>?bi5xg=Ew@dmW*2V~8xpz$RQvMvtNQ9=D<p
zcOQXwN>glEuw8%N{LDy#2AfS3awuOCX^+1VCl|6O$W9?P0zoofH;aVKO;kLnU8zK4
z72#PQ=~>nv*{KWsShp!*$GZiadjoU>j7(?dulm$Xq$1pte%g=`R=-rr-BKJqE*d3K
zo$1rS_2J#6_<VcU!qw8(lffbg8CMoe2fu!Xjybt4g?5rdrOMi_vo@O~03_D%UIS=b
zazk4kwx}$RjBWL~r5y5kP_0<<xQd>Nyq(#cYDm8tdi`75gC7I`W!<^)#DC0M<+ips
zaDgD3%bv`hJmXrh_b-M~C&1OfVQL??Cbv=OoQp2!jhtwOG+`c1`=WKG|M9dhtSc;F
zsMO{%e~6Ff1a-pYjpqdBCLFrkyFC=12kV$h-8nv8f1xF}NW5|z@5?*W!w-VGh6DNG
z=LTc%#&#n^^DO@4PZA0Q2Z1}1OXb}`FrxMEqfB8JYOi;VTz-9tVvxXd{vDcBnqP2U
z^SU;fP&g!=!se7PLa=yE`bfpaN>_C<Yfun}eJwC`y&XMOp_U0r<+B@<2+$wAUQCG_
zv4-<PkD6=I^6obdiT$uCOzs+t?4eHi>%!##V;Nk6sI&0|J2Oj1N)odrii?SrFE6ve
zR>7!hx@J}O-h0f0(dH&Gd+gDBy<)SfvnjrfM9rYe_t92^wycs>#OG~d?X8`4Ab%3+
z`Z884{@|*z{=B8PgjfFNvoB83@LDQSsrK6V2+-<!x+8`P3Q&n`Qu&OCXOCr$mSi##
z&WCg%xCH)u3HgmrAlDQDI*z0FH0Y{P#)=WI2VhM0Y~YMK)Qd{>zoIGS8)1Wn_w9d5
z+n@8E;bOQ?-iP&a9DP4@{CR>C(b{I{u6pNagc)m$;5?4XJ;YoE&9?1@upL|SYxhe3
z8{huTcAu(%=_tMzhT`HsMlgRUdHbd7>9pTjf5|aEuE5a3?Z?FswREWBrU>%dEm+^b
zAD<JFu)E=p+dO|CBxi!9GsOO-%v3<Yc~Km2lro5Kq%fl0*kfYbitLqaUbMY7WuOWg
zl1gD_m5(LnoFX~FDyg>+3*7FnBe<eoRcQsEMnQarQzdZ-4kDFE(jg&gpJiW!Yto=I
zRk~=El{P6#Q6|Y?IYx;=dWjtyr&nrI+tMORZ**_{{{8!$%h(=y(L-!W9U6-uy3XW6
z(D}W%<>qsrz4WcV+r3epuT41_w`uFIffhe~xB9`_mjF3-ju4;Cr-ZVcP97J-AKvsl
z;!2!`!w@)xI-0g2zt*=pm-poS;oZ@}0N#Rg2rX^W;N^;X$df^-_@caQWl-o0^+u&y
zOaRKr(9l6(as5YmdAZ$HNxy_^nLto&Zo+?oRSyXI3Vq&wD}V*g#N2zkshk{dWb`D}
zvoQ4eOW5x7D8E_TIrwVC{o|w`)aVvZdWKy8FaKWUKMw~Gw@AZiQLS@NpkLg^J2Nf@
zNCZTJkDPWpD(Aa=M+x5Y?oXFQCj5&a_JpzbTu5p@_fM?&>mTwy64;Z+IwP-ZXca$$
zX;u6ljxutR#F9}1$EsdEMpA>KxeQ-3Dx5sZ%ZIgOh-f>DF|708;V9j8imL0~BcE7c
zB<<|uU>=$wK@Az*+~SsnOuYTLoE8eBRy38-@{zr=TX2<LjWIUn1s5LqGGtC$YzjU*
z$_UBshd;CK_%%=e$$m1hg(ou8CCMA+bkj&^GRvouzJzsZAQPDux2G(&E~E#wU9TMV
zjsLR#R<A~8zfLwr){k<7|5flqj?GVmu(?X6Uk#HI?kZgwAlT_k_YN8)wO)X0L5tPf
zfJL4IPM)BamKLrv8u>oG07Tt2&1aXkLu?c1N%t=~B0*S<yH<e#ShMsmf}!eohVYHw
zsI|wE0^Q!kzXf3!d?46l03Vex^kgOKsCmCiR^o1FU2uWx#i%=fy*MJz6$*^(+$uf@
z8-acOUt1H+i)WE{`#2%+?4v+#O<CvFczXFWmzkIf&WdsJN&CS0e-2!aeu6?UKi5}H
zS(n2jL^KGmT|B?KZ7l@5XVO`BWW|wUO%8b4Ab7$NL)<zI`&ki~*AHUJRU-*k+0+%B
zUSK%Rvg#wqm+zn-(G3bBIB+9W2{~RjE{rqLXnj_`s^~n`+yLb8_1*$JNQBIOnWzx+
z&HXIprL1O(1G{{7^#-+3YjqZu*k{&xt*8=7Or%+d<tc}_TXsKlYF+8{w0dEHi_HX$
ziK^e!_<*VF8t#pNTkkcB6LS_J>YYA(>D9Kbk?H7;64&7ta4Y(?UaAJDdmewvIysoy
zOw3Cl$j!-M1v;C@xEG<JE?4?C#543ZOQjmi4@B`R`Og<~1K#LE=`f+O-B@DX1NyC}
z?X=gKkY&nALeYg1YQl^101AN5G|$a4|L{}szU(WG11`l5xjjeGNzD9v&7-;Sw%(O}
zr*+ylczd$jf`!%95byr@u>B5hSIBvs5Y#${g<SXBaOQfp{a+}}-=ZF=7s0DYgOcxO
zo@SW2m(oP$n!pDA(5auapM!=lPxSCGhABt&czQG>OMC46IEzgcTk7`M2pv=Z6G}^b
zq?nM^n8LuW&Uy8|O?v}0(K=dTE%fSM$3tH;XVp2mLz3UfnNx#RDz^ELDrFT8Brm&5
zY%m@qZrTc&#E>}y4g~(nDxBFac53QTl*SM{d;UY){+ojPtg2QCOtvY8^@9}$iiH8G
zFP#q2_XdzWw(O(4LO3i!KD=P)@y8f1qzdXaybC!LCUzZ)F1jf;q7M{K7t4YH<;uJl
zKA6NO=Pr~<z8}{gxB`bPvBpP{vs;`ma%8G*P0w#kmp^^b^BMr|b38K`{d_A#2WE0P
zFuZDhWUKwrujFgBR)W2>Fbkv6DhJE``_b+)_n<TD8ije_pKL+MVeR&H*X_p7=DW8;
z<CCBpu89EyMFQ3=><BfZ)*0gNnA<$Gi{)Y`G7g|T|06q~hr=dDKKS9p&%ic-s>LgW
z24NLo|0q#C-zl7&hA)x7)~9C9Vmi$n?#ublIeodM%ieMIo+-O`0{5M%#ZcBEK3_O!
zm*<f^ec<5#I&aB$2qZ)|6U1W-N<6;h`Qo|faROHduH_7$tU10K=>P6VkW7d7jbYa0
zZ2H$P2Kny#8Qj<4Eu(3ozc%v*1@W(Q&Ja&k4S$#l#kw!cW$fxwFZ>Qcdwh_;dfB2g
z)uijJmF7tRoTht1QRl_NL~cqyUEmDyVafy60)>=t1gBCOF%y`HRcXDHj+la-29jKd
zX2D^5#IN_#OcAWyN*b1LkBz;RQUwmCROh^T?vm6zIrvIUl8t?EA{t!wRpo0>0XANs
z5Iy`#GF}J~ZQN+hXrUr1k#e(ry@3$9o{n$FCp0nt6J&v$P|OSuopZiR@~U6nSbkOs
z>ja&wAPd3hxePfG)4d_&skcjX#fxs{>!V`AVW(Rt#t`aDa<=WO^PY*C;*j-I3U~>_
z)=<T{K{}Dt$iQ;*PwkLW?a{2Gh9{PZlPKe2>W~Q|((@Y}KPx0p4GVK1M|>CyH@fB9
zh}r8?oZYS?MMReeB{Ch%?ZlCvc9#C)`a(JsSrTHo;3sB<fma{8p9UY)*t)ALR|r_P
zNV~W51dFWjzxzDjZ7w@pp1gbL`M`9X;hzxrFKO-R5qrIl#pX<<qakUq%Z&6rMwTIc
zs*9P*l3&7s4u8L0GV;X^M`P*E(;s7h!qzZzrPn{CNdgxmW>_f+><Dp`$jK5RqJ0%r
zPUazU%8ltYJ9=I9@IN>$&J`qjX|J2P*;UYOCIlDB;n;7-q1l6cUcPnl*(1@%mmZ=A
zmrmR&QVyXr&|Rq$wN)owb$<I=Sq^JWnyQCNl%oT~y>M7gS{Jt#=A*bIA&})d*eJXc
zH?y81-L>JWi+gX{q+37L@~H5(QiiU?@(jDr(iW{Pp2xzV%&ek2t}jQ-PAvHCx>S{5
zp}8Od)W!8)?ef+g+2DabiBnCVqz#b@#`S4#=)m^2b_ku3j1Ps2+p-w!hTjN34VD#~
znq=eL-Ne(tL4F;>OB}*#BqlD-6SAMA7d(m74WYunf~<hsoIditm9F~eVcGc2w13p?
zFX4nSY%!vz*%DFaYZlPmISyUbz<&b%u6xlheP`R?s%20N)ml3EeCNCE?2qQpmp|)x
z=GzWtf@+US#s5Yc|9)AhRDEGk(&gjY2|HdL1&F|VIG9JqKDYsP)}Owe<ak=854i^%
zx4(aaoC|M1U$a~OId7`pkF$7%4v;S>5bACr_>N+zt_EoFeY$Soj13K=t#cOn)xHoy
z@+gHe<vT;7oHz9l8JLz;{Go(sJYHOkB(jtr>Sh@WlF_>P49rR8c3&`PNQ<}~+mkt@
zEOpAmZI(iNnUN5yi&jGRfc!?%@o{%j8d+T_3`)6b(&cctvVK|>5UOI@Beqw@)tcGt
zm@URhU`2bbGw8XnAgKzgbtkr_vQpy!0m1=1*)xAkwd<%w&W+-DI92RAOLdS$?9*by
zMcgwf{U=YJJc=L-(qsoZoGEHq&|$Jsu0zbl4&<2qPmH416YW`cMb05(z7HABjP2AP
zOv<<~9n)-xsOOqfY@!XjY~)v>3vT!ws51niqEecc3~c-JPRfG!g2L-V=f<sc7zb>~
zEif5Wwlb?JKZs65N?R;G_>~<_Nlh4#H(Nw=<DD)VZ7~hr`Z7nwGv^EcY_h#GLlV00
zA_p^UdC6FudB@|8VD-VE;DCS7pNKK6KT09Mt1GF7x0wm>Gr7&9T)Kmbhrd5{SWW$M
zmUFX9?$;Q}$yG?(+~#w8=3X<ZV9Q_AHGch2y?Xtqx!tCd&+oP*|BL?pCqL*}RJNLn
z_WOWy6+Kxw2-8byKL{r|*`3v?hy09L$1mxBXAWQaM+7A0BWT>2B&!;cH34WMod+?s
z`C%r?;sY|Lzb^o>Us)lJxfKrVe8*ng%C)L6F+jI_wRdAN3=*rJ638ksF=%KPZ(Jz0
zLK2TueH}|e546EJrxCfYsw}*SRbsSJdKPANtt#GsMN_wVUywcefK7aLT_grNpCUam
z$jo#QL0l4m(ke7(eX$UMSEZ~}h&CnKnt$)@xReQ5<vmDZ+Mh83{8h#*?k+V6zKxfY
zCF(pOs@!a1eYc0;==D?)cT_CgoOYYSn&f*m%hw*#v<&7tBARwXEBt7*{<LGV5(DGb
z)>gV4opw)_Y%FwJ9OFyYlaFoiwBgk|udHjhbfX5zQ?~pxbNwvGaZhf)f^OY~#VOHL
zY=&60y-z`cnD*puP@$+uMJG}wDDAlEIp2A$lFsz*14g~sX#eK@Pc4r;j<ToaQ1hM4
z^TZ?7y!8xYOrBSJPxC)xv?#1wiq*y5p?w__4QGF=*T2hUj7nBhZTimP)Sqwz-+WuG
zRyA@>@&k82$ZO}z)b?ENe02G+!;@RnB6}!s*djW5vcWPlU-8|vMSjP8_J5^Tat^#;
z#OY77cVCP=sv4qKFXfvGWI{4vTb|rsa>x1P6SD(&|4sijIKVqBT0WlousWtNjxdGi
z&sV({Ms{Q<v;9ZR6#yMwi<2-_$hdTfYJ@E(g5a7aferq#LMuAK>e}M9Flx4E*Hv!1
zu~{tSN(PH0Tj_<35O^tZQtMhPf!H<7C%AZSv=d)CQyBz2vaL8D(r>7|RhT2G`@uR-
zhD|3GBsY{DaKjyWgq1lmf!l2Q@IO;x(Pv%>_e4cn#Rv&15>>P5ZKT@e@Q7XCkS+H*
z9d0q*!e>G0UQ<7LLF{XO))^<XBBkF$*T}Gp!5^5sd|Y_@%G}_YLbLcw(5!*?eioqb
zEGLNdUFwtKymBwUM`nFwD4Z{upPpaCex?&$KChij{&m;=q)X!w#>l6Ne)+kM%6$|Z
zdP6G1KWQbI+;(NSOD)E<yQ0t9bL;4|D57U2?L3n6c=*h<rWT?s73_cE@j*}~epl2$
zn5Tm*|4Bgk-&rq$ZLd}JmfuBAbo0mBa@`LnK@8huYl%(oy34=Dt2=z(Dt}4nJW>dm
zk~?{|xl*p>pPTxM%*+%ca`hha?N{_)#It8Y)k^&fv)Sy73*f|cf993SWTbBsC41!x
z`t~GXl_*g`8>_o68#>aKEjR7fTa|rj0MCh9KAt|97vEnree<Dn4AVd~WgzyvYcB77
z_GQGctnD4iVFx>G_tDZPVQ+@nY}qTH^3Vfk2LnetTxk2J>drN7g<G~GV;_V>ejVIs
zx}`0A(KH59cliZwKQ-TjKet<b!;_sK;#YpvC*rU*>wu%}0y1PH=V#9wyEQv!1`Zyp
zj7_IJ_Bgn_o1sze2K3~WPS6*cCLZSDBRJGbYzwW!(v4MDyOnEk)c!cPa7EcNEy^+H
zK|!&l@p%3s{&>OF@t|Q&{$b7tqr{7tmCtn$1}dKv*~QY0s|$-a-xz+(R9=l-BY5TQ
zGSmGTP4$UTPIS|qHW5^z_T-DPyyzXG+BU)7zs`hJB|G>>?w22DY0AE%1@C^V<17R?
z`$UCDFEn~@zP7|xs>)Hwq_C+tyjXaVuR+tSn)n*Y!g*tLKcU9bsw|-ZX=6jQOx#Kx
zO9$p<)kvjM4G_AQMj{6Egu2D8zAQw=dH(F7lN@wq2X4CsJ4!%^QY|~kk(2!_oJ^5o
z1dyXWJ&%v@JkCCi%lph<6mVhH;Y@N9FJG&tR_X<F7NqOe&R7-6_AoC;@&MrDwoo!L
zkxR?6*IvvyAzQNRNXWD(ia8A3-$JdKw104LV8@b%Pe%oBGv~0Vzy@lAELMH<AM*@H
zv$VH<d>Jr56YG|)dxXz)@M@^Qd@k4Pff+27&V(T-F~sdAcb{rEJ_5XxhaT9_&)ZYm
zp$;}k=a*cBi-}p0=u#sm6TGIRTFC2S<nX8x4lxE5@TSlkoK?wBN4c)v`^V&)Shx>x
z2-iV#4*xa<Sgo_~y|GPjTl{(>$Zufhh<*OV=ZpKV*|}}djLC5uyR?@>6>RC?sj?S~
z{bab!mq}xH`L?%W9ic;Xi?e^LmsU>f6SpsoI_$S;&W>|N4QIdRw8R$gT`f05Zizkj
z$%gH<RT-K*&Jx|tw+>hMX6Sl(_N$E6>?r!VnW9R{=J{l|*%#$*z41(Qx>(u8!zL9=
z(^vLp=a=109q<tDnlG2f<7R`O#(6Hn5B3eIDN5T5{n4hiR~n}8u((}gOO+It50|D}
zZs=Dd@GS4fq(tIJ-$L8^O)s~io85;)36*Br4j*0Y4K!CD=bZ)e#}Yrv7r&aB-#Iry
zp42Mn<cNuUuA9rVm)@+|Z8|@_%0X>oFK25msuXaWscEy&F(TlSp0ojzRxXxKxdhhF
zC!fbn)^I=LJTBWd#O`ow$ojxVf9zE;*Ew*BpR<BXg;(Cb8V^CYf4QB0+IQ+`Y7P7s
z+YiNDJNI+xzc5_QVltS+A7h0J>Uq0%zPx%qkl0)x9xaF00loU@ethy1>c+e`Q>h6^
z7)8T=r*8oaUKBb2yB<Rp*05Z?V)o7$tMbj%Y`)ri1l><Fj}HI@1rLs%3+cEW(3{;B
zj;EYp;^za?))!q~w{LiMDE(Gy{fkNzoEEO9uq#Cm(HyY}DL8?L@nu`sBw`5^EmI(@
zCb5y%-m8}q4^gw*Ezk<b+X$0u;zQBw?4}A|ffk<y1D@-OiM?tn+y@BSQwR?`(Mk17
zImx$mbTff7v=9k;q9xMhqx~2rIXgY`XAZD~K}NClMbPB!;RlkbMDCbjCh1hQdbaYX
z3?dFXX*ZV<Dp^Jka{Y-Lq7&9{N01onvyO*dvt3X!L$9|_Iu;bW4ltfYyRvnUG!Ow-
zYZoMzz@KijUe|<gsofnwRcWmw?Br}xKqu&rw5<pEq|q#iZ0r1<^NFs>0jT0*P3WcZ
z&7c&~iPc|8R7I>;6mZB1i!B07c;00_*de4?*itWc13q0+PqTD*790Aruf-~9MJd<o
zC-bhO&HD53R*Xr4_6}=({hNu)1BBLx{iy8?Rd>&M+iug(HoJLW4wKMFiQ<f5*TBXf
zKDO^(DhCShVpJV?B0c4`Juc&X#mnUK#`!_;r@>!7h}yOSHL4IyuaErhlDv4XoXEFb
za&e=+n-8bPb7ekbd_8r?P3pJEPtrO*{qQg}dAWJe{`?RwY9HX8EN<qb-|l(!i)hR8
z-JNgaLE}VyUV&B9_iFr_e6MzlvfoEl<mLw&wtGI01a)pc61S|^bQAsHQs(=E)<ft>
z<NJHI)vsW)XVdWN4=*c(V!C-2@$UzHlD<#XMRt|pD}QX<UN|VeU7@3W-2Wqdhb2^8
zWRNUd8_zH;X18hQr{EjMfKUpd+VueQ)mkPL?8}4f_W?XP`vP9WSLb8x;<Zo44{F<^
z<)2kumXo2r9lGN>#EF8B1<smV#iMWI?SAlSh6(4ff1WSEbw0X0&vCDEZGx1Yg_fL+
zI%b+GhP1%0#@x~}f`jtUh<z>&9o@z`ewEXj`kZ}pu)kFm;-ByFLCl*9w(@efrQ@x9
z9%}o1J*dXkep3bBv1jSGQn#x&KKCuOxi$2{tsQ#Q&@G;z`yG9lJZowM+mzMmZMJno
zfIJ>-%?pOOxyuiCc+kM)z7oqV=gkBg`A7cR60ZAyWPNp18|}Alu@W2#1zH>ml;W<z
zwNSLh-7Pp2hv2R)P@s6Bw79#b1b3&n6n8=hcGK@WzjMyL_s>aYWwO@HJA1#+e&m3G
zG{*`<&|eg^bU(sraC^cVgc<X2kp89f-MVM>(f%;rZjbV+?aiFE+aaQ)vmU4>o(=nO
z**ZSEnl2h-2U;EY&Buwj<;-l~NvU(X%^JlCV$e?wl=3d8J{uWc<-YB1EI&#YEcT4;
zIAh>deJRy*;J>uQ6NufrnsYa1t@vcw<W9jMi1wnfbeC-1x#7XM!q6H-6QDMz2G@hn
zcGley_|q?+z#+F~?t2&7-hIt!k2jSi&~|PQ!d-0W>T0`HBv7Ywse!LiPuW|IJKwg5
zib6HScd8CcTRY?4b^6EhJ6I`m`eK8N{rjLv(nFJq)i0l*4hKjfy7f}QKDD`A&{Od=
zKS0>rRl@0mL>j_n<9e2i_e734d`v*oL_@`n-2hSG`TQF|lhI*T*VD3%F{yl?VF{_I
zC^lLkGOjd9*m)!1Tyu&q!cHL*^pQ`3clTfuF}wEXn{5RTWY6bg_IxFl4`2E>C(}QT
z_49FcbEh>g&nu%HnZvF=%U`gX_zu7h?N#Js1j_0MCVs=diq=#Ea*rUoY~9Xdf8KsO
zm|F_#ZzygMBn7=-6Io)fKAND>(y}N$)IvKDbgYYb{pXAS>{Q_r$wPx<%+8Pf{c^FG
zR_iN!2Wq06w^f1lpCePdH<}ekvScX1pR*PZ_bWlw&@_A`y74@=Z|PcmqeLF>O!;V!
z1`A#2?3AC(;G=hL9rGRDzw`JtQ@Xrs_vp@?v3`Oy=j+XX0P19U$;aE0i(o<&Ixu}m
z7=ZGUh)CEHeE`f2Ag+_MO#*IVM==7nu=*JY>4lat%`pU01_uX=5+$~u&Fw~iM;VZ*
zrnNVfnIPw`{?*yfG)=EFQJgpSlFlP3OObuvj6f+gBT0Zf^f$NH;XH%@w#yYsfF%H#
z6WfG<xkB<dQvexxm`WIK*dZL;19AuX?~+aWtsoF>*=LJ{x`b}O()6D-@~7eOA8aj#
zvXX<*;34b|3D<<aaacH9=vw}}zM*vR!Xe{tv!0P+Vd5F?gCs2w8RUqaR={`OKF;Ty
zK@FA_NOSs5{Kr{1)y-?QBecG{x%UqD^K`p}S_cp^?(NVvIM6i56}v?0OjK?{96|>2
zs7I27POt=0nblNPcMC?yxzFOqi%lcdswNlpn48s+`@nN@p9@iG+|Lj_I8mj6g)spL
zeT67WJuw_gSN=QAciDpfWBCe5XPBkBASAzCVC><PcD&uZwBG-s#iMGYhk}ihUa|iE
z)L>SxRp1L}1`jlbKIHvJ?=4wJ10Hv;9VU+|+nXJUNQu7Ex!UBC9A^gP*E+Pq^(zRP
zy(Y$_z&AfN?_FwzU<Q{hA9ers8@R7Z8g7HKs@r*fQ_kWYUOCh{##TOj+Ll8vg*`ri
zJ<GgBsz+GM>&*K<q>$x~)UdPQG@2oQ^dKm3rp3!qsZL>7WI|pZz4Tb&xH0V{JZl@M
ziJGj0UOFm?o4`fwGs`GIPp{~CKjvJ1B`A?x`uX(8ji?CKhqi&r^X5Yu4G}UU;`U3u
z1U2>Q1OsExRX(ni{ne1&Qi?mNu+&>xymen|&;0v?+meOOm4*={LUgG&kTEI#n07g_
z=e&kzlcuk+4hefTljT2?+mX194CTIBDywB`(W74RD`x!dk?^PG<7|oZe8E(`r@OCs
zWzpQ86+7NuGsL9#xX$3^SAGgKDSLOGG_sW?8zpZsElR<*<WZH67Fh#iZXei|MH_5J
z3>0=qwoZ$M9V%QD0vzp)mh5YH%30A<50C3}cBUVTGLL_^aRe+XbUfr-sd!=h-pKBi
zkl?|I#d>sy9F$57@H`S>#$mRKAC>Dhgp1ZVXL?0KCD#2Q4Hmp-ZTyymaIF!q+lQF?
z2CI)d0Uario#Rb*+rGCyeRz)szYwRI0p260A2-2UuNQvoc$5bD`$xCC&c46BjD}j?
zJ1IjCC}m9_1{!Q2{bk{x%2l&5i3=6qGVf~+<)a_d;~RfS4Q$|dC|?LWiU^5|>>C{G
zx069u$_|<R$OR6JzE?G9q>H7k7q}FZZO{to;BmkXG+*+eQ&maQzdC=jw0LoDOeoG7
zjio6n^H6<MY8JpPWxtTZI9pfEq^IWN?Qg7-w5Z<*9BQZ|(_=fM5J!_{dTOqT{IKa;
zpK;8r)^l+JV!cw3nnK#Dg@DsAy<)ju<EX%Mn~Lts(`$vtQ9%KIn}t!N^8`$pnpL}^
z4$-Hb1?!m-xCwH$?q`f71kl_p%sEvya^Cyb9yukf6v{bd>+%nF5UIwuY#1Ij*-mEt
zvpK~@K}UM<Qa*rUADk&rh5ceM$-vRX%!~aixj2-+2^An&$G~T8u@d_n30*|;lGlw&
zGS?%CE^ny=0Sx1RRKtufW#vtKt0#72$Mddu@S{;EumkGpfRYJastv<3VZpEH=%1kz
z^;zO9gb7sBYOm!AbnqqMS!5!;%FJ+$!ZO9GWVRwRvy+;bs6(g%U(d4!?T}l<PttLS
zOb?f)QTZZg$9Gj;i0N~gcaAZK<5HHf35)uyO;UUsInh#M1ef&!ix4vWYGaOiy<AnI
zsg;REJsW?jnLh|f(JL-e;TQ*$hQkkcn!6YM;vP}mo?%=V;3#2pQw6$)3TBXzmU#i$
zA9?Z151(k}^!*g)QiY-_><ouWD){R!7dQbTaC#emVze51bS5_vZd{}D^q?E?G=Pcl
z5}610dD;{}E!x{i;b~9u5Us?vL#spwPd0Ir>PP9EUo-G2{8|4h=7FJvaBbfN*xB=N
z+%X)Vx<yht!wyTZ&{{noDaE)o$EQYJukX<=ubQ>V89r}zx_`O^gKzFtH?M~Vv`^T9
zK#jj9e#7x7YMijk{BNw5KBQHDrNMuleu!In;XuEnv~I9==7Zs1*y5u@fFhC<vSHA?
z)G)Uc(eLH!{WVSk5r8PQDPz40kQ&tH7;BNLskV*<MLXIwT6SQpyogd(lXw68b8h|>
zBjx_g!R#W;Ht;y-)OVFP_jg_N4pW$OS7M^0PMNM*t3r|6AWjgbYSfNbSm$#Gh9wqH
zPtqxDW8J&9`UowmB#rtkw(-%)8pd6x7)P6@hMN@PYQVEF&&auZZZ)7Iao`oN^77!@
zEI$TJ!VI8KSipr0@sVifvfVY1OqG1-E>q7c-09oVqco>*J;YiqlazX?dPD$yqkM=<
zGEd+)C4S{Eyx83Oj2w?!_D22N@6!i!P45m4e*<F3(l{Ew39e*N3rl3n77}y8C2z?l
zeoucJTeP!#SF4cgJvrN8^Ial6VoJ*64Q6HRwQK{F-PjST&~W|RFL9<~IU{Z>{Q8%3
z+>XHP>|$Z^5esMKAVT!CRrI(j!nM?3-b%D_NiO+l@Waoeg?`u}yFtxE#WLa|tE`sx
zYEa^0f%#~w9|n%3u6gH)x0ftIP<e&GE5i$Fq>j7Gsq)a9U9Eq-*`z5CJe!1gS;AW{
zTk6Pa>(@_CQ<IPO*27X*ChjVoT1OV&Y6W3H9T#k43!2H&Icpia8)%by9_}a}IfsiL
zOcLL>3Cj|~9|rMYt3z}0Cq)RRj@E$00KYS3?bTVQAl{EH0XM&9w3jk&a`P(X9vM1~
z+9#|(JuivYzIV8><)9m<>L`QX$>|iQPjgj>EEiW>uM2DY4YytD>Wuwi#yzAK6d&?Z
z@IE0N8m#54Nd}@%R)idWED-J8yxeHIw9s>VpkV9dz4~JAr*wxenEgrEFQ;D4VvCNh
zbKdzP1Kx>Sdv&2!PesgISp@p^VimLwfy{ed!{MFa)5k8mKl&rp7th}W_>!QtLe++A
zmbn@IfUKh(^Vnz;<a^`k!YRh^3sI1E+EInEP=`x*7CUG5_wc&UzaDQgW`=<inF<G$
zm3mVdaKnn6C&t_8sGJ}tB*-%}U#MJzHGa0SQuBSV<9a<>;P}Y{LDap^zt}kgxGS8@
z*j8&S>QPJdnV$DBL%3vb%X8ShbKs1s@b)K$&pr9E6vymHz*7qec>d-0k?#X`jrC?l
zX!%Y<6!O`nF#-V32}#Md!00F%vR-v`BI>jTl1x&x8F3V3IkA*cguAr!il>R<$PKV5
zRP4Sf$^&&zS{7T8b$fdoYa&GGJe1&T1loc$1mCSV(F8qrSK14?eUw^E8dnHIxpCbf
zzA+mTYL+Bfe?{qNCBwEePAAy6#{emc7Z1-C8i=fCQP>$sefbEn%&F&6&7zMx1>=Wj
zTt<cl6Tsazi03fNbR{qx!!SWyX4%~9eL4A6&dm*m6pE5uUETT*<O)p*JzvBx&gy5c
zbR3VAE>EE+Pr?Omk|g}F0*ut|Cg((=B5es+p7M(T0#BiVJrZYx_EG8p5y-fmM%mwJ
zzk!uiWC&ph3Supq+Vw2*>m4yM)wxWD&B?`W?}>&{6Ok%Kz($SgDLRA4K>F22<g;8R
z9SSGX{KAl+yC$R{LCpRnuJkz!(l<PSdkwPn2zh^Ov>)19aO!KmPO%j~227Q!D#a2U
znnQO>+{!q8S=hOCv>tJwJYjeZYbm1`$Pxh6Wo`_AY*hnrmio@v;jf}5f?|7F$j-`R
z>mBA6Dt2J699?@yk*Buy9GE#8ym;t<G-7FoLoGcn+n9eWHqjmQc^d@;q#b=a)i0+w
z|GMre3=?JlO`hl+JsmbT;W^=J^YKW@Vq$hB%_^(290|_UoAt^sy$N~m9iND$iQLHn
z6AT(0K9WksC`9lbt56*nH&_@TkFEu)kKk*p?y=CrZ=YGqL_51k#craWI~kM=?RqBA
zaoJnu1nkFD!ye`mtN-XUF5);Wg7y%kS?^0%XLTCZ7E(30ey)m5T7CppPc@vV+^cr*
zI6HYxGpJJ;pGn(gw04P@D7FhS{~2vN&1qu3NjNpGX%=(eY-4Y$oS}SEzFrJ+cen3g
zXp@NlBhpj&#(W8wPctT%{y|?Sa3AQUypU1W+>ne!oT%2B{1yf7%%bn5N#BkoPko;(
zyWv>jy)G-^Z8J##8T7n!8(}<P!8(YZrFigD89R9){Klc8kjoGFuCUjlrq?epENh|K
z#Oc!JCf;oy|72{ld3LgfZK)yUv{UL6apl=L9t&;GX0e2O>CLbWx{@Z6q(kd67XgCR
zZaZ8HuV!hN)haO-9xf~%1vpwKya2gFKP`URp#Mn__G2cVy+N$a#x7A!0Y_T`?~DVG
zxR|A0cHxI;I!BH{@<iL5D)hWz!K^=k*2lr_<zRKbjHBXaW<1*^F0VF?rzxJyU&Y5i
z9c2sgv)r?IM9#sw-F{?=552y}3R~-Tq_f6*HL`8-w0_yUr%UkW<&PinJV|B1n>cWB
zS>WVl(WSO+DLHYJWr?mFoPtllzavXfSO)u`{NTQj0!2LY4pDh<gd-TPRhX%nO9mVD
zHvf<ua3895$T=GO$a`GA9e{jc%rm3+u+)_67-oDYjWFqIJH7sXue`dztfeC&1aL7a
z+y&)I2#x+8N|K>{Oc*Xaj5Iv=QxfW+DY;GY<36)fSWBYM^IKCQ>@@LLtfs~vz5N}%
zCYLC&e-=yg>gg+LVps4>o~N8i)AXWwIIgIg090;NRINJz@Fv&=Rl@?Ab&DkoKuUS(
z>7w(%YWJv;!@;&FrP4(VitKS}Uw9bb33WfMdFs2RpbF_>KATY3BBL^AW#WF{uTX*!
z-7o!ay_kmojP{%QCeGufofd7h<XW`n_;aT<bfhCC{#Huc0JeC)(Zj$OAM*{ZB91;H
z|0*W?&(kVS59nm*ZQW>wT@)yS*cQ*Y<mS;1F=3S92a0&f1QLXPb)8*WbI<**Hs|YR
zbPV_f7BLT57Bd;hDv0k_d{T4InVBxO*n{THOT<At(cLdMgzC<|-RaR->^_TJ)mVh&
z=S50{-4XT?9TLJ`iz6{~7<`@k(;5rSe6<ZyC&BHwsk=4o2}y(rx>_nMF*XvyTc=pd
z9w5V*7TfDP&9MwsoPfG9iC@M|cW8DqE327-$CHiW&Tb_uksk3;W8~FaOVjZdysNq^
z+;;Md1o2SLpdk1c-(?3*j-gBg_wz3eo;%r>x>}d-oIPk_-R!XhY0ULpVd12!k`oLP
z;|oUo+(Ql3Jhx24IAk3x{;8s#Ez%zYVFpbg-W|(@<IJQ0t@ST+?NjGyO4xg^Nk-o6
zEfy22HW<x(iA+9>>-aW`#4dqH>regJeyRbX_B)nN9g-wTbz#uEUk&Wfxyy$gZ+>z%
z%;1crw0L5ghYp=enR+PYd<`;p)$9qv#<W_O3-|HbF5vOI5VnH-{vlxSyra3Jx}vx5
z^Udn#P}Eb6UCVpX6`O;x=BCwQFKxBz$NRO_Yf<RhX8PkyusgrIDxea?4dgSp0)Gs8
z<wv_;j#ugM=^<>cx!X<O>w!^zPc4(R58wWF_@H8UjP1peP*6T{x%u@l7-6QQkLp??
z&&|3BEw?$7bLM*+RKuF>!IJ?B5)k+DU{EJhJ-Az`nYm}u)6PJMS?Y}Mv6}>V$<3-O
zjknvcsQBG-r%qXVvalfo1~@yb7PU|2ztL^-{itL&a(HxGblu~2-eN*UVq2ILL{SgS
z<$4Zy>#lhYssWx(93JEqKb-m;3$vR{!5`Qu;U_a(Vb+T5K1-wcenVse3x%A=lw2o6
zV}Fq7w>JOc<2420t6Ud!7$(@o++;2xvrf)fXnFQu=1C~{$@SH%W#IMGnd^TB8Q}Z`
zN!W6I&dhGO8ssH{@VP8gff_|fnBp^fwY3l*z)N~Nu0AZaq)%H}+(hg`b`c8hAgmIa
zUC4_|TGgoGunNF)`bf#VF6JWK4To$)IqCv-nn`8>foD5bXx!1_)%Eh*acbMcEqe~}
z$h!7c@!c>la}NO2uTHYzj`TugRn+VW^GueIB{t_ta0Zyn^Cn81=OuDj)J3aN#oj5I
z2t!4h7ia4y5n&h^8iea$4iFcnv;Ef$$W?(_@)}OS`dfCIVT~r$>YzT~1G}6|pYx5E
zlO@)ieO_DgX4yV>bkjA=9PXD3#5FdC2iA}4kwS3;oFYj*^S5!lW7)GzLDRc$I0Clc
zIyy$P4tFKB$Klbh!qdMZWdXkM8IjVn&6Fj;rv$*3NT<)&C9)gB@`|5shVbIEF2O2`
zXIdvDnesX>t!Q5yQp|-0IJEF0Q%uu{S7~V!n<m6Ej$oy{m^^sQ`S0h`YNV=!cIHt)
z6upJ)ukYV_8_}?p7_FaCetIb46-%(-T0$5nZYCujX(eJo8sr2#H(s$4oMbuX{kB}I
zXBfu;7o5V5hJECe_V~A#4W)t=X!*lAp?rrm1v?R^y6r9q^)bIKL%j_bBg{mMw%Z>;
zRp&7iuuL3d;)MLsyH#;R?iX(kIsyX9j}GO6MLO)u0^A<Bks~HMX)k>;Rd7pyQTe_7
zOhjP%XPT9d{@cSCUT#(bbOwXaM3!2b4utIOuomm#0DGwdg6I0QSl__!_xmh8WBL!3
z?u6j9r9hGnlsTu`p2hv*QTLGd_{J6{ljHJ8x&<&arBHYIgl5@^S6cY}`OmSD8fCSy
zhmfiKnC-y-=zGA1^0W^>=&mmfDZvc!NWM*T;~|oZ4L}V5XeQCTCM2d+8BdEMN74$T
z2$NnA3KT2iPw(<To1&Dmab;*kLB{4TZlACNE<Mljwy?3q>BuIu+3vK)BRynQc4i{B
zz!KPDxJaqlDZyEY*eUyR244YXv_kYfM7WB20gUsCqThHvl^DX@H%=_Dn~W*TZr?Xy
zll_}dp3<L7jUqojuXpk6XFYF4TPe%DUEGZ!NjbMMGn{t#7o3}^%`TKGV<=6$VeJjk
z4ia|6pWn<1_!gj)+b(GWp>`bqg?XDK02@T1KyPn~$H$m>u*w0_19zgrIs1wd@ufN9
zHnX_xJy?d<)1?h=4LY7-u4CTRSRZcZa{ub;ywhOV1=D!Je+9*rxJl~>oJW?MYe2nr
zk<u|FieZ{1p!)~!(b6K*;BXjmgoS0ZtW@b{>V^&;fwFVGy(y3r@a>DE)cef&SE7Q?
zglhEo@WJ7JK@cS6^44eV>m2O1KPU9XQ~(tdde0*(-X@c7pium*HJ-{bP-Ve8<6d%Y
zg?&M=ZS0FCNHALjdR|COZ}aK-QvHR^B3y&v=vG<#>g-`hPYd&vf>JZrT_-$|d8Fa`
z`1e=2eqWZP%4}N3Pygy$)^=s{Kx6ss-HS6vf<p<TtS5T^#m>RyN5}R(ewXr5rwMZx
zdCOhV3j+GvoyZi{BnfwAl6qc1@Kvotes>M<0J88j(IH4h(Pp4ZNANtYImQ-oa^N^P
z3G5t}T4?X!K>UozT`W_%2$0N1t3IBC^f8n!3VDLDj+(zy{ShvOIWeFzFEyGcsa`O)
zryc|5xM3YVc0p&cB%Wp4h&mKnz?Oi}i~DAyGMY#Cb3rNH<BX8`h_EB)AknjK{jO&~
zY;Ijq9~Ws7aoS*j$uGtdWK=l>)0IvyS;BXC9nIgf+!oqUh9{M7SWSW7BT|1|_t=Gp
z`^wP0kMMMYI_jD#keqB~ehTv=_8hJOEWCnU<dCY*_98>v{&Hf5-!#r4JP9*!Q{W=X
zTp5#Mqkxm5zB(UbNq@Ou190di;P<T?$Ra0!5%Ig1f$t;O9q2yZ*;h)vFXE&ew@BPD
z5N4%8+#RPfpCLs!4yGyxMekHRqKPY`h5zN!m6<ixF9(&J#qh9vN{bn`Sf>2H#V_FA
z*sio4fu*KeIw}tpTFC$=!hCbL_r5%QD?q;*pt^zc(7qR>*Xyc!ArUZ|87K0g`*G=y
zO6{XJw@i)x%(x?&+RHWB;r`jb=$Z?1PU<uzl2r{?`su%=HY@B@_T2Cx-MZxZv1zf8
z({HT*f6;Zozh#2`{GKfjX|4WC1WI5+Z;^mqr_IfizyaU`F8W>>ayBSZ7$1d|ZIA3$
zE!^iSjIhk4!>|El!?9c_QXynYQF!zwT#@2iPot#hg%k-=(9s!^>1Y+C@+1?RlDJM|
z2{xpjNt5+FeTAFMcS<l3?G2B>?aersA?qgdg+rMGPnD_T%p!pS*a%2h`enab^F%~&
zc+erygOFsLG>s2@JrXqdX&TdS3k^R>bzCNrlDwp%w6uC4m6@6QY(y+@Ku(tZaHF0G
z&9CouPn;($!a@$u>r8iYV1lO1B>Vs%uiwg*KPGO9@A&#MoY;%RELwanj63Y*aPtGn
zk=0HVxS~J&koZ@)ZDhaTrlG}UX?^<XyIlAdMC>+<Th@o=V|j9id48`@L!BOtJK=c|
zgXl(nVL%uM=-g_Im`sz_FFA7#JuvsLfrS1+w>)`D$=$O=X<V4#CZu?*>0Fz}0IOw<
z-sI)K1^>SVeSdjZZY1DS3F(CjU)DX`$Hy`Q3+-nQi7(eQDuiRYk_>0l(x(4#zI`Y<
zwFN4wgbsX$s?;1XgiMD`KBM5}agVt$Z5aM4)*z;A1(MkxlF|!Z8DK$w>(o4_-@~AN
z(tPW5K(6;m%YouJmZh|^c2Al5t*I@JL@Zg3osVkL%j2b<xs(48?u1Yk2*GVGK^*Ms
zj~>YNugH*hyR_8}v?5cG1bDEuys}Hcz6-As<`c7B=9I$LL@lDg-ojNRCS;Vg4JMh$
zY9e%2*{hYBh>w0r$WEIE38a(EQ^1QxU7asbpa~*m>Bn>nSA2IFt=t_UjO{hA#fOPa
z?*c)`^B=;z82s`Qh-4^}<ciUCzkceTWY(n^mWIrmhfuGU#e0tTrANJ)m#IQY6=J^Q
zR&ZQbrIFAN_2>_$_v@>tUeqVHXy$04UdzAlBB!UECZ$Hon?(^l&nPa7WU|_s4Jzs@
zB4OCB3(SWEH>qmowDayp)xR*q1Hc_H27z6d5aLeOc|w`yd}V_W@xVG0a)n4>D7*Wi
zC}LOBL1H2?*6~!$pxI56ByhgAEV*8~AVs2FxucGHHo8*C#^Xc(=Y}eXF=F`FLD9*7
zv6_>X0K@*ru32}JINCRzjd}`jKi&;@6K>=Xy^-wb=U>+Uy)^A(7>4$xjdMNwQX}(8
z`jYY>Y-r_kqA@zQi0I*1;OWh)?w!iW$L=wo^5=KZvBZ@yU@bbhyj=Yy56?Ht_r<)D
z2Q$PqYEQPOOYf6~pEK(1^Rq7qTfXyU79<op_CCPQRXEgd;}e|S{P~aON9zyC0_KB%
z0e?MZ^H^R<;)(lD2B<o0`@hXfA~8{pr<6QSRR41DdZWA0u2*x#cFwAy)XI?oyI{;R
zu~VL6m9O?lL$kMue!WEVI_QYIxvp61ElkxUFgeOX)XR+sMarZXw4%EhDf;M`9<sq9
z^C+dLA!ru>WZ?P5ci|{0$i$vn61RsY@fMDG)O#>6u~3GDz%N-jgIS>2Z7^Q({U8z0
zV>i|XEWUZJ<TRfe>c$k*x9k3cG_bnRI_|C;hfZg<0$Q<d@K7H(JE|l$4UwT?D^dtd
zD*gV&CVm=XuHxDq!+k^BfDS}#4YVn;)mZ@d{%DSL<kR-x9_ooUaTPm&@hwcTIO;x-
z%Os*EJo9%S*$O?NTm7+}s0(O0q%+TB2&!YW*HX?EI<z9k5pVsE7XUIp#SxZyH<Xsy
zb~e;ekVJM1Ff(*d%Bm`{&Hd$*!f^H(IrhvC#p_r|O)BrlG^cgn)k=-LV`;e4s2wY{
zpt>aN`;f{h@+MQ6<Sz0bYTf#d;sApKXeuHeb}B=l_faahkvCz2P(H7%W0tARk6RDr
zk4BvBpjeF!fOAOeuctt0beENLb%^?G7=hAkrNKuoU@IpQ4DiMPki_@<`(R!aCk#K1
zqcyYQKl*8ZPN-=qo|XMpVfUySyttAnVx{*mgw2~9Z9DuQ%4ct+q@HU>z%|26Bxjz2
z+kg1i$O|QOsRbdl3>qB20MGjw>{fHzk3~d}>ZT5hFmMjrb&Nh!U&{-iUac$p>lpid
ztB>?+sW)hNm3lB+QIE{uc;Qh?a}S5A?hHgY=K2cW-~I7=47j%1Grk>AW4g~>@h-H;
z43GN=dkzjfc`fb#nw3=<Kt+G2E~6j(s=E#OPI*Hfnj}SGW{i$Y<V!0rIXb^`!EO>I
zN3WBILlp9ohd_X2kvlpxo7P-CgX~4vOL}5i1|m`UMCz|w(lpYTvc&xqVlU&S0;-^m
z06GX0^IabFC82qZlcL&IWxa)RH$l;7<gkg!eK+Je`m@V{c<wtMRZ9OVV>bp)fYw$;
zcAfZLDr?JD7<&h8Y@UqjCD9(&XdAXyb)u}WLQs{L?3}7!8uz-C!Cc$5kS@0Zikhs&
z<r&z8ituuN&M=hx;|_V~{HyM?fDk<K?OdPj;+G~{qgPg|T%(zFa~|c@Gl6;WrjdBU
z&j={cX%0&=-uF}b!c%F?B03KH$&9c!w<6<RwDCye3K{$y?QobkE}3`f;PPY>o_SvX
zD(-Q>r}`FF&u2gAld6(0-eO_sHLU!P()@3I5&8}oelsLKVz+%<9B^(zN}b9S@;*Q)
z{0H3~fzir^-tY=#TT?B!venB(3y%R1au`>4cPgx;)T?^H_PVYiW%9T2375XnNE-&M
zWN&W46*^M8sg_|Q;B?2O-_AMnzZ-}`Ot4Za(?WXZ<+rjh+S|K_iq^vsXJ4(^rjIp5
z-zf!0NppOE+`|$?YnZxD4&6sx>D5Tz+^*cfq2ksAux$G024B4pdtN(VZ2DwYsEYw{
zcW}RQKZp4G$lC~?b@$@r#?@M#SAS(uyprC*+33|zYfvEodjQ3-jCoWvK0QAd<^+mg
zvLbnC6^b5Uo#eHuxir#p6D3(Dj3Jr5ikaipn?6}K4gbSxt8Aa-!VG9yCM8`dAq|xq
zLz)}XU7#KCV(UqDw0H}oL>^O&@Gc5nI-Y|Y>Ay`J;6G%H&UsGDu|<rHAVI>|<e(Vy
zOQyzPZih+4gq6x%h#1RO*e)U|5uh1ok!W0j<zPNS7>u;<p;aNbSyqh8=NiwhU`5n?
zvo?lng|<6x$spRHUu-HGA;L-ubE*7QHWN}!5hO75TMS^BZj8e5;!qh)03c`@;I-I0
zpOgyY@d5<a<0E#HW#SkwcDdyz#4W`}6$0-1nfa?f)$Ul9c*D0oa+5m#t9<@%5j#>x
zLEA{Qi~VR(CVD2u`clu9w^w8D#$9dr2;)!xe(|R;&b;Sa1`am+K{MN)qK9hyOGg_@
zn@^3_<NxH?)*}^o?(SLq;ZW9NR$+TJc<yD%PXX(j9KB=oQZ`Aqsww}4z4hJ-uhv&(
z<xcxze!E9BgWVtS#~3|HBF4Td=q%=7tzYIN!^fBb9rnnF@xT990xCL3^VYL0LFXKH
zlMXS~EaIK9#*1DhMUGI0)9S>)>0@lmn$euw`)IGG&YNn<b5|cLK-Kjq@X>!$!>{`8
zX!QZ~`wGS-pXawJa!_GpQT65==lUr2J#wE)NJJ|xnUiN^Y@CCnuxtZ1o}wnBXcEX`
zB~!@4yGWmf?T|-Sfw`C`Bl-z<?cy@#3owP1jvzvJgi5FeiX=&Kr>j{-;C2t$w>@Oy
zEJ^a8m^<pN`|M&#q{{`oAX1-1K|9qGFTeE)(XT>m0-~aH<JMzcFS}9w%vmJt;|rBF
za`olhm4Z)0UCUpX;|0EMW-HS}XW_DL?}_zvXHOz$T|6TfNAyO!dNiy4sO~LMRBVYj
ze{n07=z5c!fXMuVIQSSVeDnbpRe^N0aJH8{=*2{5txAeNM<+v0OvDQlgSHFWm1)dq
zu$1i=HbZrXr<gcDL#pNoJl>BNvL{njpDBsM8Pm>_D*hGkhLH)^GLGp@%z^}*HuW9Y
zE31l6!7a*8w5L57&Q)YGKAv!>4I=JE8+bTA{N@evky%@#J8fMmKRA|{KXef}gPbzv
zeEZWm`x1()3K8<|UY^RA%|p%`asrMFKpt?-nKHq)b@d3hRZJrAU7LRvcSDc%^pT3+
z{WarHkU!jU^ONiLUiAvzA=76P#&6fZB?J03OafiwLGxek0V~2*$OaU^OM@(H|5rbZ
zJTf&Qg4@zw_`@?DYNjRb)(LE`+NUH?wMyRA4so=y9gvjPj)^akav08cf)kI4QM2i<
z#?c6Lb($LEvLAlvEc}#45}}M)Q171=(K1UG1($V2t0nQ#a=BoW6g>stBnim^!2Ob0
zs9;noMSPNE1@HG%fpYL9_CTvpH6nx4=nh3awj^SgacG*tl!irUf+EShRCWk0svXjn
zXI>goK&V#4=C59*<y14(GOl9=+6vCYM3Iwg=20`0^%_hRJMCUH4@vjvRuP_J!W^mA
z{$;HSGXMVW9osBD5BV#>_#vGQKFlDOcTF(UVdH`wL?4`sX~GkQGrn+$KTi#bT3?0(
zK`~(m!AQA>_=~yv_bT=~EH7me4R;+*#%(%lZ0bmnV&==UaD2aTt(m&97f>2M{@D-4
znb2GqG05D$eU~Q!h#=W{txo~MvY3or+L|*UAzs}|sB+qH2iA|Sc&+?xCd^{GT4rYb
zudK95(}Ij&|0FXujpaON2z0j=4C~|^FwuFafZc&czSQ4H|9v>}Xvgah<s8~?b^2sx
zSU5Z1-0{1FOiOjoeicHkBmrL!=a@f|6BX@e8XQ?9iaDo`W!k5|qjHAEuQWQJ;gLRu
zQ{llrOD5>ITj-=Mw*9E~(tlq0^smnb7YPyh#H{@<yw);XEjW;aAn`DpYoe?ts3dZu
z{xX{iydELBa>lGTaj?2@QghK0{XNp~?U3WNL4~Tg=xXnM0n(GhN0Go~i!kE@(Xm#<
zDV|?TFH6~%TTKNHhQ#cj07H)nxxP}Maypw*T7R@=T=O9Xs$Pj=_*nD(<@bPI5LfvP
zbB=AL_=%VwS5gt71kj$V&fm%8AN&K5g2gN^r5#-KWTLpHX0idal*Yustq_exgfuvl
zm2?UIgw`x&Nl;JsR&p}WTJ@p&xfuWz!;&reO-3(r@ImGp2H({j*Dw<!J&(-85RO)W
zNoD81i|0t3u^17PcDdZ0u#<l3N?4xCPg|BO>}GZ!qel7G5t@BD&Zw{>_ru%>B2Qq3
zPeY$(tLpIF2OD_DLKejZk5ED7etM|OFjwBvB67y^sFX)}kD_Y&<KT&SoW`pb@wISU
z{9(c)yOBb*ZZZqkZf)OTsI-YV|0sEfM>|ru$v5^OIc||Vh?Kv_2js<?9P#hWi^6Si
z{p`yz0?<u<YzkP+r;rH@YFSKvO9aT+!(I#QLil}mhiEP?KMautoPVZ<$n2^A3xoZ?
zPKT660Ja|$b>TA~V2&Uo5$=v<pW7dJN?+9V%~yecWV3X$5hjA=b?JSh4M137y5NGv
zlGpssa(qDDMek!&u8>*rO;Aig-as?BV6J@xR@b=D5})OG-Gn9F*s<Q0r^K2*Z})gn
ziypFbS_^M+HduLnPmCYItJ>l9x9`CwYTB*WCLAP6t-&7ZeE2zFb8)mzov!S}a5`r5
zUU&Pue=CwM7I4*EEYmXIbGn0IukE>?BHT{@si{UPQiQU6qY>E!m3VNyUBeVl+VYIg
zrLH%bV-}YiR7+gs+<p3Cholsf#-B@Ac>i*$cXQ5&5{xF9XA(k!PNaxW&nBt;nuSG)
zI7ar8jw37z>3t9n5U=~vE6y>jMIP$$iU)uoER_8eGsifIE?}<}%O!1KiMt5-@Q@TK
ztJp)`6WuqSZ#vXMt7-O=a$MmwIuDv*Zl_Z8((#K-e|(SrXM%eF0w`q&P_CJ`PppXa
zyca+EA_Nc?6m|t>?2zNw%#;FX`Ipmi_gB-j%=UcQ2Iq323oZi4vDA!$5HDqz`(gxy
zEI0#xrB?LqDSLd~8xnvGc{__bZe{s5J&+jL9BwPa2J!A!V2fJ2Jj_Ojo=}0n&lLQ^
zYibtl7jf<oYBSu&H}ff?(roK=1~aa>=ZqZg(<HFk|1zoQI+e<63;g-{aM4cg+q!fA
zVtVhFW*gVI-BXk<oqznjy13+NIqxJ~cc=sx`$JrOkBvI3?ovthSNL<O`=W5mzXffC
zz3>7x3G2*FUo*Wma*vIFv2dcG9?Ie70N)s^L>;iF^*M?g5D3Ek6VtLKf{I~+NUM(+
z!_4WphZ&musQhcq^Ph?WW{By+{FA<TU`@J!<LQ@8>nU3aNyh20ouEgihl1-P9-+u~
zB!wZppD=!yMYFT>fI|Yf|D=NEIy#-*$bJm?*j8>(+vj!o+ivsc4@b*2Y5zSOld2a`
z%GARN;@&;Yv-@K-z+Q1L0E{B>H6-K71j;!o<Y~WP5{OXMgGfon*YW)&QNIWwRTY>P
z+#Um<8D3{#g5`x0*dIQpS)KeAuTE^>Do6QjLArSjN1DWyBugA)gAfLhrlpj6v#~{W
z$2BTeeOB0@_%4oNn209eaDy_8!1p{_yy0>nR=m0zD%eqFks~}?pev#gBfhy622kj0
zMCuL*yM*wR6BR!N6l51l=rYK`YgG`2AFIB}cd0aKK<8d-?B0NBkPbF;sL|E$4Y>d6
z!%0d$Qb@qR<uUl#FTjj(Zc)uBMeRL4(;UPGOj<8bpK~G|hiIVcCbuzdB%ux1!nVK|
zXvQL9ITc?#BXtcEX~m>O26Rg!Sj@I&zAsvr_~#w~oWo)(TV}nEovP|71oZExza%K(
z({WuiLmgqf*BRk{CjOyqTWPl=2_{K0te4;ZT4{jklOB`R3<hTA<}DI|PFVtuRWOZT
znER7QiO<l_TU#AB_Wz_hE*fJ${)y1u@2o0(C8cJej&2>UU=i?`oq9x*Yh*Wc;AOd|
zN=-$LGIb&0ISVls(Ncc;a345Q7Y*AZwmJS4s`kI#8UIzMKG3S~K876;7z*bb&>XX_
z20h-xE_ui1GiQE+kioMJ4s#rvPAlKQwET2prM9zWdaxfZZeu=wa&yJvS(r%!lSr$*
z&o;I^3jgHzjS7T~m39#yR<!KDVa1AUQpx5Uunt{)VKa|GW@x$zO_Wefb{`v=QUh8w
z{`UR3{PCoHfVkGTGN~jUcz4>nw_bB9TYJ%))3y=AdYJ7#s+8!r*+n>StYdS%@-Y{9
z2H$M3pLw%XV-?FE&Rp$@;e<_@HR!tOh`8MI8XRC|hUbz4KFcT}H%+qoc{_ZJU+W;|
zbn;Sb>#}43+@u$>Rl%2%G@^=lgn_hXlKsJL45&mP*)S|>2L)UNNpk3Er1Q?HjNt}v
zLQsq%_3%2T!MSE2gsuaiA?Bv~(4Xi%FCA0B^wGb{iTVR3cPLV17Mly7@Q2gWV8eDx
zE1_A*XdS0AQUp|3?q%WJvp@OQaLh(}GP!5t0BFcLkREX$tz7j}0BB1Q_cDVa(vZZ6
zdA5(g<U5`~)a+Qj8LrreBAHG>dAPX>f5)x`gVEx7zP|6IRW_wB*{|s)H>mNA#!%%f
z-D4cF%;q7{VpQMhz{(`~CP`Gxd>?l`lRfrBy%&t=-dpXBLf^QJ^Ba=|i(a9?*Sq$d
zsp+${e$;*}orezFBW?e_wzNNhd~)H;4u+j8Y2ku)|MVnvM7scI8;R5ScLEJ<;~YZG
ze4or&@t}Rkl7~jGZul&xQ_GynYI|0DgLxP)Dpd`Mo<0HZZzRLokz)Rn3SH-Gr$uYe
zk<zK6;z)(V_Gc_FG)y<&#yI-98!Wc5{jR==6`ne(O#GTjeeqXkiED(O`GaXbGnH(i
z2}xH}7al#|p%nBW+x+s+bLt{QG~YfX;tY3eiY7c(^Ne8bp>!`h@|n9Ouz5;-0AS#%
z6{GAJU$Dx`Y6W@ZFZ`|!*XDmSCcx*pFeEBk#^`y=#X4jWGICji%8G|D2w6t-0kz4f
zo`@ET{ifJ?;U~qEPD+k&plI&QaO34sdB;B1$g;k5eaf8xayV*u7f+>x&k{d9rjueI
zbqXw@U$su8Hp5%@n%!UabJWt>3H)*Ao@J|VXWY*&mjC)Uk5#)vkFNpm>QwrVxlr0J
zU{CAg$>X7o#O)X0)#cRei{texX1vyqOD&!UhXc$xRR&HS!p)5zXZ36r{8n#`MX%c~
z5WFu)E!NyDx{l{MAL>E4Mc;M3N&Gd|NL)DUIy%Hzkj-4R=t<kuc`r~i%KGHyXJ<_;
zN6TvgirJ$QRPKEfr>1UW@AYtWOcxQSV?)usqEtsOA@(2kJz-&vPygg|5RPkO(ywl~
zv`i1C8^2XO*nm{{Kxv$F5nIfWR~=&G?=IKS2<8cbE*bpJ2bgc$n7UloTlk(n3?ElL
z&PAlQQv)D_&H)Do0v3I^enSSU_9s5qt-T5jm4_F(Xd?kZhqp-Wnv^Upo1|&5DOzc%
z7}u$8lWJf{9W4^Pzzpk%!k@IlEkNTBCqq+|NTT1BNmExroXb#}Yn_(TPoiHyei4L`
zDQYCWF|FIk6$ZT9O~!FO?Iw5CrEt%#fyl+Vn`R5eZAJ>8Gm(NEfA;J>rw`;#;0Q!Q
z972Q74er!aBh6Wv5pYxLN4Q2erp2I?gEb}9I-V+gszuJp;%JTcH+mbP8NTYX9oZx_
zg)pCW6%&{#ZZY?)A*T_|`sxx>LS#h+1Alm%M|*TQe*L*dF87A=r}TaAZ0?P|>S(%T
z&g}}<ORJw^Uo7GZvZtZU!V@|aRO=8SdeLo28%*qCm!)ANg(6Ju?!0w)W&30A|8k3S
zBLhnSryT3AKuUTY6%&DXQ2*O5E8@e+sl<n%Vz=ub@WFqohxaysB;t`<BJfV-cDHR0
z_TDITx!tODgzA29AZ`Y1I5Xb@Elwz{!eP>0G_X+t2wL<klc$@7OG2TQHAL$dU|t7%
zN1d5h27x&e`wTh)lJ`Y+^6LgrVls~5&sH}g|6D}q5>Usdw{-F#O2z#6>0%_C_H&<w
z5MPI{Q+Pb?N&Cstv^|+Bd(AY}?DXIPn!)bS7~kR~V)~x3BNDd2J9zZz$KmzF?~$9M
zOzqo&8fyoF#-R@1QF7KjI>mU_(Z&<DmmmqfMaNFVWqTWy*L8_G@9jU|`<EVpM~+}4
ze_l?#`gT7MAGb<Ejr3V<X55>POI~^e+15IXt;z}}?{A2;!nSxstLH?#>O2<&LswD_
z?gLY*UX}$OPADA3njliz>)fmg`qpV}B2(i>MJ`lC*TaaP)u6RKsgRNm(`}tiYl9a6
zALcD!imwTPm+G23W4X?`tq+JW2#;=?Vdmz;b|NZx>vr`qP@2RH#A`iTH;0Vv7zIG8
zeJ*wuwgysDZ##xvb5n`Ejd3Bl3e-LjYVT7iKe6mV8L8`*v4Bxk;MM#0k>BdV&|JjH
zxZoh##Bd6zHqf=8{eF=3s2uFrJ7DEUtSW-Iat^c*T=qz1H)$ANxqi4_`bB2qXmYnf
zHl-%Nltqx+Xc=To6PRv-+xZ7@y+4MB-k}qz)l7<>@{+l~J3y+bv(s*>7~R>J_?V#R
zjp8Sh$HpKe%6u9Ln>dXc_}Fn^w6_X|?7qI8K)R(%p-iI6KJDkfWW0N(C7VDf72>l<
zFJa30?5RbYE*FzIRv2b7^%G*MU%$vKu@IBcMnoOUFcM>>Rg{W*b|%*VnemN6^NG%E
zieZ<2y2Z3p#VBEa1H}Z*j@f<KVlbH}qzG7Cix3ym_!R0wYm%&>8*9#`YT!^LUlfXW
zK|=KQS@>zFP+TgRuw({3L74bvuT+?F9F>J38du)<9})UVV)6Z-vqJS14DOdm6;dI2
zU2>C}L7?r=%LFj;ECqC+?3wMR4Xx;}34PDyiIRSrdIY3^y>n2@)#@iOvK+<_y4u!4
z%tL#`eCT$^)9@@})JE@&yPwD_xPE46`CGyM_B74$l1F7Xv-+A&wY`n_eMl0dK%lu3
z=e&+hB8T&UbUwd7y&y<BH%P0-qm6f%==*}BcH^(NbLBOITk<~jo}|SO*DrSZac{Oi
zCcLf_eT1>D$EsKy)~C1U1X)MkWky_}o?o0F#hn$Vy3q>xAH;2z)8297o()>T>Bk?s
zXnG>8j|5mG)#uahSIpoA8?66>^?=tCB&CcSjY#@66cu#5->P~$`nU?kU_G0999!|}
zQu^>Js7<l8icR>BoFw;Je#Z0X?jj5b9pUHhqc*A;qux6*K(u1fsl6{_T54Q2T&OAT
zqh?^Ep3V;EKaICu8S#!8UjwdgE<JuABc`scciN9lbWHY5i)$6w__Z@tkH01pBn*x&
ztmWj+>M>t;y<c6TnP|`V1MAzIs2W+MrdQTGeZga!GO5-nUox$2_y1bCx2CkugsB_7
z(|lcu@D$t%{RTVCu<qngA+U@pscXSIZLo2g5_;mj?VY<jiLq?>@L)7~<cA3i5Eq3)
zt-QvZMsU|i`E0gD43b;BYkmi4$6mGwK(1~t(uW*Xa7mbe1x;P3W$~z#KED#VE`GH<
znRiFnl`Ie`P`@Y}dG115B!r((wV5EgXez<OvGnzG8w_y(yR0&UW%<d8Jis@PNUkF;
z|42Lv>n3v8e(^1}!$)o2Er@kIoUe{Z|Bm8!0<y^sXt-j3x(3N)&g1QI#lKsB*|u(6
zlH-3qAUT?=DmJ0&Rw?>-m@|XSSpXLcogNSKjsg-Gx4yTzj|oGc-nPFd6|HlN6tHQ+
zP2lw(0A>T>U1%_5q#1f}ZhoS;y+MmkX;e(Z6Nf8;rjaHZVe!h)(A!a=`^zp065Ui;
zNM_)l(XXLZxuFvk?o49AcvK>MPf%98mOv<LN625P^9pRU<k9T;j2Vm-9R6hf32L{H
zzuDO?Dq%FAXK>ULJ^(xqB?T&kYVgD<7(M5S%BwE)BMg;-95+cKd*wK>{WZqoIKJv&
zsnq3ovOMhlaMMkNNIq_|<UF2)TnooK=ZONjHM&@LR$=W#o~<cTE)v-s5tOPm6b)8$
z+@#A?P!X>My{SZ>;_AHm)u##V>aAx}WCA7514Oxwot6VBdBAgzBTF62s6>x9!~zg`
zGlP@}4>hyaDynAEA7lg|z}5`)(j;k5PJdnw-*(XqHMLP87Xhee`&6l8^6ctU<aSEJ
zi5gE-sf-T1=Q@3>w~*FAsl)tCbL>7+<y$iWi<U`|71|)}K=Rl3oXsd&l(~mPS4j{3
zsUy=&F<Mio%?~po_l3=~1sZ5U`3{!19(J~0t{gC{c=;=B858aIEr;4S939rj=={oc
zop>96k1`#q`z_yS*VO+sylX3?mbY!3GoGE4(j@xUfnm+bc6e0h-YVblVGeMZg3#%>
zV=#Q|A#On2V1jVSq~IRoR>wXC4mTL`?5TsO!9}yx3ngY7(HTXW0-D8d*~Z6Lof%@-
z2=!=dBrU=HsL2@gII^Ts$Pe;tal{Y@Z*S+@+1D}lBdB;mkqvgU6N!@tYYG#tDn_fb
zJ_n|o_pe)&JJ;S*)oLH(ZdJGIqZ?--J}TmLz0(P7{>^bOW3e;_Rw5>?tpC|*Ozq(-
zl^b$3PPfW2c0D~^@*$wreXbZ~ERCIl1XWRKX~P4}*3)%W%!MSlgrd`kbayBjOmMXD
zTVIrXe|jbUFiTiiph}_8-(}i?SbL#rRgYL;sm3Ru^;fsXi@=9fJp*g>O|4Z8LsPI^
zAlt$o{;L<fz8+P^foF{A=v|bvn}gCXzI^eT9AL)m`VM`3_;}kIX?Cm;y*c`DZ2bA9
z=s4r^J=W_7kNnU=A!A%;k;WC@pH~-!=P?DjK@SVFw+K5iZ|v-zN8#^|d47t%lkG=O
zHhf^se-3S~xx(4AgaY#g#%((Z)ZEpH;ZMa#fV-wI#hgE<Hzi`?y|l!n_RcffxQ1N^
z%Kjb*%oRr#@`ixU!0qQ*lSf@M*?z~(5em(G(B|xQQ;BoDj8tVwUhCm1WaZYoE@N|8
zhZaWKVLo!K=^e##c2v<F6UD8D_ELG@WGn@>zPfytO-PIJ8YLJlnBKe#?}-N&i76n;
zEf^PqDhH^OGNxth!bA6tM#0w@i;poxVNb#$@bvb0)`WRew9Xijgr<mgI_V)X=e}5H
zhjJR_6`b&0L0p9E>7wRnBEB8Uw<xSQoN>|s*R33cEWRgUK;zW~JBpA-A6a8);Yty6
z#nWmcP!-yJ7}~uSzQj>=DBP~MX`Kx47BumqfZg*GufXJP^!;}Rd8Rvb1kn?mK)}O$
zCPx4ft0~3d&s!o2^*(;uzMNA{j$1-4e6uMoGv!68)3B+_!Ui*+{u2BY<MqJM7+tEP
zp4I9qF+>{MqE@7#k;!D;=VSlePiZSDc$_uw$ApqG?NlJ>`?e)~ASG6L{32KpBL#K5
zl&_|O-?7Fdf+`<3C1GoZK#b^H1BQ4;$Knph5LVB8GwV?)&B46zQe#7C>vprF%F%>E
zA!{v$ww1y0{7JmYJ}EWn$3tyvB8||$OCtq}Dx7C2Gu_WzNB{q=neSAjwxU)cS9eZf
z{62!cv(^IDh|n?K1BpB@T=ADye-{`35^7KoGVRGHsE2!6>4I2AoF=P^<4qi!`UELF
z&?%vv_nYa(3%e(p>MKlJcnwU>(SsuV6*dYtu3jd7S$g=YEmybOvwY<h?@Tnj+1ZSt
z)v*D112{TxN>w;{t1#=_V$X*Iw6&Djm8iFq<r>6d$aG*d;2-Kf6n80Qqp!>gHmx6j
zdT_c|>UPN1h0>g1TD8u7Pw)E!F6>Wq9iL-9Hrrow+H3HMMa)8cmWi}w>2TOiE&5em
z)bSnmgZ-XtTUuF#3k~L=%er=Bkj;w+@~vV77R?e!+?g`ti%jGMt<$i4tXp8t^*1L&
z=;~!j+(j{<x_p{FiLLp))*XsQ`NX*48If_ys8^zrUo+Vo%fy%bgGYj|h2G#ZbC#<7
zPJH6}M1jg<T*_+<cuj&e$nuj)zg40152cIOQT^Q0oR(4KPO)oKuSE6h;F~w!Zn~AS
z4|9eBT}`Q>FU65)3nTSstl*Q+=_U;Zl$74TU7?J1W5=y0PR@wrNFC#V`hd&zNEE;^
z*0m}|@Q<4BqZNL7Uw<v>20%5YS+=pzDDv^2usyL{i^Zz~^U-2T;sldC;g8xtAwsI@
z^YC8faAlwhKcRsT(14~JyMRyp6>%iN9Cbu4=Habp3{RYTvWP`+!cb_|2v8)G-hJ7}
z0MP9)5$a-z6Wybh<*P{IT@k3qaB|(F{s<{zZL9l6<s~B^yLz@+9E7^P`<4RLD#`I1
zpT)1V(|&5mCz(4ThFEbciG8ccV3u$Ra?u72Hos@FJy-LZ*Vx-dy`Utty%u6t#f1GT
zZo4zg=y}XG`m-b*imP{Dw)PP{l?Nq4Wn`cp9=ovs=HTpol1NWvH7a?i#^L$SOf~w`
z(Dz5Y<!fyT$oiAcOPe_LruapTiicjSPYR!WN#XWh-8htm21gYhtu0F{QU9q3v{&p^
zR&#x@g_P_&uxfVm|Do!w0^01N?$6*9r?j{`#R{dk7ARI6iU)_{Qard6iqj$miUxOg
zE$;3T+$CtRz~udZGxN>d<tBM@ah`MbUhB8l8BV_2(YfmP4!n7aWF`FX1UWM3f>151
zq+5_-e;Dg<Jd;0|KV#-_4bOZa5hpjb2>Qhtf3YNJ(Ci!5{x~ku(7kU=uSkWR$8Buu
zQ20!a6T$M&2ZM%;qpM&X6P)nhF%(}yycl2G&tC>icJ^}jRT0Y_MAo(p`l*Bsal)po
zVk|L^+b_7e%?FtvizZHG$7Xc6Z_|!mf=@b9)9o0nDjJGMu;;60`6&bV7zV_%lFUIK
zinZ8l=~^sVwTCBHFiHoONX|Gmie9F}c_wf+ya)~LyENbPsFW%y;RaU(kEnPTghaYG
zAvia-V&z;3+dlhpnY_}~njwU$3`G!2mFw+k{<qTd#s(euV238Wr~@H3u8oTymjSXh
z5=qk1JgA_k?iYO|7{}*;G_LmRX)LNw=j-hu4EW8T7tc8wArhGZ9bX1UwTnI_ao{{5
zoBtKDjzubHK7W3^TIgagv74VhZ@U@?uwP`s*0H;EDcRYq-@TnvRZi+z7U|;YQrBB+
zUHLvCZs)L4Us|=fVTso@S1tgte?i>v-Sy1vy$ZITz&d8&VBMdP4z_O*`S5z3)-a3J
z2pUINf~MQ8jJzQSKnI8*FJtB*eUn0bd_&-lnuJn7i)kUnvAboyJk=N|78rt=2#$&n
z_4~d<Wz6VCB#Gs}lUsV7uX@KO#~<n+_!A)a>YJ>$7bO&>C07C<AB~dxS5)p2De+T-
zo17~0tu%|MQVx+!7F2<IIj0r5rFSKgpTKWASyzU7X%PIE6rO-UBr5F`{F=%w^7Kdl
zxSxla4{I0J^~a8F;4Xx({Q5fC&PZk|o)l*Opq5!zfl^&jSb^;w0RlAFMX}U}=sZDL
zRofplL1)-#XUb<}!OI_*A9lodI@C8>`}|Zm2<U8%HkI8fhgD3e<ik%HL3Yuzc81Zi
zKAHs=YSiks1}5{7{{Y2YV-(BtXTJmxOY9Y=+QJKr_lc>=RCB~(a&dkx>MSYLqa-J+
zy#wBR>-_+<^B#ZE8{Y)|Q6@K*mU~@nq>p{lW@+jabiCkH72wy9TQ|ngq5mzYMdZ{9
z_r)z?46;v30dWd2NJ?!q%_mNy-~D0ew5GpuIu~AkUM0fuSjJ$#25&4_B?TkNA{M@y
z3bw@q*HOWky@QXmfzGf|&h81DE%E=qOe<0+*6nZhq0?{7an3B!VX0`?@3%1UQ0Z<V
zl7mvsfpf$rJDaHvf%%c(id6TLV!!m2zJS5HkX231ta)5yXRTn%vQK7+qoSN9hvnV)
zu_nX7=Z~%?iI|dFf<=FuY+63ZB{~J!RW$cAN7Fa~Tryo%xrG14d~Au#{d!ivzhs?R
z&1KYrni+p~H)8xw(XD=1uf?s{XlK3p>toBK0)jgW3jy;9#<GL@PZ#R#bRMU~6xRG7
zPfHY=?;2JdYp2e%u~n_`C9c(;^>572=1$U`I+k`Wp8IyWfo+0_!o)U;lfTggPqg*9
z<(Oa3qXct9P&zic_q&#C2-t#Q9_B>J%Lr|1@Lj6uTKW4I9#JT~9ZH?2h6zBRYn;f=
zp#WS;ISfH$6I4c8l(u+D(!>w)==ikK5;Xu`R_v=^1(f1l^1YjqRF_B(&Uw_Q3E?0i
zN%Gm>#7A+$$h)X+blhtP(i=?u(qis|=qI6Hx@dc$uV)GYOh;M50F8sWHt(x0W@xsE
zCG^z+y%Cz|t2j*nyVqgL?SV8X!Y6{a0)+<-&@we_L66q%dK+lols}{UfvQ-}&w<uH
znDeb0{-DZYV2g8RjdNz*<zPs>blUMJqs7}TZIQX|(qmH0pMFJOZlFZhYxL#MP3wbm
z_4p$}+?nxg-nI<hP}lR&8yyMT5C{1l29XP@NvFWOqHxA^-R)cz8?VIVFG*7Zu2Tv!
z)Z#aC`VSE-|8*XATX8zJY9ju%8dtVoyNmeG9b!5%EoR^mYZITNi-PgMTqV=UxH=s5
z1yNS0Mwh**4a^ftR@z4dH`?{Y2c7N=6zi?2@b?}z{)qC~K1lh}C^&_3ZF-qM>Y3;g
zmx$?-qiW&#>q3v4r;E75(d^*6|LrFJ@Al$j7qXDQ$-h18=vq9^X@^O1<x^FZH2teV
zkQZ)~YeM9IUtnxS!=hE~vwK+1o4a+wH)tqvf-mz`Wi03Er4*FOrDfDv{%9AN6Bs0+
zJNf7Eq1=1aV3r%()nVt!Ful$`BgWvDo#)x|Y|BUY{kN}LxIY<+@8`~;9UF2|f$QJ%
z;V#a?de3IL#^!}bitdHFr+mEPa$g4gN5)eKv;L$TErP%L`Z)}DUfr(TK0lsgu+Sk%
zcGqHu3S<aCd}lH}Rb|WOl2RO74w()1k0Tp}vN7T(%;O{+|C@6_?C(kq5G0I6$JW!Y
zz$t~PGY^cFQ5RJ?k+V9)1*$^;iW33112qb-lj<Nip6H#)iuZoT?|*o2^Lz6YexKYd
z7&p3k@`kW{Sc~fk$8SXR477mr|1`h<(<&TYy1~qD5n9_KN2e9O&>hy_-bp!{T822$
z>2+_Ikw{;=+s)xjp<#_?{*+Vav#qM=MQFYdHCncotWT_>1z6;%(ez*FD<sxf@EfP5
z%^x!M+Q5dXN@^2y$yb5&l8a;M0^EX%cwM;ePik1uY|`iVMd9k0gVIzo9FrlEZwG_%
z)#4?kFC=<x3;#3#-Z5f7M@L6zi}7OaAW5lX(D)$V0s@h%B5u7rgX4TKn@1?S`~OIV
z9vz5#CCILpXtSf(qgG)}s0DVRD&}(c)2$UrH}s=?Q<P-#n*1p~B}GeWuL23Imj~A_
z&)^4nykk%{K(?`$i5K1ZO=1$|@=Mjy%s1L#fkNbU6EW)2M|<pgI_9o$oIb@;pa;`L
zJ?&RA7eRS+xO&BLq1|V)g(N!_9NJjWn^aqrxXWLK;>gS|*@iM{7Yx{`!$b@iR81Jo
z+>=>>qRXtrPEAhBuQ$M%*$vk*tIix`QaJ(g$vdo-{O{z%Wk3}qj7w6?5Aq7w_H)eO
z$tGrw!G=CNjYn9?WFYv<|MbkZ)o4eL{T&pl_LkwN%gb4uSOqiwX;q^tot`Bb{x4rz
ztrn3%&={uMqeAu*u=<iER#Y;{mxejB4|39R|J!^c`U6OS!_8-ZyK<{9)mB&gvnSOu
z%Oo{j*A5Mh0^FSw%s7vey#9tJ?y*AEYJQ0dRM<F=CdFfaT8BF*T6?+3gvV-B$Q@Ll
z3r&8@97XeyEBw#j`!$06_Y3SUbL=EWT7Q`^{bOAgaan!gKv~UqaU<Iv>z#~~$tob#
z@p2yYmkV5%e{6fqU?*ez!*r0MPC)xA&?Qa_A|hWBG;)@BD9_AdZ2RuR7#<ZylTzzH
zU>V+!?YETMu{!pKiMnyTuC!)eC#o_qJ|@9z;x3x0lf=XMtx?({u84*HBaY6yToc^K
ziq>;H4kXqbzkygO;eiF(m-~r;Wqe40Qz0I;p`+4f!oDTOr}pEn-)x!%sIETO2c{`T
z9;xN9`rZ}C_K&(G9G)0+tImlaS7a0P%#BLeIz?Id*9Ju~&`nL$n)%z?Z7<$=WzC;w
z&Bql0HP5Y|Ec(_kerEFbCV^Cb*Hb?VbH2k51Y$pOup^+}L;&U7T3mGNLrLjnr4FL8
zxIFvbxa$X7)A?G>HXva!NMLY&d|sI4(sFaU(l|Y!Gg19vk?IR_*U(SA<dSc;&{6RG
zMz@Fqd`-Neb6w-z1zLrH*$IFUzzopn0uxHoAO*1!MbHtN1n~lhdbb$RF{Kcl0UjNV
z3U--P>JGtgxmW-R@>haVpx<lkyf+Pi78GJ6RHiNh2P81y2lAIs8Gv_nvdN;tQxdL+
zL}K44n2-3Zj=srJ(_7&jNK&~bpd>Jdt`q@xq1~!L-X4dfWSn{(nrOm|J`zR^sPwJQ
z=-sPf^OW}qI%EwY7PVd?;VVpvrR~cK6_<m%)PY3|V{lA>JT0@?Aq51()eX0p0lzZ%
zPFrOZ{=@fS`#JO=gb#vrl+WT`rw+O#@b)65K`<x7e9zA97D}I+xWvmlw8wQ{CCX`u
za8xNh{<u9%3|8QgFd71qWjR<o%`|zC(*95Hdrv&5Y-dsWL0`9pCPSk#^F=~Vj&*N^
zS(^wFGf@pPPJ4?Jw|(<Z#(zBcqZI37sY_-aU5JRm2(velx9O6)@PG2uPHmTu#AnYF
zuJHSUJh$Vu+QPY|GM!5{GMhL56-M>Cw2z|<ek|ItH?VaQw=;(<ngEsh144KGSHw_n
zmMp(v1MQfx3GGkZ34@GPT@=Qg?0J1yF!JeEyX(KeK~{C<|IxsbOoBdk)XlV+W)>EK
z1xBh~DSbap2|F(H+ZLYZVsLWcSlBnVsu(puTy%k-K^ak}J@UZsp6X<4-O~S3AtHQN
zq}?N`s+s{`Y(H*rvb{Op8{6e`jn~$*S>3AtLBbY0LQ&_<kltaI0dM$L?yxU+2cJUH
znB*OBhM!p)Hx*t0$JSEAXhuAfLJFwwyu(0AyenY8G~AP^!G)U9e#^>wHZBJGT9TIw
z{Pc4bwTIKH5USe&iM@2jF{&8n+4bTxhj50m|J)o3etTViKbzM&#(wqf7v}*61tLUb
z4}%QJNl8hN6IK5=YwUQTf8HY;o>KsrAy$ySvx1q4+W&47P~`S-Sg6yIl2K)(a&dey
zTv~~K3wmjXooq<s<LWFd@(mo}da6Rl(mvK+jf;7LU7mJhMeNHfu{(D&$Hv9>|8|W`
zeKl}uCp7l$7B<G9Ie}ezLj3IjqnU0KTF5AoDImBIEctd$$9O4@K~5iX+Nh&l9x3s1
z^5Jne#kfp59M9vdd0%^J@!n6+Ws(-uA7Fh46x+xweV4fI-^V|iVc+F<%>T_Em%FoC
z<|0g8d>pR`Ni8^#A`lE%Bp4Zqk;IOi!_LMJlo|vG(gmVO;vh3Rqqjkaj$Z??k#_?H
zpwpoOtka28BuGNYPnhrN^=7ek={6$F`lVYDwc}EmUafswR-@t{F_4F|)_O1+_GSBL
zp08b|$N}<BW^SKV&D@CIhqgViSQ(<a5edq=egjvlq2srab6b*!+apWZENO3RfGh|$
z5(?kBeZrGFo63woVr&gP?fsK9K%M7K)y%&b<10D@-2B6L`tvPgmguE+g&>Cx{2#)a
zIDZ36Gdt6icj)1P_s5|KVIfrM%jT3Ds=a49@uyXm)$j5BD#o*H2$>6{<{P<5RQ!c!
zceBo`r+O#+rXVx$%y(gmg+mT5>vE)LG7)RjV7^}vhJO37DC;ptqGUf_5)3LQkWMUa
zC)V9FY4HTse(r%Zj)Md0_P$%s0}b*v1_#cnJKF=+l^dX8u=33ah;Nks`5Q`0zPu_l
zs}?&qVmVAgSz^{CG$5r_a|m87=UMC2!^@SrU5&G2ahmLxA*UeR=S*IYIJo_^f4D&g
zb|tUK^}&vg@bvrtugZ-00p`*1D-HU+%#05X_!5pHxyv?4dp-Y0kIuH`sG2WafTXp(
zJ18p_v1xHgvAW&M;XJ))HxLB8bILflT54{1_k&#@D|fiKn5Nis!cqAYSps;@0%=39
z_ne%qeXOt3!M7)@J3o@U{@K#R!X|-K`9&s6HE#S-?AsoEZFArzXJt%8U`S&6yqJeQ
znx}1Z07V664$r27AGDL)I@!E<RO+UP=teSi)E^%&N=sB;e=OiFj;|S=5Rt_C$Dl4+
zZ@adfd2QanzheL4@o0qmbo9eSv^5)VKws7iC1NeMI8!dGw4P{0q<rkFH3$6p&%04G
zQXcw}#Xr{J(aPtr%m;|q&R-9LkBf^OyjD{!A&^_W*cF|RBe37{P?qs*9jKMY7a2t`
z(xhI$$ORk?=^LJS=Q~w}?{@%{^?k$!sFIyt&TmFdR8)iy^RizSC>M1M4i8BQUQg0U
z4wdB9!ZRHY7B+k0#B0K0hl(AS2SeUqA?XsxW5y|(#-nYZ0kBcKunBAenUQ%}(Ki4R
z##TY00GoJpLsUs=S3t3RW}q(M$1I*6>J3UDax=j$fe#7*$@$+@Aaq)8rNIG(0OxXO
z=BwXK>}{e;H0^eNz|I$!lt9XSjEy3MdL3kV)6;Zkx^pMC6h)eWSqH{HN+|b$ZWKrA
znw$S6c$liZV`IRl5mw`(m<oRYe_23l-8zfofubO*+P#34+|8Z{8carav_F@DWIQk-
zwX?rd#5RAX<nW$wE&Zw#UY6Xn1D}8L9rf$Kk>Xx1mbm+duRv`%o#1+-Dz=Rqm--my
zYG%<>3X-pB=4M;)RPa-%un(i2P8XgYp83oen0mCKEc1&nc9TJMQjo7ghBvonwq6pY
z^VDg1h^EMIV7PTVMUYp*sZ3%V=vdg%xVrpNxwe7K=8U?r!+o^uTtspBwqUN(A+0_#
zBO$ZpYnd!V`p2G{O0gNiKQ0*fzZ)n_l$PB~WYHsjkpEBXzU&Zm^PxcRjN6#ppZOK8
zD;6T9!|_-K{Qp*gB~^g$bQ!{)4OX1`s8`OTKg-<kZ$|IO{_^y=ntu;sIt=~`y?AYZ
z#8K?fSXL&<;m3I!lbu_!*77#3G4+FWZ25(3$QJ9`K4KctDQlu;b)c#!lBJ9Sk?cCQ
zH&)$$41e)$j<b#`A6<4ITwqgt)6jL^`9f{~W39VgYL-vGp4kb{*vGmBBR;~m60Kp`
zmR+PTuBRD6uLak~E!+K6CT#{8moGFJY%E%ba4Xw?TXH1|l~w+|hJ0D01r&P`GEG9s
z=GT3}v+*|sc27eUjXQh;9^Qc*?WqH??$4DCEL$6#OdQA&K&$ZGU-&76yH@Fezq;dW
ziEU+tPo<Oy@T=obZGS(<aQ*iu#)vPTSc~KB8e_T%UNJ?iYvB{a)k$_KE6pnw59EQB
z(tV^Uf@;T0)l0Wi(QO~+4M{-vBub}s<>V_JM!qr46@r|Cu1<@8wE36f)}jVaTF*Ob
z^*&gm1B$!=zfWcwj6^{H#1u02*4@P)uQf#`+HY~(yEI+g4w+IG1D>yF><&8x0w{!M
z5n4rS=t+3X0$5~?Ve()K?65pr-2dHHId=gVAO@v&hth@Xf^ojml_Xa|s2n@9{GWK7
zAEtRcM9t=G)vMcMNX117ye6eE-m>3S2Egr{aUhT9olb$anb)91FvYe_zDRiY_gQX#
zN}9|9XyknEpBE|n2gE2L8Zd9ma_+|y{Q%<z4_?(G4+_TD`*?#p&{fCdncaEERc!ki
zZGBB3FT3<oa6tDTx2{*np8$hMlQIM*ktCkMl9YL9(FB1QU2#W*>`;_IoFLv=bTyej
z-!KAEvg39DPy#PxC^`~=0I7YjM@i(`6o>3NMD`0HI@gt}&BEqybe#DuBer)mTDw9I
zHvz__JkiL=TVc%t9wp<Qs1zRCvcW|0!5l*=4|R#1VrhQMX)<qwRvw&(;g>718?E4j
z`i1Q&2b5I(U1Bh~ck$Ife$=(_7OTXe{=A+%(dr7M1=Ppa;g2~P54na?3-S(4kjdX5
zBn_H6S^VMr^ooEk(ueF!;d*Tj?`aZhc|Z0El<0m$Kp%z!qfU#Ooc<+~jsxH4L+na<
z_d2>kyx|@U{Ox>>r?nisR8!zEy`nO<uL$?R*@I69B-Fq2?K2ko{gTr6XsD)8c))Sy
zX<1T8-v8P-%ho}4ZF0NI0?=>0+O|l<gDa0aX66u9BKuz#o};@R&M_F~hRZ3*`VV`_
z%pF4LSp3hUoa-XQaDJFkI)Mn~4afSdz+D}b)-336*yXPXSy$o_S<}>aQCmATl6UQ#
z&3sXHKRMLu(16~%Z&&Op+siUu9FvpDaXSS6WKn)_Y_k-o>_9WwV{15FS@5~Xv)rR2
zn$pi2pxl;|e>xf<d0(hbiBz-fHz9VV_ZbeAp5*VOl`?C5hKDr2*;51hUG(|J8o?GB
z0@lO+O%E=Q#(h}ha_jAFsx~jBq@A;x_Bw_^_Cs7YByy)>@C`D8LV|s-tE<rj2{)lV
z_yWW!j?KHor3#Y8Mvb;+FVA&=#pmPRUH*l*ueK!mMG&uEb%hdrlf&u!Zn38WaNGkA
z_x@+e;!DE={!<P(SiM^WaEwc+-G+8YV-2S#JJ^~?amefMdf`|viVn8HQxm?<KqmNT
zDf+AJjQS6z!&6daxswdA$tPoMX#GylGH1JV;?XS?<b2ueV+{N^LbQ>0+yQjtY>)nZ
z&AX~S`{f*CLS=GA6U!emlp-r|*O`%G2k}QpgpdjsKZ&p0ssaOUZ*!9>MtVBS>T)_O
z>hSRTZMS}lV?XJNQ5q&e7q#V8O0560h)g1wg4BiGk1SO05-bXv<x=CQx?KM=vl#p>
zfb=bsV(NZ(-5B%z8NU!5LXpG6wL&j(;>BoNTnHWTX8#%X@@On_Tvi*oquh@bOCN&w
z-Rbk36fPx0_v198+_hx*99>1EDq1Z(%c_~BS@5OW!{3P!IdSZd_!_fvn#HN_cyZI?
zXDV0JmHYiHr<d<e())t)M~8)34O~X8CBn{LC{=Ohj}Lc#cZmzzvju{P=;oSO|4#6T
z)`E7rN~GOyMfX*?V_!~x7!nGyuabN8DPq=+1yHO?6`$$`V0%8fxx|YP?h2*g0lW!R
zGNr~T!axn8+?eVu!{MOzf1`ksNE>4aJYWM;b-fJ&ZzOOH{=o1LWIFj}_D?sPd6|cY
z?!Nax5_DJen`7s9G9I8j{wFnR{*=)+i?YxuW6Dpb7$u%3)ax*OF+RM@#Hc)m-8!iT
z#1XFc!`=+-5=s~4F&NIO`C>ca_3w1nv!t&^0`~369cv}QWyHUCG1iCB5_LGDI)hal
z7k>fJ+EYp@zN1LY!>C4lb0BYEM0110?b-wHQKi18za&^l=M%-+<Q;m|C@pXz^D@oJ
zVbJ!gvTcW*hK2zp(+jJSp<Mi}=8w!x7upb?Pff~gC#=RMNYvsrPxu`L(>Rc&YWl*B
zR$a{j!b2kvXP1+m><MjUQJ71GjPjb)y6Zfjop3;`rNyI6Eceq&Pyk<PzigF7XX#RX
zSmBJ3=i0|5ZBC8OkO#}VDRMp8c_&3#u>fD&1U)EDJzGs~Ly|dCW_Xmo30I9=0Ibvi
zBG7Tif7n(y(YAj|f6rc>>z>Zv&<=8@<%--B8axa??`-VrFeqhQMo0_g3OLZj<B5~X
zmt*7Z4*}|@!jJzRXv{?TJ7F0n+pDFhv(@Gp3w8rfR*U!bDaLr2%~#mlub)y*4~dv=
zq_LkX?=7qp*5$T(?KK+KTk0G3ui{Y8HHCW48Xs29&D<+rMwo}b-1PfV6}hZ?@1y-z
zGVDQkc(bV3&a>!5xh#(^m<z@Me#L<r?pz)H^Y`e3JeBQV9i%O4I$zd7DF?+FkSX!!
zwH??~>4b!;piZuo{d{T4FR8X>l*;}G70})%#oXLI*V$MYQlNgH`d8RlpSGO4f-TY+
z{yNH{p7HNM4Y?C!$&wBEnEDRsvkD!OB|qzWWep`cTkSA31hz<nlwb01Qk`bErpp{}
zBmI;@dEAr0xD}6Tn2CXmV$2gT`!t7Umn3`G)GH--Q73O%s^tImr;|aN=p{*?H|3XE
zybZ0)!jfMD2}OSI65F@V@2H+SPH2!<><mYxBtgPe9gJJ43E<0d!EKz!!65AvL*Hs7
zwO2FV&s2AJ+i@Ds+rs}(768fwI8OLH*}##kLgkN@KT6=6Y`6vZd`+DD{&NU6<@MXn
ziXQ?CJRLRw7Pghg4}R{1-`fFsQ1q`v01X;JFFbCQ5OiFuSSB1Auz=6#pRQ=X7ihRf
z`ju0_V|vE%M$snTiO>v^M_?W_ukXU1l2Bm2l8|~-kU!Q=o7~L2+u%$O5n}sXj2Z7U
zb$$qZ_D}tW(%0xuBjw9o8s%}ZEjaUJbWZ}2M`g~=gDnlue%Q1{;?_fk0GOv4(}Cmp
z?}t(P)tbAMd+Q1f0T|PD46383oT7ozJ3oJtR`OG-V$if=F76IX^U@Yph0{tZ?;yz&
zN>ZCmX8Au13I6`IJ@O6T)xWt_FCCA5a_x>T%9Sc#uDwk?M^)4$Ogvt81(C~wfux81
zge_0?RHtKbxrlnd7|vtki)>X52#mp8OqL|OAu!hwtHB%$ucCS?VZmSg(Seya72}da
z6KZ(pmJ$5#&7+C^;<pNJaa68ezmvp12$-CopLnPM0c%uMd;Rl}ruM!mJO=X)MaL!N
z(~=popOK(l%Vi!V6pLDhItQ4FU%U?Y?t$e$5LLzdZ>9jJQ*YIDlPxreX*QC|6Y=(Y
zERU2t3REut!Ao#?b}uej<BF=#@I;}KD1IWgvqk^js&myy;#PH?HbNXfz`N^aXFEbH
z4ZmlVi^F^#T_D>IR_8|`^-Wk=<$06%L({fBK4fmndd0!vgXjJ7+JqCNf2Q@G>kJ$#
z#LUV)=0t{HbVAE=vKz(yMDtdTBRZ<ndUXp+WKdf!8puQ_GKgJ&;iC1?o%|7e1BYfM
z^IQ{;big%gmn@qEO7xc)56)MdI&p#aw+Nq@{FuHer`6JFt<EtYk>j6)Dq!>KoijcP
z^ezP@NKwh1yM3HI_OA_?+G}XN<(kW2Hb==O{VOZYy7cXpb=6jvx?9A`13Uwtz}{uH
zU+OEeT=zm(`<4eA3ZSE+)0kMc&8dOAn+b#gQ6NE=d&^X+F#`7kF9owD3wh?2XF$OY
ze|YjWT9wc?M^!i6gA}R;d~?%0`~PRBsJcj4J4Co~%|?ik%Whv4-s7=a(ad!HaQ~4d
z#C6UgTF|ZQVTl(Tn)@@a*g+yB@nxTMhr~5j#}jIl?YNQbvi1-|_F(<yid3G;(4RYK
z>*!1I1nM#MfbX_|e2--t7Lne#K+|gO)WfD{Wrr6kEKnfm;AW}ed~U}6Xm!=%tJrio
z76)zitd%%EkHuiTi60oNC8h=8i<Y5%eH135O@!kM(gNHkGQ?VSpNn{FB(?Q7Qj7y|
z`Kc)PucnRt?k#gF3R}VteZ=L|R%^dCN#SP^q2KjBp#e@LjQzd+Kdxs1<Tm}nx5pJx
zVskHUtH%t9{NUnC5wJba)!(=is$Tub_6%t!6L((ANRpZp`x+%t4NF$X*>+X0J_a4#
z3%|OH$~iN#?-EA<a<Ek*FP~4arD}+ZsC-fwwcnG=@%g!l<|?|iP~f>h=RQW3kF9gN
z$D#{@Mrz(y#Es20ZU-UND9UabZK&d@iS!B3!v$R}a_Ch4zNZ_6vF*X@)It;o17B@<
z-Ot6|cmLLLpZ@%zs7+UjV-Ts#X0^%1*^tZ|iL{7>#)k7QOo}OytxK9JDxw!__KE>@
zF}&kX-JAvW1}=aKAbODUTvUhyo%qG1ggp7?30aMnM0{ED8Y7{Q8z3$EjUnB`5OaAF
z=wMU%ypdwREQ!oJMQsF-D~^f14~gTE+GIHU7NQdUHXh3-p@8a=gd0eCUHF`!6!=Xk
zP_?DV_-HE10}4f9+#uuHoeaAu<UQ2??(}{LT{1?2efU*$7Ab{GIo<F-j|7}0g<IUK
zgS3Md);PoZ6vKLk0@a(5`YlZPGo%E!-)(oK9nNIlLpnmj&Q8ww@dbZDI7`>csq!$Q
z7B%jS;zA?eFJz|)EK3*gnIF763h3U`zp{$B4`s_};RmV^=_PJ8u7|}z<i%)=7T7w7
z+pDN&%tw7+!fCdN_)X{pLnT^Ixr_8p-w}c#!to$kb?(9aVFxjL<@AXKry=P%S0Mi)
z!c337wkFpzo}~W0xaC%_1}?ZM%mr0OWKhYnpM0n{;8gC+^)3KW0y-xPydZsC0&B&z
zVJ1wcTOjc_Q-V&aZxQd>50-==6b#2-Z6)7U*F9~{ny$w0jVvD@F^E3aT8(06Ud{M{
zaj>_W@b**FJ!_E8fY0R%#uEWqBTke`y&%FI(hW(Q7mTglzWsX|$<AEXx1JMkKhA4$
zSWmyno?Y{VI3UPg%W3b#;Tu>x(dVacoD}&>K~yVV-z-yc`O|(W_9+p?6r*Cl55Gdx
z^%TH-L{9aa*0Yvl#a;Ch)atjuwUK~0cShO?pt0XYnDS49lWud`ffD4rX{RG%0|!g<
zRfad`Cl{!gl6)%y(EzP5pWEeC80H&5d9P+>lM)XwuO%$Dm^JPs(zdotn&r_Uj-gDt
zwP8M0_P6+4BT_+bzt)^<6N_E2Trr=qx$jj0m&YMw=A_tjpPV``Th-st-r#RbuW-GC
z#oBVd0nEc@9*nq4MXetU!*iRcqcYHje6rFuhH0I~-J~P-A0b^6vFk}`P4_<(F;U(|
z5$CmL`5yx+M0s*DE79u61cWzWsqP~-e{HZuEO$qAx%FIJnIyiE{(j3P9pY@f+43H9
zn*wt3??H#&pM-$OxT-%%=QWFgZ*@Jko5~<dMsF6a%AzAf$KswdyP#O!r0(`zxY*T~
z_jvQjFE>Y}xAW;v=mUR&8*&m!bx2I#Nd5Ba(B?UeEx1Kre4S4Wt8RDg%zFW{NAa=c
zDG7M<m}15Q3r8ETj^x9=u##pzyO(kM09|K419Savg_1O>_-ZOAgbZDW`utIWA_kkW
z(on8Gb%*@?^>z@3-2Uca5+tu>om=u3_fvBt1+d_WwWEyJehb+?>FgLPSkD^i^%-xT
z%KSn<ZULlQcq2bN^r&yD)=&P;*QiOk%)k9=f12wJa`R3*yol5*ls6XlDGM~d+#^8x
zy_Z%+GyC-dYb3cAT9W5+@>}_|Jv>9f>3E3T60@$7`x;x1e@D~Ngmwqlm()hHRcHzB
zQ#44Lqac>g;h*ml^*@qX29bUiA&S-CkHP`|niLlDw`G&)US|^XzR!p0^07BU;{MV(
zvAfV-Sm+3us8^LAd``zBnC9c3?L}Ns!i&ez<?nh$?xkOm9s}TM+%6nw5T<hDE4YUK
zE-=3F$60ysxgfkax&cUCW>h+Q6%>tH6Pa{=df!nR;Lj55R%ZN=mvzT`Ih}Nuk`-`E
zYz1iwzu#W_T83|kh!$)pciLHiRmrjOPhos|cZewBRN|fw<bGVWekHvA+LM%=gaNO5
zYxnhDL}n;GkUuTv-yZ^FR^$yT<Uoo~P}!&%{wCV$s2(&d>T%IMs~}cv|3z~FRBz^=
zNEiuUv%fHk#pg>x0u=#Fg!?bf4<u?&y$7!jyVZRNFpRZcK3!RcrtnTD&ft49S1J&N
z4=7RR;$_&Vl4JSXoB0&sG$-~Hb{vF@V-Biy)~Pq&{X8_@y+)yt63t`Myn&*0brb1e
zcfQ&P`u4koQN-2w;7+#xox!2c&|Xck{U?rf8GhMhFNG+rAEMK!4}tOsv4nDIRStkZ
zmE@Du%}q6%*(PsuC$q_CxfmoV1gWe7w#aRQsl{f2z9nF8w4KA&&m3YK+WKvSzen%*
zo{lt6S*68g$Y3JX{Zav;5sFSo_AJ?Neq(-Nokh)_rCTG{fB7ZJlP@4mr$Wu4UqIsP
zgr)sCHLn@~SQtR0RZBcy#7(2*m#BmMw_l&q((Ip^A-P-`%umxaHC~3foPHU+47;=h
z!b(AHUu2Tn-&auWED96KiVPCz<y$4WK=SF;hQlqt6C)Dru;-Vf@9i88hu!Dv9Z(Qr
zW#;3K><Hq%C4{PPUg;KqSFE?dG2KvwW{ag3x&L4!%~gD{A+jnJ*<?ymzD3H4@n!7h
zr(~Pl)DPHE=9zosrPr2_UesaflV6JGkd?nL^Lnv<Q_5~%)YIaWavoO)!E?DP>*2i!
zN9=QIUXg!^xzOR5ui2m&-2&`BD7t^HQUiB}7ALXgsH8O5{PlILeulKS=N@u;t(hgk
z=Luu`xEK8iuEmBFs`dPXZMITPNYQyoqa;EwStY{5BMye?xAZ309h*r|Icp`oTN)vW
zmNAE?N>Eeg3ug}8mG9m;_#51xV&uJ<n0S=AyU$wdz<8wp<sV0Z2rGVJN>?<ht%(%1
zii7>O$j&ccp`crXWC#z4;xdK+KdNJ#b-q`#J#UApl$bv4w<g^dwlY6HB2?zyE_IFg
zr&*yTsXMJO%V<bPy>&fG-e(nEb696)W0n5j2L={m`@do5X?7LWR@LUoQFbYc?xPuR
z<N*1ZKZ7(Qx0RBFRnS@4>s3pwi5kD9K~4gMa8IW5-_|bv;ja7B+v#_Ok;6}TZsR3^
z&gxzVijFJtFrf9ke{CFf&uA0f)vplCu;SOahMbYc@1S9|1{J&!z}Wr^8W-yJiw!(`
z9=+=&QcbiQ53IM{`_KR*BK@~PiR3t3X2SGQGD}tBzE>Q6t$BR7UPn9cB5Qx-Umh*z
zhe99RfPjfC=>lylAD!&dI_02_0OqakCI<Ww`buH%DYzc&a+YpyN_55rdp9RN9H*Kg
zR+JYi8X6t$8xk|0J3=4l&Ht^AMp{MQ<`P7%cBS*|*b+Wr$_4vKcgDgGmzH?GGhd2x
zj4t2Dzc`j+#csdv?<`-~Gtm9ud4f{z66h{F^>8iv7wHmkk#rsX%Ggi*H6gdw)6&S>
z4;xI#yI#fJ5npNuSs9!^yi?V;4CI!{dv)n7(>%Q8WHl5PE=*N8K=KYJk01guwT-PC
z6kX+eIm?4SOt-B0naY8TLPpFRA)@T&F=2}!eKyTa<t5}THs_TQE>_`<58)z5abVR9
zH0EUk(_lK|dMBI&ouO7xaBBzVRi#)&NX(fnA&WR?IhW<y|15Xsw8SvdS(3Uf$UJ8C
zF45r-H#3r+BP`9ho2_({6n8hId63c(Wz4~cz4C>sjr-C0&2g=gSpK4z*Rvel2F9xl
z?)c|rp7Zj+r<3`7ZA%{$#H$#*@8o-k3aVmf!`zdo7!5q+ROoZdkwvk{rE~e!^~|uH
z+t>Xc9YC-(&vIr!KKZ<-b7}a0J;NIWW^_)w4{NC*>2Gee{fBpXXaKvCtsHOIC2lWw
zXNM8bSAW0mV+Et#ahD%T7S}%)L5~VY9D2Fg?VsrbfWw?#{r!|8w~op9@ekgY<gNR=
zOY_yr(jwo&tvsNyWXXg`<|AGUj3l6+lEp4(55AXPc8A!rb|j;XXhtA?l^)S9L+$8L
zStn&yXxqafbTh0=-IR)0$ji14#r*M(Czyab=QUE^o!?$3V9SfyV+BsvAU~?^&(YxY
z?gFoiYhMPK=bSWA>*GEn+}I5wS+^ha1yAJc>N1u%!haVB1ZwQx<u!;ZW^A4vl>Qb^
zuWeF;Ej_OtW~Meam#fKQS-Sega#<U7%NV&xM`2*SSM4%XDS5rk9vO+OxEp)3y|b`O
z3X)p*$O>1v%Y9%|5|Y$LBJmn0&j1{dClC1Ii9B2!#mcdD)m(%MyyI-)aEZGd@x!=s
zgT;@iuZ;#d+wV<?+Za)sjmJM0COIwARz-BZDE=Y+9qW!38Yr-0k&Q@H8l^_?Ef#7_
zHtr_=m5f{BZ`Dc|*GtN1KKR9O8YOtCRK2??8to_;AFC#j$TZ&L`qLRzWPx1<<Xz)0
zyNNl!DZLHIwiu(2UA91q<rINthn!r`s5a`XmTJk=FCcfMCAo!QBz_Rv6KZj6r>6t-
z9krK-<U!Lo<GcAYxzB&R8~eWD-pWj1Qn0pqJRW+nja@vlENW6&qgJJ`lr=QGT%kVt
z3db+$7!SMLc;rfV!;-auxs1=H%IbLc9RrQmO-`)q1jccH<^alT-^A_|$PZV1;{f>0
z=e?zQu|C9IeN4Cc6E<R-)PsTwrDeTS&eo5s=mRw$Jp+kgyN87t$)0UMvdSUefc+Bp
z9oU8z>{G)$ME(Bd^&HGjcOLWs<8ROip_P1%BGE#5y1+gwmGcSRhM~<e$-vCcz{~xP
z(t5qaMmlF4Vt=4ty1L+#b-`ThieZjN%VuT{>@mTaCNM@vV)c3j_Jg8baEDnWVARp|
zT!cyS-+%T;MUr^`g&vwaFd|CCY&(MgY6CA;1xzXSP0l!RM&L3J|HhQ+AlxU(6o4!&
zHJHO`3GEt@3G4qv_K;8*%Lo6bDzbxFm)8-2=943altISuG)T{BG%0PKpkiA{ntB~>
zXrJryFAg}k|E>A<>=!u!p5N%z5b2IB2eg2mNBpE)x0!hHRjbsZ0Y78^*TimVk;}vx
z;Gq{qcOgNACe0=(*=9EA6D-mTsmcYuDa$XMgR+qFXt<qCtsyd<s9<as0ZIk|#Lb4b
z#Y3_liH({B?!;$UP$h1T!d=~%Xw$QK56Sqh@sAELKE=9)s`CDv!|>fn$cSJq<n)ko
zswlKz-ijNJYy{kvAx^|CIik6FX6!fEiD`o4U$b~7^~;T|G4{%;i&su;%^%N=^Y@Am
zsCKlg-^T|1w^1;i3X|om06|NiUOM&L|B+z4B4n!TGL#t3XWe?yk}QVbu5d%&=@P}R
zKbl?Jg7!jW`R)9Nq9mgIDnKd08vzdzrKj!Z)2GS5m-Pzg(YXDPq_5StIn>^$*}~Jg
zSB~~+>Nsc!H)o_$mU8io#I!%oa_+d8=c=ahC}>i~Pza2ZKYo8P^PCZ%Aju(8$2Z0#
z8M_`D%@?;}4Ak;Gcgo|g`;e+mowj)?EK-F1?&yzL%&W#5dwv@rmgns5R6d78fsk5u
z!Pzg45kY|@o!|4aL*4Gpff_7-oj<ADCjrGjoPnn6y)}<#a(!Fn*o~I3jy9Y+qXj;B
zAG|!DzYNcC<S+N4<a2$m(p#}5&+#Cifh)emaEv)F{Cr+9rVMMBQ%)14Gs!rSvqOmw
zzB9h-i&NV4t`OyOA4bq?vMUJIkFrQ9{X<dKQ76$5a=aNHm=8dE9gBFidm$26$#;Yt
z2wJ14n^?EQ`G@5=xs7t`nMR0Qg06d8m}FYkg0G;9Aa2|x?pkT?nhTOF(#>&ul9>B#
zl9cc_5_S@asHMpti~0bYIdyOyOA)Wb?;pnQZQ!+wlupze9YdF})02VNQWc)lq_31^
z_`S&w$Fn~TjxjIx<0KB~R9S}9qJ1OUhV$&y_U!O^e(1j0nkXG|BB{e40_*esr13tW
z1C7V2h3L+T4F;)`7`v%Mfp6b@rm0^>PNk|j>89Xn<fPIf@{+mS|MN*MuDW4w!c{7D
z;V`21M!c|nKFh&fl*}RIqu}yDzBm7`(f9bf(m!x`0&nTB1z*~(%HmMcZkIvlCVuMf
z`7z#jB4csKJ?8HQj;cFGA6^GOraQJ2*O^`3>uz6jCsCb1sOAZg767AaI0pqC;XY;7
zK7VyoPR1q8eCy1OYO_1C{Q&b)pNsrZ_La4`ym5~fYA?3?S1ePEGb^oC18(<P+g}ju
zZv-Jj@b>1{A1bC@A#ZXxLKisZy)WJvP-9gQh|Bdn;jEWdX&UwK-CFoQ0ryhErXO(G
zRkPBke{5enUJyCZp2-HPgC3s5uR$kk-_*(4!Esa~{7@@?(5?N7br+x&B4*>l@fZjG
zykx@@uca58AcN<I7%*93W6Amp%eWSW6mB%9p0nF<Gfa5$L47~#)AcKf->HC3lGeu;
zra1Af<K>AV5Mih{riC}nX3f)Oyl)bkaIt7Xr6=hDIkb1UNlqwOt(S5`$U4hGj@n(z
zd39X_BVvjLQtYq_a?(2=qi|}Q0(#$D{--LOG;POf3Rp&V-%HwWtt>3CzY8X({LsZe
zI))<u(TxC(m{!!y(>1k=H2k=+WdePVCeZ;LPht1U{@J&MG6_)U?h0MT5WosRr$)sJ
z#&9Fg#$Cp|L!}L-4McU2|DYbVxri2q^Pml>3&($rDdVTfl2Ih>S^qn88o&8z^7S_`
zpD*+PDdrD_!cCny#|{~W3<dHA_b{Esd;XhBp7-L<UK^TZ)ekZx^4qk^V%`^;KY&L{
z*}Ny8glZTu*0mVzyX!Z|Lr?!Tee}rx*|VU1v-pZE*BMdCnO~@>=u5!>3!%I>Bq~b8
zmIR_LKMV!VaL-b`%zS5axE)X<e%YV7N4QR}hC?;+1ta-RX&>m6zlb3<*W^qieLq2=
zmVf$1=veB~BS%g>q#+Wc-O`O+?9@S$mu~=c=-y918^1+JQ5>vA`NcoQ6Z9!fd%Uu>
zT%Lx$hVS19Egd_go%8+eSdG&e*0Wwqn{R}R?4Jr7H(9mE7}tz@*zT`S{ZjT`2w&UQ
zOxcIaCFQNZrcwHHcTOSZ<JQEwJFV$^*}<xanRp2~xDjVV-{fZGWD$|qnUk|bm|&w}
zE<hQ(BfLvQF4O#vR_o*l{xRey7y|qYN7BX0enV~lD*;OGekDP?I0#8eQfO~h!hY!(
z*Q@HUAmbl!q?{f1@wjGaDM<xVc8rn>yj$y>DMJgBGTp!LP0m~>@g_~0hqGqQvi|%h
z$p*ZQdsi&Qw9x2QNkK0fk|@KG7Xa@gk*(S620N@YNcmNC4s2}xv=;EbHlgom4aw_I
zsagbTVSKk5=MznW0w+Iq^MDe!-(shU`}tI*=CbN(>)}-$Blw5XVJ^)U?m&%3h(D-4
z#>M@Bih_y!KF>`i0Zq|7$TwT7y~RT#oS45?>~lqB7BH>gZ9hY05T>C9Qfq8#?>>xl
zOUPk(hfRuHRa3O_L+2nRp2*F`J8kN;v)0hY^9Am2!@XW@uun+lB*I>v_mThB^M_|3
z1$Bq-Hw9KMXI^x{Zk-Lk1#TM<b|APRjB_$q!S?00KsIv-+Zg(0wEnO|Q!yCaq$PL1
zWtnPo#xc?X&ek454D`Sz)i=;SzD>abK{ha*h;=nn0bx*ak8GNBJTH2j_bR(Ca^~zA
zXh*Zf);1mU^Nb!AT8H}DlnZj=3oyirLA^q=9*`uEw^si1NbIbs7ZC#}Y_YD(qQY#R
zn>Sd|L?lsl-t$<DDz_hOq@Z{F@H)L&NYS@wK#Bgd;&ceedZh8no;@uICZxsg{eJmh
ziWTMel8+PPXQmWpD!IZxJ-v#+AmsBmGsVI;Y{gCh`@bzNQ(+|NlC|#x3_MSe$1d-S
zr>C7dy0;BGVSD#`8}M5QgxY&2A^UJfRosR}sOjf@Z#Ui^XfurDveHPQ#(F^X(c=PE
zh6DMNVcYxp5MSbs_FSO5-H{5h;JTNEx9xYorz)+ny<P}>IX}u3s<!JAwUkw-3h-Q?
ztg>16i~riv92^7CX#0cR1&|MY<qc=e{w9c<C8@9}rN$*b^!g4tUYhhkvUvv(Pm8&M
zJQybWy9YW5dUNxu(toJbc>2rbbTaP)m<D-wmf`mwVXt1ftU!gL5a<p{U`WFt!w*t{
zHh@n2t5<+eR7B)QBzl8+obY~_vUnMLi)5rF_pUZe25TXHpJ~@C;at@w2U<}y2~se!
zc|UVYEIF6=$<Hymxu_IUvKH>ySlt@!A2@egugDB3)&yF(ds&2zD1gJXhhc`pbT{38
zG!t@}{DT?+b(@)a84t9aLeL&|v4~9~i9=^e;2$j#<IkWNS$H?L0%CS}KNJ7Th7$Vw
zGk1l#f(VAx6SM1D#NP=120+7GHAOj8`_UvyrohPmK<l2u$~y0ng~c>?ZwQf1BLerl
z2NY-R?JpY{Pt>Z6&75YJF?)RbDsA^yhRA$--+D#wg_9;?A0;4rY1vrbRwOJ)g9U!^
zzIByniHhGRvY4D*`x@_Z7)Uzs?xQ<~_nhI>@(lJ%IUpZ%NM5z0vu<>QwZGa-+cfZv
zC3ecrqVmK<H4E3F;pxDs--BkiqqrM8HZBE!8OI6R1h$&(Mpw`+2F8`4UO))dRBx6+
zsp;%ym@Dy_KoSD&PvAC-q>1XkBIg@gix>f8WMYqC$D7Hm@ENn7Pt{simGVCSX&)Eo
z`S)E&=4GyVP~E!6=lE%*2tXOXO+-nUAB*--XYbgfLkytl=QcEpTG-K#x^BBqoF%j2
z`!$SU>9AYR8ar{@9&8`8{|4qLPemzP(XT7<C)L!{1RFsHO@s9_F}=U`M9t1u_gN8>
z4kMmE@Q4i4UtEqGkCL#8<E1(maKFq`rW$8%tn9R{|7euh2$&noI`+7E5miD~{gEzO
zrT7}EO%}U`HcPc$49Nz%iweCZ-6{)7U#W(t6Ti_iDG5>qH#q4j|3=Mmp<Lvi>Bn8+
z)%qpq4gZdPnzV_Ay@1emG95Mt-Yqa=NuZE?bfn7vZA>GQ$s|F;M31digXZCauNO|^
z5_$olQnv`Qs0B<!!r(WR9?+cdy-KuPhGVxte6R7pAML&Zs@N1fsSMC*OC}mtmS^x{
zk?-mUs6mT8_!5)|7rNj_VPR=SO9`Wk6nqg9U~T#lDx+P3u01%rDp0cKv*O?}$}_Mt
z5nl9o_Uxy6<)U)oooM(T(9vN`b!|6_a$?nCTERR?7fxw5(9B{{jvem4J{6#J;d_H{
zCeZ9Ue|r6Ul6X+x_L4ULfRXt~?L47QAE>PQ=*yOwCpiB!+Z}S6YMrI+UVSl0<Bxgt
zn8(`W`5hMOkqwd(xvuKrZ^6^ULq~NJ=x=w`CXtMU8SSw&ll%dfu?~B0woAS#^hk1Z
zDxL^+!VP?l`DGNVD0};c&!xom(ybuS#NW5LFt%F*LlTqhB#<7{3pK7ws;*BHJ)63V
z*0nYa=NickRX3Q=kCr4GL!?w*^K&3gYM_k#40ll5(Pp7j43YTd0dr(Z6p)~Xj7IoV
zA{c|3JYGs->n)93T6YN5EC~-1^>r>HR>WW!bC98cQw9=PGih*YgN<)VVB-eHMG}t@
zFp=-Z@8_8I%i+C5JqZ<~GaHuKv1D1|HLMiAv-C70Y#ZEoF#CRnbmsEFs1|3q?403-
zuU^3CZko}@-8`5K<5_DwQf*OWsn@0!>oK2i&-IuE3tY}73nM>B)fL9_PD9zpNQa0v
za)69NTTIDchVrR|i>YjoA#7A+Qmry?sLgX9ia4s#Zdbpn_SmU<k3T7MRGTSDq$ip8
zeCDoBRl-`8B3|e*gmIyDzs$zjovbO?c&s><mn&IOPgXTGs2h#d0I*-jb=>{(+v5%-
zd4^ofv(;9Yj~D6JNl@p>d>F1emU8&V)9LVdFI{@@pB45T(2XhgM6+F?=#tvH&DDz9
z8J>TkMX$H%Ro1y$qm<DXf4bVvEi$Zg_kl!a>Fb#f8TmcjbFL$#ZDPPlFD`*5eeL)~
z`NO<!fdtLwxL9s4@#K~3*7Co6Q9IYR*8@Bd_XVRc_EI6(*ty+FMkTRho-79Jb+vS5
z|9p-m%HyL58uqbif5_v<2B9cdhplK)lJ*jV8jc0vqi<M7os%uiX5PhoK~pnIq;T_!
zqp#|PnNd>ciicS3^!cSSs{Dn-q!?;^8tDpM#EH#wS;Mx92-*t@2VT7kx8^RK<`2KS
z;`0tlYvzRp)9|HeD6B+C+*>~bP~7s2H+c@<NAWAzcdYu}aO&XhK|`<uj_W1%x}K6F
z(&<fd9i``*EFEEAhzU3_U9wwP5vaW)4sk(w8F=w`K<M2L3%5(7%ZTEMtFy_J$^G`9
z188Z==&G$JDepjnd8~m&js&3nd{Y^K)+(lCtD||0e<S<%7hvV-4EJ{E3;_`G({K!b
zpGcnr#Oo72lsl1J-RS^V1Yn-8GEZeiqi7B{ATp@g?yXcFRqDH<$=kK667^u;DU9W{
zeZceMS!=hb3SyuGZ+J0A%;TE$c(T|$UiC-ZcZ1p55y!UmBOJ;^oFm43Jg>;E5h?ZA
z39)dOCp3}%d(Lilry8{akTz2uuO(lZdAd8RH(kBLU#h#Tbq<eSt9L6bq75-!s@1~|
zzW%{r13|~z->Tx(Nd9>$h{CL71X(2Pxwcd(T=@`Yi{Wq^2WMWpTyuT2a$HZd(zPNo
z_MvPDz_#0=T?T!ws3H{SrcocO0sz60x!^eQWoj3b86w26r9fY-7<(j%>RM&TWn})N
z@B@OmGlj#OdA@+|d29fMj-5AODQ>T4`qgdQ!p{)+u4zNev}#Ks%{P<RVo~x#4S2Nd
zM`W)E775Sh@YA2xPF<ew_e=GX`c-`(lWfl_T0}*Hs<!se+v4QnW-=x^fm*VJa{GRV
zz{QKRK6RVki_oiZfpxDQLDPd0_R6|-Jl@XUxNDL9)$iMSS68AW+s-JeQN>5N&`&To
z-8b;8%2&Q8EbgM)8#O4Oj@4x^bUxpgR(6p89q(69HWgK|yw?6T8VX<|f8_>{4^l8N
z8<r1O@a;7<K?}sd^k<HL%Nml>26(%ZfRA*7B%1Rap`DB=AB+>EhAK1VVsRAW8U$vQ
zRvkVdfdj$_82&G+-ZH2Swu{<@;H9`1cPLh%h2rj3D8=0=?i$?P-JwvtxVt+93+_&E
z*OTY_-g9Qo{|sa%x9zppz81Mh_@FeAKNqIsv|Rge()w_rV*A$DR}Q+R4`e3!<l+Ln
zKPJEbrH-SYR%?X1u&)eXH^m5UC}249E4VOSr*F;LqcZtL7R#7j^MEpqiIDIXvg6pV
ziz=QJp(mo7PdLs3GGY<)B~Xtb|1nLf=N0@K!Ea(RHp^tWl;dBDWNY!!F}v|HUE7oj
zerzP=Ivl;V$>>8D(3jVhyXKswHLWh+bRXMudPzfe+E)fN2m~je`x%ZMrj(^=wf3t*
zSrSsUGB>ae?31S1sN3RHhw6oqUdKFC^<>?-)^25WGT%Pj)%uxOTWgJXp-B1YM&mpO
z2S0SMZ&ELAs#M;wTr-26LwS&D<Pu7O<a{B_n|jG_EwjvimyCevv;E*=f8Xp$xO=!R
zfr`Hyx$|h4!3JYG;2M78{6oWoU=WLUz9Zlt44lg)0V-+T%(uobijQ-pvlSlKW;BTb
z>xWskL~Op>NU8TSAAkwUC_b~HABj{degh<HegJ;3F2xk<?ofnvb$r{%Mf(uab9R@A
zVQnRKx(t31C)IB8?0_r!aNJi4@hpZOKr2hWzQc`D8(CH@l(B*5zFqw>OVo8*9nz%j
za!l_S3~Qp(sfFs`sFw1pP`AguRHP~0M+C++LLNsG`xj%&kU^6V_k2^F(+8afD)aU0
zxv_oycHc9-N9BIq3Vm_A%QB!wf7fotejm*3^H?_nBq6w4l)WGPNMWA)BSPwBvIwbI
zpUxjCP*FMV3Cib6rcsGQSd?Sy_?k9ns?yd9qk{I5(^jX?=H)Km+QB{&8z1$07@Jks
z9h2EMz07TSN>LiJKv|gYVwq*jGCR7j`TJmDqViW?ZG((Sbp&&tGU^U1T1_B_e0!=@
z4uek4=cDbx6l6lYD86QMQ4G{y7E#Ha>B>Ren)~cul?tvNnt`F{jlo?`*HJ>V8`?J{
zU7E<aa+wM1^qcQXv9Sg|#TE6Vu%;xHz3f$2RP$iiYtQ%1R|+Bvz1a1;M+XA!*ooS!
z?)J=EdBld_S?Goas?k3MRle8yX;XZR0`WVVr2WV5fmXQ_ui3KBBDsTHW;58=9VWpH
z6D1X}QW8m8_}|9`hGF8Krnxuu&aeF|>SznBs#^y2p!##S{}8qBvvy5h$q9+{!17ik
zTZak3Q;oA2DNRG27ox1VHQn>72y2D%IkXfF*GOnRMZPCN5<X@@+O+RC`1Q`VMPylK
zhipQ|{i~hS@F$E4%qg@J;oikN-(!$P2ygL%f&@->uBYJ8EH~^pgcTZ$FQlE~za9^|
z#~%_Gl}b5~UJ1gsQroIRe?aKJEA%=rcPiQklsn?ub=;~=lj>EKWxiC~8Gi`GNu$_9
z!l;&ndh77MT2gnz?Y;cCFfJ$8p%X;D!0s0#4&$fs!H2)ZMffN6YIy|<M80SYYtc4K
z5cceqB7Zq7n-XsJohJ953*QW$hGqMCBSqAH=>KsUnE_DkgZxLx1Hju8xqeZuvy&gy
zsd-vVKC1bzzz+1;_%taW={2$lZm&q<!XC&Vh7_Io79Y-kQ-;=`295Np+NB&t28-Nh
zp5~tvd4FfEVjY+Iv@#oNbgMuLbYwx+fwuUSnLGOB<Vi(pd!L}KszOhrcoStwt$VsB
zN-<B_s?i!^8{|U<QVR@aR=tsF_fn~xCpwJa>v|E%Gz+CJaENC>7;k4P=;z*6a~m8I
zaznN|Oxj`9Lu5e*kkVl|juKY3jUmT`;r-c;Y4l@^=s8{|u|N)f)p)~6SC+XF`E>qr
z?8?bfas1jw*RS>nlxtKH_%Daq>5ndcFd|JQ9G9eng_;hyh?02snLo0kBFeP!v?zK=
z&x?P6lNtEPx7}i|p&3TzebT->T08G2d-;}~UuLHS;)PG!VJL!G8jJ%-cbN-laD>bF
zccPYYH7LjkfBi_*DV_Q06hM91tCj63&Btey8t^QnqpteY?L)M{Ust&H^dj&3HoVgt
zIj3%{w$ceC#+$DIjbUui&bBtn@MGa^JVLwW!7wUKEp2u=!7AyU7btbH-n~`dBakHV
zhP0Pl_{a7i&h52MTfg20Un#&!|C)b?@pO{^spQraO~CA@$qaUC&v{c%G|cw~X18*&
zS*>io<^jvq)Z9*lRPSgNG#{ZmNd_jW>`U+Bbqbh8DC(Q<$~cl#seZRgT}mZ#!070w
zjW#)QB<XK#R&`-Q&;bXUFQ1Ez@9sz|CR<7PK-Z|ee0lp1PG$1XEuFK3e#|+#DAv4~
zo23l2T=&M?ygb73-JC=(<8ICv9$`JGRM))WVG7ve;qcvRzOK30S6|Y?y8hN$q6Yvd
z5Nc!!!lz_J$LIeMQ4?7xNE&O;HncUE|Lk{Np$(M%CL5{>kgU}2>Y<w)Y1k+^0Bc}e
zr&QHdw*6UQeBL7+<F+!aiq85sh`z=+o2&nO&NL#4We|Y%rHpV#{8v1}=V<^7@j^pg
z<4C=iRI-h+BSINwc$p}w$s<2US2f8j;_v~(Jp-PSdg}q{F^mE~G)zk(AiEcGA=$8S
zKmdcGyGc<Bjqez5bURPJyg7B6(u}FrK8H>m4Q9D16q2q+GFjm#zG^-Jg%DGXKb9wi
z401&{z#EiJ!*)*2Z{>JdSCC&0GJaYmri$U<pBIn<o=aUgIPd;MkLev#w>-3Hn$hxA
zZ<~Sj_U*XWv)VkF2vibi`0<nfq}H$k_<ZLYcA`woIKeFskT<HlJh#>7Qc<UkCgP;i
zSowY2#f=B)`3$)+w(e`!Y#h|bJcf#pPe?*Z-|L~e3PAjezpPf`pCVX!VJLNhryCIl
zJqgEs3kg|VUcdza`7x+SOZUV`&!!|(Z))Vnb)5TEeB+p!;R1sW<uCnT7DybW0ah!@
zgSeI-_VE#M%diNS|A@`#yt=zQOfCaK_lM2+Wh`p)f}9bI<pSaz;~_$iC({?%pGv;R
zQZ_R`TYC!~C?o3uuk45>-Sm1!F(-wJeoXE^lmC(TH_I4fNbg1b3VEe?Fm21Dxa*<8
z<*$#lD?dTNDC(4^#wdvL676?EM$#2}z(H6Jmuf9QY6qcttUArfg}7T(e|1R^Ax0)j
zDlEB4{RnjHz2|m`bDu-!`CSb)8*c2D7BchGL5~^Xm`=G*xucJY(=k_JqqhDXR>)r>
z;(>dtJ{xbqOUb}fF#npXMXDhw=kbzp;fMB<AR>&jg<-Qw4PzVvT*Me=2POIUPq{WF
z*6G|RB(^wztU?eP*F8_V8+geIA9Rt!-d9@cDsw`kOe1Y<w8X|NHC#4BBbG&S&Zgk9
z*CM6Tdj0;|^H}=GEr|#YlJAd+eKSRRJa2>8;kX^z@j;pkU2YJ(tv2%c`zq&cet-BP
zOK_x)Bpqy7-N^zy%v7GUo1~j`8*Od%Ut7f`xV#Mg7n@#s8I-C2+l{~bks7zkrz*#?
z(ig0k95es*ku{dd%)?TrJUUcY8pAO9)xt!}7U-@9b5GS>?~m6o3ib@P>HT`g_$Q4n
z`~koaxvp6gjo5wyNUF45-sF06+_YR6=N5Be-G9hCUf@iBR^l>9vi|nXtKnQt=_5Rr
z67HfRY7%^uEMobWRF2u$g|qEE6$?In_@|54`#_!q@h1nG&)p0&MJPA{G^PM%M;!x)
z988b<_@>f{*WoZnxFQnx0Xmdj(?>g@Os;&*POIj2Vm1Kmj8ua*#V<c)D@QJ{ZZ&JI
zFS@=F=Yi=;SynIcCOWw0m!R<PhpF<@q&7P3d&OwTc%arYPbgPYF_DK;_q>k?6~d^s
zKa0`O)+U~iZ)eH4N&(`3w3(y<037e1rGq-q=}AlQfnh}A3@H1+ZlT`4!DR{@`O588
zW+$GROuCYcZ*@3(zd%))Q>itimApi*#r^NGoD0l`ZVk!OoKo|8B(a+ve-*j0%r-*<
zmYR=m>tsTcgcLL0lkC=NbG#BO?XH2>7k2Bc&fTI8#}Y0B8k^2YSvsCerifs~SQRwy
zN`;`IE4Np$UVEY7G1T0oxE&p9cUyROu^f0%>a^B|J#=N$25Ee>t_bbAc)4cAU0ym+
zT`LapM)=IogB9}Ml>eIB`)gyhR}a8I1@8trVzfe_??VJLrYQKhTXtA0`($9Vqay^s
zpqO2ATQh|^R-xXS{-<+ITdQv6jx{*8&A)1u+Nm-nP(CfI23ywvtanzH-+^-#5&Xlk
z+J@hU84M|g1vOsi0MzM*&sWNgwYs%gHNJaKonOuy`tP5tQ~0*}ahP7(>J3MzR{ps+
zUdF04i5x8ipkk6xqQdB4=25AVFpn#<Qovin(x)-QYzk*XDV6-ID){IyWXLe=7KNSg
zZ&>!CwR1=|FhSD<j`2#_4+w90*<@uha^@rXQ9c7=B<N9^#;Di%cJA`K@w5qK^;qOC
zhw^#yPvvOs@XjUZ9v?0&ytmb(Ixl7v@8Vb^7?E}XJ#4Qu)bmDz$RlPEJ7Zfj0wT;S
zK`RN0;uo@d1yA-wuVPyoM_=^4weI!!**uQM8fdH2zEbLOCzr=)^aHV{sxKYv%Grzn
z;=aG>^+sh*I;mNFKp@I^{*fQdM26{4td<|=B%nIoU6n{TgJj)cjTc6&4&jTxr89|~
z+f`%b>0dJRzqekN1IuXWaVj=j&H9;$3wGyRce$43Hi!49%tLOZ5$l=FbBK4Ne_h#B
zK=cuE=Kew54BxUZ+Dyq8H#9yB5gimk5W!Q=Wckaou}r^_JnqjlAG`XPRv{NU8^3bG
zgy-vq0k>Bu{a1h@3w<2dI5=183Y+uIzmsb}_zG0C@`_;#zgE=DE;11HBi%8rMat((
z=8u!RO)^9EkGk&w>EQ97)Cm%RLZjvC_cJE#*KO*-{QRKwj;2Z#zd37tRa?@`3d@7G
zDW7Z7^a16_6qaj(mdj=CA%E!>K|^qd*ZdpC_^pqJkO>@CqC}U8nN}M(roHrRg#D%J
z$Etfl>u$YpEub)ARx4g@#K1oTChEh=(PFMIIv4jBUw@=h0i=)r*o2;$VcJ6;wVvL^
ze0|A7F}8VpW4!VL;8kdKu{SM250{Vcq&GvpBij;@7Fu*#q;TC=zVp2T*Wn30R(Z~)
zJ5GwX3;SBSi}NJk_nBz-6aSROY=1{}cFJ4HJcIl<=@?sH>anCQq;0#1o@te6{^-Jn
zpl}CGrrk|vf-IM)EgPS+*5BJg%sSd|nIBR_<@*z8@HdgOGZbA~%`}(MpJ~WKrmCf&
zxDid}f4{SH^LB;~FP$BIaz7GxFqjt>-fg>7o{!Vt{Wg)hw>}(_JP=q6JE&XYxjUR<
zvbMB-?}N=IenQ;)sNHDR@$D)E%}DA_L9caTnNRrh#9~2b4m6|YBo6<Of@foLq{dfC
z&xZSSx}QUI2$of+VZG|ztW-f;`O8o>m$Hs0Dv9$j<z4?zr6r3EYd+`0jni#8R<ut$
z9Z2@LQeWODdX%Bh0+x<x!UO%YW@?Su{ox9}w<`j#yCw5$5ezqX(xGmQ3-j$S_GRPQ
zs!oAW4;?7SEsr24^vuN`siawztJ0aC)!&?4f2;$_pRQ0obPgXqYx}%Q{-4o3fqZ6I
zz~dRxv;wW#u`5`dn90b7n#nzsm@%>syNUh0{N2i?YhrvaeJ)l}AC7}?6d7U<lPvq7
zb8D1u5SUV8sB3SC+_znJ@U91m9Blq}wz|qKmgDnu^b2GzlTG8jf=eOl2k(y`h63OX
z3a*3W`bYtgrli1yp+NTVBi+OijzEEf4Na@~Lwa>2)mFv;#}5x%kYG-GNhwylCDMt!
zbRkkS-!&-8bDI!L0_w-~1R*9I&|k($ao_lF{)u`srC(NwLHo)R;C&=YbUSQfjcGX=
zFH#}g;`~RxJ&8#{XzOi0^?07uB2JXtkk~=tyW<yY0?-F=TwZ%s^0-LLF|E2wD_sO5
z#x;Y&$9&W)Y%KmQ-iCO=GI}=M_`<KXf9OqjTdHZ0^yYCwQOF+VQ<qCw_MX;zrSdRF
zB@AB~3Kb!LsP}Fveg38@7VSWpxfn9VyNJOvQZ@MC6>{Ft2XbQ6$Zz33x=B_f;RMCA
zPrToM*VTF;@HShCS>>!x+P-<mY}9#xQY!S!^`B6|Z*K+o%2(QX8V0Hff2Zt>ptFwW
zGgj9xSg+fWCyB;<W7nw0;qv<3g8ci4>CT~wq&9csnwQppMxCN0>jT?Wi_4Vdzu^(m
zQY+=^Z@yM{Dz)C8sa!C#<<$1CE#vtNq&F2^yvu2tNVNa5@=x$n{#C9p(=l<e0TQfP
zi-+*J*2kP-nJh#Rp%7n?wAonH##5)H@O}SKcKi;TQ&Yl+S(Zxv+|TaC2dS0ooP@EU
zE=$LU*>w#(@8*AhxxPO5@q1sIOW_1D-)&wY5x1f018UZO+fuH_K9SW<cP>apt4jdb
z$-21DoMtz+;>>JQ_hAPv>!ytysi&Vei2HA$cYtp#xl1fTC$A5Nqc?KmBvs7bGblD8
z_*2UP-5aMk$)YmJ#PvIB_pkI$O~COU8BEvfe&%)tu4d?gC=8T9VrrJu`<(y4#CP_x
z!f<l0f6uX*1j+yZjY>l!d5nWH<ayrQ--ak>A-+OMXgAy}h;blr$;3z_usX-Y4t=DS
z)KkJquvQLNJs$yGX!6+WFW0EG5gL;InNyz6$M05J*{y1spOuG3Z2a{Jc9&02Z(ZH*
zL)4ZF@80SqsQDVr*H^t`;3|q)^$FbLf2ZgzKJiVL>h=}8E~SoL9c<PJ|Exxx1rMA6
z{8T%5iLKr2;9wVi<HG%b%n42)Pq(7Ec+b|sRY2FMS2?d`<e=*_#*BlXzr*Nbk2k!P
zgrDf!TYqBXGudl&;m-Hv(@#jxr<;1pO?So*qjGM~4wn2=p#r__P{Y+-h}U1ks;iJp
z+t1a9MHZp2wr&>$Iq3Qq)Nq&U9-ID1Y8kfUOD|AzSEJ@xiz}ay>r=l-={<jx0<`84
z<QwfVuYO>U?g09HFqnnX@bF14hqZ2tr)AC@A}KtVdNkNNUmX3`HXj6@sv-R|8RWh%
zH9!udpx2#X0_*AAH`P+Q^qow3yXLp2yjZ5{=3aZmxR0vtw-m)#`l5O%2)F?IK7~HT
zpD{Mn3LRlzg%OugrYVHs^T_=a>uKX8$?~WG6dxNz$beL_f>5#>e<B*-c^bV1nma>D
z7i-&Jk3azzY}zN46Re;kOHHhK@%v+aC{lrtjviC&+b~7<lsjGkVm^j@oRa>q2%<-H
zxa*v<Bq?kMr4hpDe5#SV^emYu;_FC`LdI8=<-}C^hhEF8Qps9I;Wi0DhXDbx46Eo4
z!Y*s+C>x3TVBB(k!3CF|b=i|;#_I3P7%d5vX`b#9vT|g5MK0ODl%hV4_SN?8<l8jI
z@E36y{ki^qCCDfyxI@j7{-jq%dKc)NI~($h2?8QjpGH@K^f`u^;L%>;I_>p$cZL|(
z8Cb3TJ4gv{lg<(UT6Qx;Z!aQvz$owCPySXjvrk$+TdwklEH{(_d0f&XtnaP@-YcNB
zLA)ydUzF~@|9ZGwka*g<UEX+6={da##FXoMV;wVgq&OTTx#nj5f9)Yi&j*q29xaVk
z8s5oTeK9R~FMVLWVqLp-jVCb8)6$zvUn_ZPIo36n>YNjIszCL4wK|nDDGGH(OEKrS
zx+aUi&fFi|yJjj72WjLL!xWl)!5>)y7E4^bd7WA2J}drxD7<r58&m*3r2TQ&8ALED
zJMo3aAEF6=0Et3vaJicQC@$FtVU6*8<}|w|pL}jo{yciUJ)KKwtW7j=5j$(){r+*?
z`ZO=H;7K572EjN74I<WCX)=BWRN16=T5c}ODl=ni6(Pj%?59zkM3nt~-(_kXJUHz1
zhBOpteD3g`yWh~(%^d^t-|>yQVPI5;hdpAx5#x{NlB?c>t?OPq;)78MmN_<l>0?LV
zqv5XDbOE~;)?$~*cpQr=Z(BNRu?-l6Ycf(P^~rclA>7-s%s-M4SiOvF(|eSMwd##_
zke*vm7r}<y@{Jq0tGyYf_~W<u^KPnSniPMSr<=p)J5E~-&;=KBkD%D2A6W=F#F;jH
z4K@Z4;ziDw<!2+;4B#GF7PfN9po_J^N0M*ACEm8ZXVab<&uEf1d(DmUZk)M&mgDD;
z-LKCEo`ys_ssEAvOth_#<f3oUth_?oUj!azBmBC|XIt;Ob0)4kWI6;)8#=xUuGGG=
zALOiMXie?k+Bjig*|Z{DY&z$V2i`-;BTERKrTer<kx=6I5@ShL?NP68FCKa$Ch&Y}
zbP+D-qM;{~ByFC(o=2Gf8w+B$Q#5%UcmI%6a3G(CFVgtWnf}(J+eKaTKGJ#+NMK0t
za8YjR3~x={Wo+lI?*wd$j@TE8t7w9VPC{70tzKm7F4M}hpMQ)mgkLK)m4nCFIXL=z
zOE2rJ^VybfF_yT0g_vyrzJex{c+9(R%JPgo!k8PDfJ`^W!*7(sx!0FaZyWCckPreR
z_hkeX2C$7PC&0c(b(vU_<?2TUQ~t#RX4Bs>fbME@OAZDZB`)DfI7Jv|80H3m8Q}@b
zKJCvN(2ZL}xm-^ElqYm-0jCk(2_rHS^EM*qYt;So0;IM{w;L{Hfbq}|Cv@9?oLPV4
zuLKDYeUD?=!l`~J6%ztDj)b0=h>cpHBBSYzVEt#Cb^eUClVH<~?|oGK9E7p>(T*4#
zXyp%)i8P@$T9RQmy*SYd_4pMk57L2iN5M25&eyxA+DpLO&sS(x`qKw2k8;Iyxsl=F
zpcUg)smc@FG3}PiUS?3^GYNSPAXrhZ=ABP8ueiudeV~y8mSY&LH9>*miwYZ7%h$R^
zAO8`*{3d3=16rH-)4%GhXUG>45ctINGL&2A6mAdsxZg)XdB=y-SME5pzjB`4a)Eb@
zulMbmPI5vfag`bky=GIo29dnZ<yV*1P?y+*D8}xK{!vTGUw`=<1Q*yM-e)l;zG?Nx
zYdH^wCAT=cw_E!Y_ABE$^DEal^Y)kUrIc@8q0SyvKP^{ITTiO~m?~&>M%{Qe_Ft{D
zr92f3Q6V)nTC2PbL|<~EzL(U8)6FN`Ef*5v$WeO({xYlH@76Kz#8KPK4J$)!kIdTZ
zWxARf>d5DRp`HP{=Zk*Tsa)UtPhTiB;19S3);5?tj8D!Yw@x!~VapOqijSYjy94pg
z=iCMHFT3KPf9-vV5951VSsA1{UH+UUaosC02tNT6n<nUoD2^){a?AhA0!Sjj6IaDl
zg_n5!A%c5?TEHetP{jCImJI|z36hqh!~=02rR#AfURA!g)Saz4nbYAVr1N=A*iG45
zRd(DtP1+pro7=tWPMiC8o|n&cRKI{PXNwZ)r6HFW;PWzp&kg>Qs;?bsqojCjWd5#E
zFpFI4UI(O0c}hL#%0{Y9WCFc)hI~kZ7|oqEzxQM{zx(>wcg)~kWzW~Dos|82rg|W%
zREl*HfZurZtP6su+|#*uuFWMwJOx|}ZLiI_OADG7%gduH_D34Dq5nV$$G_x2b!2VI
z`Zts^Z)?c`0W%X#vnW@!dFj^qeCiZrQCGA5n~K0hh0#CLSy`cH`=}oN@{#~3)8hw_
z!N&SY<srcTzHJ1P<ExIb4VeZvEo-Z;-aLh;M-R_RaDHOCXRbmSR#v@qXtS;4<`q&0
zC2X6>AMC?3-Vme)olRv3U12mXgpR0=F}aQ`N05T(nM6jNwL72j=Z9maZInFJ)lq{e
zlC!O`n)SRNg)t2pw)R?q*Qjjm6#CQio@cX2kkD^c5uOxpTD9o{UX9+@nH#vBZhB?q
z13AFEC6>~Lu|`xE1XkC?X6U%@-Ur&`)rz7VCC75(PT#t+DY^H@*rKCvzJrX+%+{$I
z5UXkH_htq1PlnHJ#7U|{4fW9eYwJkI0q_>Da_GF~{b#_EfZH0L^Xm_FxlE|juwE;)
z>N@Xue&aD_6`E{w#j-XUKi^QC<v2<oAyXq5HZnTHbjuC-x7D`?&9%FwKYlmJ{zfe=
zBEPvyN%`_ng641Ci;71D50TW!i;0P+YZf-Ylg#>(5QarTaXF2|7%n*a5J+Ll{-Z=*
zh=;&YQvwzgbOiqX&~RB9ODKcScd5M^S3tO;MpVnQlgVLRFt(beWX~=-79~>4wJgIG
zG635Gr@i0TydMkxXG(HC<(^cj81p4&V&Ka^@L|a~_;ATRY9K9c*up%9g!9_6#)w?b
z=o$_u2KL�Zr||4{d}a*1+<b`?SJV<43G*Vf7y_t0uw46nDZvV_`m*NG!f45*zck
z@sMj|eKRRO8)PRAc0M=$NLbR*{aI^fZte2KS;?2I>MY-+cnk9W09MZ^f7X^09Afox
z?v{aJo$s!@?SiMb=yh{)r_}$d`0J0!ITfa_joi<A1Yp<IeRW!D8yX*DIhr<7G1vgi
zCvEzm%24O0TZl@&HVN=wE;n>$Bdddf{19h%b@gziS|7~EEqF)Azc=AUlY8CnhCau_
zC78u+XJ^K8_e?*m9W=mACJA3Vi(?vCH?iK1x3b4sNWB~Vie__mHEjN&5so<3VMAX?
z?|qu0+iTfJF4dMQqpT@eX6X!$eim}y90OYp#q)1RI)aiI4QsK=MdP&+SUkm^A>11u
zJLSkHr(io%t7mNPtRoqW{+ta^Z(?0QEk<{wu#LZH9Y|?d1Qb*feip?bvEF(j=3>k&
z3SYd0*c$)q0?ZuQVF;~`SN!2hdV4!8oN%t(x>eV!F{;Z76?!XM2_a!Eofu7yhLb!w
z7Sc>8*Jka%J#=Xnz1;B`MOIyeq}Z+M6;;O%3p2`GUd={&4bLC!${LAgB^}`_R$>(u
zlV#KWj6I|G^{J7s>NL*oJQH1%yL3HA)!u7Gy`qs*FS_0SKi`xez(V~m&pnjCqfBvf
zq4LhrCAefv%IxsB(u6uWaT;lB<9go87St;Es8c$9ybVI+DV9sPk+*xxX=`oYJGnS1
zqdH%DU@BV(HFhrfb35;e0e;kBa9Lkja{xYWVj2t(4$D<fq*3b6Rjz1k9VJpd^5+FH
zh2h_NLY}S(^caL5dN9yg!CV%vy1I}sD;_t-BSel_zNhK_+tymu4A$wHjH2U?r;ZUS
zU961cQi}N9!;aJcU<rAR0;?<THl^>-j69W#A6rT;OoMHBi|@y-+d}V43LF;pV@x+^
za;N0J_42mOTaxK~ZnhWN(RfAZCb!!_qVkRt9iIU_*aJjSGLup}?;T=5x}(yf<>PDS
z$8(8m0Td>3eW!_UwVEEE0&jZ?o^J)H<~{tJl^R3Ks+T@eh+K$~fdXiuS@RSW*<U=4
z<0xQ85Iz8{XV{H`<LTuINRj#C+^4BU&U(;u59_uuxM61kF(sA@lv)!Q#Im|ETC04C
zH(13rvTMRag9Js1O_vD?YCO_e?`V!!<7m{pGkuN3F9s5OczAdca&yUPM&lCy1r{08
z&@Nwue<3p+rg-ROGv30;kHRD;oKY-T+b$p%f{G*fAt5Z#&GxE++c$o(3%+tDSvjf6
zHJ!>4fFtSdwMRhWOG&!iTCj8)7!)3wuQN94)D?2AnBf+PNuHl5*;`Oltlc@io3%AM
zsGA?L6bp60jP!q~uznk^jIwo3Z;ke6IlwWxR35B8is!5{wD#OR(}9Xw^(#D+R+$ii
z6c%kd3D27NMSdkIqjTCuEN~0w6&$0kZ`P3!6%s9979_*Ie(lotM@&bJRJVxVeKDLz
z0Kaw@HTc=G!)MWue7^(byt4c@G{I&|8Em?tE_bOEcaEmJ;oFQwI42;JQV}D5IbA-$
z;cDR~qh#rm&r&e&Xt|Z)>9LiaK$jipa^(Ks1qbI}3sUjeS~3s3-xrb9QLxlW9#O+G
zlvisEo<2^9LmxO0*3}JS7`9VZU*Lf+bzIvH_h{~Qa|}8^^g*Q~Akx2s?qGC!K%4EG
zZ^HBsHmTz>w_hIFrUS}2=;q&h!0k$EN$SYHdgvtnHozxT%iX`qDtSz19^tz1>sqQz
z*5L=E6RCFz=UJIMT#yFurFs&>Wbn0bqe?YXNEK0S#2ijt`ECNEXA`u;lUGRbL+>-6
zuXDo%hx9htVjiW+`>hr4=;<pz2Bv?g^K)BGv^GRx;e<9)E+NyStPG>0sW#bVw+tv=
zSsM?RoCCh`FZrw+SJn_s(%1k}KB+CF5_<sAM+q0K3%%`-x(zFEuhIA}+T3+k7iAIT
zcq#haB<8~3D`A*4)5@hV<nNjHa7|x&>L(;yP-yg*c}vOA0H1H-bC0sf-w%ZLTxc^z
zKDTTJpwh??1c#?IIsd9$2zt(SdpT0N+Wo17fT4MTxk7H&r3>aMphnTIx?Qa7ukN_h
zN-rrYcA6AWDpts((|<iyzqd-R`y-b+3Y5L^5Ft0&&CqevJ-T;m{R}!BgpBfG7e8~V
zhxzg!qWC9V+l2_u^?AW}jcFwi+ct}!Sa8?R;nn{-{x-eAsCh;H{vc&f60>Nt{(dtg
zB-?7Mu(VJ_BoYA=eF_uq-><;$b3<_AXf6qGKtTuR^+k)?Zc5<LRBTvJe^Y^PT|jR^
zGIXJ5rKxk2PuKfbSy7Rs(qUo6M2`t~N%26X7qq%vm*w-K9UyK3+^xrft4R!tbfBPF
zDwFqjdNsGf%#iLkD562uM}}aIT*#0=$BX9Ff-UHk>A7&0FQx{p21{am`$E8BsSnaO
z)b-Ri{<#Ap#JmQU57YSNMo#tDA4qsYznfEn{TK-y<mJ23`K%9D2x=VUQ$3w)1sw<f
zISdn_LP_iCKI#Q0H=4vPOEb2YZ8g)j0i8>ZiVZ~vPn>*?-o|pam)eSLk7A22DqJE+
zz_FR3)FaSj%M|<e?IZ!*-Dfi^ff#Dwp?kOuJm{x0&syvBvsTOgKx@5iC-xt7Rl(0t
z(B8sfUSyTx_5k&TCFQ>^rMOfEtQ+B+X&?360*aQ}!#<4QxrF*f`r$~gyHt1DZ+EB5
z3LKmu3CAn#j2yd2O87x^h2)ZKD@1L%<Gn`TR0JhI`tzZc`;;r)t5EvJqx+cQd&Ji=
zx<Ie9&_u%CZ^b1lh1yYS>as?r=``x<G9B27u0!NBZBH<;<v%&F;29cKWEZ|aK!(nn
z#a}1K&apaoB**d}Nc>dge#<o4uC)$L<!e@VttTu|TCOZHis-$L-X^3cvm`2IFt>gk
zqa<3yaD2c5H+ZGmt|_xwm0OuA+`2N(Zuq}~ewu1cO_x~@xS-Wm6vz(Z?T0iye}7Sa
zMn<OvxV(D`vpFAm<OkZG@YE|{7NPIGylRm7YX}=G^cJrO<p`Bt`y>{A46TcqiaXl5
z+AD2p)J&|HhTONi!}H<-TpDVD`*mB4LyKI%=wLqY9_mesA0<tHeY*0reCC_y*OLn~
z-uFMO4S=fk+HPSzJE{s$WxmC)4nod9e+pM!Xc6n0ZK9hI)SA79<dOkaQv+Y@p-$^z
zHrq~3vqQW0^j6SqX-UrK-1nD*J?Ko8mE%UuAL-`vm*VCBoCa45+I9%LIZ#I6XtnPx
z6A`zMw;Ub&yz6Ak!Kkuqt3M5g`6S6BRE%S7VuCfvFmfNAD}`+wbP3D(jw(8y@l7u0
zr_rZ!T9IADZ#mW8uU3wG$?7OHjWG;`Qn4uqxxRX~i#C^JGk6ZUvq)GnaLM1l8TP{S
zJb{f(q2_&%L^waA=R|a95BG^n<=$8-(tU#Y$zyZ71FBQlZ~e~Q1cSDr#zrG$ZGpyA
z2E9Vbn2%~c+VvLk-ij!`?J1gDCr}8$$7|I)-1iRy@8q!pbnl+qD5WZj7(-j5Ow}J%
ztX2aB9{F&ASRsmY7MP#h3wO{Gv-*SejwP4b6X1%^;2Fi<@;CaQD0w%L-X$|OCwf^)
zR*ruhCTjRI{$l$omf*J<kb}h%hC({nlbWony%G_U#>k2P+HVd|dmU(p{aUQ_a+7Hg
zh_TI1x^47_%2BP0N^d`m6dnV+pZ3ObNC-F{$mJ%;h0f!76R%zC+85(<9gN;fxH;T>
zNO?HWN~kx!bb*8v=4!B;6Kh}yP9wEp*)jpIJWjF~M)Qsw7fY=tq@nF(pzpL8Yb`CW
zk?=Lyww|>!BR*Bg+qUAD6E|5YtyX1E^iBSBzBFx#4kR-@>64NRAO-Hy1Q{C-p*m8U
zNXHPFP8voIk(SeLbXEeX<NnA=R=O2`QNO3j1v#C|#qrW9M37tCJc)1nS^GM*qX<Jf
zYJIX#U*=<JrPyAgesSy-!}gg@U=5MoX<q8>=TZng^!+oukft;})&nm+sm7#&r8bA(
zWH|E7TPqk;Kyroc9e?FJJ8as#pHIDsX3VDukTqy7E;WB$ou{u9MKA$l<oT)8L7&)G
zc_uo+lUU%}678b1Eg|G7&UiDQEiJ~Qh39{h5$9z(Md5l4=l>Ar$&eN|J|uetV^_<K
z$#sc991sQfUxo8Te0KzE9Shsr4KuCPu4Y{;+Z?^`Ck+GUkpAE~&bCGVF;^_knlJw`
zc{DEp!bZnn2Sv}M3jE(1)L;Y4U0a>SVeg~_vN5V$#j?jVHEtssr`1x*!!EEqxcI-%
z3Ga%Tmv^LGfK~BY=`%39wCP;r5*A4SzZdJj6QvKiWUSTxfgV=(Dc+BbjiKE)N3&4A
zQ}~sa?$KU2eQwf0@;|ZBv9Uz<){FzpHS9=T4-J?gzOmJ6?Kam+X3chueKrk>3hY1=
zi+2eEUEDZ|)d9YMNy4{9?P~S+y&fdfh4FO9gKW1^q}Hn$V_9^{9?HY@jy7jj=8r<c
zNQNeXXB5g6_v|-xzx_V)2h>dp>32Fj?ty;Y^0nH(>6B;_=(#SLI=;W&k8#Sz1>A^x
z+LJc|LW{Ux$2s>va%{;j0Rb`S2R>8pe+y&L!@S&1@MAHU*5k-!PN%<VZQw|LB&P#D
z-6iUDs3Gb5t*vua>9dhD%0OQ4H!#-8zc*>hoN>};(+EvN=$~&OZ5s|GvVmssjMRp4
zPKB`qobZU)yOeac*b@O?x8cNTZle55o^~ZeHUS*HgqV@z1mzO>fov}q@T@-4ak__n
zDdPP7M-lOi6#AHOF4mV4+Y%ud>)wQ^wG%`j9HjZ>8Ez`vovoo4ZSTYFNZwAFDAX3X
zE__bUev@lDvrHrin0kL>m|DcxxGlM+a3h<~W<0nb+=xd>$R4Xf(1#~~>K=dqPyFwn
z6M?`iEDjU#TL_Nlg4qF<%P0*W!FP1q-k4S^JHbLH=YoOd;Yw@XxwV!VTVDL~hJ16=
z>)^?CY_9!Ck2G-e+(hnx7%|tO#@X?7EZd2j@fU#@jG@TI$Tmk@8+@czd}6_letbZn
zQ-bNp{%~+D5<%+NrEVfEb-rosT8`Pfp+g}5yMrc9GG)tz;917avn-!jUae?uoR{Q-
zBuL?>%|ef^s*QQw{gx-Z5jt!qqCW38DW5Ch2A`=g`DzQNO7Hg^sw1s9<_=zN+6h}m
zt2inB@}ys$I3|0NYh{H+jO_Sq*|z~}$+LCT=)jTxm6XdL0i0ZTV{7D<i3PQzzz9oY
zg+n0OE8kav8nePx$8r0k#Q0XFJH7utZ_$RlTdhwG#BZnesMAl#_B5`V^&ki*aq?b4
z153GfBQt9fw45dWl}q@SaN0Bc*GGZ8NTY74n!@FayXOj=@MgTw*j?B9N`Yb4I)<#L
zqgf=dbB1~vAJd?`V~3Yswkc{@KtJ-NW81TIW<DP(@a)!NS*n!nwXA$9YD*}rB2Hw)
zEQxC(*-lz|Y6nupB_>h}&CeHgR<3NnDufHLwO@$-qnd&$EuV`P-~s*8x;Fdrw3<Lg
z6hX^Y3u&q3@YYszOGw7It0L?P4dn<MB+A72j4-_yW>pYwIjsUHG(brR^7WF51eyr)
zcX#h#?B4lVs?jI2y1%Yz-A5b$q)bvm;-pBviOZaT?SW=#X$kbRX@)reU^~cRXIGl`
z^5J=C;&L??Uk&)Q(9}t*|0`ixh}>5v2jt`#ZmawKI+{D6o?aglpag}aJqBRYAsznK
zhlpsk8-KJ6htKyXNdWA^<BZ0MTO0%hb{iBSIY&hS*paAES{GuJR$TpDUkb(ee0W+e
z5oVC{F&weNU7ZOD_)cY_#6C(EbrY~*THAQfau#4ikl`6Jkvfm)eO|<>deD1H9-DY}
z0&AjigDVGmBUHc?^_o<y*U`S-m~gF85;4Ax+*rgeMh%a%W;N4@E#r-Ifi%IJIN322
zoBn<X&5k5JggO*5{s@qe@}o?&@9`CDj(y<MOT%U@TSDy^rMg&F?KMb~gJqp~Ru)s1
z+xQ6dO43KRslvFy@}3SQWKFsZ@7zn0Ss=hIc+(VW8SsmEq_;aFG1VCiCyV4A8aDbG
zK(#`(e3I~Z-Fk)nj7+k@qM4(GpMH=M;n{K<_~CJDW~m@j4pL`bw#L#l!GdS*Mk~-H
z5wctV;LtSX(@`MR{eKp-cPnxfQ5>^=7D$uK$HS`5pX*3gg+_hBx(rLjs@LoN*DRl$
z+3%o>wTvpeduT1Tj0t$kq$+F|2e}@9z_d~e7mTJ+0XLu_9E$O1z-xRv4n6m!d0zv+
zy!t40a?f+Ul7Kaosnw@68M=5}n?H`&cCrlTct3XDg$pll=%;F}8Qp&CFr3X>Txjn{
z2Uyas6x|mz1)yBYDhd3q==TUvJtZ+Pr{l|hP3GXlrV-fNtkSfL)%qgqW(*Zgz84{%
zvNEM)OOQ8wH}fud_W@CT>8zQc=mxAd0X0~<ELVR|p;gEW#gV`KmF#5a&~;4M+1uaY
z{3F8FOg|{K-ju7ONpEL;)2~X<bf*y#=Q#gM9ih~Hh08;#$;Ws;2g;p5jC8OUihEuY
zNQ&aHn5N)SYuXCT+NKSFVUh(tAD0J;pO_8U?(9$Wyn3F|*I}z~tatEzO=k{R=uG~8
zE_tFBLc<n{4;$u1pzrMv2PH)ZKq!Q?8s!)8@DTciM&n*SxQz9^wW3jJyOP<esID^3
z(pFW_Y;O0{o~{&#u&lF(H2XCNAu|!91Y~i|2JiY28q)kd6ndJWr7Tt=@`Bz8VVH-)
zQNl<<?ME48Jf4!cTM7e^!YGUv5zzoue+=q6;ePue$Duxaffr*bBSA?sBl+19`CbA~
zaEnt>w!=i|#ELaDS)c!-k(kjfNgp)n6&r=OU!vAsK9|Ii2nha!&&7QeLlM%n1y2`O
z<c=jazLVDaJpje!bXYfE5ug;q8158gEY0Wgq$6;B@Z<n2v=Uv~{>HcbT^%!1?8l7n
zVS<I!w5j*CA<$OfhxHt-o;CAKXvah>FOY6Ow3Sb3*@hYV1XDcC^4jgUbg~>ldNmUh
zI^cm9LbiFZKZwmCCVtX0LiL|P2<OP7P*AguTs_U{XcVog0JAQVB$tofojr=#gyvAc
zzcee7>8nuvLd#zaqjWaPwU!a`Oa42)egGB9yYnUT&f9rFuZ6)0*B~KNsRZnGU(ZPC
z^UhE%WT9n4k<H$b^qyv=@z1u6zt-R2lQOW`2ozc4jFFx*OP&L@y+G0Ag0cxpJo;RW
zHJy@OLJK7>tMvFQvkoh%8hnsd%AaS?9(h_$nbKDemY}1muCvI}MN4|*7l)DMS2`*7
zhKLZ-3A5(PV#Ut;4T&qTc?{_MnMH5?x7iNyGCLqJbd(i1#isx7FR4;*=o!`_7}=zl
zezdW6dg3PG3UMiq$<J+S<N7v7fbwo#szIUk+4(XTK>_9x5849grk)NHw$wd*v;ImE
z3I44Vu4sOb!J?t8N8ZXEyk1w{B)4?8c6GaDnNW?9JAXdy)$+vr+!~R6^%MfR0$;WM
zYXd>@PcHYOJ(`WlKSXUs%*#=qt=a0o;bIwB)y_fz%|A)<jhoLvm6>tSLTi~1Q1R96
zxU>W@yd}$R-2>La5wttvILUKRpMEs2?_h%|SDS!!pDZ_hstzX(kJBWdd4dZFth^G>
zJ<5oY+s=dj@|Y}9{QjgaIMm{Kc137oHNnm%O!6~l%b(cc^y3WVFC=Q&DZ>d-B3-Gm
ziQ)qtD7vSm_@hLK63}=?oTE>JqE}+jfMFBq9B(=T%TW{(OHb2c)~9*fFJe>96-16G
z4w}~Woa)XpHLRP~M%M_`*o1WfxI@d3y4oep7&g9QwMM6CUQ=o*Wx8}S9-|h%kfS^$
zodqe|p~M}jtgscurOHg-8-^8~iP$A=AzrZd>Ml*>h~uo=kkf(G()K@(wCDkY*Zn{r
zoWUOSd~!<kAg45A367dag9E6}wpr0^ZIbeGc$;CMZ<T2iFIaW8w#X&ZbV_rSyyWcl
zr}EqF)kW<m%~FDSwWRqGGhLU@S|92XcD1EKY@4O-A3gabp$rb0%nr#^HI{@WyH3V7
z;ti?g%P8+dbU-?-c9A_qG``_UKxPG}?B3>tbSRju(HNdWCYzs5_Y7DgZ|18}yaCO{
z;V=$87~U&6kWO74wu^eb-fUPw0GMadXMe2(ndWGYu15a1as)Bmo#aYiyarm$dhUit
z<97HCWT$R$ANvAgGjC_@)uu$3@`C=xMr7Ll(ba0weQR|r^tQ3D#)o(p{_g^*Cxz!V
zryVW3jd#~e*QO>=;<@<6p!Y;^6-TN!^wGj|je#S9%>y3Vukw!x@5VSdea%=Fh_nit
zOK2};5|6*SNhv>z`n=l0Z0p~A<EY%xuA>S{K;?N42H~&xKKlHMgxJ}XrE=!jEU8B#
zLSH`rSJW4Dfj1jZ!?Jc87KX0Dk}Q`6&8_Fx+^5alUYH4Fl_*#dxPlNld%rw`-4COf
z1X1;VIhAFBq~b_V(s7g)g}EF&H1yhgD5S$o;4iF9H=8T8FPphFH4V*SINFiW=dmf?
zCu#h~UL2rgy+JbF6NMW)em{L=qM%Vx4EeTrD7y|QWMcj_4DFrfWhPeOH6@I|@E3C<
zrm;^^L_`FIVkWim5<epjz42_UU<<9$pQyaPhq*P^NCZ-&a2Iy#?%qkXn4N`6?LS`k
zXRvQ1KWCjTVYsmyU8{Bix%d`rZHH1wvJ_6!nqP0090te_XgVQT`PMGW@)yCRT*$6r
zN8rnbM=-cp>|<Dm53`ycF&&psB*8{X1%~dmY0VdWQ^%X|V4fdMi_|Sk@olZe#$48@
zBUY&<8X8f%;0(5=wNzsy|3VkM(#TkbI3xmVE3VS$y<1U7Qx}=+I4zvXH#@R)ru6Wd
z_w~h3i0mtqUPh9|r;2hIYJ?F>9#7#O@VpFw?7E%3r>TRH&>_>|Sv&O+N#Lk=xt@Qy
z(d#{fj7&sgZ#yonIPD0LG3PE=Bp!*jKovOuzaI~s8sPG@a*={CmdXDyzGCC3lh?>k
zXS(SJ_V%_5Ets?tyC~))i)F}#jCQJG<#Si=d%U%q54TW*XP|e!ubYe<IJTn&gBsuG
z5Ki3Sl{dV)GQK{7W2tn=(<{Rla~~$&+R2PKBqm{7kb5+)N6Z9#VMyhHV8bd&K20Ck
zJrnqIFPj#t;6#aRmD&8r?ICOY)%QK;(Xf*LqIaLppOjm?Y5)K0$mD(p&?>mKz4gC(
zh5nz8D|FK|Vl_kESADQmnjKpI*&(?aB2>RO1DVA&NWIwoU|>a%0suwG*zib>%gkGd
z1@JVuIMfn(k-`gp^#KC#6xR}@{v?K`H-D+4{Gsj?cm`^${Go?_D;iX4ujx<$qign!
zdd=~E<GY(oa%tV0c*4Fxda2yz3z4ii{$ZPMO#3n&?d|*;J|8)dgP7s@zEE*_bCTK6
z?Au-bY|^mkLuB{9RN@d?kP>PHZwf60KWYKkq^-0Ytx+0n$A*WIFKT=)061TE(msV4
z1W%izke;|X^(8f0^dRQPvSab#_e%uONM#{+eh-y4lOw*Ht7Q3M`w}YDKZ!D!KQUNy
zS5Z-6P^~?w&S$Pma$(56taF-GL&(f6aBODIwLkl%zI9216o}nCJJ@0iFL_P${r(-Q
z%@G_8aKr;1)uBBv2nJ?p_~n%lwWFzZDl(D1?`cQX|IQK;A9%@Ixdi3ct6qOx_0t3w
zZMWf+E7F9-85vIk8V2@=b8|yH{r{SJ%RGR#oBM%`c$`}|=Yku^lH4{@ti=K6)fO=T
zRkkdM<x;1`dk*GNC%4LWR&Q+9%Oz!X7rsTWX<=plfw--rO3%#pbG)Sh-}(GHQ@uX?
z?@4|&w!d>@s@c7bIbwSgsMQu-iI?%*A_gUAi$4oX^PLFdbATB)kFwU3_6wIpjt4*E
zc`7%wm{Q+t6ZxOSXj|ycpHrW{WfJe!t}Kkj=;m8hx<|#vCf10(d{Du1{{KqH|Ml&=
zsqHV2o6Durn;PpQRJS2r$<4LcOUXt2u|eYDT}NuDb@S~K0k|dz)Y}F!q_uD;0U_EY
zxm;n`lIwZo+i*NZpX++)ZAwc^MTFqdpQ6V_Yy}yJw*nk$zU6Y^yPR+qgzQBbdnQHu
ziMH2^3A0n-8Mp3ma2e6ySz894_NVPHI!59EM<s8BbGZgf(69-dns_)7cUoNr*|RL@
zDIW$ft|dZiQC#@A@X3Mwq0@ukcWampj8OO)h;uM!DK`8nx8%CE23lbBLa<D)O^4F6
zOpNZ-=!e$5wb~x*jiuy~jU8kDm8E4HyN;JSjayq!Qjb2=IwfCXEl-!^vR{Z~2Tf0E
zxOQgRI=_G=QF-gzoVxYvHH08w4Gw{8;~MR@M=G<J_)0o+4Z~0kTUWtY8|W3Ma5S%b
zrm4zkSH)^~mGsy-{_>!w={WFL{wQVPMQ4q`t9lk4)6byYzKZ7L*bhOXyoifayR43y
zCc^a#9QigG=EFtNktnYrh~{Xmq{(Ocaq)_5Ga|fdbXk|f*EEns=&R;k{ezT`TKn$F
zLMc_w%O@P9Hv_uM(=L;t@~JVuZS12mj2vb`hYZ!dUGgX?yA?@HFzJTOzT$I*$-%zt
z`_b%*w5ni~62SlQ51ztm)@s`cRIncJK?!c*sFBKVSH{)kDPoB5Hf<IBua^4nPHYV9
zPMWcF?Dw;Vj%o6f%yHV;C9<o}j{-N2C~>6DJP`hdkhbk{zca^g2y`lcE)!c+pP1pn
zxuzY{N8<SIefGb`@&AsdO{@oTOndcMj!>K5c%_T_ak`o6D3PRt^u=<c{+LA+RKj_F
z{UVOx7|=N(ctk(@6$)WOp1{?o3sVzh$K?^rq0~ptf{4d&3CORKRo$$u@<M_#QIVUd
z0{h}bhWzETL!*kjG~U7{qSOBEMn%|tfcKkp1F?Y3o-+QMtv=_&K77Zd%4N8vVBTvA
z%3|}KO#P=ziR>Dz@}G_F6&w2i#6xx@bqh`*TQBAmG<rvS1RTD%`gr+MjcG)|xt^q~
zR@xiJ%*ws}p{E*7cG5vQL~is+gh`PTu~esU5&<uPW|wO#V@>P+NOH#;nJkPOsa&Ke
z5prHG^+B;kqmn83E-tc`NA9Qq#aDCV%$SC@v>8d#ng@4*e;N1=OC$h<_v${XA8DWH
zg#_v3_V4AAOVs9u2Qv_tRBCj~jx%u{Rh;2&SeHgXv@9~k_e#k-A6}!c(}WxrDv~nB
zCX+s9AEIQa#qh}BXL-AhE4B5>|GVTA6=J@DqrUQ3Hr(HUrGGkZ*J`<6nn=h@W*EYl
zO86Ngn@A}@RX*b?Ed7#nAJAPzi~8qno}NbgR(dQpn_(M8#Y<$u%?lo#4S+gUmH}_?
z%BIV>PqE8StI+~qiq;yDMDU-@2S^q7uR{^UG)CjmVVjv&3ZLF<OQq@E>lOYD{vrJS
zWbDIU6!}q~ef(HVZNXnH#iw;CP>EH>N61<dO@pgd$IH!@&6DteYkFI)Rs>h5tBHNh
z(G1V8v$y@XO6kMp=i2|B!2f4a{%0(LUUMY=DQdACENbz*&!=^L_b-(l0=q0#6cU@5
z6j3iQ>9ACR@qCA&;{GvMq!e7jcN8;{qM`8Y$N>1aC}blV<QLKPFC2(8C{NR&b4E1i
zj{V2@iAy3hhSr1Kq49Dc5C{N5xNN!ZnUto&FOUBd7Mf<`eaet%tq8_J-O=D<pH(ET
znT|cr%E~hE;%=Ph+&IkBS5w2!;cMK7<GVpKKjB=pW_<|9r%orNL?uTSydW;`T}>;b
z{Q_X)%l*v66bLy5n_-alI~TG$IMnV)EB1!?V0TuYE;ql01@=2oI*dz$w<CA@9>*@X
zNs7>bj{PstA>C=Snmpsqe<c{X5C4qqcy^Z4z|)Ggmp1<oQC}I=2C!}2Qi?k*QoKQn
z6)5iRP~6>JgA<CC7I!J`rMN?|;_mJa!7W&@FTL;G_pOyb{K(2$%$b=pd+)Qig6pI9
zabkH4yD4xlWP??AqBB6gxy_I`*sZOgfZ8jwZU>oy%MnV{PQHvnnwCv3y(i5Q=~%i<
zV^``il>EyeK7o|A4H!{a0Sn2gDEL2d3j7``e2;nAm9a&6eUeO@5Iul0pxR8FML$1=
zU5>r}pN47RS2&93C(!^ktZ`XkdUYg`JZEz?2Z#xG@gSiYM&Qo0Fe|2yfg^Czr)1=I
znW1lL^W<joBQH!C8(Hno<Vy+n3X98r09A4kWPgkDvj)ZYrcHUl-J-DYjJdBaqeUO6
znrqyFg)3zxoSW>sWb4N)&1t#5$mIw39x);cA7f*wQQySOpkKo_*l+Q`x8Z{}Fxmg~
z^Z!#O&R#e~!l2excq~!phBSqnLag%q`}F{y?Wi!i9A)bnLA(N@H>SZfD8#YEI0I-b
zCa>cWcgPX*kg1UYN<_`C{rxnww2*m$fN*2vNfiFaiRc`s<~Ys);_j%|HE+QS_;g~z
zLiuLmgq<;Q=~elTj3*P0SHNx2u=|bazV75|9DDldFfy?H$#<h=!^*t8kF7~WJ@58x
zuO!EN5!dh7vy0@@H|3b1dKh&8g6spGfC=L57apY_CVwwNc>1ezu$hii6P7O|){ily
zT?xhQu-}pNf9AJy>u)3)kcGG3z?0+FUgUkrI^+HYk7ykQNScxBEsvU2nKjwtvez9O
z|B&OcB_WUXV^~FfdVoF8`boXm>SK-_o8-Qw6(YkL$;jNL@#h!TJ>1{+3(RhfTmwxN
zvhht}4;yb8F2pYXDSvuc5Yl?}sLS)z<ENpu&eyJSX3_Yv^TmU0?(~S7PGz`L`K<ke
zxWTm+LlE_}&jP?0y>lS_w?}h91h#OIjAswxhaK_wp0gb*iUcL;$mFKdiipWYk20H+
ziS-|}#Dm~?*KHt)-ST=XNDJawvN(GddDqC5-A{m!3A#WonYKfscsw$y=42Bli4j4T
z*q|c}ZIEg|F&b?gTBhJ(pXM9H7FstLBN~QH!F7jZIDwA{vas984Idv3jgZK-*Ky}+
z|705f+uCfXiM(1vW58+pJcKJFuROoZdGY{KbR&=gcu582*cN>6tDwwZrxBU34nYFk
ziH!W68&I$eCO?tt6UpY=L9>q(2vUv+Ucozoh@8W<cu|Vky+<&1|0C(JkW5^OA8+sy
z4nC9nJ7!3IqS729pg<%!ybJke@q7@L(LE$Y000wk`|JUTv|n^q94YvuqFwv*R`-l=
z2s-o{*<p847O%1Qti3q2K8I3D)(9<O3cGivYzHmeoAzg~L|2lf2%5-Jh()uya_dlR
zfAlaE)vwv!hJE&jMG8nR&eBNPuW)t4^XQE}S6yNArwhYoM*I4T9nE6%t#;V?g>7ni
z;fYeqe!s}Utyy=tR)}0MdnP&wqev4(Lqk0}&+%2-W*-+j{TKyajTCi?U(1Km>KK5{
z;s#}uckg5lwc6<{<ai9*DrDq*po?@Il`}#(5dMH@jJpj^+_u}^+HRYG0Qhs}_YH1q
z%q70GaoF0c7h@ZpSXX!S_Ui7M9rcEdql3nchn`MHX{_xI@5}=K-B14lN6T1s>8=MA
zwc>50)h=?SsBk)bQ``+Wg{{k9w5fTOv>c<SE@t77tdv|$BBx`pw4wey<vj!PvtfIW
zP8^y9xQNTJ$}ao-?)QJFJdrT*SMf76YBf7czKsKFPL9(@-yV=ViTeo$&MF~HPf$^Q
z*8gir{&nmBTN%bixYie^wQzMMr^rp<-&gmYC?6*gkpUVDIZ`u!``j<y9t3hBl08lu
z>FR!%IpAjF|ACKk^P|`J`TWi6VJHI4(|$ZB{7m1%2UDYv5roC1Rn@XhnD~_@sBx1F
zK}7aG2YQK<NYLY1gm{>8sNw@dT&TcR<*5gG!7PJ)#n_Kg3V2a*2h&(w)afndcoG!P
ze18@$kbGr$PMhB}eCO9rfOrD`jxG&!PQ{b)6YBAMOs2Y#j$nIuFb$TEc$>8U#H<EK
zZT1;EI67`me|N0YPVepKY5l&^tkyte;2+UzTIDwCj3|F)x2>DWyM><{#dp&HCS)vi
ziNrv!z4h)F%c-+0{>N9v!uE=XcCrrSlI7xBclP#O`^Eak8&DD68c?6|&&$s))z|D?
z&!@ZJhyS~XpK*Aun%oCRA+#KOvjeR$j0GlE!|Qs(@Kt|#Y~+z?YdN&)!A)bA98|qe
z16g$5c+Oz6Z_RvJ4WLu>FgG8%nF(GlI-&$z?7>&%uh7TX%v5U0=+l<oUG{S6!0DQR
zUeCrPZZ@L$-F5U|ueKQB%I6RfplEIxpKHy|{;r$}ybLf&r=s(2bICCf+R*WLFDR~K
z=LvU)UIcYMF>=kxPtMsgRH7N+1O1B+Y`5CbrVE_r&6^KL^mz>e#Jw*5e+j_fJx#b?
zaP+W1=^mlY_cqxDOr~RZ3AZqpcTiO1XzPt2947IJF*|B-nCdVR=6!-r)36E)H^tbZ
zSJmxJH~PYB3HiGSM^IuM@c<P^toZkDUcp42odYCwx#DE-Yi4PL01vx>3UIc8Eocv?
zi{H2+fvh>YLKl~L7Qj2`1HuhtWF9@C6;XpkQlJ{H2xfzn=bDR10dX=C8RZ>n8I3z?
zm456CsvdD+pdV57TLXX+F8)s;fM6Q41dh9d;okh7^%b48b)z>Py2lKzYebxjz5cW)
z1siFQT;!9P<A$-<q8~GI$Iy`6;Lwn5Q2KQ+i5GE0h7rr!cZTQHkoHZxhP`Rmh{0&&
zb|y{O@AQg#J*AGMH#ATE)zJi3Ce_*<y|zHuB&Q&3mFuu)@ZS0)QaR7@`A4AnJ`FF>
z&nzN~AIl;h^~o}L6%HW14N5Gh3?<)=YjxL0)FhlAGWsG|L#Y&foW-D)M9-P4cS+v3
zanxjepZ4!RM;>?+ArkaFlKSNnr;NpH^W{>juD6+J5)<nmN1&=5_w11PuV+}eV<$Jg
zni6>?3ti;7-g!M*>e>Hn;xOVs6JDDdYdaU7aNm^jID9lWYC!d(MM;&CUFbbc(j|20
zS`i#(^>C8(>lI#bQl^&CC)8yU^JB}^_|lworZgT-B&$WXV)d!=lvt_g4>Sh-kq4=i
z?&mtnLLU_=KwlC3^4~cuQ_0WG4uhF7Ra2*BzhBEuCFS8=FXeS8Srgr_0C3=v9_(}B
zr>3>LMVg;5VunzI0+UB-?CAeAuy8`CPU?}jK$Ntv6ZDY2F5}`-ogrW`M2@QZ^02&)
zbSox&l{Ox7z2<*w&<q_m!g9k-MWQZED%Bhu5AaCPcRTU4xZD{GbMXLK$B5W}oH&+C
zl$ZI;j_43etM4L9f<rOE8IQe#H-L@OfI~@%*o=ru(ytkr8?AgCX>4)>jiaoNg?Fzy
z7(0=u33z$~+L1OMI}i>bA1=b-$@*q?q@apkc?yKd!95uYk4VM5F`mS<H;k8*3#!}{
zI~C;HlTuw2rb2_|j%@5$V6OK2+KWNlbdD#-RakHz*7^7zc5<wfZDOX~9h-mmLNv`?
zQ>v~ocn{rPv23K-+yaIXFa%w^s1=&7CtauJzja`8<!4nOq^MZXPM)@UhDXFVb7}VV
zDSCj2u=a*WG4VYyuB46RpUc)=xoL;|yf+q7ok+%KIrOqU1rR+TyqEA=pxmu`Yjn$$
zHb>X%)H#2l^j3HqE9mduYd`^upga=v0039T`UF<mA*VHG)1vGY^bdREPX*Um4l8*d
z1lxN1subF@EUtYN=_K$Q{<QN(@%E6wuOCxGti0k=ui`saw#dIy@LP3}`k+)xX00en
zwSp1-bY|A3wJreBgO+-TWBuM}aPIcSrZ_uSxacFcPTk8Q@R(TPID;-Gpp^xqYwyp8
z``$3d-68<8+!Tm(U$K69R|{QhfaNL57_YL!E^PNMNdhBAd$e)fPc)mTG{upxv*#r4
z{i+$T9b^a>mr3N*xfZtyG)j#Tr!dI+E?CrCi3a^6Is_1WMRQu!Z5`YB%H9dXByg2u
z(Y80y{~e*ZqB%5M<GIAm{!v|Rrl1xN)vTbwku;9()?l}Be%2;}cHl6<{bgZ_YySMq
zf8OYRT__TUq0wdl2BJ-bcmxX>{2eGsU>6MDV#ZOvy)Z3SXo*YL55jB50mP&-^Ge_z
z^&`3Tvrqkqk`-8pg>)l3RFGNtZO+4C_r=Gr+=0#(R|Qe`sS92NE%zKocjCwa)Z8BK
zOiIpqjyDh;CN6>-uIOIl?Hn_%qO-ldd55C`3x4v3t(46--%*}^_?{9y*QE*Z{^XLM
z^uH?@_uHO7Ua9eY<QaoZ^fkfZ=QYzFWYg`IZSBR``a1|fhpZ#<!pvcZ+Lug~i+{ec
z*V<*m%B^btgTtgtD(f^eBP06p?}Q0YhoGOZ0Jf;VP*6#HiYk3&S!u#O$)`e)*09m@
z5d5|xOir=IJ4i_RNOwIFtL8%|AIMUFTKI;l$};JT{w;9OVSy?E2>hF8VrG6{jn9ws
zs#<WC*&hj|WVk>!zL=VUm5eY@C^U1&Zn|ctwAga+VfYAWQDW;IVAHHVE7@Mz0;zht
zE`OGY_>=jLHfn18NB$Iz@Nj@s6Hl+0_dMWMcC|YIhiijs^s%cRd?|6Jo%$;~_~R?t
zPn4|(!ODiFa8Mkr#@rbc?We9!kTp}5H|qKmukd+&;I*rSp=DEBVHcJG@#bU_$*vPc
zdEB4xMWvsVPZuSpw4*2fu_^zki8>oZa<69n_JDD|gN$@)E3$056vg4>?eBNv4_y8e
ze$M~b7(q5r)DKU&kHW*vQ`TCJ%jf?rmRKvR>Z*QB=W<H!3`XYOeBa7tCY9h@yYgnW
z%0ztQlTKy8%42P|OShI~>y2(Isl}~)bHLHRL5I5?9BZjq_AP`SGQ*5kNaTz}CLw?i
z#iB>%ecdz?akxfCRj6Me*&qtxUW$;PcvqFnftmeM6?YI)2mw?UlGjb(`=d>C!jY?N
zqsQwmKbsHM+}te64@Lo<%R1_9PmVPa5I&8Bd2do=Q=_O8aGSAVU70W~O*Tr9_@lTO
z%V)Pdkw4whHM`|0h(W|-E`G+4R`Xf7<Cs#i#E)`AAg^BR*LF}+3knP0E`R*Mg7gLm
z8)R<wpOcDCbz*b_O-Mykl;$h9>;_g}jb8_2^~Z)_4=a)x@FQE-YUd3Xt?QmHE#5`K
z*|#Elkb_W9aCv$erook@FbHSNA!I(eH(I!-Z=2Nes9Q0e>un|KmCZKd(6dpP2&Ia0
z)SzH|u)1zGO=Vem0(ZFW&b9+*`B;FWZ04inG=QS!tr1w^Ac+>g<$?K5_bpR0pRHYi
z$-zSxe|QBeHI1OaPugUx%IpYf$`7yGf01&~Gspn1dD-1e%12U-0O(qphn__fx3$(J
z4I94N)9Z$-wE0MuLy2r4^mgk=zt#T~1*sh3U6S{1?3(N7F}urp^L&!VP?FPF!l@4N
z6kYzor?2UDN4w(`goBrIPX=E#Y104va%x+m0VuW$*R1Degi{RvTjCI0di^$ir1mSY
zxljr%{}c!H<`knOEocv)ytl<p9>tz{_V$*L>#dFUx<a$FYlA))ryh8ZvqfVvx$CC-
zKV5jBC5p4KS?_Hhx7{r!8FJvAlfeuv-LPQuKc#~n*uu(6{^j{eFEYW+Yj!9pJ`S#C
z1y^g@xs}YW8ij|?$vUih433YdkPG^J(XuTJeK=**DvKo-^!@IEW-VFC%*Ao$1UI;W
zi&+^Jv`w#zW%22T;F+#0nIFyqb&uc66@YiPB6d5@TJa%pP5O{B=o0__*h-k>w*ucq
z&pCmw=L($4`)m@FNS}tbSkdAfMZ~N69gFOfd<qNZ(w*mQl9n8|APa|KfZ)pBe-7eK
zz{FN17J!u~?R5=;1kw*)sjM9Atv6yF2dev+@7aQI{fo@n&&7qQsa&rjm_dr!+%C1;
z9Xr&pG7N@pQ_5np4iuMZCq6Fn8=NE}5`WjedI@`}t~~qa13D>t1q^goKlY+gY=8Gb
zwQI~cm!EIf?b)o|`p5PpNKgYJqsr?SSF9U(%GRs$$TRIpBY2Zt_xe@_I|(+4M5YD@
z24Xk6!-jK2gA3gD$3)ZPJpba282s$}$%iU{dqiiY5bqnUBkZD0tOw<x%X`{L;Bq%0
zIix-w>8tM_d%qXNvXxx29meGvQqvoWIoQWe_BqdfpNoq+#H%b4tNq+$zqZ~*DPLYJ
z%;Tt=wFC_2yHhpZZ@;%7b#HSQKy*#m+6f9Q5!E$_1F(M5vAy~JX;B|{34!xN+ysH2
z2KcQ}(n{u(Z1h<gL-CDl(q=O=s7EsW0sP-%41lN#<p^Si!rMge%bP7e@BMr}oz|MX
z-lW2XAg5F5_o86s#tnWK^JE2z6spzBM>)U=r0U@-KOO?CCD-{xPL@$KZ)AD#qxU^n
zVc|z~h={`Ybt*#JYoXSTl}0a%T5reSzwzQ3OXY?`XMUaAU9^B<>QZ?xYwI7Rs%Nrp
z_WD)TG-$?!uF8(1(G4Dc6y5`OnyIytk9=RcULk<4J^7yWFSLUB#)zxU;o#L>id^&k
z<~6&GdxgFLjoz3rmEAv(gw*{}0sdqSY>56QZ*F~54ZGvRLKqi*C|h}qFbWyu9%+@I
zuW${H2P$Q;)4Vb*UQ2H|x$Q^^BL`_8UJH?T9m-f`fPU#eW!eSty*R?19KFw315i)Z
z^vu>lPdleAr}0Am54vi#t-6F#cat#zOss8>u2q+KXu)9^ae9u;G}<~PX`&?dO=|l*
z@{NDD*Y-r(Ixv3P%>L2nlKa}xa>CVo`mpYVcikcvX^P$k-)#ey2BLqI22HqEz49}@
zc-I&%P1unj)4N37`eRS5q<bTpQUnHx{ZxZfW}-4mOU3sMKJN!*Fugi+;|LJ;qu`*L
zS&p4g5&Gh5c4UOOeiQ~NsZIj^!@2LwUD{mj%Tn1<q0w=Y3WTnxpaTwy$qTSqZUiRy
z$Ko$xk_e$IaS|9D(Snt1M}USVc^&RPXD6H4nR+#Yxuor*Cc2uEt9nz8mYMVWTde^x
zK$9#KUR1mfKa;vC<y4T6`R7<@jo&8M5gTeo_l?5zhHL<fF4fiNo5`ca2h#K9tA1(1
zhbZm(E@n}u;^U7h4fEmklKT)!E?Ntclp01$`Sd9HXl7F(YrvxK>wS+X_A1#{>2Do`
zF(O|O{A>9J-vg$1kAAJKFX5CiU9b7Hp&G7V-m7Xhsu}7V5>Pj6q?hTPD_-lJOLi|k
zA6>|aC=R2p8GV4ZnOhZ4gac0kOhHJLpCKKU@|BWjvVH5Pu|92^Y0w8`2^jX}h8N;5
zs@Mc_)?%`B+J0>L$@hivyMcs_klUXVw=tGOOJD!!oeOfdXlJFVecYBZAJ$SHj_W!u
zS})eExpHv84!qk|ihXAgY`F@%R@F45swb8!xWL0labIc{?|ZqrF=pX}-DC@Zez}8>
zD%se$g@5887?BhCEZtSdq&p=w3KJC`T&F~YT_;_HmeCy+ic`jBm7G1oHgr_2Iq$L9
ztb}FZ(41I~$-bla?o&Lafo~)HmBymgo~L(CCQM!bCU>Lw@V{J8V6b>jYpcFv>jABa
zUQ<vI!a+r;V7Yd?KGzWo3p`f7O@Q0Sed0O7=t=Gw2FdR<zJM)%5<OcjH>j05ZM|Y4
z=Cg0g@Hs5`H_<TIcDpvL*lw)HdwQ0Xlr(5dgX+Kk%Wmf<U6wWHf)nUr`SRv?Q8eIJ
zg;KBo-<V|Pga{U}jG_OvhKWXL1IzfQNetq=o(leFXt8l<xzP)3{xwpY&$TUq(Cs-@
z1{rWh<&ti;<LTvD`7r56gZ1RcwBM!DlEY=(>xoAdElODdr@VtR4i9_uhWyt;2X^7u
zcOM&{PCYNvq*V3l+A9COYN-%EoId-09T*w0Pa4LF(uhE0K*3C=6ec1Ln>D6PrL3!l
zMc`6V>c`@s2tnhfr5t^5s1^<}IwD20<q^bUz;QH@iGjiZ6gH$MIJi5+g~e6J!KJsI
zhHXzok%!S<H8?nk+Hl8?N9cgk9Mzd|(X0BcC$J)_GnE<;h}Z`kuL{I4<%8l9nhI&F
z3~hbvXKg2<Xq^<Vs(tVOJ3j(WNcGI^+4AzyPO|CxGVO9XMH4nTXG~vrr_fOF&b9KC
zjqgP%e&MQ1Yde^>Jb|hdjl;^LJbIw3?0gleSk1m==La_?X8Xt%5rkwtuo3L1y>a&u
z3O4Ps<Ym%{YOXQ8o<mKKey%4GHn|ioer1OIVV)noefI|X`lMw7y<QOC3zWK$SCtuN
z{B%eqrvn(?{^D(|4HZ&CE9x}mtLRqf?Z}*h@am2_Os<e%=y+=+ztAw;&?!y>9Fx;&
z^e7*>93E~M7H=xfQTxM4U0Oi@byH7g>IN)W^3aYI{<ml<^Rh2U*Ff-(RC0OA)EU78
z79#bZMj@sIY+yutrNR+YF~nBgaETZpK`Gfw@o3Ep(DZG7EbpHpn@<2Cwm|7<QpeE<
zEVZWHSTdOG`+zWpDhM8!?S{!Teo;<bBy4T{pSQg=BZ@xoR20$bzg_^A74EaM?G)bS
zVwPj9NnlNK(6Xi`>upCBuZ;LZNW(@8G|~P*Wn!y5vVmQYBJ6)}HuyLpKLYDfxMFO6
zEZm@7b4+NTbal(I`SB*eDK1wI%D~8sc{={|bVeoC31Ktaxs~hMTsv6Y&tjB3ni~}K
zTqnaRSom-T$&X1!Y^(Hqd4r93d)w|eNb_&JN2l1Qe9-r-jgVG46L_<D8IkUkED49S
zLgCZ38sM54|Jzl%b$I#eaX4|DJOK;KFk!~zh{dvS;r~A{3?5ZAxX;og60NGA7phT2
zDNWT+>ZZ`tq9AL>EbDC4a6!)cWWU%?He<evnObo8<gjg_oj(5Wf(w?;={0&kS%3`S
zoybZ|e1`+{5_MGVEBT4b_gftFZNm&KRYfU{orl}JDDueGg=j{!iE@c(Chi(26Cwzd
zPl8E>GglLN50OP^pG#8kvgKl5{BWVP#m$dv{(<hhM5#F${h3@ZY|;woXQE7TLY>ZS
zmNsGe#Kpap8@UnY9dGZ;ZjXWOUS$>C6WShScDLvSq1s!1OjZ2Hwxsg+6ERUrT?BG^
zof{7h52FihbmFf3jW@G0)rNcHV)YLSR!;N^x38}FX~Xq-NS5N6=4QVH!I2g|I~8qY
z(;}iTX>GI2{YceruLT>Km$1QedwsTC1aJ7LO?w&_wL3<KkISM1mR3CABSB4<Fx^Tc
zmSiR!yx4cV%FLkSHZMe7aY`C0rJ{pBUkj#Sq-kv~Est8$d7~PtvpC)c>4JLaqqEE~
zncD(jylh+T&N0Yl%0P?!;=fbbC&+<tiNGZ|K5tF?D?ol;{qQ!jt$(dj`WH1_&u~rs
zsI2<?S^iz@m$icRN$_AITU-RT@9iPye+dy|TuA&cYVH_f>MI}5q=2$TdJ@});p8FP
z66;p1ZJ6@GJ6411Apgd?cf6IJxsg|;QqAY?8JE6J^2NXS23d7!_~B#gt!|Cc9JfT-
z2VfQHVQ`%E;DAd04vN97NJmERT_pc0G#=P;9HBIXHkDjcak;_x{&L}Hmp#B6BIPJQ
zaNhP(jIBAuO0L67w?yEF&;HivzYY@ro~_p5uc+zqH6y_~gR;`3-US&2h}q~u4?42b
zZzU`byDj;zL8Pu1gdOw-EuPpzF#^!o0*Jp8LAppgJl5>8m(^c(4|E>5Qp_kjn$E99
zz@&`>ojnq~nloQ(_1>(oe<HJDRNN#CnhT2(>XLJ)bDkat(WV&s*g!f3+E513^a|kV
zCkiZ2Ls~n?!grzx1B<%jA-)r|-|J`n{|Ox6vLIb*8Obn<5F{Bn92%KeZ%Pvp+p5`k
z{V7DaqL1?VT*{cYVUil{FqW>WAFG3U2s){>dVDTQ!c{jZoz6IC#fJN)0OQKWMLbyy
z+1K9Q?~$DqN<VG2m3+V<nxZ{I(SB~%Zqpco37a8C=6{R_)R`$enrTO4HCQV1`}|0n
zWb=ou_~a1fzoc@Y#=oPG%H;RN8c>n?Rb$rK{EC*E|8&E=|CQiljx@>OIdvnSg<XG^
z#Oe@I?JRRKnP5a17~WpF_Oay(gB`)p=KfTVt9~NdIjPk>B@$ErI<<zCJvqlpipOn2
zvJh})<s=)NuRs-q*=*PEFHL>Nz9M#q1&{}}QO1@M_KZOiK5m0)hf9y*cb@N;bq!R|
zv<f--cLxg#FQGN6X%l*lS|8FF4AcN;+rGqOaAko2LfEUo?*-~5^5Yt+@~Y>HxiRCU
zK`Ku;4(|iWU@M(^5q57(XC7mQEquGzbX2?-f3iDjCujI5fH5Ehu;$ua9wS4)4qd}c
zJr)b_Y!<3f@To5-;_x;enmBOFa&+4aKUkfuaTf}gAuME`MM9KGOybB`@lBvN*bc8i
zb>h)~A&&}FY4)Y1DV!R>ni7we9a=h>GpPDORx<U4%d_Cp=fmILqw{04#a)LMc?ebI
z`(M4Ux)>4Y7k*7Y;WnvNBB%A7%-IR~rCBE$_!yWrx@8t780t-<Q|200*g-fc6L{LB
za4YBO6+CEum%EXBEak4#9oxyaE_t{*T)(c(j0^Kb=r|On_Q9;7au*MWu|!!l_F3Q+
zoAn(n(4PAJ;QY~C?UoXc?0NK1B#Y1F-js=oMKF0~^F5R!yUjE%M|;5!YU7!*Kolhj
z_4~VEpTCz+80z(FylDH1NIK^;ooEHWFsn?2rc=tpvl@%?2uniVXg%{-2J@ZaEc2br
z`lxI?M?W0*EZ!Tq+hl<IKuP^aVDS$wa+g4)%rFAZYPhCUAiQM@D?iz5ee#GsZQaK5
zh8+}JbbrS*v37NLLB$8i>G+3MPy+v@43x_0=%wD(hMn2#aXWDVz3<|ryDS*(7i)UM
z9r&!Q^5DY;Y}#ygb7eZ0r~GFFrUr_(DWI^Y<@IOU_0~%{SuKQA)Y}C3xd^Wf^1zUn
zb_P|g?n^t^wt4_Xwtz#WXx*J{DCujJP83Y324g!um`uYeDS+tW2Rpb`^uvAUk^*ub
znkga%R59|KC|Nur-yAmazz@PYl?EaA1L>-2hwBmKs~k+x7@~>bowqLNqG3%sNoXC=
zeEE(=adWXM_@<EGknGr$a=chh&$|Z#N9n5B>FgnNiy1@-L^!a?w*l_wHcb<brv8@V
zj%KM!Onb(iF6O*411iVW!*Z&1fmDOh#JoUz4Bzf~hsGK*<x;b=@att4O6e32`HFTD
z<dYs_2MA!&>}#V7!rLdScD-rfQ>vf?*Z0Ng?@hqzTrxe9;1%cmm}W4zE-f`{G{(dI
zq~1c$5guUoJ+~gtGfdzikl1<*l1PGm5)B3Pz=gWwrO9==j%Ir8I(ATo-qEDXS-|D)
z%-VC1?yyY#<Q$qo@=a6<b$ye%4~xM^jr&m5xk70Y%(<}o4iqNA&39CppKaoyr_)ve
zqirAz%D)fyK^ows!^aQaP9=E1{(R)7(@QrDT8rd&bxy88gyT`)kHpDTZs)-Jp7E6m
zqw!a?<^wM+0kvhT@0u@EOLi8u_T~l`w;wfq%Qu(zmR?}rz6jhlVU(1`3(0}Axld|U
z(<r=ED@S{e-Yi)mQnD~*ewM6pbF)~09z59ZzlRHegA%E!yf09W84<Rh@2wc<b@6^P
z`paa&*2+z$D;Sf97U<4U=Lwbgl#?lZO`D$6kR6ktmr<f^wE15R!CNqAqjhh$oq_ph
z+=>qVCw3iSbM+=r--)m$o;swna_dywd`@y5<u=M|Tax(hd3X~>V)U4m=y5_klVJB}
zr;Z<lcd946!{hA<liY+=k3T7b1E7XY&yN}%iLjXo&&1ZKqn)0PXO@-Q3pO>2$(`x$
zYt(=iy-O~C(iu+=xj3tedtg)LgEYlZ^|%mD#klbORgLXPle7O&TSwf?p@V1}dt*zh
z&*E*j|C_aR8i17$ZpAq78QS7ZfOV756ZU-Qu$~Wg%3}6|PZOvo(yCy+b-w$W*yuSo
zk<cCU&?d=({W>eX%ip5dUv_7~tQ%+q;X#Rnq%$=d7}9q<_~btBKm9SbC`nmkjJwnX
zA;Y0(qvu+9@9AST5Srk&GgY!UH4W!XT$DZttNrhWw|*b!<o~ej+Ihb}jA6j=iBkIM
zwnNoWlh)!Fx!<`cF~5iXLc{Oh^!lLt!W8p^sX}&h$k(*8d5eJGjMUY;<hR<DEvI~G
ze84}C&)5O60gxL%kIK*X$M0%BXyxP`9F|rlZvv!v+CHWYaatcCXd1ObKI)WSQTRSy
zxK>u=gs%7;mAjq^-pXXI<r|eQzysip@w4UV+E`3gRrQ$)S|9!FF8`h9IH4jdP5~cV
zxGV80Zu<YJ2Oo~|zWmbCqm~h^2yA+z63ESqA{0wRnJGdUq%9#Fg6vLtDS{|EP%<4&
zfhY(_`ubf!QmLy0H#rig$E%J5sTr4c#YNv2X3}{{8^wx;_;E=V5|1C;Yl4X3swauY
z4~v?Rpy>H@oE9;PYy`=1(#9nT!9FA)@##^naar6Nl~IAFe{EnJ!FGV?>Ceb6ko3*q
z_E8)51z|`;H?bjOvcblkUs-+L*0*?Da(sOJ`MYt*LiOj!0ol7(;xNS+k6W)PNB9%U
zo{4x4E(5|BIsnj7oh}fh)HgpRj>hyt4ku%@(|yUc)1!!|A|H!aO^YMj7q$7963fS3
zKmYCX`z#FoO#({9?H>1=9=n$xtFAFK2T$@KDLo3uhyW0hz^(&!M~CKpKC6Z{Zi;t-
ztAej*Iw&>bp^t_Q_01s3RhH(H%yOS-&*YLe0a;sRo`#ZkOi9(N{Nd1L#N)j-*AZEh
zAa$uUa%KOTcbhgxzsgo}Mmh?XK5TvukyZ3?+$pj3#~x-{v^AK@Ia3A_FK+s{0+Xma
z#heoVtUR7iFA3~ss%Y99_`CraG!z$x9hD^Q#($*%dyAPB;(-<TIA`Q_CDSx4e0_>$
zis`4@RHlaIp(VFccugQ7ja>;42914X!pxrmO~m@N#To#^q5@66di6KV-!&X;vehWe
zF$!(G1m2r;d)`w*`PA-ih^t7eT?H(F2vbUi+4u1frhNS*Jsc2TcJbWwkeot2F~JJr
z!HkR&;nN@0G{Vl0qRVRa>(ojyDJjvRe}6IKB>l0t*^*nVW4`>gh>zCRQZ%#!VBby5
zpQ2{vaDAQ2nNGTDc-_ZJFl5nKW|LRrs3)EPuB>xF^e8uC#H46vZk>+DBUv}&h-j2~
z!Cpf{?y%XASzK}Y+9YItV;8SlPOX>ZV^HomSYWf+{&-zlxWCoB3HzneFQ>CTYX3p{
z0MpP{xMFN9VYgoq?a0HRS{q!ZPf#%(hSh}z09Xl-oH8;VQ`toeu;`!FpZgCkH*x73
zxo_EY&Z&GUP){0;Z8So^bvWVp%7qcVl0k9)oDKt#!H!L7V&1Bap&nT}t<uB2l6813
zTh$39WXz@8Nt-7Pa(8Mpq@%bs4yx?~-<28QR<M!<goA`aS&V6CXNRhESie7nx7;S>
z;hJ>;`^@bV-e&VMwz8vTC%swAv@xi$`Rly6B3$$MlJhc#5<uJc-ksnAvE;L6_1E81
z@7>lGHCgam;I>oSPFLEgDcRkdaOpapy=$V(Pikr@76(_Fk);E9Wo+CCdVmc~aQP(0
zlW11I#ra_>^K56qb#(Fs+W?p^EgS~*1uv)kS<g1u{4jgk=arDjHYq!`&Q#%$2XgYH
z;b)Zpz|J0A6gi~Y$k=3gutW>_q($ojUP@X{$?Z_lmz_62SOMKiDFju>?cNkhkJ1?g
z95p>vtV<Zw51q(I|5E-wMpuCwF|(R`)MuFuR_VxYyZAHkwQX{yu!ZxX^$k^yk(K!p
z8?or0KeNo*R+>YA%B10A>|t~1ZKq{xo&~vwqOd<ZDKQl(_|XG#tjeQt=t=`q^O|)<
za4bm38*wW-ld^;#Ks`wdfLy!%j|^$!E5P~Iz;n<j5mCd?kR?Ft<Mhkp_zU7KluFWJ
z{-~LP)riGt&S9JH!1o=af}r!u6FdB&dQfz$7H~goG}ISGQmDM%?s_>$ht4QNC7$#d
zXp*1J>n1i&d@STCe?>ByJQx)eGtAPW&|ts%X+O)?v1)s~c|QXnP3B)$Qa6}!@`_F|
ziz(pwD*K-a0sBP=e$c~MDKJ9NyekAX8;uuj{Vn(|Y%nGOns1!+$8+6({@4Go#t0~N
zcg&bU@VEtlW0mVWb9{i1h~u|~uNoTf;ZvXAKAXU#@+gG>-zSi$5Vr7q602)nes5FM
z;5|pZ>fu?)lP9t0tKl{}9+^Y=uovty!QB-%I;@G$o-H--*%bGL8cG!$hmQ=xf&Cz<
zNga*J0GZ~-xX2bM^8-x*9^FHIQZb;tA1qy|4Ic>1Os#HaM#Jull*%lf*tA=E@p)n<
z14gp1k_xd0JIBWqVSir<Si4nouRxIX;!$DW3+vO!f!*2c9^Zv69+%j5?<TF<N9pdL
zQEv9Sk5~9E*rR(*s=<7Ts~%NVT=T_t!A6JFLnuNc$zYSv==yv!7-Yh+^0lEGDV=LA
zr+)+m2`d>%Ad<hlpWXJ&HzH0avqF0(>W*X7q}DnK;&l!KrK8MGZ-9#)=KjD6fex^g
zp^h4r1z7h7YUPdZKEd#v<0=in<i#*%uY2DCIS`-XJ|8bljJsnP9X1fjYJ`tf@q%1y
z0a=bqpQ&ZNOSV7YpoD}ixLv;Kz&LU|3@8}kJ)3g&4_U)FZ95-ze%h^E#c99KQ~@h0
zJr7rQ`Ikm#P+uwYa_duEbqN<bpL97t4W6y_kyw9a)ozsQdwLjiUeEivKhSSR7k1Py
z{A@MIWI!=vH;uj{yng4>RJp(9%?_LG@4CJnd#rhf8TI%|@St-U+UNJfTzv`{JMy@t
zVX`WEHj*fMCgtcn4u<G5oUXyu*N;1<)g}{cmt%`Y7|YQ<C*Ez|<jn`qUEGdX>pbHt
zU|CvWS9xvE-buBc?u0RndH8}l#dlCid|a4ikj-k}Yus$_njPM0Dt;lY&?1B=vN}e4
z!RNaODk@TZ4e68KOU^!gnLmzHa$TviPy18wd$TQlO=5)**}#ooHs#TLQJ<8`L_=L#
zi14O2Y~CgvMTF;s`z6ZRnKO-75k*xn8APc=BPlAZ=M6?I87WUsUm&=JEQe6edUleo
ziqX4ACwmYtz#S%AKnD2Bd7EZk*|_%OJAHL78<X&_hnA+uR~9;jnHh4Oa^ME31A+L<
z24MdL`PjpoNihcPjxSZKWGNHy?^l>pd#;SW`OZlt)9kP36JswH`$@a8=I4p6fk_XF
z>QA{bpacKGoCb^NpQE9|Dd2AFxgUX&ka%Mz+G>?DgOiV6L$M_nmCUt1yHQPJIhMWM
zrFgciiFEtKDXy2}6IZi;XMcQ3JQJP!8Pl|{N3sbo)s{g^c1HGJDoM&+w-osfiWfN;
zXEb=*=vMyePhU8wS7?<+tB6KsG(YZK3WMYmAoKaebA{;!lhR;Zao)poQL4mW$kca4
z#VoOzO>pJwt(w1cnakS}zqY`wrBmI@teH`)(jinjg|9^?D!E8cttI!01(jBfj#K?V
zZTjt$<(~3nUKom%-VZXOaE|yY+nz>4F%Q?3zw<O;2DFKP5TkJO5&YCHs{rd3v3(}4
z7BuaV;q9s}Z#n6&DP^C@&cz-CU&W1<7Vho-I9b?^exFw6!YMZzQ&!Mgw9-o3(6ZOg
z033TFb^J2;H<v0HXjiZ|_ru|`HBI%ZJT2a*`vlMPneZc7`wSKCfK1dpGqfhSuCS~{
z;6k%r;6bUdwZ2M07j`;E;K#9-jCA08OFj5D;c{2cYIh*#^x?%fR~+G#&fgd%GLzEf
z*MGz|aFQ<QuN83j_d^o24o=bNdu@iYJOBOm4T{M7$*@w|sARdT`T@&7=J3X|RI@xD
zUHIjGqTaTxGE5&H|JK<yItT|*$Axb*T3?<YT=&wfhAdb4Dr-JMt5}}p<>hG_<L*w!
zC@v>{1EOQG37KN8vi&tJKXqEStO?V7C-cd6U$QA~K6GgCJK>y{iE{m5IiAIAmq=8v
zA!|5r0nVEl2R&aouUtgU3p7!z#A^q*uf?-CbcbQ5xNP?iFY~c1-@6Fci(bgjJpBF#
ztpzG01x^@j(;LN-P|}I0i~JV<Qk|3!;XuH*K*6S-uCJL5Cai4#HcG*Z+HuEbiWou(
zP(mEb2!Q0oz(+hS)|`(O`(=m-_+fQAWj&JF_1nR`R1h{UQ@$An@uKR~k^~}r;lxo^
zi^tdr>4|*!kPEY6V<tPQ5hPBZ1eO|hE#6ggK^y~e?Sc0s(R5#bz|U$zeeDw_`D_`E
zkEC8CScy@e-(q|xd0HSTssnhC(!*Uc20s_A&PuM^j}Z1_lJR@MjeO>)mk50-oEX#F
zW{HP~94WS!N&mhkGX)?ml}PuSftW0ZS8uHKU^3s)*^I4Dc3K;bd2rDCqcXhUODghB
zYs1fAbbGcQczOC=j-j-6c&>aqA!mOvl^h)YqXLmrl=Yj6XJJ^yhs%0gV%O&!_xsND
zWphPY$P-0oILzSk@)sYkr^p|>t>F{Iz`N(*0=<|C{Ju>0bmjzVf1kxAn_*kO%O%(q
z+GU|yzJWdfbqS^lWw4n~E*>>tIEFi4{)(-;o3Q$ReI-$Emri=UU8v_7RnIn-amj2+
zu)~|CqiBzRZP_H7$<n+78#N%-nMKW9AvM4)brG6by@gWy*R0i#j#$IK__sUCrq5WX
z=agDBcX^7;bJ1`vo@nCv=YBEYM|3!d;{K^X1ps`A7%*#vIaK{zZ!M?#n7UZxpA!-&
z@nph#!eoF(S%(*KeJ{{}qupL@Qn^FFHYJTY5T77jJY1ksvc*K;B4TV_Ks#Xh6Q=;Y
zD`{wlyfZ;U#)W7Ly5r|H9dW6VU*A?^zc!uh_?U}vysQO2cq^C?*`FA@)HGlaUSaK$
zH=UxHPga>)H1NZuyP#uw{U8PgdIxK!UxcmgEfrNx@h_f9*vEWuOSGqOF`vLOD)dZu
zih;gAtbtTG+_Y5w4B=4<oN98}wy+X~T+z6M;>l4&FYFzleXVd9y9&@i^7<{T)G^-T
z--%_RENLGDS7so^WD?m<oNeQ6*2sw}dEhUXg^roeY(dOo63G0K5gRhO4!_+>J$jri
z>9CV#)L8KZ|5-v!P)b}K%pITDxtIIy;y8ADG_8k5wmi~*hPy~`nW@>Tc{t85ix<sj
zxRr#mN#lV-U$V@iT*-u{`m$orTj(2@k!xPkx-;MzvXEAuJxKRc<T9irPkun=F-OW(
zG5_ig5EHkP2>n{Fw$msw@l~6Ca9bHW40pc8s7ZlJMeA7x!a+YnQ12kAU)io`Us$E5
zwv0-5<Qft+PlwC}v!3}u^0}a-y!KBy(sBVS=nhx$kv5Wu*`+@Bj9x>(c*d^KWb$m~
z)NczJ1Itqd^2M^O$7&zhp+zmvP%YFTeL-&A5MEO;!YaZ5aYprC;K@>fvd^uP?eM37
z=r2EQ7Ep?%18y+L=VrJ@siY3)0xQai`+#RAlG(n92#9FcosGle19QyP(?H3RFZUNu
zlI0cjp`CM6KF!;MQLf|D={qAX_pYUmtMkS3Gj?5ON4P)UBJz1A)x9vykU(0Xv>?SQ
zA1@|_;c>Y49WIYEA{z9=*_mUD;we1uAx+pRi%G9&{=NFb;O-wMkeWTP_$c76d&BTh
z99!@mP6`w$3#6zRC(qI>L3?oHGmx}?_x+vIGVK|IsO^0}1$ui&0b(gXJf{W{0Jo;w
z2|{OyP5vQ_eeg+Bc!tfg^%<;AUxJp23!Yuu)zYW(zt&asjaBklzw9G%!Vn83)hN?j
zy*^mgQ@tjyW306tbuAV39!X`Zsl)X>@)K_qh)mXV9~M%r6FC3-rpBflsYC@!yaV*~
zOMgFR?ivUOP>f&T;{!_H<oCYSy?Twn{$_chItaJmqo#b&i1Cw@{e~s!O`Uj%2&MSK
zp<JvcPK!+#nU7SgF)6^6NOZ7|u6W$uASg5{hP6++A6n)1u*&YT_(>zuN~~Eod!Z1!
z+U3Fl5&J5JWnsaOIEMXV(T|DTnF{xf0|5z$*~HGRzPfze6qz+Tv={En?3d?c>Obdn
zQk~dxS99E4E{FIGB}N({wwL{}xu3l1Mo7wgQzvN_9%W+tJUH1c+&l{h8BDjg--wR)
zxa(&ZZ6D$i0-+89TI0tK_OL1!x1kKRg#>*8NbSW)%7L+&!(D&$%2v+!AvDcsX=Kh2
zePO{4hyRd(4#-M|9yt4+f42}=e>n%UT~t@;P)<1eUrA+&e>{yaFLY@w!NgHD;}Lr2
z*9^VCR-BpB^bRt_7g>^}x$axGxU=e@>?YY>1?4*iENhR=j0YzuW#MHFp3+btKH%te
ztK?rZoV!GEM*xR&J{(Q7MyydDoGzUDoe+n~<yvm{M?g<WYP8$SemXeOh+G5hJq3JH
znyjv5)xOaBJ7E+gIO3Cjdi#Qm31JN;T|Q|nM7mPJ%FgKUi)aOP^_+X+FSb_vJfVMz
zV~BQOUhn<Tv9xp8K+ZzYvNxKHBKfuPlex||8;qJKPj9E04S1qYDM;4A50osMU`~&P
zq2ep!Oh&i{GG#3HG$_XyPK{SPcx_<!vBoNaPr^tNj;(z#whW&dXRPJ;Rn4vtOK{rB
z&I4}87+kl0EAU=J+!;m|AABF%#nkop+7TpRo>$xoH}6{-inPw}S)T%K#@+~p+xova
z;Z5;V2|_2{N{W7sjed_#AHL8V{`AS`Y5D7;PK9CH4M?NnX{o|gD;(vA%o$}{f*9Yy
zeYRECV&<ZKyy5_DOZxsU$4CZf%#%ycGRxaYAW2oQ+yOQDHYXKA|I=Z8jcZr+y^l=5
z(nIOFs7(>O%9`-VnZJvJnWVyQ7(965Y4nTs69GqKxjMF0s*%bC>d<tUW4iCal#{mh
zZ;3;8&f4(U&%?g9uHoR<mECA1U*&w1+|WyRN<#w%N|l%A)rz|QECqDJboQtZ-4q(%
zzS1A48<i~_wV*oW23rUoUQM@V6_{U>JEgmY7BWk8sx+|IlfSPSd?b$h^^5lKkg6!B
zM}O#z>~-8lfP#8K_dRM|CWSivi(Mfi4+@^bs5j&dVgvb5()`TL9Jt(7OqHM40JTu+
zGh#M&hR!V$)kbfc68}xdwoviCl;t`;^~UV2E%R-_PMYrU1{K5K9h$*v)4x||JeG@7
zpnc`!SM|L0zaf14)blGJU)u&7`0b@#PIW2()u?O0rDF%chU*SrAE2Ps3pS4|!i;A6
zTLDYeD?UxWYbMmMdJdX|W@+w)h`&cm7;-f-9WuF0E$skRK%)}B&)i1s2Ktl%jzb#r
zB-h+EFj-2b^%Qh=i_=5>`~F1rZW4{Ba_wNm)*s(sQ6Gl@?Xd1xx%VfbQV7%{2$Zj|
zl@^?{F%e1{4t=EaE_MH$P3L0u2><gNS;=jsc35vG)0q-t9^ERWZ`0%U)<XS_Wno2v
zTe!sNf^$5#?8tEg6SnH2`0*pl_t|?}XjAjOC99}pNst-36Q|c|*IY(&NLZN-{FfZ5
zHq<FYxn)u0DJ_G11yGbIjlWgBHo4BZ37L3}#P&uZf~GF#wXXrPFG`%TE-7Ll_LA*I
zNJTKWJ@Z%5tcqYWXh|qOpZkZUe5}0)*AbP7^2pb-0R!_EOX$fuJi%1~GG>`Rq{k*c
z{y3sGafFw3sc0W*>ZJXlxuIXqO+&tLMV4dwQ<m?S_luIxS+aEV$e0D^vdo1Ejv-%O
zXMDzjNsu2JSMFEz-&_mMVW0BE<Q{ej7<RxizJ;!hkbdDfpV~*<<%!{;Sb*}1ksn%%
zlMvl3xEo}~cz`i{3nmR?N*<<w?nYqJ>+1!24vHhLg(Dab{D<c6!NN1h=ubZ4B(2d<
z7_>Fk350pzx7v7$ZL>;uGth5`^K{hV1}(f&p?ESC5TqST8FSnD(eP!Ru8OZz!B8g?
zpvOMixNU3X=`QmPLv<}r>5S-;|A)<M0L_R|d|nXK$FfJ&hxM-~fZ84a%I_*nD*Otr
zqOJY&8TSp|1nR2aRbZw;7@+Gc^|`@ADwG<CubY0!Rrny09N>ABF#ehjD-P#+wb#<p
zN&OI&RDL;djs)95D(llT8Q<bBU92fS?WOQ;{2kHaY&wlJ<lRvk#@*X`;o;Ix(T@%=
z>EAv+YXTLd7pJ)o*u0)KcrqFRFPF+kziXq~UaH@eJH&W(v!l(bo$P_0e6@T<C&})z
zzvzpN9ARV5OMZHkZmmDLpj#TAQAt;^k`eKVITQ#8tIQ3L^VZiB56`!!*Bus>_|bs@
zDG`G(sLa3z*eNQk59?!ixCoT!sSSTt;F%(ri!{a-kq13B$9;R5Rim+gEI7oi<3TG4
z45m50lDvr{6VOL&yURjMB9A+xuTJb8BV{$&Bk9M=rpl#wG+y!gy!884w)Wl#)E1KI
zA1FBdm_<VDkg0YTm%buxi@b9EEyTbH-1NIuVo|}Ha`NcDLpH%;_7Cv#Z<e4GkIM|I
z-=id|x^g&)B`njlAS|+~o^y)9!{q%G*eHMGQK*9bvG`fBi>2f_gL8=AgUh2YV9;i+
z;ZUk$5rN!GwF`}+$|_Cp^Cwo;WEbGTZ$N!;maksW%$Z$8nqZVRTIx9s!tNpW7TiF-
z@1_;a%>zyR{C?_SSNDUgac9umOSrM-a5uIv(^>1462RAl1Mgf3A?%WIN4WNCXWQHR
zvL8kY^sj#gLJvr3s5(J4_GXSzFkWv9*Qbt#QoESUouTTvGnVVykq_`K`tP%miRk=-
zxw&-6zLIh@RVZ(uLEYe>XI5IP${Y%z5M%s+8v6>>3-5IpzuK%?Hwi%q77n|Jeb6Y=
zO(2g<#7D}Izv)a8nAXU+A^Ev1h(w=jjm6Mu-2Lk$vrgXtP56|@<<U8-=F=;}z&@v7
z@=HqdsVCDI?|f1>Y2AGr7hKr%)(1hPK53)h!LW@SNm@vMD_5N~@8QACfVh7_T;=i>
z(UWPg38_8Nx;<F`mN;L%1IKT7f1hM_f>q%*rY);hWuyDW280uBQysPU1MVt%ZCE>{
za7}*iBhF9c)%_Lur2fFnpIoJ4oaT^cSUUT6alf>nAlFTF|6Vnacj?gV_}UMsw_040
zzaJvVHL|?Idf5GSJuap>>(uBr_{$v1G?Q~#Z}33(3e*eC`dLKiIU2f-gBSzGHo*3u
z)Wr@Q<=wpw&z~8Vg3(A0q+;c=f25HJB0aU^O4RhiSu-mYPM_eC4c*B1Mh)ySzsPnB
z$29-d@Un`Xu8I`~=nL;y%J>BvDKA&fe{LrH&tI2W>Q>9c6$!dV269a5T6!JnXI!SW
zFC2oaM6m3kT^7?`lJ&SmoZ)rRNQ<oOb^n+WN{{-5RE{1a2PH*s*l#zbLmMvvcn*!~
z)S+Xk1ZMqMiPk#bG9Ot?nvnRbvkO6fsUiCJVop2Nxp;qCpPHvD1gCwob=~L)D-ZaU
z1na{_CpSuDB0iH1$7+g==wgadAXmo}R-o<cmi987DJ*0VgM?=|>8{g?$amf6YWT|H
zZ6AhNI|jN6ha>qoEJr@~)JTbWAC%4}S{3v4m)j{_$Q6W+rfqWD8aO1VW!C>0Q^S^=
zS#0nxURti{IYL!{#2SZ8KA0$fXIjsup^*kz@v|JqG9Vh!0T7KcGLk;9o<ugqMc1d%
zzFw#K=1ksndU?@LL6R%TS32?G3*NZ}0?lD={W#&?FtXdF`RS#&E0JU3QT1IzC;Rh`
zcJm&Th*zkM53Q$2kmj<_)B7?zn}dd<A2cfc#5-u<bR2huBrT(`qhk2=kCh=4ZJ;F}
zN9<@YSA46uUF*Ii?CsM`cL0i@65ur+<$piFGrwsB<{yeSz3Pr}MtQ#xhrBryi+zEL
z(S<~WD4<In=!!xyfz-c{uU|}6@J$9y45e;<3)PJTTF4Ep?4=$EK4><H5fChDR@grh
z9QurGpZcb6RtzK2CJ^}s$1aDAD>ACt6k`?1`*+wV)9GKn>lG;YEAo?Z<rTDP0YKp<
zMs`ew`DDV8l6JI3#Aar4e)x9U!8!7B&fMqZ{6lDZpUk$O$s4B(z^KvutY1d87}?DH
zW3@Cn#yyWz)+zoyq<^t!xN6=3nhKGo$RF`5TTXj$PYx4;u8i#R^c`WqFEKX}A(+&X
z0QOEW$;kg>>@A?;*qSxaWHJN`?g{R0K?1>bu;8x23GOa~dyoMVAV6?;cS&&9;1Jvi
zL4v;~IsZB5zxS^9-de+~X4Xje?%GvfRee=eAv-d*fO^IO&rdv_Z*<th$2<(bn~8dU
zmHL_GIpQl!NEBk?YQd^Q@(C``>qn3?!30NxA3Ix8J;EWS(Dz&vBq#nzFU#d{!NRG|
za)y~=xM?UtMH!{j2a*AYm^%4V<*CaNZGnBW;+TIPRMe9d3QyuS%GGqS?5j`gw(>Xy
zk52nB8l=4o5nQ+&Y81Z~$&{BZ+d7{@CU`Uu7xj^+X4tC<sqG?oZb`2$itN|BoG&1K
z%11|vRc0exiNWv`JL3I#BkVI^&~9bMpr$6}OW)!6qdPuIO=@9f>p;+x8)LD1jRr%j
z$W%c}5zxE_N%3~tb4GbxES3r8(1cjmqnN1IMP(^jf^Kmh*IPObMqcxCjb;LfZg|ju
z>m2iaw(is)mUH*@Z<VxUAX-OS_+~P{RqJmBdq`n`0r@D7jh;+NNr_KR7>c(7rDJCw
zM@JF9QBYAaCbV1)!-W_WqVsRJ#-P0o#Ffk-S+`KcIBz60F%S6>lNY1{$wCRl<|Q$~
zhyN)?5QyI^>bdVs677!y<c$^3yvg{8qt8;wKogD<{6w9o&TsMI&&E)_DN6Q?WUU%)
zPs8NOo>&Hbmam?A5vZ9Tz&47Ta){-8x}X4WUBSZ02z&F1m%!pQ?5Z(}81HsR;Os_m
z22bu~zy)op=n%%1WV-F~OxbH#S38nbJ8JH=@aqqo!9V?sCPQ}aI4u{^c0UBUNSM&v
z@3_#8_k1%V#J^&DZ)RM0uORuN|LupS;hDWoV)t)*K6b@3VJ?j=#CVq76+WMKTZbG)
zc;;3)L%pgya`ioV(Kb97Ba8L`JQl3*X|C$Uk~S;z+_sAcBfolNL>r6N9w|D;Hp`QV
zOc*V_*j1h5h&#y`oisgt5v2g5P@IUWTaMU8p#3`{v&bSRyvJU-e7Dt8+KFEyU`<a=
z+bBmP!I(xVZ*JxyCpHx*M;XD-jue5>dflB}KqF+OT1?^?zYFwGL{wPB>pcAV=^OG(
zQ`g+V`6Rk)<iak>*QH+w{f~Gyq~4i3_(#fE()4^87xkxL$n_5JJ15Rj7iCdO24A#_
z7#y5qUS7O!IQiPCGuv<vsbBHHoyign)Zt^7C>7w8R{Fl%_6CMNA<^nn-{REvt|ey+
z<@Erg1vO+P9|49Pu^lrLO;lj{9g}G9!+;Y5Wx2%!wPtJa9L4bmHY3>2Rjp=V$V&>l
z^F6D=;>@YU^<j!FrMx(Rm$K~5(ycW+8*400&UU9Hmc-+!HTysOrLCaBnB@rBqxc~B
z&=;g?5Gg4(Fd{mmkTGZ#5sU;`3nqGYR-FL?J+-3kAvQz^B_s_MYU}L+?pfBStQe69
zn#kcPU)r8vdo}?u>oz=yDm19v;_=g~uOAwt^pN|9I+6HkP6N=is~||7EKH+*2#B{?
z^3^?T(@}YK<&pxsyC{y`-?s6B`FB4PIy(!9ktM@#`g48DB0f)!J$zzeg^%}Qmx_vt
zz$k;Md-OojH0{X=TWMO+STFqjM5KM`k46!#Hq+SYGVOH$y_e1oS`Y?&8lJH5Mva}m
z%|KhXoW8=hkbGIkR@ZgOnK!0p8+|DuewKc?%tr&;Y51-&(@}BhoU^x50}_4y;#nEa
zhqR)!5WOTHZ!!{RPTQ8J)#Cc*Wj$lJdu*e4BBI}qP|7S)Np7t7+F1D2f8GrFcHswX
zdV-IU_CMlN8_k6bY;&CX935jDD3|@}HX=`MlsT6K$-Qac=a+VU^IHrkfil+4@6wvp
ztr~CkG3XWFP9J$Y*BR3xx}iR<x}fD}@Q59!tiON*>5_M?C+ZQco7UXM#!@UF^uYvS
z8uaqITXv>AUvt<k=WUGku(6nol)l3>!pVN6Rv4zaVzPR^A7^NH#sa>CqyEAa^ivI|
zkEo0&MpgSnM@vF`W5;tLL*t`^R`IhPtnW%M9CMiagvSNgR9uK&Sq%*L7q#NP8(4Xa
zNC*cl<W5WKDid6-hQDvOL)2^>upoG;jR)oFXIQTq5wK)zwDf4y1}oL~Ty`R}^oDN?
zYy_YQo=Ufm&`Q*AB6hW8j!Dbr9^|F_z|{5Jryy^DhTHQk47$Vo4{#B9a0CjWKqrAY
z@GQW48s&aHA`8%5f7IcRD{Wp8u(StNl#F1A38}{A-ENHbT&yi^vB|o2P*a~_OPW$A
z*=YRa-l7rkX3t@A>P*p8FV9`pyLfN6x{80doi$jXv?T2i%^C3F(5iNLH(Sx^6}M9M
zuKL-)oLVYR9;0y;<84WnX{XCnX`t*$ww6B0-E=G@Na1-)xSoZ=g@4{iLb%tk?UG-1
zEyDj3f77K=aFON;`Z@Yg`NF6)f^U%~+WDh9BR}fZg!gt0DT*n2LPtB~y|ppnC7Sao
zUrdn8w^8mda<ng#*1p;6<rjVRIVR5uibk*qtH!<gIOE5Q4FQVrkBHJ?gp5zHlykhn
zEup`bW;ep0%RD*?iC)Tb(my57k+M4~0*6Z%Qc|dUo6!brNqcQMv?^^NjM6eF>woNL
zpwG##mv%s6fRoQD`^6w1F2i7Qm0-bMp5iJXyyG=BXw3vBY?0h?o``@^ceJoz_(eJ^
zctuta9nNc-9t|auoSXKC2WTp34urIyr|P&XTOSoKQ{Q5(tq-0fvXir^p%1^804+(k
zk9orM;9DRTm<ygSp*g-dGYruu7Z>rLy^YpS*v1^jj%g85cpKq??1(=jsDPR7Mt5bX
zjBNN57}CxIwak@`1!nF@9|?G;svlSCxXBT-JLO}1?H$M+-iiAy2X2p{&p~tu6q%Ws
z%?7)v<cQx8g#sNr;P85ZsQj}I)N%Al^T%j<GpwOD3c+4X?H$Ax)>j%^DHn~=OM5=2
zdRwm#0b6$o$$FOC!-=^o_^E8-+syR)VeUHn>GU+9wC$Q}hMQ!Mw-F$S<eMGJ%#3w1
zws2pB@b;3h8u*z+q04oUcCx?Ny>4tDX<nN4NG0q}!7n*ma9pgxvfJdBp2@|dlk@k~
zC-*B8(`R9Z@h~8pObmncwVuz2fTGC#Qr2q-S|C#r!I!N<q$;;Qf<@lV6&o;E-$Fo$
z@oER5rZcNDQbwpVYv77@rQpOjJp{zWF=<+7pAbqZWZU{WjF8fglu%Qul|fHf+N0kK
zeN7J?rdwsVaEp}X;n(7&nkEq$QAPEv++WHX&GnrzP*15Oe%M0wQI8Jaue64=swV*L
z?SY=E3B-$yQhtqP8|Ih09VY;5T_(qav_-ThI?=O!yGTKN*YC6Z?Py~wjiW}w!BUG1
zDqa#qM<+EqH&?6Cc3Wrk6*sr~dH&_;_R`{xmm<7gNv+K+d4YumCKgrzW0EycBYU#o
zIMT;|v9PA}qS$k-J8B$o1*IU6{ITxl?k;+V(X<xNZ0!r;rz|ii&M@3&v-O_ZPVR63
zswzU-c9Ub?8_YsN#ZibK7<lkS2yIN*hQbmH^6hu|wOW#z3^~C{r5uJtn8zHTU!W&w
zV&4emd@-B`%^FP*RtD&0J&VSxLZhRnLvUBIhe<|0qoANYhhzx36KU3$i9Mx)lX1$>
zMVM{3(f9;T{v@Qklyz#nEiN6uvV?~~yCr|JBD+CnTxGe+<@I(*=LzDZ_@kvPyR2re
zEWCobUe@IqgfC$Des}3(4n{IHcUX9w`t*%zktAI)zxLv<%cD)*lUe;H6ODJ77>z7+
zFzznz1LxJEy@DZNf3P$|<kAInc?(`R0qzDINOASC!&R8INq+$wdt;lY4X^z3F*WyO
zoE>x$TB3yG-9hMTt6#JwV1SPuK%&9VVnt_X9e%9$Rl5H))jZ!y?{-}<aQ59hw3t|r
zlVRah1FCgP>K3VgZjFRo_4Oi~Pe=r(JA`8>p&8PPf#({GLKJI#7I}9jn(Fv6`B-)H
zMaEX=R-2BRjeuEYy}ZtcjAboXg-(1Y-knno8Mnz(MH}~vuIW}v{$%(ft3)$xZ3{s0
z&bV$^jzQ4nrNc5X>*KxPtuaDl%uI_UT9-^~X+ZPUzTiu<emMH?7@e5Yrs(e$ftsF5
z<33C}rA``+W=zOpQJ|_By&2M6^y_ntfR0(SbSA-@Us<q37-U4RCA+Qv{y6J>Hc-*b
zZm=SBxnDJs{T-5>a6QSp+IQXuj5#r+N<C3WvyXhe$9UPJEWcyZK=a#~I(`Jg`BWrG
zQ%FvsJB_74y78DMTG{+L2z!hJMbh=+d%wFwPIS{&-<@`v&}U3;6YPRxX&hDmyw3MG
zPm9^uR2hG{=NHO+yefk%nh0i#FqRC_Cv15Bj5>cAOw}KT=#-`Rg!m<2;%{lm-uX*f
zns1GWzbdxckv9ZWQz#?>RvF<x8?#0H>&Zkbqh<yl#JYE|KpB6d_YwEY&BG;@0hQ3!
zlq)CBW<{Rqx5;3>u;1Uww<?YVDsY{Qdh;C7jp6YbVRq^_1_|aDL`?CL;!-O4mh$f0
zE-fRAF+g8qE6E41z=q~{6`q8&I`m6Vk4$$*fKsNw9>8R^2UxPjl$6x{4AQ+E_cQuA
zQJCU63}q6u%eB@igJ~}%)7;NBKjwZBKK5HT56$;K!nJMw-25*8=xgY6Fq+h{N{wz!
z+impg^$3>Ma*iTfX85Ta=B9o2t^v&F!L;_+e9TN*^$NL*&PiTBHj=vWxtBePVncY}
zuC6L&i|L`O$f+5!@*x=eC=fdnGKnm%Fe{KVh&_K8x4sS-ey3u?>{Zo02Tl?<SN3Nl
ztnW%ze$~A`f`O<$G#+P0L}rFlM7G{t?d>-1`i5q4G?esPVsoX2?2CpxV@tS`i{=!9
zJzbqt*Y~ChfJaM2l&%;5<4zyzfv_R&M-6L(Qh?HE-$GEKkDXATM;%~knncTsd^giB
zPzPZS1@Dpp?wnia08SPjojPqrbRJeOT`?>@(U|*^!_zeyCd@CABOj7{xRc_=>H-1D
zK{5o66g_DFU@a;ToXN(??UL-JiEV9hUoSgxV7{&Qj<peG@3Q=~8F8VapnQX-ui@oW
z#jg$$T3+6|6{%pYNFi@P!+iOpVO}6c3i5~y^HwCCLe`#GExFyjGQ%W==F&V`9U639
zPQZHlEG93(ER%cLYXbc{P&(g%4jD@39Kf#*IcYyVkztWCbg1O}@dYb*w7~oNT!KJM
zFdl}EAVTIQeznqaK`9mln4kmsSF2!)tS?IUKG)Vz^WA=i(wJr6d)?#FoUAGCBw$$9
z5_Po4enroE^N+QSfQ#iMNfPf#bnDw}J_)iw60mYPFb_}tPClVTRj-6vF4;IVt-SSh
zf?EOe3l8%bG6jEDpq9eqq~T<DeQ-(5a{PNfGW8qSYnwLRM1U%I!Jw!bjmgt7fyF1G
z+4-FgV+u>!4;KUDZQ$4wJZtVnr}%};FS_Unw?B9CqO9CnuTI<U6H262T++1AtO9V2
z3Fz*A%Qxjp6qq*Vq`Rt{CmuimerT9}N%_~f8@=Hza02zb%wb?18sKKWj;`-6EyDp!
zt7PcLl90h_qahvb^vHqg$Sc-%c?m)ihMhhrCdq_^-W|OtyQ8%Afcx=@3LT#g-DOCT
z@!BD#W=n}srQ?9rhV|s9V}@>t0X8D{pI@H_!Rv_%2^Q@32LfI{j&$+b4!D03Yw;&4
zVI0YYY2UavfztPNI=3FqFF|kv<?CRcChylf#5ixUkLVc_%Up?IM=poz#3nLUax2#7
zL7dv_urC^62<s-SFYt-oo<W0lCBE;^R_P|AwLlsI!xH#lLR*wR+!v$DT^6yA5UbH2
z2l@oOLJC%JMMv*g*+HBZ59*q4GJ9XF+oTr!o{)}+*C{4mo>z)l#Xn;$;TK{st1H6&
z(}1LWR;oLwuFK_yL%~UTwN6;%l1gSejx=BOyD9ptViosHhw7_M|6^~~z^n;k7`47b
z*s5XyQ03qK&EqZ~uY^z7f<={$jV;*R0!VHt>Lu~t9Ctm?g`R&tNK-P*vEThRhI{`b
zBo+r7oATne?cNOGb2sgXc3YucV0wP}mtzq)`)NN4FU1D^HlN_~feM!F4(w@}u)*q2
zIo&8osEKL(o=BY}odI|v6ytx{_Yk3#ONyM`b}pXc+;tLcYBJx$Qdb8i2b1j>rK3Yd
zl3lylP&DOw*X&roI*gUQZ}Y?WTYEde084E#dstpN<?zS>`1yB!Y58EiY!1Z3mW1}?
zgY%Sc8lrX?>3r_i$xbpgo<3JOq94RbaY#1GnnmGAG#x%*Cw%E;sB$nquN0NgYjBYI
zHE2d~axz4RiJ83+JPK)-L8#y5mu3n1^PN$T#AF7lPS%_FKUSLi2Crk!i<j4w<V1OV
zzYffH<9W&SCYYusfy-yQT=t+#)cD;9Dtm)t<wCPti`7V9jWe9Q627Hj9;G%h5Pnid
z;QO@_Ku`Ycj$Y@bN325>mSqhwi02aL<ZC8~$5SD@I-=HxtC|=|hYk>XZcNsY6l7Hp
zp|z?cF}@p6%d}otR2uimEk?;(y`6X-*t(SJy52Pa;NowTin;|Ro33B&l_f1w+bW`C
z5#$D{M5`_9uvb+^QY(&))#joblbxp5Mm~W*{d^<oUhcpVxa0U$Nl9tF`-M%6nJT_8
zLY#ogo^$nLQ<1Hos}ebTRZ{-i^#wOrDIZPE6bd_U-1wFx*hyFz6I@QZF(sq@J(iC`
zzal`WkZgI@NpRs|uL52h7L<PG@X5F3oyAuyUIxv10?!0hDhx3$xa7YEb83r-_-E(k
z4KUR$kOO)+#wQeKPofKc>*{PT>Q$=^VYkIh+wFNu``KE50)q(vC;ic`j1r{1`eAgz
z<8u}irXJwO@}5M#8Q+!VRSM7Bg|3IXHlGIN5%xsAFnfVps%opCuVX<ussIX+E?Vzl
znv(D4sJ7NUyQU28%P_{#E3h88{WQrbm>BCv0iDtF95OjlY-IZ5yf8rdKTfB7H+o}k
zi6Qw`_m(%L*MAqDD9Yrn0-(Xu9x2Ur`rob(|1_d{)vMon-Ex(=2rb6EG<*xy;Uia4
zUrPh?r3r~POYVFf{q0e4qkl`*r(CB0JkCcIK9{tm1vz-bHpb4y`KCLE8ln?=+tOBh
z)(s#^!KY41$i8}p1(hX_MzF>q?NtaIUjVGn^9NFZpbP%ev!67ngQ!%Ml+>sac7@^1
zf~~gU`swL&G;~G#846)fcyKBWhTeW`BA-8oT%Scc@faD@S8Wddwvq2C{mNbD3R4&G
z<mBWMj@s^*5F)ts-6ga#o2g5bm$K=n_|+F|Tikb55Y^L?Q)ibG+>pNKA0OL&<g!lw
z-S41mEA8bM(c#06Y3?K!datF&&hl6(wZH^h3JwEFDy8R6-+^{eeBu^O2etFruIU22
z7W)+M-Rdfhr{n53=PWn&%39;vV57*xT!bjY+^k6l+I7;3!q6ed>2#T-xk8uS$v&6J
zCshbTIxBg6cEbA`=4AbzFfODlU|0V(>(_|i1icrl6M_2D{`J0il(#C@qC%XJD0pzh
zis9R~ZMSNKg7xn#=VNZ-;xM6pPbwT(T7P{v#e;NobpiP<mim~_C~4I|?K3%g6JLpr
z_>$XNRAZ<o8u^M;*_z;F(u%e!<N8C7MV_tH2}4kUw6p5>(xyX9xAF1dHn!)89Kr4W
zEmTShrYL3D*_{+o+k{MuD&{%A%DGOypD^|@mWRkE`$)%OC#phP8$)Z!@^sXQXkq*0
zo;+!+ZH2C|)2O0*YH6d*K`zP_R1mtne&&L53JP~C(NINnk-P!Vc^3zw8wi>ihcv)3
zOrY~KSK{b=`r@SI`rzndvzt4WfXT8BuW$Hvio2stGnE2MJ`Bb+wX`{;>Mjc+mVA`{
z_!m*6+&r4#4e7^^AJ1B!PyMGC0I$XTYadib<+oUNKT;(?Y0?JbF0FiRK228r4hNq@
zKg5s{#1XQrjepOvE&iDz;9SB<#V8h+*=gJ+6_}u|^|M=A*SkCtoE*@e%xM|9m%CY!
z>U}RL95g2EhG+eFSAy6Ex0Y>DW6ES9-Y*Z!4`DZ`%wfG{*1ggz-ZC9Itu;4z*meu(
zQ%;UzJ#s~9fF2qiTi%q;{)L??hf3+=2U%7Du}5#unN=^;Tmjsc(GjqLf<uiKd>v^E
z@TM?)Z1NnQLSyWX{lKCcJPTQLsxSB;9mltC;~GiAujAV8@0w!a(aupgLqDPK+Ag<x
zpEP4?byVRH8a2Om%pIFrdoRe`L*$xTq8*9WgAV_a-T3C$j|Y#Gz<0^LcGRynyoPY4
zX_(@rX>ZOhFCF-8^t66d4P;T7S=I7V(0Tdd{PiLF4fEYjUV3C9!rn}gtF6q-Duh9$
zcJPC)ixDa!Ur~cjI^ay;78;x3Iay{s8+l{Y9mxh(HN6O`nR4ay9ZuzOsBz$F@V<6Y
zPvNj=84<W1M30}Gwop`zl~sLb`vqNKuej=6QTGil=?Yz$5kg!kP>CuktFHLACHC8b
z1}k3ZGdw1($pukIJ`$?(xeiZ=onJa1Ke=LK&zx=O{W!DiCf>I$W6Ho^QZrxBU&MBv
zAh=tYKyNBbuD)wET&<p-j$64~tM~_27u>~~7Y#naJXQd`7q~Ky`ndW!P*VqG1X}+{
znBuWMncX+am=pEy-`4As1r_Id;OSYaf+gLJW}DbkBPz)Pd98(D&1^+lq|UTuvMz*-
zNu_<9*QYo>Iw42s5NYYcA9v9g*FdqyU}Y~N7WHJlSpl$0ZlusC_aGp<X=k)mANY!g
zqUKYiR9L(ajl4a#v4tX3xKe*#uTZJ$K;Dl#n`D~mu`6c%VdEY+W!CPe{nv1uO-6U!
zMFjr4Uu#V6vJW<qmZ2=}<>7Kpv%oA#{b{!|htsg{0bN3RB3|FZ6fmAk6uK`UR2Gb4
z(tGs=Auu>hhxe;pnOZL2_NZ6HBd|c1WMoX;l84OZ{Z&ya5NQ&lw^f|Su#umQArpOO
zj~zWvfAofxZ+lr^ytrX)__fKG-q@1^?&V&k4MjvnA=Lb1*GtLFYn8%l8MkA8F0N_w
zt$trO^ZOLdD)0rnzR#tJ`(=W<6`sWv?T<;fv)4G&b^3yuE$<Qrm7IN5e=;lT<)SB-
z>L7l6Be?cVJ-Bj;M-TDKi0b!37J>VPM4*A^>q|+h;490YrX4lPR;Jh2Ef><}=J`A_
zZXxad0f-OzL-QK&yQ^nb<k)b#$ON4mJmFdBYqjZDe$B_7$QvgEEOoTuXQK+Zuk{E=
zO`;ZkfwE{O>gdkYboWhyfywyzcoZ^hX8#lvy=?O|(e`+~9)&YthW34IBCh1YVw3ut
z*9RPz>+$k@TVnRqdh!PE`(9$>zO1#Z7AFj9_CZa{z$Gac*ug9nQ*cd3r!#Yj5O(f5
z-%98=({>w^`ni(!J8l!(9t+sbf|@8K*dv-Y1yDjO9ow*vk*9ueW4&60p4X}UG_ryA
z#g82;dA_4mtcR^+4zgg2br2tUbJDjBJ-Nxya|k|p?p*zy_SCz4lsQ_YQCme0EvPpS
zK*$dug9Pe%_i2p&iz2#M3p;}(HGiu*uWp5u#~V$Sy5@ROfs`a+dzGIp3$HK{-N26p
z<G1xg?2%u^<y}N$O<CXNHk0cXH$$BFC#Be7W~(Lisf1tT0um(Y+|<WizT&8V`yv!N
zrl6uhs;pX_EFAC@v6)&}47>Pns8cGSHr@Ab_SK~0sHbk}g$MJGA3wSSc=<h?%9we-
zPZi{y7@YmWlHcBC|GDTIbuB4O&!?M)X}sEs3s#Z8TO%2lD#;QE*rKfh`l$FOZBu-{
zJ{Af>MZ+)RQ|j2;q@@B!M21&#cC^d@T_B`}m6r+$hZIS3CLV7hE9XkwQOkdtlSwN0
z5b}q3p(ovL{a$I==lY;h(|bF)q|U(%0UEM~C7;f4#sXG0UF~@<cxO(VoH8B(b7vlM
ze^!@(K$4G=M&fh1fhW3%@?H&aByoT)g+DikUZ!EbY|5NGsJHb2c@fQj+}hlHzDkFY
zIj;E2yzDClra2*$Jfvx1%SBCIh({y6*2Pwdb}kti3Ol2Ib&_`;1*_5gT1nBwo4PZj
z?Xccy^LCdSS^9y)1w7IuNEx%Ej9(n>?2_x*AFRH|H1ap`HLcmouMDT3Ee)*Pp_?9g
zTFk^gsc%RQqRh+C60Y&~J@EE;jyxRlAW&M!(9rfrJakmT0CMU@JlN;vNOrh2nx~Nt
zY7~XxKc|#>_nL0$j#OMZx~uh%Z=pkB$C{_Jbh(;gnealv((VYKfSXq7&7tbg=Za4n
zd3Fm0nVDc{Tbq4HD5vOfENo<CE@(TF{1`3|d=^D;s(ZwZb%+0R$NlBDHmh&-V{s%I
zhg1Tmqkzj0jHH^BkU&MGik&>lx>xh_bAQw1Q3T!#P$377hc0q3`rV;|ORfEVMaw(*
zrnB1l$d<biAgU(Nxqo|9REDxTVS{sp2F!s-7z}RU&pRX|L+Vr8ct%0SrkE)UqIT5%
z`9oe3ti8!cC@Cp9U|7Rt7F?O6@2y%0)36-S_a%6HQ_aof=<T-UWI93Oqf!3xK#WSj
zkq?db-lz}Lk(2lj40mIDT!)aoVmLPT>FvegV$`p)4r$JPa|8@K0DJeg-JP*$V~F$#
zy>iY{;z>cd1h_!agXx0$=SR!7_0Y)BP*xq`@5tb0CFh;-uhYi!wbmy`zW3UI8b;G5
zMqN=mfx;8Bp5)4<m|psI<G-B33lY4(Q3m=dP`urJLPSoD(1fm593{nRwNsr4zG8nN
zvQ5U!`Ye1WR(kd3Y)1cug;z{l5%x4Tgy5Cdsi3?EQCC!9)(W~EvHLnk$1=8FO6Zyu
zSF_KLMt^OfDfJt9`6QzRPt|(mAj7uLT%!E3e!r$UFzw)>nM&6~mN>dnO62bfKh8;3
zR|KlheLUi8#tc-*bBzSDp$OTiII6wT%??uU?FS^&UfY$+|5l=^;Ld;pd3anLb!k1t
z1hJr<tzY}JaWI;W1mD>S?Q>o{$Y+U+#vjZ|Ofhq-=-pFGPu~9#tmODdOSA98-p5a<
zAhbOEDSQ!k&RU!<ihtp0HuFc|QK*h*ZQ<h7&7GgTAl#<Gi;7HIn-I+w70HnGwtw$I
zn|TH0KE1y@E$t3K(*(S~o^HdBP4x2Yl?O8c0GD9Bq6pJiV>uB=@@;9^(A<07mZ!ql
zhepp{aLIN@5*WGpTWsG3d%(X9nieKyQ0~vwBUjzkM^{;Gxsu+g4$<zJ(_W=t?LgGF
z4!#z<YRzgu1<OX0Q#-mFMU%PtYZNjE)D@rFEbbPL8=D-{Pw5zO&SuYI{J}F3fP5?(
z)^iT_^;IzL0vs9cK*B^><?kf@ErwE3VNMG4n^-D3vyY!VoecFwj-6=OO0z`Bl9jU#
z(f2+}crZaHt05%)z%=-G>!K<rKr^Vh-R&o*IpWHIvsEHTS|k0n`xA#nT(N$eo;v6G
z@h`CFVZ+usfYTC7b(@}BsNc=Ec-G}4XZ=FV#9@jN`Kl}JE@$ZuN;~BZ7J4_9DbAU}
zcQ&akv1Ghqs*0Qql!60AO>nEvS6fx)F>~yCmb-!J*im_?E6Jzk`+ynaYHpp%T@p-g
zovl5?IL1QAEM$qRK|zLby;+!hT5wGxr6^p<Qpc+@-kOR@r`&e?$vyb_JeioDw#`<3
z5Qy&eGvWt>;}?PfQ6dI$A6ConZf%IOzlGHEuk|Xv)83O$LTQrOsD3=1N3Do66I2!L
zmMwBsj;TBLM_9BH+sgwrcq}F#?VyjHaF=T*f+ZiK*3fAggQa(t+Y_Utz8`O$xCy>3
z9`ajkvS6HScr^(XdV6K0zI@gBl&OqX%ErUK&Fj_cuHFE%S%$PFsv>>4B!msAX{;!J
zT6sPO0rtETB|7c4FSYphtk=uoO!4r4Wbn^3E@-;xSz=~4&B|h*)xmrno?BSq0J_WJ
zMv{JMO+WMr{5lY5PYi(HbLmPcSr8WytFfA8&wIC{1+M^SSM|!c7hSvlwVCs(JJNS1
zw=H{B%Akgnq5V~)OTZE1!keLgRHb6rc>r!ca|RlHCG`R!2^9tsH%QaPk|&C5QCZua
zN6FbchR}$JmxijM=FY${%t^JVOUhAF(A(Wijq}(&WAU8nJj5X)s$jTzuu@I}eRCY9
zfV1Ow0(vEWC_YwZCj*Z9MUCoI)*r*~Mh$KK<ZrTsbiL4eNg)|bj75()HvNK-2(wsF
zH}PJTl*$BP01XeF?<IHjCwL|>F)+`~4Xd?3`OUlnh*>ektvyX$Nd5gAGfzT{%|#-`
z;HI86?MAcB()`&0YY0|LMi)1M#LPq~1{jiLw#LhLR5ATuKEJKP#e>|Jo{z-jv-)0+
zEYe}-HjlQR>DKh*>cXQAywoR*gQw})F=cAxM$VLodQRom3nR|dI)&NU{X_c#@LT!A
z7W3!#JpcBz{c&e6Q3-#U;dJs>#b2N8fyOp3$Z6d?MIeF)Oz@zXdjKRzXu7%8lQu68
ze|imfIX72hY4#;cT<cy)_(L1ez))A5^{L2i2Pajt1eq+sv!#la&+tobI&43i6-me*
z4{d&sV0^1JfOVWb!zLE&Mr5tXmNGOg*FYO|t?<Po_!bBkbz4s9zzcRtM&gqVB(WqN
zwO$Xa1u8@%u!ITdR6NA&_*&#X@v0uXic>&yDbZ~LMG6lqAPqtNOhs|#AupC_<xp)Y
zay_?C)Jq+q*t<_VHzqNllK`EHn-lieF5nAkN8Amw_r=jyjYvK>OYE(lx#(BT``R}u
z@Hvx$^$dbs8K^b%N`+qj_9A9Mu%X17ji4dn<ZfTImIF1;Xz?Vzhy5wmU#gWxK|_xj
z9yLii%;8o&H?XAU^17%{$wc}>OHY-Ks^7>#(LINdo+%+A1j%@|RhF(WKPC_UBB4W?
z@;?a!9gw^DH2ilPewNy9V|lgZ`=MHgh+B_VV6<jI*vb4HXd|rlO6!h6p9Iz>9~qxG
z*-uXYHHv!)A`H6}`|;rcko^7Ni}cez#5E_2)nn+~<k2!U&BZX9XIoAs7VxUa7)N!-
z+4aX0nXq1LR6RL4i9CL*oGV+QD%os9^c3wQkdEc-U%_|>BuO3St?LlQ)aIz)q?-TG
zID25|rt&yCRRjhFA@U&*IS#T(EPZla0`k4|s(7mg=*HIH%7sav7mSpvReiRt3JTWp
z{lzoQJ+oo|6XW4tgFO(02`~*^c~CHqVWzm>9&&O15j}Iv0LLQ2VE3UNO5L*0JZ-ln
zfr#MgLAb8t5ND~15fP?2Hgr9dboRqr*P6SvXv2>mY`ks7dgw5uPmL?jW=$L<B=?r#
z!Uv|7h1Gn{t~tlWH_qFTv;TwhL?5MP2A@a8>Z{KM$$jLJzB&CN;LgCSV5$>K`3dt0
z75$+6)AR6H!Rc91cNVP_in=x}`W_cVEax_z&62?cQFk>(E9q2ipA$D!q2a)XRc;Oy
z%E9FEZvC0z)kdStWjwch(PEC6O8;lCQXpaz*80lZW)a6g<V$xh%gnAx2fDCcxM;V$
z+4)QsnC#$Dym0GLBZPz|Up%Ww-jg@qBDd2xR1+ff?pyiWAx^>`vi}e=|8Sk^x21oK
z3?yWbp4mBPSYABVDRPFH52d9zsiT_}>oh3Qd^*g^%(S|c=lc<8P^2P{+*yUjx7!}0
zS8w;ZBu$`qXlN>g`GDvepW7x?jxY!nnC4}?J^Gp5^V|yH{uN!F&)e?dK!l48&!vO9
z=Zq4@Vy{M)m!AuKAz;x>x=y56*>83m-@`Af;k2CAc=X`*CUS@cX^{qj+)^`R{b<m)
zXP6Z>ZK_6$r>FUNs5r&qL`vSq^GayJDGk8#3$w|UU}nw{L-Wd}d?uwY*F;k`*507?
z@FQ~2K~GPe=+udEt9g8EM=ybB4L4T3lS($@{09cOuhJXFmHP|qh{sjF>H@yzf#Re=
zg(<E$1Stg{56X3AW1j74E8AHX0c!Nv4s`r(|Ca))PYMvIe1X{hWb%^?SO56!bVzTr
zJyYsS#QOM0d_P|61vq%skAGULvc<G}@puDHU%RPR81q|q+VtgTVV`jUG8LEoGM`YJ
zc|9DprfO+_)z&OsHfk-zY?s+eby{0xCc2nPxxoLjWq<-063N)_WQer)r=W<+*Digw
zoA|mrEG*2w+;z_7XsHD4y$71SlJDJR-m35};QPcl?@bdB=McUQwkss71-NiBXE+RM
zy+8nw$vt|{PF9<qM2hr=pfeT+7k4h4h{b_W!0Ve@a6tX5k<cWYk17nZZu5I?*XIX`
zGX!r4_{6465avTt4x=+$8s~tZ*n;p%V4#5_8rqzg-1SfCjbV3q<9@aZ2WjbYFmVI-
zQk8WEXLDTP#g}m%_w6s{^VO)trEwQr-&!IvNa+TB92eDM;VD<Bz-ncB2UIUNIHWRu
zfR*NJVlCp8)+;V9AYn*Vg(w#aacleZ-;8*lO5GnSR(dI?l`3udd@tANki-3*Fy^Zd
zh>`%c3aw0`G|@$$1Yz}Y)_-aTI-RYav%LcHNnmJxIage)?g2(&e(xQ6%$g1_)q}15
zWnHcXpq<f@+>7J-{6}HRXkx~6M^)vj^XSndKHj~(>%RX+5)a1)<SV#P6=~^k3xrqz
z0BDUvgg7RffSA~;^2o?%1q{s#z^n(j&xgB(S@1{*<;##Ucjp5#X0{h#<>*X0N1kYB
ze8C>C9*(pVcI0eAIK}9lz(8;h^9P&9prvzOiC#&`>*FpoqAErU1Vq<^IV#{cYS&*1
zgNEb)#qwqcpXq1qe9OHFF03!UjReNs3EHUme30#^@*)iNG#*Pi%lL=Qb<~fb*FV=2
zyEYuGyvmWv)@dAF3y-D6aT2MrCF(cUhX1xCl9C2Ob)@cBR~+tVDWa7w+Rpirf!_=)
z){blLKFiGY8IfGi98+$*x}v)^+$hHCBB}z|<G+1}#~>am6Lw3_34Xr3q)`@uH*elV
z#<p5hIN@IWSutX1Qr-n(B^G(zy1u~@m<#<vZI-hfYMeX6egngu19OZ>l)<goCtkg}
z3F6L?rDVzfwa_#QKk-6aipM0gB%&&-7wc@(n*m6b*u1@MuG!>LvicZFFQfdE<1kMm
z@8%a7mKGop<eVT*2zMk4^bn3?C~L0lR<1!o5uCtB<Bk6ObMPl83Ba2kPKDiGUuci7
z7;(8VXV&mNw)G@}hll4z(h}`b<a&qBMgABML1+4<4!Qq3A4U_r1iB+7mm?Carl6^9
z>v*J^i)H7K^$3(%&M%4wlW_GI{m3{`zinarPedmvRG6<$Jn6{dq#kv&IZc4abn|MI
z!gsVcHVjz7HSw@t$?|bMVsQOTYw_zxcwe9E!Snl0rII~p@?Qc_9(YtK;g4rz!K51^
z=f%SX4m}l%8(|$i5U=_(ej@ja^idg`y&1V~CrDyF>z?Zs*DiPG8UNfsj;<1AGYU!Q
z@vS<BIt1|D55SFuM^~nvTQo%%wz`RFnAvDix0z?t+sKXO#m&;FrmCLVB6_V|T@H=J
zf8DOXSHD$?)VzBU)9HFj$$pWtpnk_833n%v0fMLN6uSU~SqyQ}A6Ps<VtjI2L2pFx
zzWGDf<|n{|IRWBs5$7jA0gt1<ShLD1qEw#L<PbTE*JYm;8JDrVrVZ%XybgFpl{xP>
zGc0FH4bKm&?AIh@T%&i^*NtU(u$V;KU+DE9@Dwod^Xrm4zs|@YU9)9n9rgIpp%jLQ
zc}xmApa88O(NOh9E~7%Ef;gL%r0W;GA<OkbJL#BY0~czZcm=kIcxRD`qg*afRqr#+
zH@_L=qbJ~TWe}XU_0z?D*hFR|lPJG?tA}=2JJjl8l2ppZ@_sPCg#YJ5mc<X8oYWp}
z)s&|ZNw|N+!t%)>^)#~9^Xj_QI&zSyT2LVoACpoPt8jp(kV||6=VZirek0_-Gpa91
zU9~S#HDz6}Fn#o?2L?Cq%3rEuNZLSb+3xaa;o1BLmXZFruskv89<jZPI{~6%P(ga)
zmh@Zu%)k(hvIDOh%hnT0n>HWV^P@y<qT0(7KQJNUuTPGj1JC$7*qV4tTKE2<h5Z3v
z3o;q+zG4>emdl|#%?q;X=rp`sNwl@Vz`$@=a^KQaYcd(YCE#-_w_0qRihD1nr<Znf
zVCx$%FL0SHLBNg<8f04_;*++p$KGYstOUPKQ7u3=F?Gb2O-RySOLg6w9@5cmy}$E}
zAVm&R*Rc<ZB<9z$X}vUcKbulF0peweTs2frQJx#qT>(zAwMF34HDKDUEeM|x@lM8>
zW||<l&Fy{D{MK8^-n6%UelRbiq5?xvv#==e0zf_hYe_+0mRBFT9|Jj?a<99~El$0y
z#HX?`LNBrK@kwY$qNU?nrCi>VA;t<Do_w$w+>RBlV5c3r^C+`U?&1|6sDJah2Iplu
zxY@rQ$bL}+XH;d~c6U{zK9~|u3Ed<|FY9!O<YiU~zCg)&RoIZ>m}*Ml$9lRs@`x~n
zq}8G}%#IZEl{Ba+BAkgob&TgLd9=ve``YQF3uII+{)oVW4@GFFd-PjQQdMU)f@Kha
zB~v`rStK_hke;l;<;e!;(se|5c(a>$Dlei6)+=wJ57mlGPihVp^eIIT7wC^h`r?(Y
zsordebIIPwot>Lv59G+JMHm4M;S>8#?+HhxjFzeT=!+|{*@Z!eq@$nXdymJ4>y#xL
zjl|BS{;7PRw1a^c&Nblp1cpAolmz;z8>M^)3d}YFvEBgj`rzJ*9%2S`*9Tw*B)?4&
zKbcwDt+y)<2JW-8-6+{K?obQQ?v=GMr+c4&%$Z&7Qk!Xt)^qZ?zjf^h=F#=o$zeC?
z$A;n+e*eB3;R198WjoPx-@t*=q;W5~iJA_lT12t|z|H9XA@oHUq?#y2C?ynmGAfD|
zhHEXnA5`uaPb1HD#Tl{y^ihd_$b4vW4_ZEc2v$hoLjn^f%P3adE*4#L<kJN}Y}g_y
zZ&F_=E_v)Ry(oBn)I5QXiAo}%1K<|x&}+3~4YCM4mc;wJi?;C=Pgeq9a%hPFKp6l!
z<wx02$fHFi4DrOC>@6v6$Q4$0SNp@jZdV!R?l0Mee3sQIJstL$rxbap8HGx>C)9@b
z%~2*4vnWhs{P5860aDn(HDYeW5r^CIU*U5=-;((J`rLig^^mCfb-Q;>SubKsqmbV$
z_|e+1hthHU)ghjxF9by3OZOxGWmA@`q-Me=mU+j*7A5ai=l43}RGR!MtBGchKFLAd
zDOc1$057X%WiD#g?)K!@+Iwy@z)fn7e#9sK$x%AhOMqp9>*Ds~SBJ0h^XsMD{u?3p
zdyKcd@G`H@9(}QRGT-kSQDt@3st2QcVne5jCLx9(vggggEyIqg>5F#(sqx6H<lKCW
z_=_I8f{yr{V_HMEA)G^Ts-&*06t+kHH-LvyQ><`7RjjOvg@giGO?Y_D^X}|!hPapU
zwRMZyhXfbU1Dy{SG{3K{!O4%j3&5iZRYxV_HSU|)01)0pNVfo8pT`gDgSi@qaVZ)C
zllpc)BZQCD2uN@b!BH}DL{!wo{e7#f(z`OVk+k4msl2_pi|sK{bWHS2W^43X1IKDW
z*7Snl?Y~|o8Oh`>3b#+_^#JJ95O<OcoAO!O5NS*!;9jo--d)`yq-VJ!6oUd>b`Q|m
zrjK!huF!#?Q9Bl*Q-lS@w|6i3WZiVr5sY5D)k~|HBtTxAc^ndB(bBCp^C-RG-uz9A
zV!`smn(Yx7PTc2fs{3X+5}J^9;9m75Y2@_`P^Z0e2@VfG;I6oxay_3sUU?pWh>bx{
zTg3Mr!^`vYSqJyxBa!^7*Oa};l3=CTuT+&K8IVRvO*l1R0{0(#m?pXsE3>wdloLW(
z*G`{0)T=h*B%9=rtBY=i0eK6@z_mq?QZ+?cgPa)FRg#AAY08Kg@fSAqL+$ux)cYvQ
zyBs(;ff^3x;D4Ou1@ZRNs;r%mS3m?0?(N%^oLn-OZUQrWy_|;{E$x?qPaHYi497F+
zBp+ipE0KA9s{J`4sNb;Z+`0gC-NBzi=Q;v<YmCZ8@?Y6uOXlR}KL6bJoSYGw#1P0;
zH1HIakQT`7S?a24OoPCDJI;KjgZO#iYI7vR&WAusTuh8XLC8tJbP-+}Fo8szx$VY5
zhMyYJu8l|;c-8lz8l52axvV{`E=_DW<+m~PdqfCh)cHj5!Ug+V?Tw-3_5n41u<k?N
z7sx=x_@(bS$j+A=oEZertwHqB1*pV>=RKBEQfN$2+l-u7KNi?ujefHoc4p44R5A4b
zy$l&&5l^j1D39)2;sc**3c{l+m04Eui!HI@2zVObYb0%5Um#;q(NE6h4Mb_i{<Crq
zCNE<O-R`>VY2dvGK7U|-KvM3Um^d(jL86x@5D?&b+l>!K(G{tIFKn9?_1sqBy5eCC
zq>PM%$2ZkBi@X$G>C6KI166JXyWc9Z#stVauf*O0QT-f|29ltl;Na)PS!r(H5h<L~
zvS{*RV$d4^TCwrn&E<HJItxEc#?!4`r0HQOfT5uwqcuRpUr)9yjfF$TMToXYlWbSk
zd&o4^r+S_A;R?BZ3J#fPAr~DDI*tX<Yh3-SJ#izBK*^T~=PS^Xf}tO_xdqo||I4=!
z0xe0BGWPmRH04(pPs33L0S%Fp)YST>y|y~@AREqn)L)ji5<y#M?}$R-{zxd3#LDSo
zjFY{p#f@x%bIWzQA7q$>#%B)qQ(Xrek;0s1*1T`LZpvNQol{#_dK_~9V4Oej;o(^b
zE1%ybwsZ{ai4Z~hs8cON$J=JUR<71|nfdbaGA=HT+`_^lE-g*(2!c=q4;=$YN*nFa
zC;rD~2=8Zd<<8jeZhnsNUVi|(&guZ2sw=%N9fY?l)}mj(X>>Wz${GS#@D$&|!^6Se
zxOjNTfhxy=&Kw}>L99*~5Db@lNR~9Y9I#I>m+7}@wP;U$AbhD6=||HX{i6;<M%0|v
zA@f`>>B-Y)K<(8dgtY93q+P%iFxt{&AdyKarxSbS=$T0kBF+mWSKzBqtUIqLS6o25
z0^xjDW~IhBvt0~}qbyn;U%7y^$NWss<A1Ijxb*gU1Tn$}n*evb1XV~Hrj$4k1@%qN
z1uG56M$CUur{vE(Km&gar|jLX;&oa+zae+Tc>ibl_)>-LOFP)_(}^}lhQdyEZoKFi
z3wF*pRS<LUE))u7JzqjvulVI_4{9VM+E3=bg8L|v@_L=kBs_n;<7vqySBPuUcBmcX
z-#sj(KbU#;zOgECUr1cAz+=lq@kba{ez8iS(s=g%AR-eFMYP;MP=HuIWA+HQD?k}w
zc|?`4Ir#?!puAN!n#)9K7US_)DJWra%X`jcrD6RwYB=Q8hB-cmhXI!rJBaC)H&~Lx
zs#GJ1Sr_~w>LHitfFtM75zRN?@N<Bi38<a>4SoN9PJS;`h$&FQjdC%R#@JBr7)O_P
z#mdC@Ev}9$PiWYJ#I{c~v$;@-9r-{n3yI!qUBWCXfj!`1uX?4l``M8BmvD)0xA^KM
z8Cbmr(9VLY_&9Yt5cfiewV`v7_phljYqiDboo&-BcaOAQCa)|>yd+IM4pm-!J>{>%
zb2GPA%Xaf#`J$3Grt$|+X!S?k0MD-|M)yH<X7L8UD?9#v`JV$nTg+F?%+fwf4I{Pn
zBd+_gu~HRr*za(}oRQF~raC&}BCY63w@Jz9tD9$m0w1+u%>5+^Op)T-4|^wnqMGg8
z2u=_L*g%h4bX`8twIv$C-it2{;a3PMLszn>ZFhka@9v<hn3YaR6YQpcigW(_d^O=?
zDx<%^#Xl$XTTwpz>*tB^%m$C?u0N%=)|=jGDGFJ=PyV_xyI85gkKeH8OQ7~H5hHC}
zh&uhzGwRoN^(sqqRxPU^o=ZSsh!|*p<ZeM|rHnu4q`#Q3Vf#nR_)fUo$@vBR0xMDz
z<=+V4_kkVFV~-dI8r}1M3D4gi9&k{S3gD6`hzxAD>3kkyjEy_cB{SU@A-!-q80MXA
zsZuWn2iLy>18dgz_)NIp9X?^tsaJVQ%c6picN@P;(NzCs8Vz;4Q{->S@k$XgF-g$*
z<?Cspp12q>(PjgQi}#o&E-m{k29#$#6Q2B=m|}o-FgMz&(Pj_6IzM1|po^?MZ~8_L
z@JPNU7yoXq^4FQJSb%RD&oSG!HGZh4_1yaFAa}??VR@g#|0Kl!7S_KHydv9SHWnva
zcZ_`uRm+3jm{sA@jCxtjJE>L}kk7;2gQDHRMv|6z#N<YC^Sq`n|18c*7g)Mqyj3U0
zLi6n#pJpoW>M*z>G(;|SMrvw_%q)JIn{)=p^I%?kdwT~UPv_C4!Xx>wWCj$MRN?A<
zQj{VE1;y8q^@irs0@g2jx=tGZzLeh-fWI!wZW7EZ5!f&K#6!Bg?Mw+<<S2(OHon$q
zUZTG}+l5#2=&te<6?2X{W67iVKc53Czy0%LDJ1)WW?Ym&flCq)Y9e4W6jOE{@%j7m
z{~Roc`4ML?fD;%0Et~(A$8U(w!r#ZLhY-<>lrEDMp&ApwA+El~+$K%F>hU9XSU&N4
zo|e7OuSybSl}@?#nxmu#y@&0?pRf<<V}#7Biz60b@<2SBQFrf3M?j9;izTg9hToh2
z?^9`xdervSbDm($>wnstKYFHJ2EceoNlEZ$iBSzf>{~*$*y<{i8i-(JdBSvN7KRa}
zHlo4AMV`w<-U_k~2f4C04mQ_d8G*kNAe5sjV4&J+BN!m<DrQOa0i`jr!Yay86TyEA
zIpwGr7}R#FlU=Sr`AUOQNGhGVwQ?zqE~XF<gW=yR^6z5~Dk$;TDxkl9Ui?f-@pNIW
zeG(#L;_i|L{&g+ILRHQlvR+iulDB1slhhwKsH2(qiDlVld2MH6<L2S&Go~=MXUfuL
z(*;_lN>Y&jvuc0jrHt%&8d=HNAcas3>m2P_5;{_L^X8;|=07LB0&D+PVU*L{yr8yu
zgzn${pQPR6bc&e7L~65C`(P8@2XmPH5y+nkk+N|>-pE^TP00+=Gb{V&6N;kOsJVZc
z`xsB7s@FpPW=R+PFd-2@NQs9SOA-I4MA|ulw}p4Ne^qe)qh{+9+5u-~WBP8E=l6?#
zTbTH4hJgkT2}m~*XlEtTrH_0r<3K+aQfBkfP0f0wKG2Xekuc5ZFX9ExW8bo5CYpcM
z?=#8yfeI1@o)D)07ozim2_-m0DJsr-MKBa)utaZW4aNRmM-v1UHP!wn19cDy`!H)z
z-Ps|fw;{Qeh5RcwAZDDzFBSx(g+zEd97nguvP{-{W7EpP!_fbHcY8b{WiI4fp=?2M
zF$_o%%BRt7{p;$9Mg2f@?|E=i5&pSmKnQ-}LTg^jY^2S&UQcJ%KmUExptw|znD(0&
zubCWsS>&p+`N<TMjMb4dgP#$Q10=^-V2Yy*Ku)8PO>z8_J^5eH=~O~c*hrcd^I#I<
zip4lkin6F-C?xbK9sS2f{7}uGobJCx(2esku&h_7_9y=c#(xw6@QQ#)#uQ@qlJC8D
z!|nOJb>7iKVOUXF>s%t>AG$q(R#vjpF*9@JB0;gEz<eJckuz{wx>0>`1X3qx7;Zm@
zxKUhpJfwVld>jsHW_Bt6@^l1b4fdu>P5@^^TkS(y8uLTUb)8KZ@S)tnjZQn{Yd!Dt
zLTE=`a}o<?0YN_<_Lnc2e)8Sl{G?)L9**nziwep>qoBxKY0>gaRZU7r@iRm$TP54Y
zd_$q$>mQAopCVL*1xbLxG?>}VW2F{9Xh{DF(rJJ!tNPbVd8X0vqYfetAQ_yP^&5%W
zR+$>bf~>4Dpn~aNA}h&&LeZjv1|j410~{J#&Fer!EC3~-zx4<6#kZH#Ez#F4x}=!B
ziLN>^2LWAFVt{0a#^x)7_D4wBjADOP^T!MZC^?PUJqb*|T7Kv^@#DL!|G$%9`2i27
zIMv}V{y**jP!T2SmGNw!fQib4X={;%{#jD^Ux~l}@y>t$Tk`P{oKnf<zT?5aU!`1;
z0HM>!wMDKPEQEg0{j+oaEeHSOxLyblT0GJUr2mrH|M8}We}&{gs{L=RG~I{1TsL3|
zb*Uj*t|ei&|HB^CrvoQ-Xf}xPd%LM(ZKF>~Jk1rCl5#hRun$E#wXLY&Vg0{<>K8lo
z(h_wb-y+4!Jxu4tf4>RfiR8Oo%M=tA$MhZ}#6A*CGL?kFWIkww{ob4z72wP#x9eja
zV^P+_F8{S2|B~_l<8=cH1p-y?`v!V;rJEZaEXY_+FbWCU-xlZp|AAN3BYw!`9KnXi
z|L13PsACe;R~>ZK{s+7=cL&{Rek8})_^uvPCbCbrtPs|B_9h%u;DJ#A`tRlV8~^=%
z;P(>L@)|r)K+`R?x+-n<btn!G4`(#Pp8Y4i$$7ZQ$ErdIZYfh-&HTyp$+QHi_M@wL
zs{O*0)BroC&WNA2gH->5dUjz#KpTUH>LEI%n%gH`RK`1{9`lN%sx0Jcq>o1^j)C@?
z5F49PHa7J&3y&rGR}}y3!~dnIY-vWB=k!~>DTlI>KEo+GCXl{^Vv1QW#IkH8s1oy=
zy%`2t$2<-DZ3X@oJ2w=l8chuE?dilu2#HVmrwjq_t8dZnfF~=Gp<br9#I8CV{~xXV
z-#*5l)lW`%B(iy1nF)DL5sLKe84QR<aqM~V;Q#k3`duSd^Zw4d{r9i`eHy?4A<^Ss
zOSENHtufLjr|0%gG?66_;G<m<k+6gm2~7!&h*t%mq?lNZjg1LNhYy?(fY|IY&qpq@
zD+`svWx{jVd56ssyMzAuPQNAu;8$TyLY4ZM=5IinnwmhRkl~BCOu-jO#Q(hQKX;ac
z0;)QD=`dl-kZdo8D*l(-Ms6xRNshyA>-|JzvqKacPSFd;ogcbl5L;muPW#!zT^KsU
z5`z5C&46x_j_!h$aBy&N-%6EVVG^LDj}jL1JVoRnfC5EM4+GQ;PD5sF-QC^e&xWG^
z-(<Kw9s%D!KY`vjZ4S2tfCmSc{Q+b*_|vx7VpAibod5OF3{G@&XP^Xk*;_qV_M_qJ
zI+m8xIN49UQx8_<?`w6#gubw;7v_Ivj;6hKbaW&&v8MCqYZf4n(oi0T;0_=F>9sll
zp8zav9+2kOoXY=)vA2$^GW*_!Ddo@&64EW*&7r$PX+b)qB_#wDDM3Pzlx`%H?nY8T
zq)S?)JAC&sW}NxX@BKXU&mf$*_ugx-ysouqx$h=0vFigjY19R8u}4hi{{EL<F8j|M
zA6^3|4i-XyFVCPX<V~?C{b#=C^(&Yj=4CgI`onn3$t?ihv}hgX=hr_3tHllliY_A&
zH<2c$I)jX2PVkCJLOY5oR6jYEUD$QMgLj@z3=*W|LK<i@np<1q_(&NcO&Pe;lYPQt
zDOrcuye?nliABAuq~+v%_FV+|O(=ISd<zUYx;GRrr;x(t9ZrN;?JD{s5H-0rc#e!Z
z)DVis_+K_%<=A(*s(ca^{yX^!Z$b-9J$sr?=}%qvKl1)M5V{tM4KTdMn0KpEv>AnZ
zu3l9iudeobiC?QqXmdB@3IIb2-J*m_Xm1G&H8u1ViW4d9)qi)XlZN#daju3IR!A~)
zJ(}t|5eIZ3z+?DZaNyUFX8dpTp4I$AL;U<Z2aZ6rj=kbuG;GHjf}q;!86vzHc^=Bg
zjFx(ardVEIi<95J<=6nOE3A04Yslq$>SQ&Z_=AdCk^t3${BU@HQ;_*AU-yD}sqG(r
z{lYp#ozmUy^4E#{v2GHX``v-4dQI=7KiS_&81M<B<<6_pxO)2wYI%U?L?4QZf#God
z+({a?7C_6;$L96WiD5A`pxvroQ}W8Oo4bjKhzP46NJ_*#OEXjr|A#(lhy!-a*n3I9
z;rCtyev-I9Rb`{8p<!xJ+}zwO0+iEx9+;_Ucm{VvW6rfy#3J{OZgF%Hx0QbGXtTe6
z?bnJAEPz@8nnk3;H3g-5E<Fb`ia~$_5s}uwUh(@@;v5WX3IAw&%0lnl$YYj{iHRve
z1msOeA?nyGHdet0f4UEp@bmQQ?#o6?chwnNTLYK!$Sa&W8KYf7aJT-tNzqU=9M8Q@
zqMVw~<d4wQK59~qOELttzl732`t>wu>$&^PwGc%?Yw`(I1Ox;UPU_@S+&}<`fjo>$
zb^wDxKzEeD-9)6<&!GU>Yym)_13V<pYzB}hBSN5`>;ZC~>4v5yrN}V^0j7UesBCYG
zN_%y9Ez<t;*1yAMq;O3rAB*mh-#a*PwVA~K!N>i_i@<SaEFy^|B`<2yKr&SUySkmG
z7lbJ*pvpnydFRb9RN_teCI+7_uLTzHX8ov^H<zeqde6V(aJ_hvpCF7ymh_KPlLs&l
z#%Nm+;rIVCt(1YXI6xO@eEuBz#~U(|AUrvgwqyAkKx2$5bFXpy+e27L0b(#%%kk=u
zcm0sV*Ad{%Ri3`Q{imZ{69|W-(Ok>|`jlXO$aH5mpL_4iR95U_LmT{LXEg5xIx&~D
zY!n$&mfVY6%04=mV(Gt=aIkM4zK0;;UoT605D<5E7H*;*jpTPJ)oApG;RVjA5~y^p
zsYG!9p&)SIV%GqtedVPRn`w^-8o@vEZu+mG?<fo9Q8KlAL{KyJ&)uB<@S0)wIPfNJ
z{#3b4A*Ow-;`kt__aYcs&jVXQWat#QM;mS#8yA<21v)~MQDKvkZt=RVNdJ9HGkEc8
zj;HK$JOD(97#;}K-y669kK_{0yfQk!Z&8XX&3ez{-Y3Toc5k2MgF>+$HrosGOp%3P
z7u26`FjiO?RTGcDy!8Drb9HrvF%%ACe;_`hqM}OJ&s7B3ZwAeK_ZW@p++d;6#MCyx
zzcL%Vw~q-8g(L|$o0S}q7`z1yJkGYO=Ob3}=#`__*Vnrqjpsjdg5qS047Bh$Owj_k
z07n#i-8wZ28X9-h83o9m?cl=96PSM1D*I;(!9UdB-`5|yKtZwX?3DybLXHuaTm?6n
zi%%C6gz2zJ*jlYj=rwidEG$+u$8+osqK07jOrWIzsJsF&LG-6Ve?2|1%U!9B@uIsW
z?j!~rkI&sEE3ICnxPQ`fN6m(nF3$kc;fAX03{~`NVUN-o_X(peuA5loGxi7tnVHmH
zmwVkSAVi9T!ANiY;J$(SCu#TT+$b31+qGmN44#yo3TEVuU5o`(aPwnZ>fE<z`S~Yw
z*Ro?(D{deBp+NtA69bS8>7WvEMcMaWAc=%RwX2XW=Up}R^q4@YiJXU1H3WD(1PHtL
z@rj76v`aK4yHirgKscD-F#WzaTOrnbr7ynJZ8Sp&9v&XSQO9AbO4X|c)?NyYh%N5=
z>ViL%^bUwbKqRbtf(g+;87rRxJ`Z|zh|qv2EVu*^j=|M^924ul4<8y$r&4JC_6`14
zZ9g3%C>kEP$RTQQ&wlnOu|7fq%d`x2=Lv;nuRw9P*tl_Omm`G3Ek6rm?gih&8OBs0
z_m`hP-%E9vTw7Toc5ohg3veR5BHR*gH#ig0)6;Wq?+^1!Zsz3V==HfFm?@71we-z2
zWoEYJKk<pTZ1-Jc+S*-ge=-~h8U%0<)w_KaSZ#WIkqt$W95C#^J<dyiE-8q<)zVN-
zC?_Q<P!MoZ3{<)xAuJQPfQV#!y>ii#xvAd<bUpp!^TTon3NUxiXk70P@8j=l^GCR1
ztOUo;`a<;h9&Gp?4BX(dyC9)gZWhvT_)!VL5f!u)Fs|oQP5Gspd;{U6r5VyZGaZ=f
zJ$Cq_4h{}h`tFqR1S`?VMsOf$N<G7j2a11cQ}Q%O-4mNC)uoS#3wS9s+B2O*CdHCY
z36q;JJS>dFSqr|w!OO&VFN8{f1oWqH-tj%%T|}UInMHL-Hq{mw^d4kbJA0(A!D#oA
z${TMkKaP&x7;(fdtOiukuBrF-lV(NBeei7<nL7Z((|zymP@1SkO8~s=tp%@xRlT}=
zN9AIy`N#J^XU)UR<@qU~Gax;n9wSGxgk_`5EiCjlIbdmmD<u6?WTqW)3&TJB#ed7h
z56Ku)g-gSt3~h_-9m$f@LeddtPKXZhz*SE(WBz{jb`gb6-YdLdb4;g?-e@J&Zn0P)
zg8-M2g@uQ@vn5#r5HMOEjJ-1Pvx$aaVZHLZIw$T1|Eldz6P*@#y2cp16!faV8H~!_
zU-=L@bds0mt>|~TCx?Q9f)g5wUFZlhm-|5V^hFSJe~FPJcUJTS4Mt!Pi1uV$W++FA
zjC2%6RUV^j6E~K!>uP_D|J9aBmk*u@#~aMx$z64K0DTG1dfPjbzRcMm*lO&v&HLcR
zk*Px{{*Ao}3ja$}NVg&Js2(l3T-n|4muun}Wg6IO7FJfI<W1y0UpXbj09LHU1B4Y=
znoX{oHMqZ()*mt5pDHb10d9_NovM(YXf--u2wT9AveTr2mCmGzCcbhGuX%p;wEtrb
zHm)@W61LyBe(s^T%0VF|dga9Xo(fN@Gl56vwcCme%g|u2TRY<;NhI?*yo7Q0AkN6}
zSi$mdfts56O!EYnnt|20;T{azlO*C*;r8|O2&3!qJW%uwcyf!kT)i*0t5ct*xqh0{
zBM(+*NcXL1!6R5jTO^7y754b%_NgPM0ptsH9)0~x1#5t=L4i<dmFCqY@W`Sc)JH!{
zTfz9pR22vmm8ssF1^Yl0Q6zCBtuccj0JTZGU;yiMQgYBc{-5IVpTG2E3~XtE2IetV
zO4M-sU;EZBJ|&I#N~e_YISgQv(f9Bg#OUbhNzOo-koy>7g=|pgR)9nAxX=>la=L38
zc*5ivS~%d=BOF?;J_P>5wmaIHZ+=go;JdxpoASKYAEe1#b{bDe5K0wd>4~4vj0K+L
zb=|9eIE(&5W0w+Wd=D@W;2W3+z;%UqT1nRk#l;n(#TI%K#Ir#e#_&j_9)a-slW-><
zL3wK942pOD+GKfPJ|r^$A6_CFTYtSV7x!cA?9~vrKy1?F<m{w5|D=jvS-QIwfNiCE
zG*wPwBw3~3?T_oc`a1Pa>FZz*r{Jz~pJ(S9@$K8JWl9g03edes4FH-^wU`a`9_3pm
zM!?QzKu9zzOJM)9F`D;WyutI^XVAFy)BcFFfC*$5nK3T}pul*KHb$ROQ&annu$~9N
zAyLA?L!fX7*#GB`o3+sHsN0?B8Y$Zw2?h_}0IW}#)|&HEMy7?0u>!4%Cp-fwhEy!O
zWA28={#WjGn}tOk5h@}U6jl$Hy8XXjnA8Ep<^~AY2&QemcM})Y>|UvYt)K8J6um1W
zIBqN>6Qac}_6-=Obx>5^=VQW-2zpVd^)fbgWmaJL`8F397o)JSLGj5rXZ+-QNhzt$
zrlZ`|SmdA^5GW<%4Qxssv{#YhlyklZDx4J_N_Ej;LlXGu!-;LwgUz*{+OVA<6sVA9
zvFeVOzk2}~sE^c|2_>6Q%Hm*=h=hJ?Tr=jx$i>16D*d$?nQ-3`@19|)t%&PNApO{$
zBZK?H84^4I1vRd)&(EMi2^Tnp!&p=`J{ahP4`D>I>`$bc|N63Uy2^%zbqSor@9oSW
zM=^A2y>U1wyfX~QMr_v@rvNC&Fu2fh)CZ+2-5JSxw%IQ@ibHg56oXg-5YViutw6+t
zoNWaLcMk!FENt|ov!f&bcqP80L)y8p7PzB*(*iAFK`*LgP#=Qc=dbuzR#tjHd~B<-
z8SPkLBo|gA5O7|S=7fgleXG33qRCyBCb$<`A>SR(WplQfsHq;efA8K?I3xsvmN9=w
zl8aCbU56q)6?b~f6F4RsMPBPl>k(=yp%d~xq6!%dvRK&{DoKh3HV(aJonN2w-p2?v
zQVQ>5d~iiRRbz&%(qKl7LCli~Qqlfnz|02RqPPkFbH}8hiUo$^gEy^y{(BCx+st~s
z^Zj}#*=U#$h-x6TU_jQuWpHG52BEL#PKN8s;CyJQ#K~s@gr`WQwW}mj+ryAQHqs9Q
zBmSi<MIq)0Y}~YZtY80#72_7@BHlV%?=cF99dao;wx9^mF}mcOn+s?{qjawfC7)qY
z<N}1;FwjkO-?;5Qv2a#uOGv;2-yjrl623HLhXB5vrnk3{GkpQTZX}ihs@7|6D@+Ru
z3k8)gR+za<kH|ziM^E-%c?zy87#x3_y`$?qFXBK^z2dsq_F<eU#b|5(`!&Bx5|<2F
zI{t8H_k!7BqGkYL@~)eI-~F=92($}uz1-Q=uU{Mx?+ZY*QNJAWFeAx(;TE<YTc!CV
zq<PsAVl=zK8M&S~;1iG<2hF{;%RmX13mJi<T92fVJBiGGVWKCEh>Vd9$nj4=`ij81
zz0>nQ?yWjR9$v#A0<U2Z5|Do-XjPzmn!t<t&l?^%h3^TP8Ce7DVwL}Qzoq;6AWnP^
zvrC-vFb0m<!YIHDQr|qrCnOA;%oW(uHZ&xQye*t?eN^T@q9%SNO~%ual;b%i$Oa_w
z6GqgM!}rr}aqGpl5b#RnlNr|+HG_}02)^-ib8De@ELo&^M<GU5g}mSg8ru21r8NjS
z9Ny@owjbGLd~$NpZ0GPfo|=dR_~5Er&@N%8r-jeRz2#>ndN`Mtmpe1K<^m&=`5jdO
zT$NTvS~~Rcf<JZuv>gXZLL+Xp1;G6=*C40vc9C^=wB&`dv{PJq3S!RmrvjCYn8|95
zsSIt-Wc=?HPZ0iB$3Yh$PSCM)?lT^S1ITQ-pRmfkeNxfOX};IhYaHACeNs-^3x>=Q
zNIxYIdlw=1RJwxRWBhSqEGO%u7858sHiA0~r3LrPr#R%~Y?g8_Hw9zytQ29?md=e6
zfDHm#E4i<)FCRErhpWksnzpuuHNwbPfJGJ$@X+!Mswo*p^^cyzYl9(ic`&4<P;iQo
zjF2pVBiYO~c#Tz9D3z;dXy7=jj{?pUQ|X<`@&*vTjpeCPlw)DKEKwHMx~}y-Rzd^N
zS<tuev&b4#Q5L0L;&vfevnV=VPXVdnIYEwN7yO6v>FiHM=!{ob0NosE9ieNft}G
zuZ;1rDnuln5;;hL&`eI~`$3hF6_K_C90EMs!kH06iy9^Nvu>QaOH;S4cD#)?C2!;<
zY)*P{$Zpm1SSv%-9GM)q6`i}9=Lv4MAG4oRL`g5P6UM~En9a-jdF_6n9A-kIT+j+c
zw7O%Xq<f=8Q19!g-w5#!aU3gc>HXR_B}y4}^T)}s-LQR073O%;ZF5YaL83+L-fndT
z(6PTxEr7-8$^oBQL>=}UPcXS<OZU|PG6qR@vgt-UmN;#F^hN@=(mU&sCnUVK*_$-9
zw81d^YWX*?6HEAB0Le(RA(lCoh?r1i(GiMM<+`TIS%iCXXLjNXx^7=vQlUT)fm(BS
zpf)IyF0F#_Ms+d*)Vn0-AD^TO+@1%!{N+&!hLWD=xoB1nNFY|-Z$%LDyR5lIOLyB~
zYJQzla5xOZOHmB7Vu_NSO~gI@z0-XXaPP2B5vbCelq7|OU4kK|YM<1(W=}%9C1{|1
zG%+3+r-+-^7dv^a%ibfqAI`n1x%DT|0);J&RbI^V7$^_);LKe!f(+HJT=hC0P?s|B
zOq8|;s)E~ttZ!yA59a;yJE<1bC8(W1pN7!+)0n3p=Jn5jm}TpTNCADr;<R2B`4KB_
z<Txm&>co@NV(rz$UZ~UQIqh5TBipkVDIdKsKr;MjTd9D22w1C_GQ`YNtUX5Td<R(1
zc=ZN}91%9PBmn-(EN<iaTZcdqaTkh}$RJ+|cAHHL4|Wm(rrT#ze<m$4QT}pQ8y6S%
z3)8m9JKj2@KDU*^#+U~#vPD%@Ja>&U$Put1e1SL8W0nskm<hTZFrQ~SA;}Axv>G@z
z#y_lCyy=A~H-28>guhdcC%2Jym*UW(di_o)l4M)4+u1hFMl*Y(8myS7rFC<J>3%+T
zKmY|&9K424a6mp8lJuEGt>2?Fz@H9t{h(yEwl!~*?YD;x54nZM$Xq_!cGL+8<?0I2
z<;53p`2u<T{pzI9cvmoEsYMZyXjcb{?{DWlQ{y~d7c<<#4zD_og>dI?X*OVr=?sBE
zWRd}Kib!|<GU<OgOFK763~w?;y~RL7J238BAWt8-2=>O>RF1^U#P}9yd8BcFxF+pz
z#osR%G+}yjeO%o<-mTRk2rig0<m(6KH}(E}V(#Srh*|#MuX#W&{I!5{CXt+^)(?Md
zx4NaU#H*^$ozHE2QIOS3S~|E8^XEgTKZeH)RU8tOOLj$Lq$!I&UiqSt6x};5ACud^
zY;(u6PXtMtSVM~Rw_k^#+hR}U>Wk*=iTNyuQCD;&p}9LXIaW$=a`IXX{rnjm63N?^
zeulJ+us%gaf(nDht6cbIY^TjYs+MNEvDl9b2-%ljvx-VeHdp5uJnmib*COcWm^b0M
zHLBsaH!fu6&*`9HYiFYFn{kY-{$Iui_ejyDZ_yx;yoA=yUa@16VUL%Yh?9u<3cw7B
z_sKVv%gV@~D!G-e1zrQ>T;k`4yoh(!THP@?^?_UW76}`h;NApc7Rb^4Py>_!NVv(G
zr+wSzL{B==(9rB2-%pv1reOT*P$(0TNDKqU_LjSuXlSIGWJQLo!4S|dN3U4`mn|_2
zH9X(w?)>51*z)`Hbwu*thEx2KzCSa$xgU(p+Xz64<&NQ<!2M}v%+i9Ks=>i)Tvy7#
zt_6~<7L*iw9H`u-h<87~O<gqb^-XE_ub*Q;zK6v!bcvmu2VmyC5nGM_dYZpP1a<^R
zsu`^}<XMgC;^@`VSWJ2ZT~~*JulD2FgdZt4P<_l-P*$mHZ90<9aMom5`Lg;cwzAa*
zyptCPf`O3(BIFOmGSMIceEpdSm?P*cDlG%U=xvV~D$7eLsP?Qe8rf}dx%}Xfq;GuZ
z?I8Z`&JQ~wfyB)^nC)jPz!1S5@@p@lbi?%?6yx3;r93roO#a5%02~n(ip~&DFi&!l
z=lb^WHZiec$c=aDQ+)S)IaDD$wD3)VO`dy8Pu=<i?*3SjZUFa7k@NMeqvJ~_GuWyx
znF0bB{Wmimf6JV$&izFq?b=<}D=B@cp7Xg^=hMDA2D-9UegsKmk5HLdX?(H-D%^Kg
zissxeY-V#YX1})t)lI9T{AH=1NWj5swmQXg$o~G!zYGFc2(~$bCUlIV_C^)&uM05y
zaGgb&9?!H-L`vKGr`Hbl>NZ3m>UQqFxb6n|BpI%ZU-MZ)x>V<h5d+2iN4r~-mj1M5
z58(gyGF$XocAB#UDR6oJkF!*1xeKoWQZyMLzn}qh!zL-tAoAbu|Ks2@qL4srBDX8R
z^0)T<JKTa;!cOa8y!|uPZ!`DzAAyI13k}e2ubUGv_^-vQvj%D=r+#!r&8OE|#USAc
z1$pY{=o;g72%k~@^I4$y5EQzKr?E-OUmyet82Cu?#864R|N7|RVWhs007lHGxQLl9
zXf`IQ<nfc+Pj7*ZbzdJ21QH(+z$j(**Z%(T%;3lS+8Ij+IcB4YiMN)TrHGjz(Oj9q
zeE+qR8EI(1`y!;Gw5{CzOpheJJupBsNMeyqN!b^QJM~+D|G6_`@o)&)n~LP!zXkb^
z%?1{aF(4pcsH6MMf9)r@$2vfo-t`&PBZs4h=j0Dtqhb-HNo6v1kSVUb=VC5l$$c5-
z8mHwMW`Eo_bsIlw6)$jgAX&@BrQ-9ePP09Q^z~n!;NQ<g64mnLO3r5Q&GPslHZK6#
z!}@Zx!?!0%{^P-ZFCQr)z~Y20GmMK%Mi$$-Us|8y#REGhc+JJd#jz?IEn<3V$(BMn
zI-z2m99p-WM+!Fhw+Yvd`vn9xzf4NGF<K$t9_)3Se$U76qJJ%mqr%uMP{fBts@;7K
zSnN-18*cp+Gi{Umy)Ws^tM95fJa8jl*`&bn`SCc=3`hhie-vm>TS}q|<W_s@ULaHj
zX+|YIFO>@mUV}HLs|lNfZ5JpDO$muW+h<^wNx`S`|8<<L*AZ5UM|+&lDGAW9SECtG
z=Y7X$6oSyA>p;`u=M+aJ!Fn|7PvZ8*gqpGzwJH-!#+cc3eBB(J;T?pkA}R)N^EWh$
zlkh#ak+_QO)-;sH9k{sSbrG<?7C=7_{wYpKku87{^2D<E7UPerbE-?p#Pl2lJ}c9z
zBcw=%5JupuoV+N1&`m8BhjvF>swExa>ip=y>SUetf5d<IKCsdO+K(GPs$?dQ4r9JT
z8{O#Cy*ON=NghdL^UvY+A^KR}@Ob=NUY+nhg&DhiQ|J`KQCvMo%yY7R62d=q&ab22
zjSs-b=CvNCBoTC>MacBM5$^XNuL4RDK@-Rvyq8`CR-wD3T;1C%X|AOHI$?r@6%Au&
zz7?X>Hx+%jVUeL(HTjnK6q9iRL2(s{fX)1APf`E<7PpX=lX7kE&Ie1aP-cO?9c1)<
zL>NKN?Mn~N9?R*ADd6@yMW^FD#rjKS{(6jS`~`nukcf-H#$HjgIK=qt@3sn*RL`l!
zjbtCyrz(ai^$==Zu@6st40XQsmnu+SL)e5Q2~k9SM}F%#FQ)(AKFOC-l@vI^R$iis
zgw#@IHuwAJ>RL7G)f_S-9yDv#mMEXZonmD<;MJ*@eQMW9$`h~~g`<icU2>p^zCAj<
zF?B<71@!r~0dDw`Itw8hLN>x;=6!VRVN}W+X}U;TzwZaQ^zMwnGC58G-q-CYxMv_o
z@q1r)lLB;V8ee^S77SskARs(1HgNl`;o*=rE1^LA^c%f-VeK9Ec~iNqqO=0W00p~u
z(J54dp}ahh)LlH|N(SE+C3Yx;D69|7`s!N7t!^Q|(73*B&4ba0A+=vzKC8~~j(MCd
zJ--e9NU$bu{m3;pvCC~jTebT;bB!Hkt<w+dBfBC=@A@Ej#en<|2^yJzfr3E=vo-L_
z<0o;bbRMX?QKIe+^oNC^>a{ap!m|Jht8q61h=NQ4FI^vfCx>g4YKd%F>I}CzKinuJ
ze9#@01M;9oI?o7x?0>#E0*t<_#jy;2yM1`_;oP-L;D5C=bQ=OQ+%(0Llaq%a+8p)2
zKA@-rVG9D6I&+SKhwWs3>ZY4w-_XcX#CH_EEJdB8Z@QjSjqyon>+FrW>Y;+#52sYu
zDd07^<VslN12sUYEn|P*rEaf_)NZ9l0D?pbJWLDd*tZ$nKGs51vjV@e&v*#7f&Kl;
z%&aBn_E;z=8vHFJKcYKBDoW-nD6xa^J=yxGxZ#meI<akvREZYSz?m7tsqzQUY+wDU
zNsg32`V~=i=c==_^Jt6K`G1W=pb$1RAdb`xGtT>X(h8AFy-;d2O_gTeX`P|$Nme$d
zBJ<$L;BvFZ(5LonX28H#5eYGImI*Xmv!$Y@eQHLt#K<QSNFvjrbx)pl(w;zDS3`M{
z&WbT_7m482#z}iO1Pg_svhytG7SVVxnWXnUHaY^CrS|b##gg<TH9hogG%-umFFS0I
z50omH%1(J!2rP6ybdHz%8Vv<%z_JhnpgS0P#J>CYbu0wz-4`ucc0=lZ<pbiWSwCDO
z?0nw98A)K*zNf=$$SJKSVq??M8p!<iWB;}cNTn?svDzZWiDqhyz8;-3cG&UXw-Am0
ze5DiB0$Kc0rm0k^?oGwLLtiY>L+tUDF`p6BNJh@O_|2DOy|XIa2ge1Q4SJL5@e-o?
zCi3));`QyaR)e?c!aF6uP!ossn!kxsXk;uo!bYu>_7oW_cp??&XHK_*pQ`&##ZE9<
zGx=C|T;a}g{VX->$K&NK>#cI${kke_Ta>(OrUy)dZ;G^YF8aktvDoMa6A2h^1Zo7j
z&~acwwAEU)oV3l%s6o~|{V4{iEU4q=&q=%``l+%FGw9(pV5Y4w{`|x#9Si&R|5`;6
zRLK0-@_3zh6{g;&%;+p2F#4qo9>XKaZ`7}Ua>*Fu9ulEZZNB>|P}5;aSUA#gq8CHe
zeNNlyY3|Vlk(P}GPAZdtEMI+i596EBt{4R#o+C{v1MSR3DV}hZ+;StwY=Zh`><j*e
ztmLOE52lARjzK`k^GT)XAbG0812?Ge)9riGd4uHppXbssuy3uX869E-*5NG#<b$%(
z2q+}leEnPw>+L|5oLcbhW-tv1u0iO@s6HBy54r+<9DXBZbGSajW7Yu)3*Eh3fgAqu
zg1=SQ??mvgwPD{wzz~CO&U(z}3=gZn=CTO$wE*R2CiX4XG4~AZ4>iYyrt!LjXrrG~
zHOF<VR_-4^dQl&eC}pFaSHwik(^XIsL}vDs{^0{o;pVEEQmZ1<!-K)0l;b1q)=@ZS
zrJ|ZBo`z+DK4z~Z?)%Q1XnLh_bksayvL0#DcRc)ou$<iCAMdNE4%BdLJR-EPv<!QY
z4C<9B?;=e6L7BBAo^zQJEYB_wG^8NmxDSVrX^3Zv297jP2&V@R4D!+(U^)<x!bg_I
zAm$1N1#_CFZaG=mPrY6X8Gk79|M>A#wM8(j)8jF+jM~P=#vy3dU4W1c3jOBRU#Fl2
zu*qptWx0`=!A!?-E7Eg*R6GjudHQHMN(b<&tr{Ya$<+e8lk&2igAKBnc$p&WP_@>b
zTjI1|n(1rBuCK-!d1@9b66Fp|Q<*m}EfgsEXx}K<>)LtWB!$_1R`oW-!YYR?_&wDa
zyH`~M8|yBQH*vRL&XbR1W>ElRqNt*rHQ1IIpW`^K;kN(P)RehZ(c;~F!pZiu9JkeA
z_SwOjmX}xU;xdbIL*i-cN*I_XfS4V1TiCJ&iwF2DmkhfAGaas%dFg?42yza<lF5Pq
z)<PP^;kN<%mlXW;B#z`-_CAN3Gqb*dEQSx6a}u7L7wcSc?te&<57Y!1YHS4WOBYD3
z7G4q~7p=bWT8L|oSE0LqX*T_ly1W1CjK(>gYm?Sn4I2w82l4VW0ng3Mhx|py3iwa5
zv*8MWZ^5}0CxatR$IHenyf((}xHZ+#JX+<wg~gJKs5hN$)VWqkH(7Po!yaek_VQ+R
zv-Th(0Uo(p>g%I7MhyIgQx|p?py<(2hJz1*X857pKW{j97lo;8=q`kklG(GAKo<2c
z-{L=3y88@jk$&*prR{-P&-!BnB>A%iMOfqt(sE3koG}j@w+Sf@*XzP}qvaQuUh7~e
zGKXt#mFl(?<FwJo)|_2__tKbsRikm5-us0zt#0T|=xNvsWtTqkcPs0Pvy}7&6qvzF
zlm(XMy_y6v!Mg&5cLlh0stJR%g(>h6vZ`7&`ffgR<totd#8b56l`dUfX|#H;nNy-%
zdN6^YHhDVae;q4!x`5a>G;}gyI?Vo;vyuM_febsIO*!e6bAANZiyubqX>(iv)4}G7
z5(2}Ge=a#=@P@LHxBBNC{>`X%eB4Zf7m9y5Q*eQG1lGXChhgSGhFd#_oJKm1>BC#o
zc5C3I_c6~}Fmc#!jun`{)89yTY#MP``_RWh1ywX)b|9Eyo;`ZrxTC4!kvA=`k;G(v
zZwXyllZjbeE~S8pzL>K!xEp1hzE8|Ua_Z}T?OW*{ryLM&C9$wgdrw7U&uXVcso8dG
zoilYaj@zYsq@AbMy(6PFj}fB4?i#)34GYamY15j@6*F0DsRiCfDkxwr>ST;LRL{x+
zN&`;!Xt5r*mmz?-6;8Yzv6inzWC2qB^I8YO5kjHba3*Bv7{B%P?-dWCMUblzE(ZP)
zHeib&V6Z{Mf?l64mfo5iB+kliVf1)>=53HuGE?w@!!A)Pm7SS71j=Z!xUoB9VXOIm
zWd(#c`8M{XilGh#Q|2x)?-a*-I{VIAiVGuewDEX@ykrQs%nD(8F1^OUO76^E2PAm~
z<U>AnP<brP1WbPU`VT{qB)pk08u}0a>zCvHf5+UE!$>T0STSdxairQ^5(l)sTGeYy
zHVonQs`EZ<r4cgMA5+R@75XTAinQxBuk@NynR)0q`O@BD>Vw084|ng&$FjyA)eDdp
z*?n_YU+67O#lzn=kF;s+3$(R$uF@OO>G79x*WH=a)a;!r4w;#w_Ds6LPt!Rk5M^)l
z@k?&PvRsevm*|FXHgC*^IclHR=0~#-`}e@USlEL>&JeT{s`Elj>WdtxFms$rvELTz
zxzIn}p*kRbfCBlf_Zs23Fvr?zw8aT#ExQ`pd*jEw<T!ME+$fcE_xuf(ZHT<;OB{qN
zs;%kD3?5lVpB0xU8GkL0?D<rZ*HCu$#kqoPioMK7e<A(O_1Tgc8!Od^gn7|zYR78#
zuZqXP^!s%>pC@P!O|w5w)5exqkk%GhOz?57p*?3o$FGl8@A%YMJ@e6b6G;#ATkM1=
zt2X8>Wy}v}!-!gX#CqR9FZ}p=DXR*8g_jd08_7M|nk;4Y25(VxAGInoC(vB<`d-#l
zZ5~gWmaDG>6cUH_->~Yfcu+L`IF9sA$V8}=XNi1g-Qr!Y$?cDY(^fu3pUmx2sjJJs
z%_-#0*n7=_4uV`^_M#VkD*bwSy%mntuwAU!x$y-eN?~%pFrRWO_bBaAO!f4X2c?hZ
z_|>Bi?^(^+L>L@Bs_zrc1=+Xx#Avg0W5%AXo7#+TSGJpH_5~94Ih-c?+Gpidgzu7n
zzPCqydLhK6qf0WCI)8B1Xu_=9Iz^jE3KGANMbWp-Gh->SDhZ;(EAC+L=dOCm{rn!W
z`8n*_^;?(IRWtY;6bfMYIULx{$d{v~Z}Tm38-w1V-90X+0acrONo52k+-$2asVlM_
zk856xwR|cdom|Gt)>&*cr%CElTybA{vb>nwIbP8<(1TB+YWhLMpDoS*mFDpi3!tV1
zbnXUfe9Q>@fR-b3COG4LO@vsRNl4UBe}^|JH1Bhdbl6J2W&?`|Nk}(ZQkvvrv6qRv
zg(g+C2eLTNdCDh@o<`x_);)sn6z&LiAuOI^pS!6zehqP!{(YY*7sTOBZb(Oxi@uhP
zxFr!2Gf1p^#qa&g$o*;MA@=YGZS1{!od#Vui3wGzeCuaERXAcP@RP16@}nh8h-dX(
z2dxn1GJK<x(<p8D_OR7=9KS;Jv}UGbe{uTDy|WV7?`G=oMjCQl@IE|!_r>#C!fST)
z$;Jnh-aV^@h=Q8W^{!1Ggp&-j%oJ^xU6o&YW@urOba-3gz5kWkLo>a7)=6<s9b=DX
zN~GyaeRtO>x5H?UWxs3D!Um@ZU&yEy)=Qd+0~XGPAeltfkH;^pRvxpaH>w@yq<`4@
zMyUMU;dzrD9Y1HDBW-toJ?uDFv0lAwY;<a@{p>K8dJreq<7YQs-tX8b8Fe_mcCo+f
z?|LF(ela9c8*kBR=F)JL0)(%%%AQCtJL4u*Vt-M3QIg}!$1^5hq!O>s7zI^H(6lRN
z&iN};YSrc@xK4O091CBw%Fks~-v2i9%6@0yRrNjHE9EHu^tcI8HkLtbOZ+f`Wao?*
zch@fSIR;Vg_C4chJQNU&x`pOl6;TuMb?Dk)ZbIa{t4qKCU7&#Ohv9xBU@t+f;M|$r
zn5~(=RD^sgM)iBsog0%?)iO)9;;^dV5xYGnSNKUhQR<jiyaz0sg=x0;c#pdFxy6;O
zOT*z?*b)v09KXn4x*k(g4fiD5Sv=m=V{YD*J;P3zNTa?meh~OB0j=xh_IRP)1HGz%
zuj!Xc-#%<Vp^LEA(ZNnsfZRK@oniNv?xWNiEk^Ef{&Y{hbY<V5_js7a%QN@IJXQC8
zVLE|V(BP8-FX7BG-5vv$<cZkj4IkK!=ME*{T@fqWGqay;rm%X5%N~>Ff>k%4Q478&
zMT792)J5A@iYJ`ByF3&Xob%BzKB;-@!XCpbZEq>$JX7SHobH$WHH$3)%hY?-)Wy6O
zE4y+no!T|DJdqelzPGBB2!V=2ktB82H0xs+HR@f%1xA&H7h}2?zNI`9+xR}FlId?Q
zSIsgy<al$~29VXKPY2D<U8jT)>&I&pls$E-GRBhlb`=Uw4vPn5R|*=-s%g~wCNg26
z;13Tk5Q@Jm104gD*v^7yIM@WCcW*s-@~>DRu#S-9<hDE&)taNUesD2wmx~#;^Fw4C
zaWUk|amlWI@l^2#MbWHXhilG56SuyLzt9q>X$b0MF-MM>5MhDnF=;QqPi!Quc1pwC
z_)Mki;wg~gQrT&*5jo8g(V3s5ApEekrS$7ed&Oo-$R4REn}^+q@g4SW+eyK8St5L8
zq^wgDiJi8*eyLtFl0F-}mir$+P&cq?mvOevuiGj6l(1OtbU6@ou{L%oOM*cI6MDiA
zV|z_cd2I<Jv?uz)x9kb$aN;iZ1tPrnwy0>cH`n^6PjD-DU2{bEnjc178|nvdF@Nos
z!Vc%(32X@3ubDHp3g0!RZWJ4$m<@Nfn?pR~wQ8^~jURq>ccPEFf&M!WH%gg~qjCD+
zMBi!6JMp`g;k`u<?v%w>mvL?nji0GOA1qleuE~4ZJBpQgTC$EerM&C^Y}%<yxTv3$
zV@GE$a=5B*BW`upIK4g;f2_Y}H=FGIEchvecxCXr;p>o-o8Wz?S6XMTo#G_2-44No
z@R|?fUdC>TJ&2pzMg4%L4X??%G=(Oi*oLMr&$>~{GI{)_eqc=8i^bi3s@s8dPJqKg
z<el2)MtQ0@iS%AIv9GqI6e7(u972w-o5TplZ^sdy3S^)IYS3>X{r=5w1<8XQ$U)^3
zwsP&?V`WG5P}<uSkDHoWa_&`|Wbfx}-8Bysf~i<D{|B)Fsq_5OHO?f3pC{;%<Zs>P
zN2X&s$x+|wx=l_`ixwT_P_(<5A{89Kyi$HAxCJST-|x)QJ<-iDuoY+1XgKm*Kg5go
z+gt3TcT2rtGxItk{;8+msU)4^^*A3UI~`1|3m#m3aO!Nr$F)2aIw)`VWR|w^H$9gv
zGfr<4Sa}@M#M}t^x&#-By;0NFgt=Hp`uOpFO*>7nPxQy@xog9{F{_VFk|q|$NP;1d
zwPN4Ly54)|Uly3hx3nOA`PYIzXXp4&nBvw9*&;fvDIv&T8HXmRkg|y>sucu^&qaq%
zo7%RtJa;srrjt*gm<Zti^hkrZu4-{TwJ%&a((W$nHX+Lf`<#F<hken`qUv4<R4Tl!
z-g2UE+PTQTRIw3KSB^R%`?w^cnF8V%xOoCyylP3dBOovJ#KIRkSQX`vKGjo?s+aUM
z;XZe@ASf-;aV`-cjq_M01X3rOsAL^{cNufMp{wuK$g>;pnnH=dzvZzSB&!3xN`kit
zJ*BPou{2`y3u{H&OJc@&-m|6(`N*%?`x=bbk8@PFYVi38uEc*)m(U=zAWF^PUW?MO
zHN)CuF;Cu?VZy1f>U_!{Q@@nB(U@1HNzq4&&XPjVMJz8v%1A(9!Gd8(nTpu;+cAc=
z<6$FaV#aB7wQ}`|XiCbdy@CW@bdjA>c#3)-W&p$zrBoazSJl6j2RsB@Mn>y&TuTN;
z>lIIYXzk7od}Cj~WZR&@Fa?ONMtttTil5F=sqX7)Z3mS4efmt?2i1-$SIC9O@6d0{
zEJiNtxR>fGD=2KLYzuqxe@i?V4S1dk;WwjGa?d-r%N?B6AuVpxwoYIZ)dnx@FG$r%
zR-U~e5r2Pt)7cPj`C?zgfT1YrHs79*3x0~uc&$F?$*5tGL_1>&;wyA~tp}#4+!i*;
zUuSFLFVJqz8Q$5PGw?F6h-ymgXSxmhL}iER0=}fW<K1tcB{MP&%}9N!9Tf{3?yv`6
zuh_bnTNP~@mLz%~YB^Lyk*Qe@yt~V*UYw|3ty7G0<{mszcXy8LGabRARe#ZAEn!8U
zupz2fh6<m!rG7l@?6`uca@!VfLuS8@tGg(21(Q-G;ap_B4cf{a{-c!v>6NZ1hi_R;
z9uLWdUbo@#>^Kz873uc33B-4fy{QCJA|~|(mwaQk>2`fj{k~++>2q_HgOq@f>DRm$
zuf31CeeGkhv$L61UZU^Z`8HAb&h}*ZLm2}zi#BW<ux=GjnG<uRv}HZ`u}%~nXAHCC
zFWNg@tPTD>GNA^0t80B@Dp!@Jx48pZ3O2VGPD=Egyvy+Y1{(zVm=M>N8ho+rlQT(F
z)=lVRXvFX8<W-whpu&gW3{&V4GT%K837jMGjCE`}d-|~MO!jrXeXa=|i5$-x6jA!3
zC}#owSijkelET4h0*9LgG_^j~19OgF9^yXG|5#JX-?#O#?M+Im8?7V%3Yvt*n&LsW
z?U|Q)y;8oB**N>UkE%oNJQjOw^}&-)GR27_e{~&x$K@h(5ADLuo7oXrm<l@0!({za
z3R{@IWwv;XcZB`t<-RwZh`vZ_M6hVMCD)F({xJvi4YUknn4=X67P~!Pyh!Eg+~sJQ
zk|pD_zak1S=8~|EoD=16@ct+%LhgM9Tb2X62b;l9P}>pM8+H*yK{?swqWht>L{L4+
zMHb;sZ~$WxCK~o`J=DgAX=Uy1)paZK7e_um+p(UV?k@Nr`G$;#NEIQ;Ch_r`&Chsk
zKGIS)cj)m;3NZg$0Yh7a13^-Dklvgcz2Q)PE>$kbKGo!<N^NHzqgch}sx>h5Zak(~
zw+AOz_+u?OH+xK<wX-2HTtS~zc+`nnnkUjf(Ij2@_Q>(5%Qa^+8F6o<i`u64mAaCE
zOt7q$sUEy@?j4XNfrZFW4hZEf=3AwjSL8GZnVTm^os&1-!&&LIuR-sd&_sZ#R{;TK
zHw12>-uU4(Jb~W-wxzv4@4taL#$s-PYEW<BLCX9`5pz=&RFT}ROtNTpgi}=!HM@$=
zZza0mtPrKY*;Kx5#iz%}+)!R9zOAT9<Gxutx)$Y$9)*g-B}Lbnf3ke`w8#KeaeQNx
z1ZA@7rJ2D<(vt@d>#Gu_*4ah#)IEF7Y_|HCl^74c>{MNKcvoMT()_h`*47A+P+RM}
zWNT>u%y_^d%%KV}GxuodXVfRRKe`qR{wAyaM|O8C309|#XtG3}^Rc}1E@DTXl~iXS
z(C%IENaw6$uFA87(WYYbc4TYZ*%VLj#?$=OhbPa`b*!E+GmCD$7j^tP;?3y18((ox
zd5~TBS$nQMHJibJE^~E(R@fS08h(Yi4yx9^&Jic&8MS`HHlnGJ*qT_jsxTjD6))#(
zl*s#JU$G*4!1M3&5r!^hbC2pB?l7CJR+rz}Rtt%AdEqPHFKs)V$oaq;V8if<iEZqy
z{zL8lmHqg!7w>4*2`n^~kULC$8Fr_e=y=|&Ip9AmsY9{A;o>tI{$^%7osG6l^dviK
z^&1{S7;_3{U#s>2X`Xs_cG?&V-=s~wv5cQHX6I9eo@9L<RRS;`yP&w@Ua1zF+rv-k
zr*BwwmGa*!qCZ?>x=^OCY+&WgrMo>e&t7&QsUU;1Ezo2PX{b{#U1QM<nl@1csd)l~
zLd6wtn>QAQ%tsHoCG`o?(LU_2OQ+R2>L3>X_2_{&sX43`ncA%i(L+f@y?G+Z1aaXt
zo*oTj|8igdea^p}+Bz5*!Kn9Qp(ZCj<Zx>)gr|5}2cH>9gb58<bT5lKC&eXPRm*up
zkjgXvo;4Aha?V63Zk9pelNjJ9?(0$2e~enT(;4@Y$;*$7ey7Nc^G1WGZj1JOd6&Hn
z-$SmGk^ytrb}paQ-S$(r#e&W~`I@ni&Du6&wkWm2hPUk&u$(Ma%LmzK2aiTaQk0!X
z2en8Gj5IyMFvDYCYfc9bj?w0ZaIDTVR-e+w#}S~d5la<37iFR=7fuTP9%r%k9ODc2
zwbZW?3{o9bQW8N&SJy(DCwU<5)hnX)_O8>-=dhZt=k9$$DQ#{1T-1DqS44R@NC=>x
zXjc8VIwjtZAg?qGI6=9E?LQH<V-$g4Y{W-^9@b`njTYS;FM9PGp!F-?^egG>$q)d?
zw9|l4dSpYzVe_1$ab{smsWXFN8xDAKIk}uA&&Mj4ZIXMp$avzsTo4^Z-|MFIsm@Wd
ze8tN>+8Qe^vn{Nv38f{O^?G^BUYp6haB10#StV0HY+Ha*QO_j=j>UrS4(5AXf(Z$G
zTw9D+h}oWDjo_9&4>Yxu=_<${zDE+xr8=4OZU38Aa%di-G;Ajp6e#P*RQ=84wPbZ~
zT=`PYYq{D#3ot<n3lA@-s$!Nhla~)Qi3(DsY5`iHZ3_+|knVovWPU~lls6xA7a14Q
z{i%TcP<AgGjvMD6R!@}0c1->yYk!H-CKRT=0={iY5hkpR4hgSm=!Z*;<#>rNZ5%YP
zlHGD=v_sw&I7^NeWd>8KHBFft-(3)m^r%imb8$Q;b$v?TxR9UHhF5jnVpf-bF|hwA
zd{U8hS!+Z6Sasa_J(|`U9ZIXAWvq<O(49z{!rbHKqvbQ?j+VC)1+Ub7&ACgtw2#v!
zC6h7P`Ns$&K^@~MLk4>2RJ?&=_kPh6O_H>mgSQZ{ts$_KZ$g4B7%?vKH2<knt8>(E
z=ly?|1U}^^a{yd{x69&fOi~N7?>Z2tDQHQ^Le?XdkNJ98=6MAxad{skG$tkCMir<&
zuV-KvcLeID^%Yb;v)1VH87n$+GOcS=tsB^GyXO*Yubxt{AwT$#vwn=pw6Nimzh%%;
zDsJX$!5i|aYyy%+AxwxD?}-&S7xxOAXe1v5w=flPM4F7_OP#gx1c*QU9$d#yL8)9j
zp_S-T7Gq1X(l^r3SY?#R6`e==<ESZ70MsTyoLkx%I;Mdnt-u3Qjhp3ATU%z)-4;Yh
zXkqunt-PN=-2ZFQ0s7z+ZC_-VRHBT|-0qvB6BCHl*0!$N4b73KYGE8BVNy=ME0Gf~
zqphzYeI_%hZFy(!JY^_;b%cZ}nrVhag%c+-xFuV_bgEIFnT0*0Qby0Ml;sr}=Ci{2
zaUa6`PevtVL(_7Xp(ro#y2x1h?y#|EXI8itE->2oW1>$#GHX0tK-Vpg*=uV;(bmvo
zHvfo<6V!yNG^o2JlJj5zMtu=S;6(`yi-^buWw|~*-8zpQWB>b3^uQno;no0<%=16c
zg1=MQe@W?2VSWRlgcg#UY$zHeaTt+{ALjKJZ4|uHCDh!gmFyGpXvj4Z+#Zs+ce$IZ
z{rLk^-frCKuBCVtfFo^|1>C}pH&P|mBcQbm?c)}+NaLfl1th!<EsZz-B|!iD;va5<
z6AC4ixu%pz_NTdhh*Sly(G8p+cQN4!VOOa33vy9wFhXIeN(=DUvH+EKLij$cGWWk9
z^{<2jJOn8OSmY;5T&B@K&NqVr6fi{z#{RCi{*~|hk53DLp;#0#Sbo$sp-{9haZ?*u
zopD}0*aX9`(1L>g=iB~cz1Y|Ykm&tI3+tI0P6DEMb&HV5Vk$z4<A~BH8W}ZjBS`q8
zFuLA6=*NcKg2CeYb6TMPZwo5{3-k5d(64@|0<f+7m8kg&;Q~KDm-567o;T_R){oE4
z=MGHaqLR|q6|^wd87`gCEPVJh9{?*yL4}uznmTiE@CNeV4#WR-3t&qPY=Pkuu*#$W
zOusWi!iOu06598yv~_Cy+DNWI%GUNRCdKkoXbGs073wBBWB>nc?O$7m0lQve4VP3@
zR6OSE>gNxR*hoa0^00hl@3zw4uNGu<J&^D|FX(OMbBiEC@FRdi<4n9p1pnVoe;r7*
zB61GBPW6oMomt<50k>68HoA8mhqk=s=&O(Kvdzd7#PijyzP5XE#>4RlGIlQjG4uVU
zpvByvt!ffKn0-XTZHb=SU$i$4MgEVi8AA?KN#+~rjb*U&Wd<dG9soTeIaw~kOkguo
z1l%MS55n*ewx&}vt|Vm_qqnTEDyltSvWUG{9*q2X?_Q$e+V+zeG5?Ege!!Q|bf0}^
zch`2lsqtUt)LI@UM0?|_0f|J$zG-6LPhPmqhfiG8c(@UvBXox39JY7ct(OXa%k1dr
zxPObQ3j&{x0-$1#PyiC+K(W{K?MY9MJb*E0y^*9B0}yFxsmhx$n!M8G0Tn^zB+GB&
z7GS(tTQhcc7d3lkT`K3V<-Hh%N%=74jnl`*7jus9megu6_B$6=v#kbG6qFKJoJZSE
zbHP+8Ex_E0*jeaKSR#5cl%jDvFen!QhVly0a^nGoP{Y*pjcpHr3gdjg&M7=J>y%Fi
zc=imy&6OG70z7*K0L#-HH~kqF{1Sf*1-S2Z$L-^3+!w4QqLw^YyCWp8oeo0$1Q6=b
zwc*qB9S0+ZnE2k~i0*aZ1qvxgpdpCi0qGcG9Kk^{L`@NUJhj4S|1&pv&gulZec_7#
zpiw|G79n7g`uoKcmO&m$q_2;Bpb_oklyQ@<4BttJ{!6fEz{YcBez8b~c|6JA{bIpn
zQcP;9i=E(PY}yq%Q*}P}J4|`sMb3C{&-y?8bX}K_nUxi^FT!CBsewI>@NN75Cp7^F
z@~7Z(rI+>rdaf$C&;X4{F7$SpQKfFjOwkr+EidcMbRFU6vkS5FuaOkJx~7>H;_O*~
z>osu{?c@tNl?KRdF;i)V|B%xE&`RphU;;&R4d$Q}0Pq9P_V?V8h^_B3jnCY2o7(F%
zyN!Ft1gKv3>sC^{h$_8eQa9F@n9SR^xp4H^IMJ!M_UQ7f+Yh^AJ(yC>$i&o}{!Dfr
zd$$OxniTZjJYEC%Roi{6IayvgUpPEFz!d4t$w~&mxrhU(2`L!m2dMBMpRAvRn14SV
z)cppGBLv3N=1<gD+l+Q?J$@^-fM)2m5<ixsL^KXP+i_=X!sCb}QHy!5;x6oV&uL4A
z=w|$13(ttUv1{-;oxLSPB53`^fd4SyN=Z;WzNC75xJk{-$Vge)sXU?)<0t05?cs>G
zTI>PAvRU;m8ke+8vwBNTu!vspHbY({Yn3&5w5EJ{ec5A812ml78qr(QCV35gWO_rz
zC{1TO;D-(cU`=f7@nK^Dfq8u#z^@1$b>2m`XTa5ltqN4r>FDU<;^T9Ks=@VxF@=8L
zYjuAlfcuaEDxwe%U}taWm7H>yqTaZpiU9gaxNVhG3yo)Gnxewyx?>!4*c}vuGVIbb
z)dCX_X*X4D9T<fW5cjUYs1!Fq=-r#I2c*5Y*RK_4>)f*d=VWsv3t2`^Zejr<JzedX
z2l|T(9dQB|Ur}D(snA2G++{^k$m3gztt(8Eb1KGz-gjA`Q-aO-##tL0dp5Ya!sAK=
zd1~pok)-^JMXEys$^3Z$6(l%LS^)V{PE73YKK)CVHPm;SQ8hgS<8$aI&K@j+nNqIj
zq?invHb9GsV?Gn3f(x`oT`>7H9(-WQAKNPpxWlQZ3s~OQ>SEr9jH#kNx+zbGlA913
zw?T*@3+u!13^sf=ta+@5;OKs<J)af;TI~(kT+4)(v7;hw2DRW&8N}67TZEEunTIA)
zp~7o?t9pxp(1f=>R)80upnK)}G&opbmSb;rWEk(qZAN<8a(Q;Jk(>Nb6XSU$yp!m!
zFh(sfoV_XWwChcU{aiy5z$IqE_&%4^E&O=%pwGn;r9;8)JoL<QIzB$W3Uqcp2Q1c5
zN4R~AWzZ^O6hIk5O{zg#)Q@a#tNq0K73L@<`tQPE^D+0H(juSl?QqTzeXsvCU18RV
zx`a%Te6cg5mslck+O}Eayqpg@kvy$mW@kT$^t!HHO6P6j9N=d=dn$6m{pjnxcK%ht
zjJ;{c<{V&NG-PF;o_d&(<IM-Gxt6F}x|<yJ9+UFftJm(nBN1@QFzlY#KbiM`HQ^PA
zSR_FC{d8}6>BzX!GSa%!{cAh_25bah`$Tlm)tMhW#k0to-B*2bpt;NS<CqJJc0eEv
zpU(b1>+*qhtg6885q$wh8aT2AoVRGt304PF!C;v~TmZ=F*=_;M^7U3^ifqsbfyA$D
z<dNq9?}R!SclQ1@x%WDiH<&V<x@n{X>j$E`qXtNQG><_mkbSn9HCWSZ_w6<?&deG#
zCYrMKO05<@=;z+Adt?e)`BVYUt;4%FM^N(QH+cD$S-;JN<F&~0q}rt(j9+}6mMwB~
zo*|vrE};VT1Mq8aqXBm{fgMgRZ(n@wRD%K#-3>r7gG|2ldp^wbf1(DsE1g=adlalc
zb^iajy_@M!KR+=5MtfX{x7<$z!>K&SV+d<i=!+z~?BH(@Yzakl0T^JWxm#aIW2aaW
z8z_+b8o;Q@Dli^#$sSV{)!%d-bYP{tHY=^LpMD?NbiR4V`)hf(H4vmjFj5iMzkJ+C
z4?uIB7b1^WMXX2`DQ~mEXm*i~YWy0_V6ez*06)8RKSzWfna(*{4>U4@Az?Hg?$><%
zAeQdE$uyf>)d$D)espfE8PMB0CpiSv3u|ib@vegI7-3swjmNJ)ca4fhfxZhd51X$Z
z33=^j7gF$fFGuos_|VR*0_G&u;PD%lOAmP@c_x{+=z$9a3nw#RNJYfy<7s&&W@g;^
z(dNs2rPrc9?<Y=foi!S>3xvH8b(|CN26IW!{cjQzm*w#R;$-DC9Mhnt&Rx3bZr!`|
z0&D;!`~Aku_wYB@>EeUK-3U|upP998ou8_&gB0IxIUKUI5sdg<-F&5TnT3=Wvf{YX
z2*Qvhes~<Rl?P64$p7WIBN*NXYIbQp{M1ppv=7{{&H}qMpKpS7pbx!UP*(QiGugNN
zTJSFP^bSsOV?^Qy^XZ_MgVz$2Ev>Ot#|yL_>5Pz>T3~eWnHwD4bqa_ng3X}_hkuar
zxKC#>;|`|mN6WWP<8u^`>x2^Oy$3&YdF`9$EikT#rJ#gWnu9JCcZzgAkiG5FQ-cAh
zWXW1uOy17bdz4j>z$8u-cOe_~($ADfK+$W{=lJUzz_%I&(vw$wT+=V{sub0#+u-f}
zz@h=a0*{lkUWpbmd&7TjShPgpwaK!+sML0~8=%J&T?*VnYQSBHctnOxr;_vxUBh_q
z4d=hyj~^~H<W?Xc9OJ6p_@d%<y7;W~tJ=3?XQGA7nHPET%ajx}E5L)o;eO)up4ucH
zj5~WiO1}{q7*yfuP5P|eblnLU88e?W)Tq2&EkZ&r_Dnc95i$j73GJrdAuVa*jqqk)
z>+I7kHJa{my93~LZ0ZP)gSU);2~yOSW+M-0-r2npFv3?WSSfINivU!oy7&!&J5<QE
zBv)i3S_>yQm{VVDgaq5u^V%c|L?R38+mf)nP@AThMckQ@B{o!E^=qZNbQRYz2a3F}
z<7#yV+U1-bq!7;iBkhjRV<ZZ7Yn?eDWAgV$b5+dWEj`GP4gLH_7lFSoE>Zv-^6Om{
z6URQFV%9jJiks|WV#fWD4wI^7(#CIGZnSiC&xsRm{(tRV<zH1<+ZI$nTBO?`B~(NX
zAz+Y7NJ=9N-JKGmsFXN@h{6GpMnGw41VjO8q+2=<(ua8Ot>cU{j?er20ndDx-|?{b
zK5MVN_PwsVu8WwK)pX{1&TscSLbn@LGdzy4)@sFO6^#-)_@aw#q5Ip>pzST934?bZ
z0_T_0DfUMr;u9%;$$ylkJ{iBCD~&m;2oU$-`q;sy<V2G2(Z*TeBusZLecMjF?Cc2p
zAqVNFE!clu`h@2>)nNY0#@tn-fTh@U|Cq(XM01={+lptsc=_&@)xf^A3Xq?uxnSB0
zg;5yFyZs#N>#Z^@bm!XE?zlkD-_pU2^ACp%R_mo`^e&4ro%Hn=<>n45upQvW)63Gz
zeU~O|mNB+&?9;rYwow+fKewNX^y#<*-Jv4Kt_MPW8&1mnk~1lBKFi4r6?r6|UvG}z
zZiNQqmf?f;W~EgX2fT?O3OD4LAVjdCy07ZmZim#+pJ3^)>4iE+f~A)@F{{&~!TgVv
zrl4_Q1^9N}&$i*>GxO>eg`=<(G)>!8LU-ER+mqNwzW^b(+U$Y`YB(^LN^X1)2o7rE
zkCEO}c?{+6r%R5Qx&Kax?}!O?GdUf7bp=pukgAZ}s`v2<XbMk^Ifqcu7_n2Jxx<LE
z`E?4e_DcCoq^+QaGl0?BXaI`w*zuh7K=6Qhzu$Iiir6blK6~YEMEIu+h)MKM-@716
zLTPZRDX|u<(o^JceEMJw3B$P%YJPnlyo>fa=Zc|SyjD^KNOd@`xKdyH#I^JMS>xr{
zl4-k3ya(w0Dxrk6rWtoWE%5H$V7dup1M5sKu<2_viPB?Nw5zAv%NxfoR-_>PfuW(D
zwkny30+4FvZ01s<5D?=W&Ly3(S(h0IqB=PR3Pesbo}KCHiZ1zjhu<4#*bp&7RfIV$
zKx|9et~Z~WD4_pJI&}Z9gAPk2`|xu)172+Sc^}u|hmt{y#9>a`x+(!ENOT|!KO>Sv
zZngm&(=<gV)jfbVZ@PiM;>0|q4S+k%{B|O9q<xkf735~#lf#CI2<Uyb29tz-A{^Xj
zzW;VnSU?@Z+a-(&nk~|G{J`csofJi3R^q4$6I(Lo(F@uU0$Z?SseRnI@|a9~gLS~U
zT_N4GDN$Wlmj}Y!(wdWBpuIZ1*FlH0W|a6$42F$1c<jsGwGX0A*$Uz`PO_AD?8NCq
zQtq6-RLOyxm3(Y`onjvQa@h~M1`;<ss{9M;-)cP<hYoxs-5k-GJ`NOEy`>(x>0z21
z<!ty;Tn(ijUuv4E=g%vcFc|QCv1Z;z(@{_yb8AqITz)cJyy0cNEnIcsQw$T7q#GL*
zNH?ABfmWHXZUwX06VTeQeBz&S@H49!vk(@uNXIzbxU&=y21;OoTU)~cWHD!C93`C#
zZJVkxy|3Ef3G7;bCJq<mVqB?Hy|_I4c>f1s*=GVOYg4TO(4xz$D8tCar1$8Op#g!U
zqzr^7L^xB^%l+1(?t_m$;}&@ShZNO6V>R)L2i>_(*o>roULiD7Xro4}DCmTkl5EWy
z?8$p-wQRBrn6b8*J{0lZcr~MpPs*sA`K2B-z^2sdsGujtC0znGaa|y%68F&3MyS-?
zkEm*EibFg80Q3GtcwUaggM43{HG$Acu>`Z??wORBc$sP=4<ddkrrwXD?AB%?u?r>q
z=KB*@n($r7{VF|h%F@Q%(?0<%j@OCNecYtsqthzF`$v%KeV_~(jYwT^-GzeSvAk%B
z3^bsQluB;WEWKBTIC`K6tBBo5NKntvFgg|9L4*GmD9fI@7hg}aPZqZse(Wj%k{faP
zbxU2{8~q!e0BNv_?aT}PZyvd)U*cml7jv0W!+hyiwLfA;F;-A-{h%M;%=ojR32E^%
z<&(zwH>KOl4K-uRkBf>>U?RauX*>QvECJ+SNix}SRJJ{bEzI)Ul^x@S@ckS{STVp+
za($alnTZiv!6icv+!l-DBwvc%=NrxyyjHFEh6Hx$Kch<j{HW}NLu$>%!?Q9|)sCj)
zf`%+7%XdO}t3mp#p($JsSGFF`tc3B}Tv1Snq#mC<!lC%Sp)_77l0~Vfu6@LU#X6O@
z{@E+fDh*(aYh5^&?KanaP2V+t_<^_{Q-kbHki)bU(4^yW4?1sxS@RZgnMv$t`U)Dz
zyUV<U9Nsb6jPN*vEA>f3G?NUv2Y^1jc21n}@=DDSrF%HXJgairnqp!t5Vf83ZEv78
zXaTX|RIh&EU#Oq=E{k9HNQm`O<F7ud)HAcF!eBhsY*ZeG06<myh_&rGS5Zv9?3F7=
z0^aBKC9T%6`0B<yR^NQXWAg|tTy&3QbN#s_A3`(oL{F80F^NgQZZckkbQwttd2_^v
znPT)RrqcWg&3!Li@Kf7=^XPEiuNJt!_K{&OQy};#^)#=To3Vr`IbCmYD*_8VJ)D}g
zCi6ofvGhf0z;NgMCNibAof;i@KZSSnZ8P+1^|FtOm^HSDujFXQ3Oh*;8Zda*GOSo9
zGNG>E8_>Ab0xlCpr`TjQ)v0Qbr5TicaiF?dhh4HNV8?N8b3h0U4um00fzJ42+>c`P
zjY-ifE4yicfreDe4X*Sd2-u-s5WS|Rg9><<vX5$BT)GP=aQIB6RSKfcu*<oKjmP9H
zdEMCC-HBB9(OO&YMh)K=aB|S#utU)1G*nPK)|1xWM6kTBLSO26X+_1b0`B-vn=Yq-
zJjbD4>iGVX{QNY1vnD12ITA>xZWCB|$8#4HNDiR3o>g>KJ0&g){kEZl`l~nphwb%`
zkK{49(gDV(VQvs+O5yk*gr^rdI%<@nz!-9TN1@kz-xF1iAVP6z4UNawU(jr;YiXsK
zXX<~Rp3I|r+`)SfEWjG=;LDQYPt?J+%0ix#FWh1in@|n-_~}Tn9oj0?16j~~jRtp8
z9NHlWS*#40x^@N$PznXN_k9GKxU@kUDlv5i76UDymqK=VqfUrLS^)7~{l^Z_dW!Q6
zOebBM463Mm*Yi;K;0K2_Zc^=%jnQD2Tu%q(FW^;7pS4EUhvnY3AQ}qx0Zot^i)|oH
zzL<5ZFdy65DYB;ssTDHRjx9Moy*z-r7ZA$1*Ub-MB8z*_Q{tR{=^%i}^Eg(kuGRYF
zEi}WMDw{u+&HogLCy}XqX3Zk4zQLZ8kx&?#D=+7Njghl6QSL|zk4nsQBiIwNiq$)v
z@G^iW>UCHJZ7DeGovxbx`}$?=iiJCy-Udl*bud>yb9v#foVC^Q%u(6y=_QLj;Ea(C
z{+bcG@0UR=zDpP`;=UBW&|!_lD0oCHJ2*$jtOTSTCejeL9m}mj4i?$hr<|vhRm$re
z{K(93M(5-ub~j!mjBnNw@t3xwH5pLt(lf-30j<aQS!=~C4D6ag?M(6v!p0q)ng$Bc
z{i71Ou?<+L4IyO`aX4s};3&v+R8;;#EAaL+U?PGLGC$i>ku$6=dCLfK7KlBm%S;Z%
zZp73*VBzm_BB#S&=kl#H1h(-Ff1tW!Ev$wvOPI~y!qd?VYd;rFA%UpxcCqpSE(BPd
zdrDkAkJ+Y@w6s18M1<^&0tuT`o_pJ`9V*w#cv%90lFwr)$;tIVthHFNrx?Q}pUtD4
zC!4LUKKVdg9Y1xPs%#%Y`DPjLm<~KFU9LY_yqXHVK({o9!C(~!OmYeCTpRSk+KgHZ
zMBLerMj1+LWtFXIkEa)TZ*MqggO{XQ#?Ou7yrPq6(YAu?15v>jHVcI=zM-M}n-X6J
z*H=lbbsIwhZ17!ATsd|@bihJm^5Zj(#a%FM?;8M*UUXu8Ov%XB`9&=VIGob*=<S<5
zyKauw+=r8x&+aWO127j2-78xkbo#p5hRYyXQYuu<{O45sFe`t3+|cr~3jxd-r6|)=
zep3`k0lU-0+1RAiyPjXd?-a>iQo9P=n~JM3T0sJef_8&3AcJv{@zg4KP4)9&^-u6$
zAp+SjqqV~pdY*jT+{ccq3T?hnvLJTBhylq-O7(KYMIb|sYghA(^Vv!p!aWj$=TAr=
z8UN|P3a;wrFk{H)*|)H^BVO=sGY=NKSgEf9)x`zRXwu21*bK|bjL5Hq7$iyJz%9?q
zXPFWo92&~jG^3_dYhbJcgq!VLOD@9}<12)ACD73tdk)Hy>RmwE@KIFK9SGlDgKpj=
z_0BI4^~lsFOQb4m{5|W;-D#6zF>r-hV@coP`o~jMQhHtRn#Oe+24yme<32X$o}qmC
zWoG4_w(%zJo#y_d(#Md<dG?dS_SHz~=?4+bDns9DPW&q=WMyUJ^~*fd@RG`=Q+n@x
zYwsoS^asE#5*tu)h7ZL=Ky`t)(*Ls1kaup^<SFPTC;a^EzW2q#en%pR2R=qUQ+0Po
zVM50t5+rOwJi;)<*pbMC;L@8rt6lZn4nb^a&uOqKzYAzCt(@B)cAV+p!c^6bq$MOP
z#&Ul$aBh_{3gn)TaDYAUtX8JRs4=Gb<WS?d;mPAdXLedsP~9S1LlC*G!Q%9WT(j;0
z3TzDhvf34hg%I6(F*8L!@``+^P!iO_oXM)%6V|5NW!!XJPI}azE}Pa6UYeMU^L3p8
zE%D@0Y>8uaedNj7=7VL#?)9z!%_=5g+T<BqPDY;hTD#Xi2R7zvIz^mt_gx7Y%Zv>x
z>)vJ=@wkP7l9Ka<kWvs)XsByv*U-=i!cHHgvdER<9$4{-(y+~L$~kvQ%Vlf$=8E?&
z2!gh&DxDwiFo9M)aHi|er-Y`$J55Z~F<6p0a`gF)Q$+UiBfi^XEX!Q>9Rj$Czmb0L
zSt_OO&a-&bg*C+g`ruMoaKV<H#_T`%)hYmw1N^hqUXj{tMir<0;TjP1c@3CBP_q(W
z)}ep{+&vswpl?+aDolDbVPT0HzM=(07*~-PKA=Lbj9Np^rojmCdIw+imw6Rhie@j-
zL?BU{H|PM!={%K!cyNhs9Z}5D>HEO41GJcdKw({0|NAIdl7Nf=s`4H$c&MUG(fjvz
zAA&)p4&gFVCTgzWMz8cbFez1hSZqbKS3r+Y^P`&9ZbT^R5|cy-rQ${<-QM=;aSpQ_
z6+BstJJQ2-mZxmaOEJ`2cjUoq2uF*A{w@HWM+nsGZH&WiM^|Mh0kA{aJO~SOO@G9#
zoo~rzN;uYRYDZK-y&|ur6|JAI=>oa}5wWqEB}%{qVOv(P46XFN+%F*k+(~{))0kxr
z7bgn0_a$|f_*Gq3Ka&4wPA>TEznA#7{gZ<$N7>tXeNg%;Q&<7UilFuh3n52?m>ZWN
z+}1bfwOBU#QES`>gex;t1C=a9@zd1J2)TS$xI)wT(0f~56>T7U_1jbnWZ8r|NRdL0
z;lsnk;?JKySKPlloe~!_v%>A3{1kSCCwwJyXGTMmVjeySh}{s7)Iwdq%zq^!C$!ZB
zo!rc_F5tl}YSU`kJvI-1W$U>w*kXxy$ux}<T*}jk<(}j|sNrCkS77hTrHS|z#JaBe
z1A1lppV)1iZ+^;q@==iN!*KlC%a<>ok{-lfimxJK4Nwd41-fMLk$}R-?$^^npRmu&
zc+_GE#c<e6-(DT!s&aAO%ZEU~bLQL#9qBpiZb{hrgfVKjr3vw6%79NpQNe@0BI!z0
zTYi`RY(%7d-v;p29!1knf!0qO_~Y0YN}V_}hIaxwH?}a=N3v|gzNocXe^9+)VJAO3
z@LP63d*x9i_du|u=CJKl284f|gCMWVPU_<*crXqWZ`X&SsB71EfuZ%Ds$==mTcVo9
z3>6PB&-VpLHWTdaXCY@$yY6bR!m(QAujD<{p|?b(BwF7SV_C9cYz?%zP9h{|zGl4<
z3yb_aSUI4U%Mh;a0N!17g?<IW6d|@aOEaO9z0JqcEPQ;*^RCZN0*%LE{F@fJvJU%(
z!Hp$ZWuT+u@+qLV43UPy$SctJ{U)(0aE8Eq^1aLYbA#C3*JS?|4*X*`-;nmRE0DXh
zSI|I5I_FxVHoh#(&0TNtO#d-{inJ~F+qZ9H3YJ&oy^SOGKL*xgb-K&$b3m<}1l$?n
zE4KCAqGQiik-SgE&b(b7q$p0?%PmTCD%A-A<2Mp>jVQ@@yqV0qjseQ!HV`G2P(O!H
z>N;xi++9QxO>N(nD3^ZILyCW+j^(B$*-N*Nc{Zh5lU+e;K7NkAu8hz?jbx|te0Ds0
z3v8aDq>KP}TR@rNS=eeuEq-5FI9l`W9d`*rum*%ijF-1SNHatFqr@yCBT9QOu{Zwu
zhxhN}flycGnv>7Np^CirXBuYOS}9>vRNlJGh5T1^bz`2~4<FjojT|iV3cjG0Re;SH
zWHGjCPY!Lj7dEEGZ+KHry8-O0!t>Qy?-LVfwznaq6RZ^r46uo&%U=?tf<K$M5%u0?
z>HpM8>6&#$p?KE7jsf4<YJZ~V+MFE1<*Vwn>oHRK5?aDjY!H)c$qtdrr<*tILjBH3
z#Sco@o-Hw?`VO#!aASP5H|s2Rjb|Z{_t})p`YRms-#S*=UcMA6X<`AHzE#i**1BWz
z%^J(VjoYuyX?4;$>?6QG94{$JG0elb%Na9JIW^=Yy#HQCqh#-EjDpR4sErMSz62x~
z^#w5>#|(UT)Eg;tfX=QL0@T;*{Lgt-Nf*ZwL^ApbLQ%hila>~pO;3*0wq7uHn4u8z
zH>~9z7KuwkRX~+3g{cx!?8?$Rxz?=3n)s5YzZv#`qvDn&N~FWfYAC}-bBoUO$mg?q
z*RCB^)aowvaKa?>;a|j5tB*|ITG4kBBq?xm5$gb(n1B3~)5O)MLUzpX3$iYfO?WO4
zS_*=+i&<FCQpL^e@-0h9xGHM|H<Xf@B|WN7SI=-xk%>h)9NQV^_kH;EjrZ=P#2y!3
zJvYNk-8a9Z`mTX1om2d-VcRRFWMA5i>cGwo+RpjPF5|I%0yb+>O+Y@evyOz_Iq+_r
ztgG)~*>p3o8EL+mX5aF<z<!wA7q9uL;FbLQDK}yusgd5)2@d`OVCdMVlb=6_XWznX
ztP{_lMj6v9aiztmJi)JuPAcD7(>c8_Cf6LZs0UI4I6zX_X%T`v?9wBUk$2NKpT8i9
z4Lw}u+wVSw`|P`GtW2;ZF&8<Vut`?vIpnB5cT4KqBjvz_T-1waw#9NT=QQ;_lKSlW
z)9H=BZMU}+5oHd=s@K~eu^<$5VBn2mCHYeRz9BJ2A={XZREr@-9zpwINzAqO@{Nhi
zL6ZKbE3*&8t<EsNz?uS-tJ)T`p)^|?jTb;qgxi@~1-tg17;M6F1vKVi4G|tRZycWr
zpYce|8b6MCbKwkK$SK_ySi>v|F<Lj!`Im1_Fhu!&`Mi(8BwExh%TZn52l!Aqd3hD|
z-SR6{XU;?7HT|Qv#G?RnkN$EVBqavhcb`d=bc&`^&jpfm8=ltNZ|dU=dPA;we-ysH
z20;o|wEij-5v+XDu{*3(m(C0M%A`7Gud7)T{De-X+&1K^K%>oTdirxnXv~d}l$o*k
zV++vKjn^+XzZqN)1oJopp0k-oouHObYhn{0*D=^<#yksc*_pke*m5YgsNuSk6wgZ<
zGKx>LJIW}5N{tT4CudC>g<8~fHp8q>`oR0ecOZ}JP$V6{?%XZQZ}{|AfuLKue!0f7
zTBlIe`1n^H$U7jtc^Rru*q|u3-mJ)?oqfRK4rIo!#++w>M@C?aL#jb1U>dR@#|SvC
z@!0Bm#a=8DEI!?i&fvc-4vy|wz^`vunoi~a_JP7fPkZ=kgY>NzMLQ8(s+?!fs`aY-
zAPT`QxxF7q(qEtQTrrOV_6A8#e5y`$OkMZJj=pFgdEC%bL8}@=Ej0CmTna~O{w9%B
z-qMOQ?)N2B^P&^nSx8@cjo#<rgk;Cu0|{UEAn6w-@57H4E7+CHpOaTqq)w6kY;6>y
z)u?H*G+bq57+)Yk8aP^ch7@<Qxe@}!x_y_uNkE9$Npbb-s-}Tey>8lF4JJ&IkS?*`
z@W!_^`H!CBUp|p!JepuB)Vlmlh9<kh(-Vp%bvVSCI)&`$W7qiG>)^643PooCldw{5
zth6iD#;wkf>e<y7AU;28pQ;Xu=mB(tcRgSfZi6hw<EIk7<*GLoxJO4vQ*{b$i}mbX
z&(Ggf9|-hGU#aUx;ERL&dCg0%R0MVN!l$Ode|GJS6cBNk;3|8&gK_y%urzxomvM%w
zi4k(3d>)osTuYkSnT0N3X}B?G^LpQECHU-anx)L?sFVP((j@Nv{Sy*8QZoRE(pgxI
z7QyXJH3Wtg*5`!Z&Gx^o7-h19?m;Itg|>yiGyu)#{oJ}KDzxJ)<a=BCZi_(5hTx!k
zaU*DC1gqAScp9{b@{-CTdW5w+s=f1K2cH()zJ0rex3E7U)3rn#cD#XkL_yvb?))w(
zQ+f#&iU*u$t#aE~jJ=>Rsb+d5k!wsWR>|2p4@EM2>mZTyGe?aoKtD6425k#V?~gxd
z5$w`^k?E>E$)f=hisv1|pOn1qqv6x#G>vrZYMMGn8r1jdMu29!nUj47<hE>A$ju?k
z>W&ob;j>z=DOOtMr@JaAH>%rIeye`?#nXlulKgPn75*ed@W(BP&5fg8mHs1C8R;EC
z3Yl&x`vt_JX4|1cq?Y)C7~9hf3Mwi#Jk@y!`tD+f_I_VOA2%aJLi_CtjGwEuo|keV
zM$i`TC|V`+)vFd0O)+UDsXkMXvMDYJVG`9S)Ytl;+P0a~@u_>EV$TQjbO<Qmi<`U^
zA;MkLzKyXmhIG1Zisras2}l{<n~zXka*$w@tnDG%6!;8*^OpI$0aJNN1{FRv7qk%b
zEhR0Te$FXQZ09J;mxN$G6`L=nQnMeij?}e~8f`_0!X0K6*5A*g%r^ir#{*0IP-zhK
zrftYpo@i{HUb<#5g@+pf5|HK3o!Gf{YwuLdpN2U0`|tCmtbnsCV-MOkr9ykgO?uo$
zZ;pOrN@JZmD8Bjm2<Aa`{gDxg(JhhjiX7iRN=F9>R~80zwy{qgDZP)s0xhR{?yr9#
zaK~v1$3VWV4T?0t1<@)V8^;oW&v*i@@xU&@D&VNI8)3M7-CR`a=U5tN8Dn6MUHT~F
zQuv$#mqgFc42|f>Ouc`29H~GVt#C<BPF6`#j7<eATwKLVGQiW_9GATb``<M5<&HYL
zi`6pXa{#ng!*;kX4k|$9R}!csj7kcS#;41TR8%+><O0*x>FzM!bz49P3dLtj?Y=_^
zp?!N|d3Hl+5$7pMUY&xo{rL3I`GP)(+K3!_K+ZbeL&*DQGDI>KDtgnKrW_2Bm&=6@
zK$A$t0b0Z=`)h;!QOkTZ<_DNJ?%;;#7QzPIo}v_7MGmnO$|FZEL7+U@cA)rvvk&tv
zgJK!K2!rcLKJ$;b;8L(3C2JCz>?8l(iz&bMIKEdAD)^;vjq=dGzunrBCDw$Hh<@7}
z4xk|<guD3E@bdR>zCYog;*U7ohQ&T-q{8A-B}9&L5`Wvvzh0kp35YJK+`CSs$~Z%4
z%M&>2ruX_va%t1j#c_sHN-Q*Ra4t|DxJ(Y~-T3TQd@BcwuHTcQ=P$%)1^KGRt7tCu
zm(vM6;(U7bsxEU#(|+2}i7U>!6E+tk=r}ehsC1R>tC%d3YW0&z519<;(4>~@#Vn-n
z#@<$5?<m`IYFbo_6LH>26Fqh6)gp5kjlTO0q}p4@^QwrF-##4N0|ex(lK=dWO4t}f
zb#eQ}pMS~=@jH<AI6D2))D%<^u|Nta?W!4@tO<(o&qMyhhZB@jSGyl4zw@gsa_EUE
z+1=eeeaiCEiJyM<=Yh~>;!-3^&o<_nR}>c)pE)1J8aZPBU!2DpeqHjGn$eQ0U6YV(
zK9-n)<rjDQ!%xQNAIMq1WR+PyJI_hEM56M-Ud8!7GDF(hh_NHNdGluH>CZz}8A|S7
zM_|M8%Dgq<l~HX_V-ZAQOB-MBM`qKUAE_8aG5T&}qAz@m5f*To`{(r17E9I{CK2zP
z(In@j?goXLi6|5>+o`)Fz*tG+%0G|)bCBkO{=VCG1bgcewPg!4gd^#`+&+rD-)~LQ
z(I1UQw+s!jcn7FCZmrp#+RLQING1RJ$@EGYE>-Lb%z13t-Kn0QLsPKJhjZs$llx-;
zv~vDU53P0k<~dj|r0-00-21n+^5@{beaPg2V<XM%DA#WG7tKZePC`r^@>YL_;`<<f
zwdCOJ%Ik;Uo92c@USpWNab#+)aiafC?cj|U)K8Fe-rcK&60u7L1_my9m_GW&qOq`<
z*Z9clZ7IhElt6-1u$lTbMDdeF={m|KH-UK50fM{-mGCfVfk7;qt_50<s;|i~MW#9Z
za|IDANZ#sbDUG>F;15}lyAZXy^MsR1qVha;E|<*VPRgOLQ|KRXs2{wF=Y@rZp7L&P
zg$L5o1CbG!+aU$scK~dd;=O^Y$wMl1hrfQ$?++-K#~buzqAwuDP85k}j*<R-4l7#@
zo{%LOXr#hFeD9D76zrki)1i<3bA12y1j7|sy(IFS=Pkdk(Z65yUnlvl9^B*&j!*yk
z$Dhr>akgP<uj6b($@8S7q=AtqH3R+q{gK6`C_0~qppCp~bCTfy;$nt*M6%)Z^8y`4
zJF`<$1lgs_^7&oo&!6wCWB#}M`sv<fd0`-FwRMtSdU|@D*Va~7)mglLI_e+G@az4s
z=7ZJhfNG(#g@uI~&auoj%73?a|2X+L8|wwj#Ls<21HE?Pzq59}J7Q!0^923#8-Hr@
zaMn)TM;(@W5v-k-0?8HsJmTkfz<-Sj&M9{iH99^!w>@jpIW#0;w9|6x=Ocgq#=qi>
z)^&38-z9BsMd-{49#W_MFK*6X1q`bE>o-$|xcP@v$}h|qs-Tbm7n6zVhx14amuN$V
zIy=>w^6kRYRlb?Q|K}!8;o^=-$15llyz#ROxBPPGoES1nS?*V$D%R4~rD>CU?(iG!
z>2GgqQYC~cnIThB-q>vvdl4~1^C#`w6@=~p-1!X;<+jlOhn4-|i5X7eG6h7QS!vL{
zFkFdCCblFA`E<8?EW_^%hWnu=0$;pG;Fp8+r>3CwUN(4(PTAAw9&1|Qx-mL2f2TR^
z&&^4ezootX$QBkGU>7cLH?+o^rBj$SUuW+3%S-<7q%x0!F%+DT<Nszre)!8@XG{A_
z*4@7U+_=7R6de+$tegG#`#WWL`Cz}1-Q`U|#b0g+KTLyP<p{Zba9Q<p?g?>!<aLE9
zn=83H_EIbNEtHJP1%A6MCzm}aRV{ic4fuMPnVGqR7~wk0)-&Xfz1Iz)pK~PI4E-Tw
z>1A#cA@pK==nzJJ!+-6t`%0@!HMltf`-`E5qZ$IA0xthBX8Zh~AKJi6d+s4zg5z(}
z@c0fuz&TMT5bMDHo<v{N$o1{6CH*Tw&FG??o!x3QyqK?`%K0x-4!!Odua|kcnEGKP
z60M;hN_Fz<-S^52hX^zgAZl}_lBaL1fs$9_x3T>=tMo^@_efM?Lt__sa6^e|><YRK
z@bK`~BvyoeJnC1*(*HBAuSw1cJMDmZZg_BdW}k|Fb?6V5%S3j^)d}m4_3O7fQsa&>
z{ke92d^+DB>bFN3LWgrm#n?2xog^<KBcu0;&YU`z*sl-&%TChb_&XeLNoMcUjbCF(
zgk{<1ofjzctMw7W<zKkVyfx%r=Kg-H?oQU-IL;sNLgl?fV3<#Dt=SZvyHF1}X;F-U
z?uQ%uYmE|+(+1-Pu<RV)#=P^TBrVJJ3ObN>kuLI&Bm5sy!ODBMYS-|oE);um5y4#J
zlnj)=xYiH%8E<mnglvAJTE3!;RmBDQTWYh7&(IV<9Po=f>%9i>PPA!;;DOw&4gw{=
ziU06|{rL#faeW7fGPTEX{6`M;g9mv7%JGg4T4rZtSh-i}D*WQ7{yZJ|U`PvQFwlRp
z&Bz~0U(a}h)ZGZgp$6REf}eQZdlbhK+-+LWF^KV41I|mbva+)1Vreo(F-%SI<xGfL
zqlk%#&(bG&>R;W`E_Sw9MTt+%&JqEB`Pd=oNnS@+r=#No@8=S95q#1R+;#`YcBI?*
z!N-_A(C!EXxlBIQuBj@hl79>D^_djt2`e0dRAUpI{mQWF)+v+vh{kCoev9NT1YAOy
zd9-p)IJ>!-DeghT#9Jtaj!v`s9RK3QaUO$mF(*s!(3&p!=razZHTF=5Z^k!}Yo>%C
z(&VhxfHc*VK;ZU-Q(-i3ArlL2Kr6$kFTw~xHLWRz*?<GHs*>FJOo;#Lvva&o(_kDf
zDJdx-r<&7{ST#VkDLn{6>F9JWjiCx1!mJvUyOMXh!dN84W9;7^!ji4;>_D>qcC$E^
z2f3@mGK`1AL>G!0ryX`eX3e9|Kn<lmSRFII9Z-S+C7bPVWB^Z$!w$(Xd;B)B(km$b
z1%#i1sd)(>u`ATa^vntFLdusgARr+69KA4Yrqy=UV0N!Z=!3;V2Sv*(-W8#XC6{=m
zth`G}mY^9Rklu1=^P~RQ*cM$^Yj4|}pbw<bM_}0LHVGqj@CO=xeMYFC2)FP>S#8U4
z#75kwf+9E(<z&iCi~|{cSK67R9xLzusM;MZ8;dpi3aHL|<+Z-h%TDjp4_E8SF?qh$
zXJIT;pM`N~rd<!qXIr2f8+?7$Voe+=Xt)7oaM4*cg1ZomY3isAKuO%~t)rYDdsen+
zJMq+dNKfb^{!7|<D@UK9fjPmRa;zVJXfcQMm#1Rd>jT)sWO?jDU))YQ+#CYvxyvr#
z8@02e$$syv7xKd^6t5q@+`xTDvi+dO%j<_FqWM-Qk;1Xr=Cl;FHn+yyB{u&=KK=24
z_bM1}IX0Kq+$>X(K~z{*pqK8d^@0_CrLF{;A7@6-Z!c9RjX_*5i*FFLdksVef!BCu
zG!{fLlg8A=Su=t)Hx1L&O6%_qAXh;>4x*YYSiWJSR|=@P(76<v3SROJ0*Z$VPNKbL
zMh9H`Y?8(hknMtj2FTGi3=rO`(PS?Z?Y3dKD-4@zGQ3%}MKsvbP9HhRjPz9`jp;$+
z-UWm9-Q8@%*uEKmT<i|=d3Mw8@NT*5m7dHhN^mYk^2vOITq!XsBKmC8FZlG?4(B4n
z`hW~;7lzg4@6;rm!Z(NnK1K9H==RjaKWD$Y2|)cyU2xY#1JsUdis&=ylwS_m*<qd;
zY85vaWh4<)6q6`tUuG+Jt*;5{s_C*WFJ1kllQdQvm*#z6!pj6p<2&(X9Kw-&O#E4L
zn-pDHIXiuj84>R{)YWNt7i{HebKlCR=X&IUfnH@SbKQDk4rbKrMtRM8Bl5uoKCHk8
z{bL5B=oi+dMO`1Fia59#_!nxI*%o^0Yl6LMyzhH`IP-c`(1Gk3zeBdpIsJa)$t(TA
zQK2+@B;owkJ#Ly5@?S>qP9Ew>+UlCSx&DD8PPeNNNm8i|CAhVgV<ExL$&_tAbQ9Xg
z9)}@^AQ8|>p_rR);}U0o8?7D7$sX%|;7!+5USh6*6mH3{XZK(iHbfetV75w~_nww5
zW}UzvFA=&fN6p(=L5W>}P|V{Q9q+Pt=N-z4&4zYbE|HB_lYs7L{-qg%*4;F|RKjnV
z@6@Z^IDB7G7wM6>^JYy-y)>bRzpqn9-T37=Mf{U{r*FQ>T_*Fr^Y{JouMP0~34eTJ
z!xdg;C#OT&8m~gsZ(mLFd|84@RVd^9S18=_2^Vam&-GkWg@O5FeV#3Nf9;fpA%rom
zy%r-`B@4YmTkZ^=|NKQ-G&m0<n{SFG;~g`?5C-GBUHV>`!&>Tppcp#;jpy&-O}ajz
z()GjmT8wZbb)tLs3cF*b`Tn*8hS<WU|H8{hJI4(w-iQ#h1@0<;=ic!!RKh|r?IK8x
zzRa*oH0m$firNGk{7V3ljz^g_-8XKC0-)4p3>CK*#C4yDxL8iRgKKH6R|>MrQDH*R
zgi!2W$LMr*+Fi=0HQuqJx!9FH?a8Sd7Y8g(`wxS&_A*04O)93&`+jyTJm7@A0b!5j
z&utiJL5N$_e<JK;I_-|3$=o~x2)hkqWU$QEWTRDK9?9S83<Xo0Gl5~EqPcDn^HWot
z1BFu&{9tamnGhem;Vuz0>B5pq_5FzM+wkNOEmTdun#N$KJBGz#XJ;7Fre<c+MdxoH
zQ6vD%m<hwE1%BgMMG!ako_<Z8iFqBG<La>UF>2%`2Ga}cj^5xi4GB00k#@<?$IRpS
z9hf+^?UpCU9|RlKb}9C}xtHAocQK!Xd0O_4CUbvvbYBts_!Zrf`Y@ra=uL{rf>d!|
z{Z<FvcB?D85SPz1=NF7wUrcZ@L@f?UIWY{iH-r^VO<I6K_aCeP((mlpgg_RW`7x;E
z(ViJPD3v`{6vbnudA9KD`_)R%vOa1uMg`++Hd!B#nl;~k!|Qmy<A9LYkX{};+2rK8
z+Nb?<Wj!O;A2lqLe6d=#UmK8;p!qoHCgi1HyVg%?2LDIiCH<V=KBRWL@#_}7vYn=z
z$1W62qB`geR`gYcl&OA%TCl<;t8tkk4K$*#HZIi)sE7aIU6Ax5Al}7R+>t=<cD=14
zR(wLqzO&BplLM2x*g$_mHmR`d*xy#<56kxNKHyy~L*HA#407r2n`<Ng4YmSs6s`i8
zOy0)k&v*KlS4ZwPTn&y^vHi;e{O6<pckQ3x>Hp^Y?`!t!s`}qL`QJMEX$JncPJW(&
rpTGH^J^A09{5%6ce-o}UyoWPGY`5*$-(<QE{*#kZyqtB>@YeqU`8vAC
literal 0
HcmV?d00001
diff --git a/doc/index.rst b/doc/index.rst
new file mode 100644
index 0000000..cc0463f
--- /dev/null
+++ b/doc/index.rst
@@ -0,0 +1,42 @@
+.. vim: set fileencoding=utf-8 :
+.. Andre Anjos <andre.anjos@idiap.ch>
+.. Mon 13 Aug 2012 12:36:40 CEST
+
+.. _bob.paper.ijcb2023_caim_hfr:
+
+==============================================================================================
+Bridging the Gap: Heterogeneous Face Recognition with Conditional Adaptive Instance Modulation
+==============================================================================================
+
+This package is part of the signal-processing and machine learning toolbox Bob_.
+
+This package contains source code to replicate the experimental results published in the following paper::
+
+ @inproceedings{georgecaim2023,
+ author = {Anjith George and Sebastien Marcel},
+ title = {Bridging the Gap: Heterogeneous Face Recognition with Conditional Adaptive Instance Modulation},
+ year = {2023},
+ booktitle = {IEEE International Joint Conference on Biometrics (IJCB)},
+ }
+
+If you use this package and/or its results, please consider citing the paper [GMM23]_.
+
+.. figure:: img/caim_framework.png
+ :align: center
+
+ Diagram of the proposed framework.
+
+
+User guide
+---------------------
+
+.. toctree::
+ :maxdepth: 2
+
+ running_caim
+ running_pdt
+ references
+
+
+.. include:: links.rst
+
diff --git a/doc/links.rst b/doc/links.rst
new file mode 100644
index 0000000..b053dbc
--- /dev/null
+++ b/doc/links.rst
@@ -0,0 +1,13 @@
+
+
+.. This file contains all links we use for documentation in a centralized place
+
+
+.. _idiap: http://www.idiap.ch
+.. _bob: http://www.idiap.ch/software/bob
+.. _buildout: http://www.buildout.org
+.. _pypi: http://pypi.python.org
+.. _installation: https://www.idiap.ch/software/bob/install
+.. _dependencies: https://gitlab.idiap.ch/bob/bob/wikis/Dependencies
+
+
diff --git a/doc/references.rst b/doc/references.rst
new file mode 100644
index 0000000..f839bdc
--- /dev/null
+++ b/doc/references.rst
@@ -0,0 +1,11 @@
+.. vim: set fileencoding=utf-8 :
+
+===========
+References
+===========
+
+.. [GMM23] *A. George, S. Marcel*, **Bridging the Gap: Heterogeneous Face Recognition with Conditional Adaptive Instance Modulation**,
+ in: IEEE International Joint Conference on Biometrics (IJCB) 2023.
+
+
+
diff --git a/doc/running_caim.rst b/doc/running_caim.rst
new file mode 100644
index 0000000..5d6ad1e
--- /dev/null
+++ b/doc/running_caim.rst
@@ -0,0 +1,38 @@
+Running HFR Experiments with CAIM
+=================================
+
+
+Running the experiments
+-----------------------
+
+The database class provides a standard interface for training, development and evaluation.
+Now, using bob package, all the steps can be automated: training, and evaluation.
+
+To run the experiment, use the following command:
+
+.. code-block:: sh
+
+ ./bin/bob bio pipelines vanilla-biometrics -vvv \
+ bob.paper.ijcb2023_caim_hfr.config.mcxfaceprotocols.VIS-THERMAL-split1 \
+ bob.paper.ijcb2023_caim_hfr.config.lightning.x_iresnet_CAIM -m
+
+.. note::
+ Note that the first config file specifies the database to be used and the protocol, one can change the config file to
+ run the experiment for different datasets and protocols. Please inspect the configuration files and pipelines to familiarize the experiments. Please run this job on a GPU machine for faster target.
+
+
+Running the command should launch the training, and once the training is finished, the evaluation process would
+start automatically. At the end of the experiment, the score files will be written to the folder corresponding to the
+specific protocol and dataset.
+
+Evaluating results
+------------------
+
+To evaluate the score files run the following command.
+
+.. code-block:: python
+
+ ./bin/python bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py -sf path_to_scores-dev.csv
+
+
+It should return AUC, EER, Rank-1 accuracy and Verification rates at different FAR points.
\ No newline at end of file
diff --git a/doc/running_pdt.rst b/doc/running_pdt.rst
new file mode 100644
index 0000000..531efad
--- /dev/null
+++ b/doc/running_pdt.rst
@@ -0,0 +1,38 @@
+Running HFR Experiments with PDT
+================================
+
+
+Running the experiments
+-----------------------
+
+The database class provides a standard interface for training, development and evaluation.
+Now, using bob package, all the steps can be automated: training, and evaluation.
+
+To run the experiment, use the following command:
+
+.. code-block:: sh
+
+ ./bin/bob bio pipelines vanilla-biometrics -vvv \
+ bob.paper.ijcb2023_caim_hfr.config.mcxfaceprotocols.VIS-THERMAL-split1 \
+ bob.paper.ijcb2023_caim_hfr.config.lightning.x_iresnet_PDT -m
+
+.. note::
+ Note that the first config file specifies the database to be used and the protocol, one can change the config file to
+ run the experiment for different datasets and protocols. Please inspect the configuration files and pipelines to familiarize the experiments. Please run this job on a GPU machine for faster target.
+
+
+Running the command should launch the training, and once the training is finished, the evaluation process would
+start automatically. At the end of the experiment, the score files will be written to the folder corresponding to the
+specific protocol and dataset.
+
+Evaluating results
+------------------
+
+To evaluate the score files run the following command.
+
+.. code-block:: python
+
+ ./bin/python bob/paper/ijcb2023_caim_hfr/scripts/hface_metrics.py -sf path_to_scores-dev.csv
+
+
+It should return AUC, EER, Rank-1 accuracy and Verification rates at different FAR points.
\ No newline at end of file
diff --git a/pyproject.toml b/pyproject.toml
new file mode 100644
index 0000000..bb5e83c
--- /dev/null
+++ b/pyproject.toml
@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools", "wheel", "bob.extension"]
+build-backend = "setuptools.build_meta"
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..3c6ee7d
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,9 @@
+setuptools
+numpy
+bob.core
+bob.extension
+bob.ip.base
+bob.ip.color
+bob.io.image
+bob.pipelines
+bob.bio.face
diff --git a/setup.py b/setup.py
new file mode 100644
index 0000000..2a3372b
--- /dev/null
+++ b/setup.py
@@ -0,0 +1,56 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+from setuptools import setup, dist
+dist.Distribution(dict(setup_requires=['bob.extension']))
+
+from bob.extension.utils import load_requirements, find_packages
+install_requires = load_requirements()
+
+
+
+setup(
+
+ name='bob.paper.ijcb2023_caim_hfr',
+ version=open("version.txt").read().rstrip(),
+ description='Source code package for PDT',
+
+ url='https://gitlab.idiap.ch/bob/bob.paper.ijcb2023_caim_hfr',
+ license='GPLv3',
+
+ author='Anjith George',
+ author_email='anjith.george@idiap.ch',
+
+ keywords = "bob",
+
+ long_description=open('README.rst').read(),
+
+ # leave this here, it is pretty standard
+ packages=find_packages(),
+ include_package_data=True,
+ zip_safe = False,
+
+ install_requires=install_requires,
+
+ entry_points={
+ "dask.client": [
+ "sge-big = bob.paper.ijcb2023_caim_hfr.distributed.sge_big:dask_client",
+ ],
+ },
+
+ # check classifiers, add and remove as you see fit
+ # full list here: https://pypi.org/classifiers/
+ # don't remove the Bob framework unless it's not a bob package
+ classifiers = [
+ 'Framework :: Bob',
+ 'Development Status :: 4 - Beta',
+ 'Intended Audience :: Science/Research',
+ 'License :: OSI Approved :: GNU General Public License v3 (GPLv3)',
+ 'Natural Language :: English',
+ 'Programming Language :: Python',
+ 'Programming Language :: Python :: 3',
+ 'Topic :: Scientific/Engineering :: Artificial Intelligence',
+ 'Topic :: Software Development :: Libraries :: Python Modules',
+ ],
+
+)
\ No newline at end of file
diff --git a/spec-file.txt b/spec-file.txt
new file mode 100644
index 0000000..d49a008
--- /dev/null
+++ b/spec-file.txt
@@ -0,0 +1,286 @@
+# This file may be used to create an environment using:
+# $ conda create --name <env> --file <this file>
+# platform: linux-64
+@EXPLICIT
+https://conda.anaconda.org/conda-forge/linux-64/_libgcc_mutex-0.1-conda_forge.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/ca-certificates-2021.10.8-ha878542_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/kernel-headers_linux-64-2.6.32-he073ed8_15.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/ld_impl_linux-64-2.36.1-hea4e1c9_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libgcc-devel_linux-64-11.2.0-h0952999_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libgfortran5-11.2.0-h5c6108e_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libstdcxx-devel_linux-64-11.2.0-h0952999_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libstdcxx-ng-11.2.0-he4da1e4_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/mysql-common-8.0.27-ha770c72_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libgfortran-ng-11.2.0-h69a702a_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libgomp-11.2.0-h1d223b6_11.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sysroot_linux-64-2.12-he073ed8_15.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/_openmp_mutex-4.5-1_gnu.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/binutils_impl_linux-64-2.36.1-h193b22a_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/binutils_linux-64-2.36-hf3e587d_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libgcc-ng-11.2.0-h1d223b6_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/alsa-lib-1.2.3-h516909a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/aom-3.2.0-h9c3ff4c_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/blosc-1.21.0-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/bzip2-1.0.8-h7f98852_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/c-ares-1.18.1-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/charls-2.2.0-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/expat-2.4.2-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/giflib-5.2.1-h36c2ea0_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gmp-6.2.1-h58526e2_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/icu-68.2-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/jbig-2.1-h7f98852_2003.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/jpeg-9d-h36c2ea0_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/jxrlib-1.1-h7f98852_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/lame-3.100-h7f98852_1001.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/lerc-3.0-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libaec-1.0.6-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libblitz-1.0.2-h9c3ff4c_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libbrotlicommon-1.0.9-h7f98852_6.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libdeflate-1.8-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libev-4.33-h516909a_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libffi-3.4.2-h7f98852_5.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libiconv-1.16-h516909a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libnsl-2.0.0-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libogg-1.3.4-h7f98852_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libopenblas-0.3.18-pthreads_h8fe5266_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libopus-1.3.1-h7f98852_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libpciaccess-0.16-h516909a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libsanitizer-11.2.0-he4da1e4_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libuuid-2.32.1-h7f98852_1000.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libuv-1.43.0-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libvpx-1.11.0-h9c3ff4c_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libwebp-base-1.2.1-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libzlib-1.2.11-h36c2ea0_1013.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libzopfli-1.0.3-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/lz4-c-1.9.3-h9c3ff4c_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/ncurses-6.2-h58526e2_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/nettle-3.6-he412f7d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/nspr-4.32-h9c3ff4c_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/openssl-1.1.1l-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pcre-8.45-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pthread-stubs-0.4-h36c2ea0_1001.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/rhash-1.4.1-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/snappy-1.1.8-he1b5a44_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/svt-av1-0.8.7-h9c3ff4c_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/x264-1!161.3030-h7f98852_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/x265-3.5-h4bd325d_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/xorg-libxau-1.0.9-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/xorg-libxdmcp-1.1.3-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/xz-5.2.5-h516909a_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/yaml-0.2.5-h7f98852_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/zfp-0.5.5-h9c3ff4c_8.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gcc_impl_linux-64-11.2.0-h82a94d6_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gettext-0.19.8.1-h73d1719_1008.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gnutls-3.6.13-h85f3911_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libblas-3.9.0-12_linux64_openblas.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libbrotlidec-1.0.9-h7f98852_6.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libbrotlienc-1.0.9-h7f98852_6.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libdrm-2.4.109-h7f98852_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libedit-3.1.20191231-he28a2e2_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libevent-2.1.10-h9b69904_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libvorbis-1.3.7-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libxcb-1.13-h7f98852_1004.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/readline-8.1-h46c0cb4_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/zlib-1.2.11-h36c2ea0_1013.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/zstd-1.5.1-ha95c52a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/boost-cpp-1.76.0-h312852a_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/brotli-bin-1.0.9-h7f98852_6.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/c-blosc2-2.0.4-h5f21a17_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gcc_linux-64-11.2.0-h39a9532_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gxx_impl_linux-64-11.2.0-h82a94d6_11.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libcblas-3.9.0-12_linux64_openblas.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libglib-2.70.2-h174f98d_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/liblapack-3.9.0-12_linux64_openblas.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libllvm11-11.1.0-hf817b99_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libnghttp2-1.43.0-h812cca2_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libpng-1.6.37-h21135ba_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libprotobuf-3.19.3-h780b84a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libssh2-1.10.0-ha56f1ee_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libtiff-4.3.0-h6f004c6_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libva-2.13.0-h7f98852_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libxml2-2.9.12-h72842e0_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/mysql-libs-8.0.27-hfa10184_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/openh264-2.1.1-h780b84a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/sqlite-3.37.0-h9cd32fc_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/tk-8.6.11-h27826a3_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/brotli-1.0.9-h7f98852_6.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/dbus-1.13.6-h5008d03_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/freetype-2.10.4-h0708190_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gstreamer-1.18.5-h9f60fe5_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gxx_linux-64-11.2.0-hacbe6df_4.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/krb5-1.19.2-hcc1bbae_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/lcms2-2.12-hddcbb42_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libclang-11.1.0-default_ha53f305_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libxkbcommon-1.0.3-he3ba5ed_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/nss-3.74-hb5efdd6_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/openjpeg-2.4.0-hb52868f_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/python-3.8.12-hb7a2778_2_cpython.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/alabaster-0.7.12-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/attrs-21.4.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/blinker-1.4-py_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/brunsli-0.1-h9c3ff4c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/cachetools-4.2.4-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/charset-normalizer-2.0.10-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/cloudpickle-2.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/colorama-0.4.4-pyh9f0ad1d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/cycler-0.11.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/dataclasses-0.8-pyhc8e2a94_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/ffmpeg-4.4.1-h6987444_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/fontconfig-2.13.1-hba837de_1005.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/fsspec-2022.1.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/gst-plugins-base-1.18.5-hf529b03_3.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/heapdict-1.0.1-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/idna-3.1-pyhd3deb0d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/imagesize-1.3.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/iniconfig-1.1.1-pyh9f0ad1d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libcurl-7.81.0-h2574ce0_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/libpq-13.5-hd57d9b9_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/locket-0.2.0-py_2.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/munkres-1.1.4-pyh9f0ad1d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/nose-1.3.7-py_1006.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/olefile-0.46-pyh9f0ad1d_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/py-1.11.0-pyh6c4a22f_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyasn1-0.4.8-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pycparser-2.21-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyjwt-2.3.0-pyhd8ed1ab_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyparsing-3.0.6-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/python_abi-3.8-2_cp38.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pytz-2021.3-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/six-1.16.0-pyh6c4a22f_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/snowballstemmer-2.2.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sortedcontainers-2.4.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-applehelp-1.0.2-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-devhelp-1.0.2-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-htmlhelp-2.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-jsmath-1.0.1-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-qthelp-1.0.3-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/tabulate-0.8.9-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/tblib-1.7.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/tensorboard-plugin-wit-1.8.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/threadpoolctl-3.0.0-pyh8a188c0_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/toml-0.10.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/tomli-2.0.0-pyhd8ed1ab_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/toolz-0.11.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/typing_extensions-4.0.1-pyha770c72_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/wheel-0.37.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/zipp-3.7.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/absl-py-1.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/babel-2.9.1-pyh44b312d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/certifi-2021.10.8-py38h578d9bd_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/cffi-1.15.0-py38h3931269_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/cfitsio-4.0.0-h9a35b8e_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/click-8.0.3-py38h578d9bd_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/cmake-3.21.3-h8897547_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/coverage-6.2-py38h497a2fe_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/cytoolz-0.11.2-py38h497a2fe_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/docutils-0.16-py38h578d9bd_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/frozenlist-1.2.0-py38h497a2fe_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/greenlet-1.1.2-py38h709712a_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/hdf5-1.10.6-nompi_h6a2412b_1114.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/importlib-metadata-4.10.0-py38h578d9bd_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/kiwisolver-1.3.2-py38h1fd1430_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/llvmlite-0.38.0-py38h4630a5e_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/markupsafe-2.0.1-py38h497a2fe_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/msgpack-python-1.0.3-py38h1fd1430_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/multidict-5.2.0-py38h497a2fe_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/numpy-1.21.5-py38h87f13fb_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/packaging-21.3-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/partd-1.2.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pillow-8.4.0-py38h8e6f84c_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pluggy-1.0.0-py38h578d9bd_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/psutil-5.9.0-py38h497a2fe_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyasn1-modules-0.2.7-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pybind11-global-2.9.0-py38h1fd1430_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyqt5-sip-4.19.18-py38h709712a_8.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pysocks-1.7.1-py38h578d9bd_4.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/python-dateutil-2.8.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyu2f-0.1.5-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyyaml-6.0-py38h497a2fe_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/qt-5.12.9-hda022c4_4.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/rsa-4.8-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/tensorboard-data-server-0.6.0-py38h3e25421_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/tornado-6.1-py38h497a2fe_2.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/typing-extensions-4.0.1-hd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/unicodedata2-14.0.0-py38h497a2fe_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/werkzeug-2.0.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/zict-2.0.0-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/aiosignal-1.2.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/async-timeout-4.0.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/boost-1.76.0-py38h2b96118_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/brotlipy-0.7.0-py38h497a2fe_1003.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/click-plugins-1.1.1-py_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/cryptography-36.0.1-py38h3e25421_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/fonttools-4.28.5-py38h497a2fe_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/h5py-2.10.0-nompi_py38h9915d05_106.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/imagecodecs-2021.11.20-py38h678ac2f_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/imageio-2.13.5-pyh239f2a4_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/importlib_metadata-4.10.0-hd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/jinja2-3.0.3-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/markdown-3.3.6-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pybind11-2.9.0-py38h1fd1430_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyqt-impl-5.12.3-py38h0ffb2e6_8.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pytest-6.2.5-py38h578d9bd_2.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pywavelets-1.2.0-py38h6c62de6_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/scipy-1.7.3-py38h56a6a73_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/setuptools-49.6.0-py38h578d9bd_3.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/sqlalchemy-1.4.29-py38h497a2fe_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/yarl-1.7.2-py38h497a2fe_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/aiohttp-3.8.1-py38h497a2fe_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.extension-6.1.1-py38hfe1193c_0.conda
+https://conda.anaconda.org/conda-forge/linux-64/bokeh-2.4.2-py38h578d9bd_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/dask-core-2021.12.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/grpcio-1.43.0-py38hdd6454d_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/joblib-1.1.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/matplotlib-base-3.5.1-py38hf4fb855_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/numba-0.55.0-py38h4bf6c61_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/oauthlib-3.1.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pandas-1.3.5-py38h43a58ef_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pip-21.3.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/protobuf-3.19.3-py38h709712a_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pygments-2.11.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pyopenssl-21.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyqtchart-5.12-py38h7400c14_8.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyqtwebengine-5.12.1-py38h7400c14_8.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/pytest-cov-3.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/tifffile-2021.11.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/zc.buildout-2.13.6-py38h578d9bd_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.blitz-2.0.24-py38h5652565_0.conda
+https://conda.anaconda.org/conda-forge/linux-64/distributed-2021.12.0-py38h578d9bd_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/networkx-2.6.3-pyhd8ed1ab_1.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/pyqt-5.12.3-py38h578d9bd_8.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/scikit-learn-1.0.2-py38h1561384_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/urllib3-1.26.8-pyhd8ed1ab_1.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/xarray-0.20.2-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/zc.recipe.egg-2.0.7-py38h578d9bd_1.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.core-2.2.8-py38h5652565_0.conda
+https://conda.anaconda.org/conda-forge/noarch/dask-2021.12.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/dask-jobqueue-0.7.3-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/matplotlib-3.5.1-py38h578d9bd_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/requests-2.27.1-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/linux-64/scikit-image-0.19.1-py38h43a58ef_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.io.base-4.0.0-py38h33f8406_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.math-3.1.9-py38h6772507_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.sp-2.0.18-py38hb84a040_0.conda
+https://conda.anaconda.org/conda-forge/noarch/google-auth-2.3.3-pyh6c4a22f_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/requests-oauthlib-1.3.0-pyh9f0ad1d_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.io.image-2.5.1-py38hd166c48_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.io.video-3.0.5-py38hfcb1991_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.ip.base-3.0.0-py38h59b548f_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.ip.color-2.0.18-py38h5652565_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.ip.gabor-2.0.19-py38h59b548f_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.learn.activation-2.0.18-py38h5652565_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.measure-5.0.0-py38h6db0feb_0.conda
+https://conda.anaconda.org/conda-forge/noarch/google-auth-oauthlib-0.4.6-pyhd8ed1ab_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.db.base-3.2.0-py38h0c7a4f2_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.ip.facedetect-4.1.0-py38h3da435b_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.learn.linear-2.1.10-py38hd1e3326_0.conda
+https://conda.anaconda.org/conda-forge/noarch/tensorboard-2.7.0-pyhd8ed1ab_0.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.db.atnt-2.0.18-py38h6911941_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.learn.em-2.1.8-py38h59b548f_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.pipelines-2.0.0-py38h57a381b_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.bio.base-6.0.0-py38h67de6e0_0.conda
+https://www.idiap.ch/software/bob/conda/linux-64/bob.bio.face-6.0.0-py38hc6584bb_0.conda
+https://conda.anaconda.org/conda-forge/noarch/sphinx-4.3.2-pyh6c4a22f_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinx_rtd_theme-1.0.0-pyhd8ed1ab_0.tar.bz2
+https://conda.anaconda.org/conda-forge/noarch/sphinxcontrib-serializinghtml-1.1.5-pyhd8ed1ab_1.tar.bz2
+https://www.idiap.ch/software/bob/conda/linux-64/bob.buildout-2.3.0-py38h1171eb6_0.conda
diff --git a/version.txt b/version.txt
new file mode 100644
index 0000000..30893e1
--- /dev/null
+++ b/version.txt
@@ -0,0 +1 @@
+0.0.1b0
--
GitLab