From 7e3956ef74a6918f5eec3bfd08161166e5d0fb51 Mon Sep 17 00:00:00 2001
From: Justin <605492+justinpinkney@users.noreply.github.com>
Date: Sun, 4 Sep 2022 21:46:39 +0100
Subject: [PATCH] Img condition (#1)
* update reqs
* add image variations
* update readme
---
.gitignore | 7 +
README.md | 281 +-----------------
assets/img-vars.jpg | Bin 0 -> 453200 bytes
.../sd-image-condition-finetune.yaml | 134 +++++++++
ldm/models/diffusion/ddpm.py | 8 +-
ldm/modules/encoders/modules.py | 38 +++
main.py | 25 +-
requirements.txt | 9 +-
scripts/gradio_variations.py | 112 +++++++
scripts/image_variations.py | 122 ++++++++
10 files changed, 455 insertions(+), 281 deletions(-)
create mode 100644 .gitignore
create mode 100644 assets/img-vars.jpg
create mode 100644 configs/stable-diffusion/sd-image-condition-finetune.yaml
create mode 100644 scripts/gradio_variations.py
create mode 100644 scripts/image_variations.py
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..5af5443
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,7 @@
+logs/
+dump/
+examples/
+outputs/
+flagged/
+*.egg-info
+__pycache__
\ No newline at end of file
diff --git a/README.md b/README.md
index 3738dde..5d9857c 100644
--- a/README.md
+++ b/README.md
@@ -1,278 +1,13 @@
-# Latent Diffusion Models
-[arXiv](https://arxiv.org/abs/2112.10752) | [BibTeX](#bibtex)
+# Experiments with Stable Diffusion
-
-
-
+## Image variations
+[](https://twitter.com/Buntworthy/status/1561703483316781057)
+_TODO describe in more detail_
-[**High-Resolution Image Synthesis with Latent Diffusion Models**](https://arxiv.org/abs/2112.10752)
-[Robin Rombach](https://github.com/rromb)\*,
-[Andreas Blattmann](https://github.com/ablattmann)\*,
-[Dominik Lorenz](https://github.com/qp-qp)\,
-[Patrick Esser](https://github.com/pesser),
-[BjΓΆrn Ommer](https://hci.iwr.uni-heidelberg.de/Staff/bommer)
-\* equal contribution
-
-
-
-
-
-## News
-### April 2022
-- Thanks to [Katherine Crowson](https://github.com/crowsonkb), classifier-free guidance received a ~2x speedup and the [PLMS sampler](https://arxiv.org/abs/2202.09778) is available. See also [this PR](https://github.com/CompVis/latent-diffusion/pull/51).
-
-- Our 1.45B [latent diffusion LAION model](#text-to-image) was integrated into [Huggingface Spaces π€](https://huggingface.co/spaces) using [Gradio](https://github.com/gradio-app/gradio). Try out the Web Demo: [](https://huggingface.co/spaces/multimodalart/latentdiffusion)
-
-- More pre-trained LDMs are available:
- - A 1.45B [model](#text-to-image) trained on the [LAION-400M](https://arxiv.org/abs/2111.02114) database.
- - A class-conditional model on ImageNet, achieving a FID of 3.6 when using [classifier-free guidance](https://openreview.net/pdf?id=qw8AKxfYbI) Available via a [colab notebook](https://colab.research.google.com/github/CompVis/latent-diffusion/blob/main/scripts/latent_imagenet_diffusion.ipynb) [![][colab]][colab-cin].
-
-## Requirements
-A suitable [conda](https://conda.io/) environment named `ldm` can be created
-and activated with:
-
-```
-conda env create -f environment.yaml
-conda activate ldm
-```
-
-# Pretrained Models
-A general list of all available checkpoints is available in via our [model zoo](#model-zoo).
-If you use any of these models in your work, we are always happy to receive a [citation](#bibtex).
-
-## Text-to-Image
-
-
-
-Download the pre-trained weights (5.7GB)
-```
-mkdir -p models/ldm/text2img-large/
-wget -O models/ldm/text2img-large/model.ckpt https://ommer-lab.com/files/latent-diffusion/nitro/txt2img-f8-large/model.ckpt
-```
-and sample with
-```
-python scripts/txt2img.py --prompt "a virus monster is playing guitar, oil on canvas" --ddim_eta 0.0 --n_samples 4 --n_iter 4 --scale 5.0 --ddim_steps 50
-```
-This will save each sample individually as well as a grid of size `n_iter` x `n_samples` at the specified output location (default: `outputs/txt2img-samples`).
-Quality, sampling speed and diversity are best controlled via the `scale`, `ddim_steps` and `ddim_eta` arguments.
-As a rule of thumb, higher values of `scale` produce better samples at the cost of a reduced output diversity.
-Furthermore, increasing `ddim_steps` generally also gives higher quality samples, but returns are diminishing for values > 250.
-Fast sampling (i.e. low values of `ddim_steps`) while retaining good quality can be achieved by using `--ddim_eta 0.0`.
-Faster sampling (i.e. even lower values of `ddim_steps`) while retaining good quality can be achieved by using `--ddim_eta 0.0` and `--plms` (see [Pseudo Numerical Methods for Diffusion Models on Manifolds](https://arxiv.org/abs/2202.09778)).
-
-#### Beyond 256Β²
-
-For certain inputs, simply running the model in a convolutional fashion on larger features than it was trained on
-can sometimes result in interesting results. To try it out, tune the `H` and `W` arguments (which will be integer-divided
-by 8 in order to calculate the corresponding latent size), e.g. run
-
-```
-python scripts/txt2img.py --prompt "a sunset behind a mountain range, vector image" --ddim_eta 1.0 --n_samples 1 --n_iter 1 --H 384 --W 1024 --scale 5.0
-```
-to create a sample of size 384x1024. Note, however, that controllability is reduced compared to the 256x256 setting.
-
-The example below was generated using the above command.
-
-
-
-
-## Inpainting
-
-
-Download the pre-trained weights
-```
-wget -O models/ldm/inpainting_big/last.ckpt https://heibox.uni-heidelberg.de/f/4d9ac7ea40c64582b7c9/?dl=1
-```
-
-and sample with
-```
-python scripts/inpaint.py --indir data/inpainting_examples/ --outdir outputs/inpainting_results
-```
-`indir` should contain images `*.png` and masks `_mask.png` like
-the examples provided in `data/inpainting_examples`.
-
-## Class-Conditional ImageNet
-
-Available via a [notebook](scripts/latent_imagenet_diffusion.ipynb) [![][colab]][colab-cin].
-
-
-[colab]:
-[colab-cin]:
-
-
-## Unconditional Models
-
-We also provide a script for sampling from unconditional LDMs (e.g. LSUN, FFHQ, ...). Start it via
-
-```shell script
-CUDA_VISIBLE_DEVICES= python scripts/sample_diffusion.py -r models/ldm//model.ckpt -l -n <\#samples> --batch_size -c <\#ddim steps> -e <\#eta>
-```
-
-# Train your own LDMs
-
-## Data preparation
-
-### Faces
-For downloading the CelebA-HQ and FFHQ datasets, proceed as described in the [taming-transformers](https://github.com/CompVis/taming-transformers#celeba-hq)
-repository.
-
-### LSUN
-
-The LSUN datasets can be conveniently downloaded via the script available [here](https://github.com/fyu/lsun).
-We performed a custom split into training and validation images, and provide the corresponding filenames
-at [https://ommer-lab.com/files/lsun.zip](https://ommer-lab.com/files/lsun.zip).
-After downloading, extract them to `./data/lsun`. The beds/cats/churches subsets should
-also be placed/symlinked at `./data/lsun/bedrooms`/`./data/lsun/cats`/`./data/lsun/churches`, respectively.
-
-### ImageNet
-The code will try to download (through [Academic
-Torrents](http://academictorrents.com/)) and prepare ImageNet the first time it
-is used. However, since ImageNet is quite large, this requires a lot of disk
-space and time. If you already have ImageNet on your disk, you can speed things
-up by putting the data into
-`${XDG_CACHE}/autoencoders/data/ILSVRC2012_{split}/data/` (which defaults to
-`~/.cache/autoencoders/data/ILSVRC2012_{split}/data/`), where `{split}` is one
-of `train`/`validation`. It should have the following structure:
-
-```
-${XDG_CACHE}/autoencoders/data/ILSVRC2012_{split}/data/
-βββ n01440764
-β βββ n01440764_10026.JPEG
-β βββ n01440764_10027.JPEG
-β βββ ...
-βββ n01443537
-β βββ n01443537_10007.JPEG
-β βββ n01443537_10014.JPEG
-β βββ ...
-βββ ...
-```
-
-If you haven't extracted the data, you can also place
-`ILSVRC2012_img_train.tar`/`ILSVRC2012_img_val.tar` (or symlinks to them) into
-`${XDG_CACHE}/autoencoders/data/ILSVRC2012_train/` /
-`${XDG_CACHE}/autoencoders/data/ILSVRC2012_validation/`, which will then be
-extracted into above structure without downloading it again. Note that this
-will only happen if neither a folder
-`${XDG_CACHE}/autoencoders/data/ILSVRC2012_{split}/data/` nor a file
-`${XDG_CACHE}/autoencoders/data/ILSVRC2012_{split}/.ready` exist. Remove them
-if you want to force running the dataset preparation again.
-
-
-## Model Training
-
-Logs and checkpoints for trained models are saved to `logs/_`.
-
-### Training autoencoder models
-
-Configs for training a KL-regularized autoencoder on ImageNet are provided at `configs/autoencoder`.
-Training can be started by running
-```
-CUDA_VISIBLE_DEVICES= python main.py --base configs/autoencoder/.yaml -t --gpus 0,
-```
-where `config_spec` is one of {`autoencoder_kl_8x8x64`(f=32, d=64), `autoencoder_kl_16x16x16`(f=16, d=16),
-`autoencoder_kl_32x32x4`(f=8, d=4), `autoencoder_kl_64x64x3`(f=4, d=3)}.
-
-For training VQ-regularized models, see the [taming-transformers](https://github.com/CompVis/taming-transformers)
-repository.
-
-### Training LDMs
-
-In ``configs/latent-diffusion/`` we provide configs for training LDMs on the LSUN-, CelebA-HQ, FFHQ and ImageNet datasets.
-Training can be started by running
-
-```shell script
-CUDA_VISIBLE_DEVICES= python main.py --base configs/latent-diffusion/.yaml -t --gpus 0,
-```
-
-where ```` is one of {`celebahq-ldm-vq-4`(f=4, VQ-reg. autoencoder, spatial size 64x64x3),`ffhq-ldm-vq-4`(f=4, VQ-reg. autoencoder, spatial size 64x64x3),
-`lsun_bedrooms-ldm-vq-4`(f=4, VQ-reg. autoencoder, spatial size 64x64x3),
-`lsun_churches-ldm-vq-4`(f=8, KL-reg. autoencoder, spatial size 32x32x4),`cin-ldm-vq-8`(f=8, VQ-reg. autoencoder, spatial size 32x32x4)}.
-
-# Model Zoo
-
-## Pretrained Autoencoding Models
-
-
-All models were trained until convergence (no further substantial improvement in rFID).
-
-| Model | rFID vs val | train steps |PSNR | PSIM | Link | Comments
-|-------------------------|------------|----------------|----------------|---------------|-------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------|
-| f=4, VQ (Z=8192, d=3) | 0.58 | 533066 | 27.43 +/- 4.26 | 0.53 +/- 0.21 | https://ommer-lab.com/files/latent-diffusion/vq-f4.zip | |
-| f=4, VQ (Z=8192, d=3) | 1.06 | 658131 | 25.21 +/- 4.17 | 0.72 +/- 0.26 | https://heibox.uni-heidelberg.de/f/9c6681f64bb94338a069/?dl=1 | no attention |
-| f=8, VQ (Z=16384, d=4) | 1.14 | 971043 | 23.07 +/- 3.99 | 1.17 +/- 0.36 | https://ommer-lab.com/files/latent-diffusion/vq-f8.zip | |
-| f=8, VQ (Z=256, d=4) | 1.49 | 1608649 | 22.35 +/- 3.81 | 1.26 +/- 0.37 | https://ommer-lab.com/files/latent-diffusion/vq-f8-n256.zip |
-| f=16, VQ (Z=16384, d=8) | 5.15 | 1101166 | 20.83 +/- 3.61 | 1.73 +/- 0.43 | https://heibox.uni-heidelberg.de/f/0e42b04e2e904890a9b6/?dl=1 | |
-| | | | | | | |
-| f=4, KL | 0.27 | 176991 | 27.53 +/- 4.54 | 0.55 +/- 0.24 | https://ommer-lab.com/files/latent-diffusion/kl-f4.zip | |
-| f=8, KL | 0.90 | 246803 | 24.19 +/- 4.19 | 1.02 +/- 0.35 | https://ommer-lab.com/files/latent-diffusion/kl-f8.zip | |
-| f=16, KL (d=16) | 0.87 | 442998 | 24.08 +/- 4.22 | 1.07 +/- 0.36 | https://ommer-lab.com/files/latent-diffusion/kl-f16.zip | |
- | f=32, KL (d=64) | 2.04 | 406763 | 22.27 +/- 3.93 | 1.41 +/- 0.40 | https://ommer-lab.com/files/latent-diffusion/kl-f32.zip | |
-
-### Get the models
-
-Running the following script downloads und extracts all available pretrained autoencoding models.
-```shell script
-bash scripts/download_first_stages.sh
-```
-
-The first stage models can then be found in `models/first_stage_models/`
-
-
-
-## Pretrained LDMs
-| Datset | Task | Model | FID | IS | Prec | Recall | Link | Comments
-|---------------------------------|------|--------------|---------------|-----------------|------|------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------|
-| CelebA-HQ | Unconditional Image Synthesis | LDM-VQ-4 (200 DDIM steps, eta=0)| 5.11 (5.11) | 3.29 | 0.72 | 0.49 | https://ommer-lab.com/files/latent-diffusion/celeba.zip | |
-| FFHQ | Unconditional Image Synthesis | LDM-VQ-4 (200 DDIM steps, eta=1)| 4.98 (4.98) | 4.50 (4.50) | 0.73 | 0.50 | https://ommer-lab.com/files/latent-diffusion/ffhq.zip | |
-| LSUN-Churches | Unconditional Image Synthesis | LDM-KL-8 (400 DDIM steps, eta=0)| 4.02 (4.02) | 2.72 | 0.64 | 0.52 | https://ommer-lab.com/files/latent-diffusion/lsun_churches.zip | |
-| LSUN-Bedrooms | Unconditional Image Synthesis | LDM-VQ-4 (200 DDIM steps, eta=1)| 2.95 (3.0) | 2.22 (2.23)| 0.66 | 0.48 | https://ommer-lab.com/files/latent-diffusion/lsun_bedrooms.zip | |
-| ImageNet | Class-conditional Image Synthesis | LDM-VQ-8 (200 DDIM steps, eta=1) | 7.77(7.76)* /15.82** | 201.56(209.52)* /78.82** | 0.84* / 0.65** | 0.35* / 0.63** | https://ommer-lab.com/files/latent-diffusion/cin.zip | *: w/ guiding, classifier_scale 10 **: w/o guiding, scores in bracket calculated with script provided by [ADM](https://github.com/openai/guided-diffusion) |
-| Conceptual Captions | Text-conditional Image Synthesis | LDM-VQ-f4 (100 DDIM steps, eta=0) | 16.79 | 13.89 | N/A | N/A | https://ommer-lab.com/files/latent-diffusion/text2img.zip | finetuned from LAION |
-| OpenImages | Super-resolution | LDM-VQ-4 | N/A | N/A | N/A | N/A | https://ommer-lab.com/files/latent-diffusion/sr_bsr.zip | BSR image degradation |
-| OpenImages | Layout-to-Image Synthesis | LDM-VQ-4 (200 DDIM steps, eta=0) | 32.02 | 15.92 | N/A | N/A | https://ommer-lab.com/files/latent-diffusion/layout2img_model.zip | |
-| Landscapes | Semantic Image Synthesis | LDM-VQ-4 | N/A | N/A | N/A | N/A | https://ommer-lab.com/files/latent-diffusion/semantic_synthesis256.zip | |
-| Landscapes | Semantic Image Synthesis | LDM-VQ-4 | N/A | N/A | N/A | N/A | https://ommer-lab.com/files/latent-diffusion/semantic_synthesis.zip | finetuned on resolution 512x512 |
-
-
-### Get the models
-
-The LDMs listed above can jointly be downloaded and extracted via
-
-```shell script
-bash scripts/download_models.sh
-```
-
-The models can then be found in `models/ldm/`.
-
-
-
-## Coming Soon...
-
-* More inference scripts for conditional LDMs.
-* In the meantime, you can play with our colab notebook https://colab.research.google.com/drive/1xqzUi2iXQXDqXBHQGP9Mqt2YrYW6cx-J?usp=sharing
-
-## Comments
-
-- Our codebase for the diffusion models builds heavily on [OpenAI's ADM codebase](https://github.com/openai/guided-diffusion)
-and [https://github.com/lucidrains/denoising-diffusion-pytorch](https://github.com/lucidrains/denoising-diffusion-pytorch).
-Thanks for open-sourcing!
-
-- The implementation of the transformer encoder is from [x-transformers](https://github.com/lucidrains/x-transformers) by [lucidrains](https://github.com/lucidrains?tab=repositories).
-
-
-## BibTeX
-
-```
-@misc{rombach2021highresolution,
- title={High-Resolution Image Synthesis with Latent Diffusion Models},
- author={Robin Rombach and Andreas Blattmann and Dominik Lorenz and Patrick Esser and BjΓΆrn Ommer},
- year={2021},
- eprint={2112.10752},
- archivePrefix={arXiv},
- primaryClass={cs.CV}
-}
-```
-
+- Get model from huggingface hub [lambdalabs/stable-diffusion-image-conditioned](https://huggingface.co/lambdalabs/stable-diffusion-image-conditioned/blob/main/sd-clip-vit-l14-img-embed_ema_only.ckpt)
+- Put model in `models/ldm/stable-diffusion-v1/sd-clip-vit-l14-img-embed_ema_only.ckpt`
+- Run `scripts/image_variations.py` or `scripts/gradio_variations.py`
+Trained by [Justin Pinkney](https://www.justinpinkney.com) ([@Buntworthy](https://twitter.com/Buntworthy)) at [Lambda](https://lambdalabs.com/)
\ No newline at end of file
diff --git a/assets/img-vars.jpg b/assets/img-vars.jpg
new file mode 100644
index 0000000000000000000000000000000000000000..80917114082254016c8a2f24b8c02d6ec299f8fc
GIT binary patch
literal 453200
zcmeFXWn7#+w>NqX?(W{=-r`Ov4n>MPrMUax&_at#@d5=}oMOdYic5jwFfh0jw=z%$
zd8haN+y0t6}w$_fAogan8ne!%@YE|apX
zthuI^hJx}lc|-&N;F4Ilxj7*b0D!ZLmxq?(Qw9S=BL?(s00W={@Bl1;-onz;O-57m
z8Gs1?yno|=2iPot-~<3u+`oDKoBsa_5n5S$S^@w_6OsC*m8YdM0@oq#6H7M_F91L;
zK+qX|yxe}{K?El8KnRGyOTV%0-}vM=e)$(hLHR309@;~_4{s*?Oc=?B?g@x@O
z{C8atIS~cR*?YP0*v?x?_aAKODT~nM-+8Sp7yd+xst`Hueh2|6qGBWu1SpwWqxHKiI=n
z>rYuW9{zoKK&=XjfcXYx;uL!sQ)V0&R+RXUN1WxrGMy_7XOyt+C}EyJUtEm$Yy0NC;v}+
zYZu)=^>y=l`ll>6r@zL=U%Ik(Qv5^r1Z)4v>*b;IhwkR3`A3Hi7An8{{a<+jIY1tG
zhWIf6x&S}GhxqXW&jEQaYv12}6p(dw^YgH`wew<-L5wnM24xpZ9xevHM|^w$@O#Yu
zP6Ghv^M8#+5E;h5X`x>MKq3MWdGc?X`ZfU6wj;Wa^}lJ%ya0fYp!LmJdV7HXVh5o7
zexV}l06sthP$FysBfti5BV-c-#DOOWiB$j%KnE}c%m7Qk4sZrM03RR_cnyRDF~B<@
z703i~fqb9@r~qn!MxYJ&3iJabz;|Er1%kprF`y(+Cg>BW6jTjr27Lt$fhIr;pf%7g=nQm?
zgp7oPM1n+%#Dc_&B!VP^q>7}2WQt^q>GB>gavMjO&vJtWkvIlZ7ax`)pavpLeatrbR@)Yta@&WQM6f_iK6b2M-
z6mb+K6kQZc6nB(hlvtEZloFH%lwOo6lr@wSlsi;hR2o!HR54U#R0C97R3FrE)HKv0
z)OyrD)EU$*R2Ui>8W|cZnlPFYnjxA4S^!!sS~gl0S{K>`+6LMMIyyQfIw!g$x(2!h
zx)*vldIowqdMElM`WE^X1~vvg1|Nn3h7pDfMkq!a#%GKUj7f|g3^*nMCNriOrUs@J
zrXOYkW&vgk<~ZgSCLD_pixo=}O9#sV>owMUtV*mttR<{-Y;5d@*rM3avF)*6VZX=z
zf<1`6hW!hN2!{hl7RLm~8z&y87^e$o0p}bS7ncS139b>Y7j8UmDQ*w$GVT=~F&+<|
zGM*J)2woOm1KuRw5k4k9GrkPI8NNS$8h$PQIQ{_vCIJh9EP(|SPfRaGZ%>~>-$}p6
zK*}J);KcBrp_k$00nG!Y2i^}pJs5icf5`e!_hIP6s)tZUEJhJV8^#pIZpITPdL}id
zK&CP#2s0+L2(uk?I`aVYB?~KyK1&2kBg+p~3RY!Sf7UYAB{p0(X*PGZJho|eGNSs2P_MF+A6I`fVqFgRqd0cbcINUPaVD2*RH6AjaXFQ=i
zjXX!ZOuRf1m*>a1=R#21iJ-qg@lDXgerx0g_(q32!9Zs6Cn}N6p0lX6h#$%DjFi%Ci+WE
zNX$#DM(kLeTij9nv-l4QRtal~0*MXDhmsbOpCnhM7^GfEeUw_2ejsfjohQBigz<^h
zlcFa(GVC%AGUYNyPai$?cv}DTQdUehNVZE3Sx!+dT5e39NM2VyQ+`?Dp@OYKnZk*p
zpkjbxmlCRys#2oToHDJlrE;nAiHfjFkV>yAj;gk5rt11Lj%OavTGWu#RMe8ymeiTm
zUDO*j01ai0WDTe$tERhV>vOc{n$NSIZ)-i)3e+0VCe}9B{;Untd8(73v#86a>#f_Z
zhp%U%SE_fZFQ=cRziPm15M(f7NM&em*kpumq-RuQbYU!SoNByfB4`qBGHc3e>T5b=
zMrGz?)^3h(Zed>g0`-Odi_b4^EHo_gEG}Luzx?p>#8TEW-E!aRiB*c#uC=6flJ$;_
zq)n2|4_iswWZPXkX}dJL1AAHfO#4#@C5K#xOGgdIV#j+YeWxmC4Cfcl%`QYPjxN2f
zbgsUx6K`bG6@jcQGItw?QU9dq4>dgA)%22g`v!%3rE<7$&(
z(?qjMb9ak)OMNR(YjN9ywybv2_Jj_sjyE0mo&KE{U2a_mUv0l`benfC_89a`_df3(
z>r?F;>{saT8IT$19F!bv8xk989u^sH91$96_$K(RepGO@eoSbrVO)5;>AUFnmI;Z8
zj!EgsuTyeU{nN_RBQxqV6SKOr^K&M1tMgX#KNg%8P9b0jeDT#1>QXe60GhT;vz)iW
zzEZU+wA#KVxAtvadwp@ka$|qfbMt2F%{K0K+7A6r=?}giZMzD)-}j97w)Wlj;RoS|
zgohuF*pBLspB#TXF*w;e^*Fshi#?}2Fa9a;v-?8xVj1QPgI`8pQC^k&68SZ7t$)1(
z_qjp8d4J1(+j6IRw{-7(fA41DVeywQfQ)z|+1LQUSutYWULc--~d;01yf?!e9DJ^uOgtMnOeG
z#{gksA>zRV01^lp2^kF;89~FsLZJsCApErboa5!$?HX=rll(w6cr-MrAA2fTde<(3sLZI!4Q#(hzcNqPyl2UbTm{DItmCC
z03i_~qYyoKj4JaSjrbuSLZ;V=pQ`xGy
z>Twx0n;sbe_z*>je+A?qyh2FwZ^KYnzPTECD%u_BQ9$oL%<(EJ}RP!7y~p4X|p
zx+lPUA;RW*hj4Lssj|4Hfty4v_JGH&b#R-2w%3aUjk3&MtTL<=`Xh5Y%e&Y?h*+|H
zrf`Xy~|dE&q~qAq9TDk7o|D*pz2so&s~NQuFuFghuUf%8;~8q3c^
z0IlP95n{yEW9zhQaGzDq;7M`G?{!&mWJmTNUmtlZeN2R~u-z$$BD*o>hy{2S652|B
zfA^PINE>f_rK5e}kq1qRp-sVX<5_P?2~J~9HX}o8C#zMwSAa6>J0R%^%6D4x!k`q|
zg9}>Dvp8rj=B^L5=Lxz0?C$yD`^5F}o0iOnV>S~3Oo#MQS9ixXccg2yvL@?})TM{4
zU#9YeaT2go5~ECBMfYSSk$CKADh=3Wr^6x^cnRpp4mrc>)oErV9`hNmbXoGrDI+N@
z9t~#T2(vSCD0tw4V*jUR`j57wqzAz61KFS5aRONv#xWfN<;#?*)lM1^jG11%vI_K&
zLro!j32=#?lc2YSEOgSr^m{eBgYw+{;(6LqJhEd`d(QB(gE}Kq>E35_1$@
z_5VP6#r7W;!1c6*Dd!0mQjhl9*n^3e$`^6y&#U7V<`X;}PC146Um4(JF7jjvq58d8
z@2HzdlensB8DP5W&p=}<$dL=DfG9J)D2e7X$aw;YWe%~jDoxa-*0>aZbc8U()n^mk
zb*U)DLX9PiPv@v#jq!1sv;@v`E`RO6)iy89k+`Z29#g)JA0&+``7g?#Exg
zZ25ACT-(8MFkos%vw_Wa=$ZrNNoCBiMib%!4JV0}dHmAul@Ncv=GWd|sPh@#A+?q_
z_u>@C>=L=&dxm1hE&KHoQ#V_5$EBd&hCt>Xp&|ED^Q#;bSC0N-D`!UQ_pc;M22Ls~
z$YO*eZR4|$npO5k3PbWlFZ|1mxOk;(pPSwu#h6XBvTvvlS2_K>6PQ}7&%9>I1krY@
zP|AGWF4)NS@=brU*T&)={#jtN!Xx*G#m7@_=_vUT^+TSke1a;0aM*#d@j6)M$2bkeqZ-H7mL0t
zSe^)Yv_Us~q3TQ@d+@NnxwY}?{Tt7Mj{{AW9tEpg8XR-IZM@oRqGa4!MKhJ!8%KQK
zui|riKc3k`3-&)Vx^>!``1F4NC)Yw$lls0>W`5=e<@eob4?!@b_dg$*N&KTxs@>Z9u
zO7lPSX{kI1XNX7mx|*#+&2Z7f>$0*HHJe4EF^E0sJ%RvQWf8!WNe|*Itl)Hb8?G>$$(vU%s!pU3~ipSfzF
z#(&e6yOy6cyQ%zZwD5}z5U=95K^j5=3o%|jF%)Fi_U`FW@%4$b4^(%hLp$W$qFQg0
zo&94yxdiZN=bFm26BO3k>5jEKQi#%E)dKNP_T0okd;VC98bzaJgVir$QgV~@taGKf
z1thoa*{bx!d)wH1+h|L&8Kdhsp&aKg-pP7n#%16=u^wNhM6vSNbbw^D1d-e>=i%}F
zbco))5b@T$e9u(Sk@vZIwEMr({JzsfQ{LvTEClA)>-y
zyxH8ED>%J66K2t(13eAoSKW9_>(JVbM7GccbX*cu?OkeSz8u_6pt+vn<+}QDx?URB
zs+wd1QAO`y^R^yqL!8NmL`>DRcS()iJ{`xJRFC+LPjHxKqZ04Wc8^J$iF!lsC+!ne
zAPz$r@x)0v(08FZ>`+4`mH@{5HCBp~j^eHx(~1+l21cp`{5@eY)0J>#4Ju3{VTck+
z+_3wgt}5yO&ZS15mRgnk1#bqcL3cC`dg}3*li$|IAswBLG-P;sWWMoKKj->9dn!dK
z7s^)=_C||OOvOsTtuX*^M#QyY>}1Ii5`tl%YMbwj3sfZsvEVcWt*NU-Z
zC2Z_tvey4JErWi(56W$S^MiJJu)v!-(#y`
z{pn46!easXTpIX3&diN4!YaicY}rGP>AT>Ca+R|$WXGRZoyvqUgWG80+#6C2KJ@o)04
zL_{p!;W<#UWd1X`BNwz1!KK89!H3aRaDk1nx&`k+v
z##VEOpvC&5Z*YP2p{6Y3b!Dd{0X3dIDAy!sNwK|^{YI=;4nc(p!?HGQ8r;kDMT=D#sEy
zQ^ExIhMC+0#2u9P0J}r(j)!idUZDKvw7-Y0R8CM_P`16AH3Z+`66`ioQVxTl#`gLh
zu0w2^kT7ZhSr_XyR>O{Nv7@I2)cf9E6Qg{Zn^F|Vr|JGt^tFI
z01_1%HcdQWXej?>oG~oer{-Y4`UrBh6BFQ=>h3&
z`{`%KBOjuGomi@rNT*+IJvdK(ij**&{f9Nd8-ojokApR!Q?b7NbM;xx0u#7&jAvPW
zLE65@B`zW1!{M?A1~)_=I!IO`yGa%_7Me=`>ORz-Ul0vnf0@$O8c}mZZX81^+UwI&
zxVi6Y>#}L}AhEf2gH8^oi$Dq34jsH*Z*2s>g~r5k`Lfp8|BR}trxo=Up?}uwm-nbH
zNrBIlARJ2UF;EZs`~&ILNWjDIop0JpPAVfOt`e8_V0qS;QC7ID_W%V|J1_dzpSc7*
zF+)S<3xRiv_)*)}QZYOlN}c^pH)A=+`<3r~L*9y58$UccTmLmDDZ!+p6@qYf^3DAY
z@BDir(92v;bZ-q$JNhl!nmzrQI>|%6@5CRAk)OmgbP!CPCejt8m*YR>{`z6(9*COj
zgNujlN^yqLUsc@$1M7DYGgsTL?rRIh7y26m7s3%CL0*zO7$@I6Jfk_wcGD$&NR-C!
z0ZP{Ws&d#pAUT4Dy5)TPGyTVS7K|5Hk;nFecc$})hiR}!(BN-1QlGaRzgS{kq?oi;MZq9@S;s@qU4|2}@jvJ@cyxoK&tQt&%wWJ+hELw!RDLq@DYPw{r)kg7aE*POqFv&(L!p
zhFtsKSzq%n-_Rcq*xx;ExCgQ`ix;adMDBryw^owxZol3G{I`VTsizUsZ$XU#MIjV)
zr1!u`=eOdj0+`~hE&iSBHIuK|R7l|#f_x8X2`Pq{U+Q+}ZXzVxK@G@RxjT*cWxnS}
z*B{)&3+a?iD?2|w{Ks2{(jv6FvLSa5#4O(fS?xiJ)t!4+cS9kQ<2kV6TfPv7+vo3n
zPVRv~DLRBi_rRGS=M-XRa>;z;*6{0r>ru0G)-@yZ(ZDim%^2Fjm=wVVty-PqK6XaI
z%1QFGMn}dp92}K-i|hg+QH~9%?Bx--;@UMX(v*XB&2!hG){ekrIG%=GAhvP&ti-i<
zvq0PAXSfku-SaRaR-i-bkUvtSQ@+;Gkyhb{1jZg|^k)v&7h+_KO$lJZQz`!Pfpw~z
zmJcfVtOINMm#`r7x4X7^6B}O>7gclT&VKA3t=;ez4|Ffd_J>2Mt~WGq(rOa7mGa|7
zu@YX4awH#D2fYnC{vu|;EoG`3sBRp*nS676CMCYeEw>!#A8@4wy=<8KdGn32&gJU1
z*wB5ps@Sb+vsH~#y)AQcS(IC@ilcwar%4+qX~*S`cw7Nz1~)qM@O
z%dHg3O{&itpv4s(8AUxDxj3R*?OhCwsip}MX?uRj<_%XJHBw_&rn`9^RI&Q$kb6M*
zqV~)ccS@S(6l-inn05a$IM<>VS^&F*&^ug9HbnH3>uDPsnFgAe;ePJ@3ah!GRiK#L
z%3(c3YwWA$GD^!-ztX*+&$E*tS6yqSQxgNvcpS{>uG4+_GoQ7#mVzt3-lm@B6nghN
z!DcfHBO6pV-xi)wng~(qkU{3RpGeI3SuzFb*O+k^on-55eO^=6otYEf8kIf>+^Nf<
zE;;3vuq@u{gL;Qdg^sj$I+@@@ipqLmOT^;KJ^6iZHXU|!4;WP}yV=~RvDb2}EV)PK
z#Np-}Vj5&a(PH`zLc!8T^!uJnp~EE4y0MfJ&p%q84MKimW6G(k`Pk@i2<
zoBwUofL>LqZv|coS-CMs^!Um4QaOJ0pmOQ{Vb|^L?0aB{ty>gXb+TT~Q?qeWz*J~o
zv!nH$ysM)lwkp?{I#M1RbCe`NiI_tDVBRpr)TcM>Dp#^BE&mGJm!r6u`Wp)$r-qE$he`_I2wOdI1#R)q
zY*|vqH8qbFwaSZ82t)DfWmqTFcUG*o(&RcuH6ndDl<3aC_#ya3)$=->Zt!hUum~gA3!R+E=)3!i2b+XWhM>Dw$6g@>`FBGHDnW
z=4`Kvog_oYOJwYOP6G9+dgip+zh7x8z+?Oj(&?zSpeA3~D8rCiji{LUw-03=AlB#R
zXs5zcNW*lGFuJu
z0EekhZ^m-Xp15NG$gq0)P{Bn+80*!C1{{Lg7WP)%iKD!*v5kxB*zQ-4{AF+cV0^yL@D14`;I9dhD^htb_G2
zww8@qhKnV7A=zHCt7ClZTT)M0fa0{*t$t)>Gvg^bd3)T}6yXO&6ym2%Fc4-xpYo2j
zlJTSEvkKY!Ch3S|m&c6&4fC)3!oMBHp%A86IV$L|;VYo_Kpt4j68epJi8
zl>xFX%9a^;v&zz~fFkZYoz&@q>!
zJ!X^Cx|9S0)e$82ZyT^FuVllEX>6W{|4kKEW@l|6q0f(=xL`wM3
zH^fryxVE#|+yc5O{$TEEt`3g-a!2dSy0TZ36!CLM-&vIU#&^~_p3YOk5(UbTSIvP$NlH+NFllyUgodBMvMiU4
z<9@1XM_sVZDkdARhvgenN-H}*RQ?2h;%7aQq!8E)VTo)|JZ6L$!sIK#0X}^}ixo#d
zO@q59jw=LaI@7MF44!1YVVe>cQ<|ry5q+voVG+iq`Ro^;jF#?e6dHmc4s3cEc*gDf
z@369Ne;-HMR)@^^j8A*j+_TA_<{w3GA3?up`E9)j6oLAEW?LHm+Q`d*n4W}x1}%0B
zZO^2?ozhxm7gTunAm$MC1Q#YNYh6`xLDl|5YPnibho5)_ZZ>_S_{
zjtpP-1FG>0e>Yn|N5VqFR}<2OqYB9Q3nB+H%Ls0>OP#g-49G5Xy}E}QN~KSkLRNNa
z44SUEN@7HK)U%oPuK|shFAi&8H)R{sCs`InM#p_VLf{0`(Bug|4o=rB9lLzEr%5-GI+)uiK(mD5H<_Tn0_FzK0n@
zkE4BeYI4{^ccOo8ycGC>PBH+t;xqe7^YJM~dauTVttrZ%B?W5lR~BZLrmCgROh33Q
zl7c^3mMtys`a^bM-BX&5UIuL_!4`^qsySBI_`Rv&V`7J$Zw2C!5Aze6s3T@|C$D^>
z1x;Cotb>lY505sZAB|osI^g7LYbyT~8b6pn$d79+IvJpbg%mhcL3oMXIwDGpN8CR*
z)ftb}HQE<=L7w*JDO0QxkU6C=vGRkkLcRRV7o-m>Jdz)4TNBxWjT*{o)uUqHT}_@-8wdDi+EwT1INt
zP44pvihmr%7{cS@>?V|oEhX>h&r{apKXVV@?Kki(I2f5uH|VEL#!3+PZZXA0eA3gx
z|E8--JGT^Z010cTB(FIPX~Oh-7T~N
zwgs)KktB`@K>gkXmZF<1G?g|=+M6g!R)2Hl4xzBbDR~HO5#mn5?gh4$9wo$>`1Sjb
z+qF4Xo?a5Z$1CkMit8`-TUD%)rn2RM>$75~iBJk^^^t;prZ?q!#2=V$igzx$SCJX)
zpSyGC6#EtKjk}iCJuV;aR%&3NK<>)K4{SItRFM-3yx6EDnifHiYYKbqZKKz|2Gu%N
zZnmaCGxjrLW!y>;M9Gzbhu`MjCcswa1td#g%@$=QW{zSNklkF=lKK{GHl>Kn*Q0HA
zN1{6)jS$->n|IgQZ+C;1#k==!bKJJtmjb08)OSum_RNVC8!}0`w=b|CW$^j^;-FA~
zz(wGBC-{C9`f=Xs)7%&AUg|WBb&U=2lWP?>s-+8y;$1t{4@AxHS_-ap3N2b5EuTVT
zz%pNh&JF@F4~|xJA|f74|8!>6;$My`u&;ck_Bj42Uz2jGn`eaH&#sae%hGl}9y#6c
zCNOWF`?!r5xXSgjxx0KvGQa5_w_7;$kZ0=!X#`^n%Fbb~e_E&&-bVIpKq2hweA|-w
zqB+rkeJ`VNbDe>C%v|r*0#o>52nCLa(P0bwN
z-g6@_REjh#M#P+Hu&L?ZETg61Htg1I7bB$!-kLuqhG9a|+84~pQj#v^KX91z7ASfn
zm-rSxyB6d7;7?P!jjK}~s!1Ktq7Q)mAPK(e#H(gA^&so+mMpWI!0iiXq${xY2*s&!
z(zd|{y~L`g06OI8;5HUUJ70{o*FtV#^uY(Bd-v(oCyAxJ@v7-$!o6J?drNogeAJNC
zCUN6vv%S~1D!wTp+P5=$HNtFG&~a2&C-yYZEu}Th#xK2xJEf-K2=7SW^8vQw?VSxozuKzd{9PqI5?zx9K!2=#~M%4oTM0|
z+B&O{Q9S>o_kGg(2HUa7^EAkhk4Gj_!lZ;YQ}nhGZp(AN>g$&^WlBVp!ZKv_Jtn$R
zd5O<}izNFzCM%)>Tl6W>(gN;Wu7F`Ms22(eFr0?-tQ*&LlAR9t6z@Inei?r^_5}Q_
zM49uO7~kfr(xYpUz=_S#gOZ$Q!nJP@zv-MmGaii$|_kag}ZOoSuY)R?{TJvYl>UKAFJL|!ZlZN-U_(S(`h@TdM1_xym
zv>c`*8yF)yMqRes8{CW?gI=Ek6NYU{hk`-arp`8f+h+3ChStfxDCas!LUea;pN^-|
zR>S1`gWkeo%E6Wc4a#U01u7(MTt?LI-@C;7zJJ{!tVKZtqU=F^p`%m%_B$$Q0@?$A
zaA4bzGcRaHemJ6MjEgoE!6S(6eC}sYohAPw}ojlAuAmJ(L_aZg?$F$_RIQtr1ulTTgf(;Ee>9i4Dw38GePcgqzhv3_`<8+Df
z9KC@Nq)$n40So%L_Di;q2=tTS$|>&n)~elW=ZUwWgAX#-7=sGJ!XX3b=-Hs9kl*I2
zN4yFAIM03H-{vY6VXk=pGFOFvnJbdN%vI#G$z0p?-yM4pJX3z6Xy4E{wF{=<^(>Q1
z^HG4_J>WTfsAYhY*x<8uoZ!GPRv&Tj@nCGD!HjX%yTVrIdn?z(r}``E$Nsp`gKm_h
zb!8qVEyhMIl_ZJYDhRS-qW&pRETI&Kz9?qpx#DaaWk_qW9X*akTcejeX?GN)cy#
zRS>yEdUj;z5cRAC@c{;OaNg$X3`s`@r@Mv-rP}uR7!}ytdcE)zGci_Zeu25zV7ql3
z&Cqj*et=v%b|l$o{{eTQu>D3%_ZWB4p@w;DBD?q4-ILlP65VB9JH5lMTa36#(LKHk
z{uPXSHz-+Pe?8M*37LVW?VBopjC&bp2|+4BY;}Lg^=G2y%jxMD<>?U@K7ftX@mzgS
zbEf+gM359DbZkszmQi#JN~2`ONnqLoEy+c{O4Foa18degK!ZNq_KTaY?jE?`V0u6VU_EDb|%o)BavTjLNXXh@w
zaTjnmH&SZN%LeTxzUv|)ota~GaVU&ST|a_J>qOJ0K=DI#6Bc#E}Ye4_bfPGoX>`84H>sLfr2
z>Z6mbIH~_eD#Aa_8f_wN#W))HY~MaA;wjtFTHlat@ZGuc$hxg@8m9)+%7UX~7n-;9
zaAY-+fV)nUo-`ojm>8nyC&fN>)mgXTwJXTYuR&$SGCa37_myafVX0Zvd9qR3NmrGs
z|5e8-t+?z=)5_DSrpz;HQ@Wt1sM9Q&u9#$tzNdUQ4Zl42&nw6A2V#V(X!u*gk{Y1y
zf#)!qmZglx_Qet$+sRc@flYAr-Gk2B-~+ujbJudPg=eAJev6fdi2}EGDNiO7=UQzF
z-=lt}ex*$V|MgCzlPa7O3I{z?WDO?gQ=$Fgy_M2mjgk8|HU@N(FCBZ=X$XtTJxPbCXY(Tt$B
z4f6|$WfFHjn!uXgfVm&o3kgUkZH~KB4L4idjWDLU3n^~6%=+B};h!RtX3yj9xQlPm
zvs#XQ=q7$fRR)O91R_P)wM^D)`zA_>q8n1)H4W#!FO#+oB<`(lwktF@v2N*@Uc7r7
za!4Ly$H!b2cdHc_Op<;fx)y*N77F*8Y!y5dejd+FbqcX484&AAh8(=hKXQs3vVzF`
zkO2}ygCF}6P59Oci2c%`7q6_)E*>}e9CDhgd-v!fn3~%rF3@qfGMB8QgK{A#2IjNi
zV+(63m(1AbEEx!3QuFQ6RIP~p6pB*}-CHrb75^$BF0(d~`CZk+JA^2>egg5)B*wA6
zWkG(QQP;c2yfpMQ_XTwE;*N91yMt-{$6>wkVkt{o#M!}Pr_^`@Cl|-@?{mo&C-}(I
z4y@@?Dp}6;?L5IfQ7COeBGC{3QrmbAnw@%7;5FP7@w~a_{0th&Uyjd0!-0&
z0xuOX76($|^4(qD5AB-2S5-K7v@V?Fo@2?sDBH-b&4Q1AXZ&nX7xTRI%7MMJEZ!F$
z7+9sLsfOZwtLe))FI&}wKu5oCa14I?Ddpn%
zj+6cc{P;+*XHT+pcJB<{j>9<>HyJW@1|Jc}>k=<=ze-kHZ5?54jZ+og%nsVl*71PW
zDAQOnc`n9#Pb4fH+Xg;eD|EOAPDUHAgzf=Z%Bz)lyAH{x-uFN>J8vKJ^@s7r9SgDI
zm;Q%!()7O+6=smzzIdOF#9!%L_TX&s}K@7hM;O`+>#_h3j{WKeOc;)?>9c
z8V);~d9}COga>lN?SlF-mK9lk?Q)pMTqk%{9zkbBB9#k%5bi?WZa#^^O(#?p_@St!
zYt0a{J-rU|TsTzQ=+R30ndN#{^@Gp%IBxao3}S9VWL|tBy5A-Zju{&yl44$iFI-4b
zHKjZ?dm#}ZstzvCdV*6-^8qxTh&*h(ewz<_X@+Zf(mtO{1IwMe2nz8jyw!gVm!T(V
z6V+RaIdR;(TRGLae0KLpg7Y3o56RFlxf?{B<6OH3M(+W+xZ9$8Ko#+YFa55-DgLhW
zsO+uu(D^+uhIJ{~dAoss);WD98G9RUULPEuau85M_wX1>qXf^GH_fg?ynDCgKO2JZqkW5bv3ia@JXTZ@Yh_nG
zZ@F72yaH@(i`rTxoFveB(m%Hh3Hh#j=`GbIuK2xRI^gkaFh*>Fk3Tri^`jv$wm_y}
z`8~G8TB1)~!|H~<_e=8+W`w0^HlK68D~SwHUeCnI5=`tYt*GqytoqZI4)(e$P-dZu
zbonHL=t-%MO%FaLi)g;YtQ;$)9)09C1`45hrf+zdaJU0=(O9#2n3-D8GdPs2k*pKN
zb1hK1FX-8rIP%eL1$e4d!;^W+zD>HgNwsw~D|@qF?m6;c2L@A)3I?kvaJ@VZmrU>9
z-dtXCKfDc&*}s)$96PO|9$0(yT-9qu^}4g{Ld;{QHN-myFC14qx&w7sT~k^-M}gN^--fXeV_49>5`1Jqsz>LA|tI
zI|2?hWDlup4~>-T%<JZy$?;oKRJ99A??EhH5{&awd$QD4
z?58wFLq-_&57A6rqW#r$f~D9i#T=&0I?v+7S9)$wt^#gi5FfWTt1|xiaC;1=hF5BA
z-$iP8-yJW3AS>v5EyrYcTEaq^iFFBUm2!?>92~st4j+=$ou%Lw|OI$`A
ztJNUK!INXWiS1MST|vHb`ytc9Des}({d!1wFg2|@#u6lE4}KihBu=}UGk%2~?ot;#
zdZ9ZlFjTd?Z79EI`qGoeVWeFYpX0&6xK#ahXPBK=F#YTG?;~Eujl+SH=jK|9J9M07
zGmf5InN8D4m1Qx(oqnIabB>_%Rf5L*6^e0M)I&$iNnR2%NGYlB)2NlULNyr8#j_SV
zK{?7Qe)bK=8>TtSyfmlpyE!NhnaWm!sE)tj4Txi}cc4v)WX1Vrb9?uuk*hol*@;+UiU@z6C)
zggn0KxtH1ZiwBm4+|!;>YU{?q-m<9Fi|GmyWyfvi@&$9dx3dk#7WU9NX-6sk9zp+X
zx)8qteU8eigfE+Tr$H%T{w@>VB8Q+X_lJArbL4j@r?`5Y{jx2N6kRYI$|&9$vh!6DDIL%yMe
zw0&vP+>^cyWg4et({{drG{=c)IRB@<)lT)e;{{5ezPh%ueUUG|cU7#>w|Kg>0j8eH
zwJH1OJ%@;O6x6ty4(l+Khb1{yN+O%Re0Irb?BU(74HxL(4ji&^T2&3x==#~gSJSR`
zBkqvx6EgoR#9qSu`I;-$)LfsB!*$h>!`Q5d?X%V8Ns)-Rb*|*Abt5~CVSPd4Uo0sR
z`&#FByoCdKbJkPqzh3Sd9{Sqk`M0x3EQlvHxCVE(=Got9_ki1cATEWTj+Oxg&2>|U
zOLC(iN<*7>VfBW7Z%t?jeDG2)mXQwv1%1h?%vth&DZ3Apg_!4^ZI5T3CfM&3eftR!
z$TAW7wKKu{3r3B$Qsc7KlDhiciT5XY2l(W>d~Yqe&+glAeFit-jtdEI0=U!6wu$18
z?|~opz~%ET)!z0`1M~sR9Cz}!ah{9gUKCBdGHn}XV;9(Et=Opc$6oNf3Z_S|(~^V%
z7hRg+#CIT3!FAtzpn60$+^B*iE7s3FsBkYg+uz?Wr!g!*^TJOrS1&1fO>inWF36_p
zXm?=+AN&cTBf;5w50IYl+Dh!KZ4q{CBGa7sc~&7WbmH}_w`>(pn?YRT#UyiS=d4c-
zsmH@BCpUAKeQJ&<4$>3KMLgf!@va_uMONjm+YkvyPCbi2M^0ljGdEgia>C^pvM)Ekioe&ahw&l7|=-SDZ(x(&N$wIl6#KLvIdM^AK{cN*lh
zGLa6YSlP=Ujug*@@%BUb9cZ&0@rwp!F?hInJH#dyX{kLQGZsVL=VoiC18CR$4~&2E
zL2fbw@-G)dl6k`*2?Uc?3+3_0IaS{owa?$un|(TFU$N)^Se?u2|kz)m^&=&fU7;_wg}4pnffG`dzSH
zBc_urqnwv9l>TiKVdZ32yuws+=5-a~WRda)g@@HNw6CCaF7Z5P!gs8=hp(H0BeO&I
zS4KMerOc``@mT^1tg(`smf5vf6`S2vZ-H!Hshg$g{pyi5aH!R^m7JvGpUG=k_g>gR
zLM6b$CE2_0wE4Pzv+SCNs>(r7v|;Ug|CS2WTyT`PY`=5QX7+ZWwfb~9>+x#C>t0z)
z=#T4tD8+nO++vPA`le!`eD;h|J0eztNr6?HqG3mG^Rplp6x^M$bwYG
z{4wQ7HzaPbEmSFOd-zoQMbq0T;BW}i6^b#mEg32Jt2545K?7|@QUY2zbzy#9dvwFu
z`)NJXI}>5GXPyvY8?Ww2I->qyOmW$?mq791BSl7CuAsVO>z?c~w_`B+!AQ$JfNiGs
zc@(!8&q%&_ltyb8V$`pvsF;|7KidBgg|??}@GB*mwMGg(oIdEZ^AdNavuNQS*#CWm
zBJDOSka-!u)@Asja#Y-FO;T2=ae_BLE5hi^w~z&5ZOMgU3GP;W=mKoeJw|2#mV1@5&7f8{BOC8ZFduuG{?9D
z?a8}A8DR8o|M)U;0@Iuy6i6firNkOZc1i(M8(Mj*6$KBLjEnK;lHG;
zy?i(&Zavl7YX5ks*SU$C}$0taJq{4?bEHx+Fp
zc%8d$e_d&WmS;d#BH#vJCf9j9`WR-31cxrWqf1$5r0c>O@mLPTTdkp&Xl6?Wj)B^x
zUwdX7VUt&$2IYw(T(u8qdVJl)(6q4XUcW|_H++{6fpc_ZB)>jj2*eUl3=2c;Iqy!sD=ISL3b!yly=@*&;^jms>4dOy-fFO*gkmshy{`)rKHlnKO$bOm&EL;_3E2}e
z(yeTI=&kRl8p#~5uPe$xo2N8{o-Zur&6`Jf=Gc7c*Y4SJ>jVf9v5bjnB
z!)?AbuINg{P7s;m^G-f{3;vw`$|Ib7roNgxNHEP
zz%m;x0jlGQ#Y!>B!lb9(D@^#V5lDyFo9zy0IxV(0k^1Z%K&^D>HS6?}^Z$_c-BC?-
z?VeFoY=EMObQKhprgWr4LPy>~(s6r>9X2nbOD=^)Z;P#gsgB#_I~#B{MyM(PnwtWP-a?857ElCs(H0nES}#gVsA2W
z5M$UlW!DLIL8$C)n_>7r^K6}r&H|&(*;(FULF;b)%-~9bGCm%bf&x5aA}-j
zTS;b6hore@Hc3MFB)DcT?p*Hde$S|I&Lr@(u3mIyM&phjmrXET`8TI-eD7=(0Rm}I
zzG@Z1;htv}h*&YTNqqvF8zDba`rtpYI5=}_bB5m8Jbj?;NZdQ;>ceq!>_$@{d9GIv2?Mz!rY(Z;?eN
zwD}(yG%)9mtla|kEWBqkdj66Fx~WhwDh|!?p8L~@yAdJPgk#VRu{@RX%E5-CNUi_TF4bIA&d?#gR;{)2N;%_%C@hV_7Y5u0;e6WsXe5VH2j9TFR
zw>^d)MgJI%i&B+-(&8~4W6{T;>{{@Ow4se5LcFmLetmudo{d^Pq)r%_@HqrX#Jf?%
zHyPVt-`0CB%x|sQq;h6`*FG1wd25tEAmz=9LEVc`wbH-F{RfaT=Zv8%&g@Mp?T_*%
z^9pei%>(G=@lxTj-Y((Ifq^K&Lg0^X?w?=E@{Xr`^9ydL8iwxslcv!x`#f*J@6PPoa~*BwTE15KjK`o>)22($`56mOn%KF)WURlQ
zEQzjMeTV++*yM|u(!+n(bUo5K200lw*!Dl`U}-JOzw!B`T=+uq65FXmKHbpacpkz=
zOe1qGeztM-A{07MZdqDWy>&&ELG{NiR;!DfpH2AM^w^6S^?N>ucSHJyXbCL`7)k^V
zGbVft+MBqlg)nM-=YX!;8uIV6JV&5JZdMIIBYn^9qNUsgHS#hwFL!RG+4`Rn(hzv-c@xqj?__-b)Ng^^cm*^=}Ud
zu_r-Z_v5uS1a<@vriIpJYS_udM0`_#L(HXj~>13o-=
z(pd5P`}gkz-~^Bw$J>tX2{nt`86}Md-tcO^z|STST2yFF$Y(tU%{NCLg93|Lb`1>O
zZ?)LB-!HV@aG%D|x4(E&*ZEdHj(;yXM9#+?^!wFPkEQ^H@G+3f1gwkaN}Zd-2a$%0p~H=m<9i&UlBH*`ZF-}NUS^A^
z)h@I5dO8Aku7AGUdT~_7Na}8cClPf!ch&*DBmb&JQUr*#Dz>eoYq?;Jlu#|8%zP6FyuDH6iB+-dm?$4B-Yx@(EJjl4+
zVW_K^Ko3pWU>9O94Exob>Dq|juintq)`NJr1-(05ucE3JS+6#1(b+D3atBw+yxZOE
zmHz_i3Zixnir#1(56C_;nR0*Y%)VlfOiegH)y{}|0&~*6cxeCUi(`Rb!hG>l0;8c;
zT8Ks$IiqAGLYUui@ASv7f+PN~Lf!yHAriA_qJf4NC*J6C>cgrum0+DL~Z{mR}@e+Orz{_B+fW#kgdxylE$6
zPQOS;Y^92$?Ue(^#X^C59$^sv;tHh+J>ABZo|ZAGvg^cK!3tVx?%h<|Jc}-o^w-wT
z#R2`Kfdq&_yzNkBA|XmS9Wr~ft;cy5%s_fphc5K!d->}a)UlZd!z8pX(Y3BahHzm|
z-W<)6WsX5j`P}eFki%SH0fk_Qs#H69Gdj!}z#`8Taok{ajj{Etla$+$Fm`ii1z4gKtaEUVEHs2fD1%k>5I9-E!np
z=ox>ST`YeNeNgF)_Mm+F8j75jlNo0prFsmyH1{N4#^Ra9Qte}HQ2&G0haXP}!!!yl
zE>V>HJomlbXtaFK)bg%5=b3A@1skU5F*JJU_TIO?hkuJM&Ku~yH{eb&GlbnX`4+s3rLk$27nA$$x`uz1}ss2a&Zc1g6H
z?egU8Da4#H^u2hZBc0>-YMA4b9w>#yVg*jC!2`Aitp72Hy8P^#7ty`>y(ipXX=JnU
z!BU)xdh>*<$k&Fzs>0mVx!mIV$Bz7QzZjun8-Pq~ooo)j(!%vWy-_+wswN~}ofeEf
zu-^qJ*THBOCk#o1{}`l=%eyq_OL>NjM>=4rcL8*EX5^idC=p5Ao2lJF0t;I)`fD;y
zY?g^?Dr_w^r-tBvm;$2*+rEghN4hYQvWDU5($Lk~aVs7e88|?%?$Nz);G>5?UD^pp
zg59`vXN?2OMAD4L^Ih2^s?AETsQe_wD7(6!%D$o=Gb}m4{-LYvrKq}LySVPpmR|bE
zLF7BixTE)^YivXhO1!wo<(#%>0KHhIpeBR;~YsSt=p;8
zzOM@XFzypWQ&Ney6|dRQh^^$ds)=cfLRglYWUi
zWM@O~gMD}QSlPiOrOHjPfpD{og-xRvO*ds{U=>+QdDjd1!F~B{_ldW?K=;{4ehN^L
zwApG0sWBfP=g`~rS}a}Ys;94`eq6Rsi(HaGh>2b5+sOEJe<(F;F=xz*K&k4kvWb%1
zU1G9zHM()~tSY|o-ZO&Vjy#B5ozrw1|!S+dv8xlMv6%-k$9O>;dJ;
zk2F^bzArcN9Cl|qb!y=p(hW3huKIHqN?cS<}g0@(;U;G
zO5*fyP)Gck?4K(`xKS$GF~}e!zb3F;R9Gnr*)g>%7g;^vf1rhmWOI%YjN+Lr^|Epg
zNPSXq3?jXCUZ7g*?v{qSv}kL$^et^G^|d{q@u|k}kB1b+Fz75g@Ta0*VI
z+23y0;8Og_H3JF~e>?&VY`H}mH!=AOJ`fLgN^&Qh6eC9CC?GRh1e9&*7=%kd2Bp*UCk|weLiv`VU$Z-dHN$^t%SOiCUaXqT
zo}YJk{kANGflFPQ$WmQTn0w@7t-H{2r-|RFTAOQKGG!{%|2cwfYabhApRYgb>TNug
z<}d)X+A3ANJ2DwA3KBE)pQ)VBYFtV#*7lB7tP1^7`^rwN{8Rt5Vr`Mf#2*J`Q`9bwH?+Q`l(A?b4q*#-ouG5GgK
zTrGR=j4cw(EE*D6hk!1LYuqi~fPa037S=L0`^72rdM-5=sg{sJe&Y6MpYJ7dLISdX
z;}{ewU%t&VM8!*Ohql2Bk$>SuN(sQnw!25|+qHAWq*b}To}_5thqM|}=X--67CP~A
zC6c&uZQ`Z74E4FBS#CW;jIiM-{#X+$S%(`q5x2CdJ`LD7W^pZWzH9Aauf(L5nRePl
zN9`d!|L(eJS>89C^ifL!OfiL!8}?H4hF1bt{JYDeNQI8HJLgJ1XKDD1kPjcWN%AU0
zj}au+CZmGc-4*92m-NEG>>qX=tvV)>tVzyu^YlO`rchaM&V
zl0P~IHI!WiT2A1+lNFi+tdaY%;ND}9em7Zy%n>?q#HC-4VQOnbf5^XVW_mEyK{*^5Ayo%07#Fpv1(wUQ}NQriG{)?T1ga)@yde8(BIy|vI
z#BjPRAX@^}pPUqR-{Z&#G$*tI&J?ctyVpTJpc7EW;w>oW4qVEI?FF
zeIuK?U)##Mb9b>L9>#HcCt({4`rcy1zHX}Q)QsvKUpQpm054>(N)vnahxbc
zmGFZI**A4{5!-|Id-u51NC~Yyf*YmBprf~8_X8wEQAu!LsiR81*V{c^(t@M;2PpV=
z8OVv!1Z%}vmud$QJ;<03LvjB*S5lcxLdX5Ca+z248*dbgagyhGFD!U=r}Z%nZMZ^l
z8+fH)v&N1P%NOn|shA2uXujZ(WJsW#ro`*4(m!m&e$j}0+s7ay
zFEL;x64N@6%fP3$6Ini}pudmD42ex3d>1kGSo%x)2fPv%NrV&wt85cIP+Gi;e02;O
zv!@d2pAn}C3B+N#>QOhAE_n>{dqh0c-k|#yZxh^!a$umGtv%8~FOtQE-{-*5F?z3%e
zSu|s|BV*v)X|aINbpW{K0PGqJcxp3&5nhr#KxcnriEld>IRoV*upgDdSM0|Q_>TPe
zmLYDQc9FiQRe#%&0~Wn7l$O1Em2~XwTjE95c5gp>bDSMPyjEBGjx?B9NQ(^+YZ%oK
zl7jR8A=@3`5nD-mn=GP=;=pRZ~hLpSehBJm8QZCX~kOG1c6cG
zgP@&hq8CsfHfxVTchJY67sbI>=tam>Ragno(88TWD0}dhl|A)I#K_cUn(jJ+&?^R^vfr2+!
zas{v+Pbcp14zFa_cHE8%G%%)8%w&xdWHFWmmT337@(*r*Ggh{ZY}CB^0JP_2vf^ds
zR=xy1=WUD7Mwt(|+XejuWKWP-Az$v=Za|f+bvSD-QU*t_v2UIZR-cr;Qmw_M6NaHA8$=%KTasq5UAeT9NpcGyw;Hf%lV
zkT2bG_s$qa&Y7r2A};-@N!nUv_H4LSI#IphsQ19q3vx>`CK>-ju#5g+YJvjGjrkp8
z)TrY=H|D&Re}Is~yZz9V){;p!Lo!0M=ql(x$Dld*RrsfG$DoYj;1iS@_Z8?jsLL@Z
zpo0t^A4TIEpH9?uRlDju39dhr()Npby7Z^_PxYLgUu^45^;X%9hLF4vWcB=Qz_TOq
z2lG4OT?ZQ>4B{q^WQhhUhONi+f-xho4}xvgNvqmNRv>csEPgYWeBizJlNK
zw()-FB6#awmncSun`igqR&z2Yr=iYQ-M&
zzu9U3SkBtl+k0guEy1TRmSh+tqEtw
zv_|(tZ%&?0X}{*#!_(xgls)<2dr2(*ocl#@>lMNL$2+``;Z1t;ioDro{coy7C;E#=
zEer0*?OTCXzSB^#*_^t_-@E1RKRbkjRZe~3nOR}KzL)w$7I6%+6KaCcSF^IFuM8w{xM>fwpGxlW7ge9D
zH@h0IiPw*jVpb%kw}yBZ|8{5EzwDQlhnv!!Ef&@g=`7|hE4ZF?RpR3bfh7*XQf0wm
zc#Q&re0kx0&u?r=;xQ;m$H-Q9`$^GD3Y`23e#MRT7!;2~u2%jN9P$_l4q;e_PB#|L
zbQJql=P0a+mX+>m7UT%_kF-xk&YVN!Jr|`m~A->yu^T!}g=_hJQ$Do;L;^kvd{i;25n20>YrE2k#
z_b^?}L{>i;&#&aOIt*#08Y|g&a(0+O@GC%MNz}QQ{p6t=wNOd4`EsKyH!m
zP48d#BcUwgW+G3S9%Mw6N(*HD-YOwXeW_(dJXP&c%3V}5dGkPHUnkNnn0@notKt_w
z8Q*q|ycrC-t1W}Cqdw5x-3HCyeL!hkwiBK{TP`m4Nj$n%Rhx%m99-n2>ohPQ^NuyR
zShKqJ7dF$Ye`R9Ch$}d6(38VZ{qx1suOb->LzIR5`Mghf$6U2l+ipVpco#T82eS6nFb&
z?qsdI>Epbyy+3a%{ZXaA0Kg|cSc~KcocuiaVYb)?kXB(t2HwJcFn1{T^Jxe=&d!^9Bx0rhu$T&>TOAghmp
z0A}RPF^K%Su^qk{I5V{?PoxI{0Y|-cG6X4N9!fN;1lM$Wn*BV9shp+jYD4Bauw}fr
z3U9u8|HnM(#Pn+ZQ|n>YYxlVA+nWoVX;uF3wZ~<@cXkdLV3~q?u`3s@`h-cdaDv{>
ze7<-C*Od(f0hPdtD|=Xc8Vg2#R(hVv;-Yq1AWKzEQXn1C0N6~NCQvP8%MUla>DIOY
zI(EYy*%xbmYhJpxeBB1X+8edgtMsi9YwA`qwRaLA5IzQQSyU}zsyUZ*n_pFEpyo?$lf)V3oUXJY
zs6NF*W2KOECCCV!){Jd5*zB5Vi9_+(Q6GNWklu01(Vl?cH`BR4Tas9*dgDQFrn1d_
zO&i-`F9AD~vx^AY=Q0u>CEL%J#!LaonIho~@iGpImi#&i!GxQ$`lI+}tET$34#wrPCG_}5zrIDz(R
z@$LW%pe;-HN!U?$(lz=&fKeNmJNC6PzWH4tP-*)o-YnCfMgNOSiu^Z@#^;12@d(nr
zRsa~ykMc}eku0wYvPsT;1>gEH7|Xw2{@Nc(X7@K3b`2gU=iB_87B~GvbQbvW`;K|n
zkNn6M%^wLab@!V-vN70L6$SBEnwo9EZCWDR8LEp-C<qPr6();_2X9=KJWQ4_FqanJP6OKom}Y`8<(dsO$pVkqibA_SuN_x5<7d!
zPdWC?TEPB^D55KRJ|O*O)YmcDUz&=BrXurF-=Fw`%3$oAA?rh__PkVlZ%&5NjMAvq
z&=n6x8^1L1Q)Bot*@&Z?`#eA3C+85PPb!`VnmYh8?>0YM;x3PYWj=ylTbJ%D@e~MB
zLzF>&dOu3ZJdmLbM&x|7LgJwkO^6Cih`i
zv1yx^_ol-1E*K&~*P8$j{{3eG#r6+&pkpMsle9IY^txWa
z(ve-7Z;&L@+9PoQDjnD%4;+IUL;I0ErU}$t11ml`r*r2S{x|N3V+Y>MPWnu6Kin%n
z2Ibx+kTqJt?5rNB{xdtamE;C7m#{PW1`~rjrDqQAsxkpv(!RtBhqAm;!Wd
z+rHN!M*DZL7n=R+At0Rj{hDp3YjcWK0mH5MrTp!
z3O?=4o`tnxmTL)~P$HJ0k!#)a_@KN!xf`U>dbL9B12JrP=4sI?p{KnV5CCb7_2U
ze_3?*g~q9i3*LY4y^exX?$b8mZ6Q*CdRzS%1QXUB;;Hf*>QZd+M0Yu$zoAWaO3=zo
zmDQyUzj?dUoZjRt1K=?OzLuZfHpum#fbct5xtrPQd1?^I{ob~J{eO80IT&lU{tF!z
zg{G2kv_uhk-BCD0o4SkelN(3ChcOIwEX(q@#;C-*xxK!}`|V7!w_Dfl5~pjG4baVW
zE^XzUorI0ani(vK?pS*hg?@um#-~3eTP~xMf0|Jwdhi;aky4J0$E~hd^P;@D4@zmxcs-
ztRcYpzzFBk4=7P4pHm-oEDB`X91Ix>yC+n-AK?5vk3lQ|E~{!W{+(kK}bl0E##kzi(o^+L0vXj=mS2teV%8<6m
z4^Xwcam(g|opa|LjQo;JDdkmK%JL7}<1t0QN!?zvT=!?^Ix~X(dReL!1d@sT3*vrL
zxZq`_xIiD{?P!$2uV@bh|3I|`7KRADoSvUqkLc4JdGVJ$T1Hj8yT13X4vV0t9fP%M
zw31$>M1}wB{w{&!BrRp@UItHpPl8TmOYfedyZ%!1NdPDgp1v72;Octw%Z-ZzH{IE$
zINAd`gLt!%TKK8xwI;zk&ok=eOf~d$2!!&2ErsB=>d@O~RkUt%x;DDPvIO_C8fP%P
zTM2AijFcN5&;}s}r9I+ryw7)8&`^PD{o~&YdW0itk9z)cAEf3a3G;&p3=EAi`_i
zFCwYd$Dq2S)A-ux+nT~~Y91|appED2R@SQR04g*sXKuaW`mg$N6uNXN-kq6;65sy1
z@k0XtKtJ1{q7p(A`^fPCK1S;3;*m2HScfM0E_VqI;UwjjN{_hz=E>_ex|)^C3?aS^
z@Ra*!;}hcKNERtgfTj=$0p@7&)YYxILKrRd#7T-7TPW5
z4wKTIe9_EUyo%BFBa&G4E7SGIwmswc6
zd|pzU;UfxSWRy`-fKBvD(Yk`bA`#yU9V5az!UZ{3G)IPmw
zmwu>N?jc2`N}yOSI5wngX59??tNvs1@1yOT=@NX{E4)n`8xJXtSkrS&oc&IE2`oKy
z^H9jZ(2T{IUqaHUZuhpErdEem+<1F}C%53v8F=$6g9L-ITgpUSD2TPty)1
zzMZjrJ2S<(jw-(}a~sHEeV}=-D$bKzzq3_iF=0fHhk2x%dnpxkKy(QaWQoA#Bx@H*
zXiX(KHyqWQ^XcM|QMOZ#%NGZi&(!ZY?!m+V&RG0LZW?nOgF1zgyLRQk8<-D_b3^2=
zqijeGqNTaVn!(j
z1?NTUw3D%`8pEZjIJ4K74M!|sp`6#)o^XXzgN_ic$Wg?O7%pW07aS*aX!0%iKjqOKviv)b
zPU}DN=*0a)<2X0RstEjM4mQ^sBH|MT@;B38r)uGpqDjfep!3z_or~YfBTTWn)G_~~
z^7f%=d_`(}0CgOBCnYp?<>(-MLnkK-gI``yB640uVW~2I9H$=T&fo0p&P~$ztu(K8
zD&ZNec4keHF>EO?pr9UHQ7blrNIIBHt?lpIVN!n?ge0wPM&AHKmXa#`kzP9YFDNM0mHwqPwI031yxZHs#
z*dI05LhC=oI4tB~>t%M=Y2$Gxd@(WA8(irn<+7UZ7wT#gzo*_45j9L4U;GpfzIB$L
zjWkByX#Xb2VIAg-OQSEXFX1D?9G}Svj2%tUM+W3uFfJL*w3?K>E&q!amX^fIxdCIXNes8>8qp`f?2FDhM4u5WZq6#-Ss4PAmUl0wE_v;@?XS@fRI-+tZ#|ocR8}
zpHu36|Bkv6E=}j0veIh1EFU&`9SX&3XPboR+Gu{bgPAWrLP!wo9G15B8Rs&fk>KHO
z`Tqo0&k**Bb;z`%%Cu<=+6ykTtJb~APbZ`H4?U3s*mC!2u{kd|co&jPNFxHOdEkkO
z!8$-7VT%X=Xk&oBP%+dKi$L1Z14BzF<;=7p_U*1+h}Ay8dj=^&54$|+uCCBL3L`xe
zyb9lKKM-?)hemFTO+wg7`IL$&c&|Tnrxcpz3=iy~4()-JM?;A$6`}7$xajzoK&^QV
z$e7r&Sx{XBwRE_>SVh(t#+$YRw2dNW@PR&PX|Ojv6`2(p`>*B6NG;!koHqW;xSl_3
zdFSmO^QuE?ig>=MW8>D;pD*nwu)bq@zqYSysW(7_0a
z!ydfY;oI%FF$5hJoB?+~GL7X6f{&Er0_B=xcG}W5osaTqOgSi>i++Lgd=-Utv(GEz
z%DA*M_MeUO-sEX#Rc-Xqh9}o^aA*jP$G_s-r|M{aCnvP=7IUsTd#31X1%jLPEdtB$
z1T-+I6}M8A9aj#ja+@+j;o;RQ-p2n0@-BZ*7Ni-%g%irjl{7h8K`n8DPRW4mcwwk3
zD-EzFAU+dV8b`TSc>hfay8qGBwd}4xq;fwkvFE>h?Hd;cn%DnG5sYbdU
zZsGW6NBSUl6?F_EWJ9~B)1bto#vuKB1_=h=x
zeh&^zfhZ^NWoA%W;b^|gct05tm({;HN7b}+vpF$Z7YTCsjPJ0hZJGsL2^Xg^HCNJ>
z*r$P$GL`aLfLz~sCkjv+eK`R|FBET#(F4FCf_&D(Pgl$^P
zo@@(J4E*GO#fJZvZy46?ryGG|!D_yQAF$b}BYa?e_
z7oAq83l}Z9gKG|-yAA+k)k0Iq=N*)=;_
zj(@~v{aj;Z!opy3@K+P1M20)*e;#pHH2)p_lGuDTRcjz=M{|5rppu(ir4lRS#FImk
zac=&)fzI=c>(wgpJiUaT`~3TCdaL>{tNPQ&1z(e%6s!bIX!2PGf7)-y&o-6Y`oTW>
z6&J1}+Z?Kh-3{64#o~A@7yM$UZLD1TfIuM0n%8x95dDtxm#s;Hr~S-nG{tmkc|z7Z
zwhkFtBAA}*Dt+%;UG!D(fvly0_+8%jD3u+_Z+2Y`6OBc*Op7F-xeXz;1$%v8bTM6#
z+Zza7I9V_>jOXqo6kiRL1=Vm2;}51fEy_IeIsIx2$P~Ys?wHIy#
zGzqVsR4ibZlhyqKT)4gbMdF2NaOiyZ?S=S^?eWZG0bYUHhui`Ss>hm8D$m=uDqtT8z^+2r#~O@;+uZpGX+|Bs?T<
zMVM`zn%SQoBYV!2JTfJ5KTwy9=c)Ou6|j
zRRAx3+gD$H4(L!O8p;px6M&5H(*3-!H*>U+;C9P>8$$Ur0*|X*)2MCSF4}95J0R^E
zYl>KXM^f@V2C-G55a_T#yU>=ctlQ
zx8ifpJoXGb+kO9<(=n(D(NYd_;ywrV8|xL65y?qSg|$fG?#13{q7e+KHJ7t|F*7E
z$tRfj=DB%)2vL7}dEeHeK&mFcBZE2HhhT@=iUNx`#q(SK>y8v_I^ME)xFW
zl#r>o+gOH!P05=hB5z2TZE5NhMIju_
zPs||BX_S+*;O@34A)6&>=15(#XB8y_N4<5%>o_?)1%q8;vR1BLO`v*YIiKlu3
zl+;C8&Dc5wWJF5spt@qTTqM8ha$Wakpro7dfZB1C3q^8I`~EE$_xHP)=`A5fAgw2Ny%j#5K-Fvt4CXH{|{Y>*8x*4=(Eqo+j%v+M@I4*CXSgI*+vLzdN}^
zOI2<4R=0x|o(ViQ83`Gw*);Q$FcGg-<7kIl@4w>w(R?TGE2)daA;V`(GA>_W*j!j$
z`>CgffbF07H}#1X{O4&;Z)1oNgvR;j&ppAJ!NiLZ5(t}I3WTD1MVXRy8V
zv8eNnmD;nMM?45_Ai1Ui@N)X1++Kn6FN^v0P`LKy^j#VoM=S4}69CNW
zM0tsf_CXFR${*zc%ENa(vG;P(J+)9z+NVj;-|Z1BdY{Qsl|4nNg=ICzB2Riq
zKJu8l6}3Dj>viF+6m1V5^VT=WyZJ?m#b}E{0qX<5B1V{TtN1p=Wp;!m?P0mwKxFMyY6+-Lb;do*Zn~|VmOfc%03N##ceywc*c{DbD+p!
z$Y#>de`<2n^lmo&EXPZchVE8@f!d)jZ<7L7fLvK?SlnOu&vNeA*!a7Qe1`&(PbF3v
zgT$5|qA)ufz&$_IwO;h`Z@{noTiDXI(J_nKHf8Q6Je9oD6~OK7dolTSbZ*nrAA5V3$vy@MLsyv
z5BaBZj{Lmh@%vcEQjadSzCvNI6CZ)H6p-*u=#3}t(dfsTkbEM;bQuy&tZs(?ioeyM
z*|Y1I*4i8|9u??%i!tI=ZdPMrnox3et*Ewyp|Eo==qp$yYnq~|Jrj)FVsX4Om}Z=j
zoxsxBSIdVcF`OP@7~2`?RZ6ijIH<=IS9zl%g9eAI*Yf-;e~!;StvtPdH@Zi){vKVZ
z$^8Z}7w`oY-{#pTEBnPaX-+3x^|OL|j1B!U?XHUuM=vokzb0qj>rTz7;M
zq6Sjc*@PWrz=sWDm3h;x!Iz+FXQ#KdSGWLyNNmi-)%6Ga0OI|=k>!TpMC+-KAZ*;-
zB>0l&IO|X$pc|AS+5ZE7rX7J_{Ii|+f$hA4*N<(yPs~TV^{Gx}4;j}?NIJvBqIRo4
zpDQ6+8!~-lUh7T*UlWuSIAvvoFNZMi&r_6wxi%-J!~K$5=X5%ak3LU|xSTSrUM~3<
zpcf?*__{W(Hc*WjA5sLYp3;`SFc$9?PEfpQY5_|hr4&GP0aKT-FUw3O)u-+PpgS#w
z=ZF^RUc~UZjNMzx?jL?ICz7R9E5IEYSW1!Ba7wRd*G?`xk(1R#uCUALb}$LPFoyd6#a*)qBJxvGVCyj+VMRg6|Qn*X!%_|(tm?O0j!
z0^iVQWj4Ydhk8~yrSkqjP+_zG2kgDZXFtb%ML8y=-nf;yOLABRYpFqgA4En?e5&*n
z;ktv>lwOf5{4NZyC|d#
zlzp9NAg|=&ze_Z9osX$5a6fo0hh5mb{uVF!P~nN3+l8MJXmQ#yiQ~onRC&EG14`rliyH#iGTzvqSg4HCJUwkytNjTjT^}O*04E
zoU%D{OV`Lgs|=RfJSUrDjLjZdoCxSst#2_a)vY2(AI-Sr)ZHJPW!3ceeRAlUxh3%)
z_4S0v1z8sSKJnbw{Q%-=*HX`vUewmmCR3b78seq@dzm?3s4ymt%%c!fxHs&gOB}Vqlk)qJk-~mzj
zfhQAYI#a@{8TMA!-HC5yn#7s%LcY2Q{3`j*0Z{hUw3JEHJp%H|oS$UDwGB5^Lc&D@;+6ogv3S?S++R=?Op3#H1-gdKD;0)
z^)%mJbe#1nV=(Tty!@tZg+iR*MFCx(
zk#pz}0{9<2R6`$!0tPM@Mp}Y=ZXDBfh=(K$tkFu7fiv0CoUs|S@J=`#jbClBAOOZI
zWbaUah!`+QlQ4nH4SG~jNuzOg>qw!#@blDp3|*es-xYJ&-Egt=;o%
zPblKm^lBK3Vjb8JpC7OCM~XE9J~LQ)Bn1e&{?RzCD&}lO9b}(Y04s>LRIea`!B0@b
zFAF_X={El(A-;3DQROH(4lNo3U~Xl-Gg-QLK&kH=BR+MBxTdjF(8B<8Hu>!#aF@qU@a=
zIG$C-UXaOjym8HisVmh(RY7pZRq*ZKBRiMOFF|hK(^3GTu8b;G6D6V()3#PuF`AYH
zNWAinxHsl&tI>khqH@(Wi-iN7abF*khs~`ocuLgRCc^vaRG&4eZao#?k2_Ex)sI6*X#Za!%c_dJ*7TMn2_{lHkyGI)P2DeDLWt%D!Y$}*%KPD*KTgQK+?2Q(B$qfx~jx0-FE4WXZMn9j{q$pD){y{
zlA_A^K(9D#%z)ma|Mw8aSk*AZ()v^RPMQ&dFC6k{AHKB7i`T57-&$<8+gA4xYRrt|
z$F5i%rbc_+Ow|KjZ?zQwouO`nraFK?it8aP9bOB)8(ePE)WR*y8huu&vUArY1lm@y
z(plzs&$zSEy-V`5E&`Ik!!XMTgc
z_R1?r2s_Dnd8wkVWS0!{GJQ)+XHw;IlNgE#@=keum1#sLzfEU|uB+OHP9{sCO(!jf
z11djY^e3&NI+C^71Oy88QnyZO`1PEQl4UH4Yu|RH^0z5|ikWm0Y;Z0n;$vaZ^0e`i
z^&xENO#*jn+9~PZ`%r*0&>2S(b*^LGHX9-;U9t_YaRPAfvZFO4qQ^#Xt8B<=YP5;f
z1ek6kpc(X4naYGUx(K)XF^H9lAj|F4pJ1qw&xZNPBMzsXmlPuatOgEf!wq0M)&CKY
zHTZEUX1)txe0ZEIL?3=O+u)QOQOrE(#qfmNz*lBsHG%7<8_KJ%FHLAmLcJCMOC@NJ(q<<^Oi}9
zN+fT#DU)!S9#+dIErS9z{0j`qiWyta>m@|&zDr%-F0Y{?rW+|*ENF>?)~#|nKVzz$ijEvBmC>lKmNO^S?^_h%I0^=2Fo_<_+rcGTu&$NuwR>s+ZPd4KkrZBf79X?
zZE&SKNe-4s4(E&LS(nnT=s|@hY#bq$4=Qs2r+s#Vkbb_gY?1Q>3~EXpAfX_>gZ=Ur
zEef_job(;j8o~~KS*({O&7zY>jr2ci(Df3tZT{|*myyJ5eRv;&^Q=>6G&6x}gQV*l
zg$#zHMC>CVYddC1+kP2GNLthaFR?k1e3H;*y4fVyRIOK7fE(?)YIJyt@0=XyuZ_3x
zd^m?(0t?x{4TgYg4#+3b@9AmcEs-(1f@VuWMBpPORH!9y$JNj=#aov7qW)=wAD9|m
z@wdB%i9S5nWNLqk-+T%~AWh0|9JiEM^2kL4q
z5tgIXA}Y!utzahLoTT;_IqMO+(UhvQS4u*gPVCJn#tY
zv=`)}d37|)2aUmO9HmhioM+fyb1UciDgI%l8X?xsiahK8U*10S+CYZqrQmEqN%}V#^6@@$H*J^xYujPW%!T
zg+gkQS!mMmL142-ZCOs!X)J%CCGnz5;#{#(lX_<#rHq^k`mfr1=(GwR`i(^mfL^bG
zNuQCs@9o3M6%^ec(3w8Gy^lZohMyEj;Npf+D=>%1h=f)WhK^p-eUBt|+->0FeA~zg
zZg%KL#E&dA8d7b^vE()=5-d5nS$d`-@AvJfiZ92YTH!6fU*SD(p6GtEiL1YHQfP0;
zrpfF~+^TpxdSejVjLzrfhvJMD_<4LbL?P$oZYzauaZroPBKU-qMlFcjl1yy~zL{xXn1k@@_i
zkIIb0VM5zrzRbuNyHC*jkoC|DY+_6u9HQLXZ42?E!Qw=T8Hxvjwa1w^-~Vhpk5OKO>*jW)V5loe-_7%Io
z3(sGB#kzlXai=KYCpoU*P8Jf(j0%#Rl^E-oXP4arcq5i)B!KnO*`oBB%wx2QjzL`I
z-MdpY>3;fuYJ+CW{~!y~{OY@!v&koBuA(zMI{vyA@hE9M`tw_qVeYA&5!zc?i6~>g
za%M#qbTM%~pzDJ}bN1HM*H*en**)le;NEOU{&2Gm+~U^2wc^gb4U_Z`8P_T1rnE(<
z9)5Lj`7gZExAst)2zXwNJA>K&aS%YpDJJwU^l`I&93U_n;6JQ)r+a%_D9ywF2Yc@w
z4Oidy3y&a@=tS>9h!(wkOkO1TzSR7+h!P
zy6^jW&V9~#-{+k3&->52*05$VGi%S@zy1Av%cqnn{ieCr5yZ%x17mxc>jzRuiFT4S
z_%`&2h8}q(A6=@o1z!~$F)|Uv`4Cwe7i<kU8R}*dHAU@H!%tcH!);DlvZt*{_qqSf+zm
zm{;5)^`0~$wt#^nqq;))(X)Y(^C;gg6{PW&g3!-GfQ-PD7Ed
zWGxHK2Jnny5klpnjsp_Tq+FU<&R4dAc9@H9JUIOcJ_1Y_dxe~!Gm(R<$LEX+!z|N;
z2M*WgK3u#ay22z*7GLX}9O>{eZ0T%?_EVxIiq&a;re&pwglwItn{0yY09z_Ofmr
zX|e!foQ%=s5+~@=je2gAfMn7n?S=Oy1C&~>DP0K5O~j+dk)fOcrrU<_jSc>R2%U-E
zvilF&DMBWKsEf#OFRJ+qOS3sozipiguXmo96K=UDo~$7kQ#2+jh)Me_OFPdS0hCD0
z8B7NxXfHiJK4wJ{<{1ARC+45FVk|E5bLN{~K+o?x)8ngO(g2h;PF4t!0M4Sc!!Xhz
zxF~nU>nSEl(koVs2Kvr8EnP|Mi()hk4CWh`e|wpF=)&e)Y*|M=yk#`_J`nk4M{}jM
z<~?|YkCSPP*;+O&&Jv9^TarXD;J*r8kY|Oxc^VQdeph^p`AeSQJ`4uKrn|UQ7qzsc
z#}4yVn<5i*#(a}cL~(U-{{3md}@?bD#Zh@$!_%=o16ZTQEIo5`TQLmrA
z6UuX%{Hh~$BXaXV1LyU>U$uW-M5}gwL=f177#DG5VCS#Vdo;AXD)(5wr9t1to!txX
zYWEJ~^Zpfs0UGsy{&uYj2XF2>hK&qx0TbLsVs^^OTft@jIP?0}Z%0&Fx5?+|aoOkn
z)KE00v&pIp7ucvX!aIu=Mo4?;=)Ia+Ws5xjx+j*OZ!7ki<^yO({E89+w?Q&Hb@tnC
z#r7UX5$s>Mdfz1vy`svPXs&Xj&KIcZqwheM1jx%8SByqx7!0ztt|SFfK6&~tXI;SQ
z^6!WI33!Fh1EJ!^#2W&mUidmwvM3`Dt)E7J$fPo?BD$VZ?S_`Lu(yL`#X=#4$nBDh0~mqolFH
zx+e2R=0=&id?9fWYF0T2QRcYxg)bfPeg4&e)ep6MfccR4&NA0vyR4*lu#8~5o%JI9
zWdOu(E{wCId|)cy3=j_CAf#O(0ER*#n
z&)zmNWxVVsl~?AWgT!>#XqmN%Mi#sTcY@#E+Y{^_)qzVBW3T}G^4
zl^W_yBGEnx&vY`xyUO11T9G?=aXi_nK|W~iFPjYe_}UH2V4feZAj!*f_%W8vV*IrX+@1s>Lhj~t$-Zm`VZ|3I3lIxlL*!Z(j-+)b2`lP8h
zzB(Y{b6dW5Qvci>K9^CgI=1(Cd;NJOaA0xQI{EfL56Je_e50wEwtPhya&<@~vL(IH
zWub3g*k%|=s>N@bt@C4OQxrKdJP#JEdx9I9Haj0do2d2u)yun`Y3yQThN*;54UCqb
zuJxvw%)XK@GbX9ez&|uBS@X|?x*R;W6p^6;;=H?>6vv{B_o&3~e{a-_W0(@gOQtx)
zdU8Aa*2GqfWCnWnSHL1Uu$WQH>uTQYD$UDa$w86JeP4U%gbb(nVo{q=d@hl7o1og|
z3@eYAt&$Mpw;O%fJuf$3$-&8sgu
z7v7zk(7-IZ>Zo3h9hfdbg@!q{lTF6g_>esRII!H}{Y!K_ooAf=O`pUn_2OgdK$vP(
z-JZz1>nsG8LD^7na}=vuf~lbEnirGPcZr%8W2hJoN^=uYU8{-~Gf8*1_N!NGK7;WS
zx%QRcca24%K;9B|M}+s`Q!fQzYy49crm5kp$A=R(2eTH(3uOYI#e%S
z%1b$k|Cw?gzE;drrBm#TGTLLj+%ZPt5f&`qFH3?Zceyb*YS=pSXTi+$(_1zLH7c}g
zff3d_q^=*O%|T%Aw*_j27fJRHg$wUrl*ntgl2RydF=nK4Z|U1W1P%YE>XexfT+77E%w3k83ARjGM~8Q1
zr|ZR#D~IJK&jt(nQ({+~~94fs3sXMc5*18{xEgJfp=Myq9y{ROXl0-vhl#Fza
ztaN`gnLgve*p?BNPEy`Mqg`gXx(@6PIx*X=Gf~#6N#Xn$99Y64{R$ctr;#N$EgMrS@>_eG?_|4OoAb91dZfdj-weRa-W4frs=X(%z;bVO$RM
z(nc#sF>dxh2Q@wqu{#KPvOVr%sD+>hL?)?3pJ%B9c@VY#e|;*H!kWWnk&~rbPZqcf;2>(K1#Ls)*i&t&G0MVb~&Im
z_Roy_zl={LFto%|4xI8>{3QKdTBB#T+!mcJx*(
zC|}m4;ijr0v;0d*V25>>AlEK0vr*CNX;Fb`FVpu9*MY=zDCdKMXg$E&E3J>HQd2RQ
z$nvyPk6-&PyhNt#{22S8nZXrTLuAG?1hhAljWY)uB
z;3JuuR&3+NPM#ZYq4!oCz&-vBUY4cHd#M8_8grHE7ZO4dh;Nag!Wg$!-!1n~s#5t9
zHktopUi)wOL{OeIZmR;siKWBbzOBSIW9|N`o%p3U>0}jQrmCUr)?iI3t3ge;@M36R^LwWHpAyoEZH*z;?iP06UJBt71evG2*pBR$
z=^TV;u(y@iXPDa}xTwq=FqiQmx9gn+supOJ{bHkD`sgU)ej`xWvYbf>ZWYl4kaV^L
z4su&EB4+cVH#}34erjF5TQd~c{v`EwARQ-d8mpVAO8zFN$g%;(u7DzDj;ofAoCd#hm<;vyGubOLJS^uWR`wgqPC&_Gz=AB9+)P!%4CK%;WKKrjWDP&zQK?0Sz;3q*$DU;#7ka
zd+4D~mqEZuTPpc;dQuq>j$;Z)_=F?mk1N`}Lnh=}yi7eLk;|6RODuf8FEw#b+@MLA
zIs!eT>}7agkf_Ad{gFN>Lg`~QZM2xz|JBX{=mNUCst{npR!g32tV(UfiI~r1^uEeI
z@Zz;7S%yFO`i_!n;2G@;A=&^%HirwCq+i!;W7KE-(^WXAmn^*EQ#B`hVWRH
z0Q;*46+pylkVh)FpB@rdz`qWbnX|G}sw+r*$fBTl08@Gwo0t*%FwOv86J&OloxX>;
zOsR2bTO?cV7x-N%ne_ckQh;Wjc4R=0;4e^B8~N^YK5YzcZRetl^kMcA*bJ-Nnoc@-tvx&5{WH(}WRey*Q&cs3n>%*wIEg3kqo5P=-v>)%%$Etuu
zY9I4w#ndm$tmAa#CkcNex}^%hvL=tZt`Ft70EogDg{jcLYfcG0zWcY$mv*q|BZZS>
zfFdLvAChZ*Gn=q!Ik0w|&2`Q00uW&VG3LeNvV)S}izPvm{Cd>3^IZ)EO{=rm+XoU)
zH*bb7@*9}UPGb2R;As6B!Z^&NtuOmS`1G9h=ErTK-|yG6;3+EOLxD4am#!oA%y7h$
z5>A#UGb6Pag_U_~4W_dd!mcLh>g(5C2fYo*2LrRQ4f`x3)XfWbEW@;$m35Ont#V29
z1meESyb&QQYrg%Otsw1Y0o~W)5iFy>F)`O-Gr+*Ph4t$ifgCC8F}!NBYhD39zo^d_
zMAV?U!#8G6$Gtuh`Vt0Ex?3Xaa%Nim8cC?%u)N+_o`*$9&3*Y{LgfSa(X@%MbrTS|
zDt6~d!#cPSPF3g82l3c?4fgskl17V$NytW-Rkum?>Tfu^EFam;9J7AH4ZfSysML|F@^)}|uk_2Uax;rTHD^EtN
zH32?ccAm6>kJ5_wX$z#^kco|UI$f&FGjtJSxf`R#Y9sR?&PxA@s*R1xuTq>09X6=Z
z`3`!qr(*$THtQ$9e(k3>Unxey(!Ahrs?{sPKJ4a3{&Jwa&;s<{cFe9v3X%7zd*Fc4
zbNpYoD2`4E(H;4Lzc*V@s1%t>73dWX9!OR866oZdX&nG%kSnq{C`j7%YG%kxQB|g9OOBcJAcVul!A~X^jMr`;h@UW;vi6Pq6up+w|=KZR|UR
z*vG_wf#e;##vbpDl}j?T=K7Djv^{L>d)?h#>n5MjceGsfJ=rW%4e<5Lxt};%|3vWN
zC*jXB_3Z>{
zdk>_m9qQhP{7^cj2gFJ;#csSmElv<#=o8ttDW;B<>7$@m&vpzeyWUU}bLmn8WkrPyx!Hl5O@
z8e;^$YatP%O>~TuR^#gIMPyUp%td4)GtV2!*PFZpuvE$WQ5z=s@f?$Kr7=-=IX)F|
zWMzbEeeJR`u)fZ6ldxSe9B5bnmO)Zg18#qoiZKem{$N0f$ZLuVj?SWF5w}d~mMVL0
z*v+)d9?$#>S=oQzJ}S$ty1e`&2>)s&xYRn)1F6tH07DuyK19)|Z0rX>2s@-rd0)*I
zw+Q-R(8t5upLuIB8`u%l
zrjgUpJ9G3*)*#OT%6Zsx71XWn#Z$`aCq`*
zQ+VKX{eV@nhk;*CzbNyeT*on~)s;wv`JRz;B2(6<74w7R<$=Z+3kHkwEV?_ixSs&Y
z1~8E3{A2a}m%ruL&Rq_DkxGruk~1*bgR}_F*_IgNyYvrlNT1tOFlsw{9h3dq?vQ5G
zOXitQKF>H*K=#(=uueC~FMCS49G>w?!0LIws`yioVc$5rzifvqdNp+IjoRpz{v&99`f0M?Q&inB#Q7Rx_Ma^M21v3-Tf
zUq+{YwU>R93NEwvmi`%|q3{Ica+Gsm$lf0pDT}g$>kA}&d_Ry_c63t{khc#2>FuSR#+tk#vjT9|3~}z=-szMO
zJinMS3r?KrHCyMjVh7V>tn92;ob8VbO
z7BgN-3za(|m?
zL+WKBs4iZZ`3*eS1Qk$t6J@Dhr|V{pFj!gs00YgPI?ws>)hG3a`(&UF1Wr5pzs}b5
z_(ZYUaJEZnTRTO{5NU)prclsx3Mtq=c4s1OC!RmIzpgMESHQPRGsm|ru4u5->$pdE
zUjM0ww)Y`q%TJJ@HoowajVXNgtlsta}X#xN~t`kRX0
z)*QhZ0jUFhlVAKTWDkcbgweORR(IFz67csRB1eVvMGHBZ0F*NGe)Ih4yn8oij2%2C
z0-@?t$Ua@GW(1Lze{l{)pOkuAp-S`)fI`~EvzOjpn*SjGS(>ze(FeziZ>yC_%6%Y0
z;O9r#?F(Dq`
zI%XP2c^T53(2`JX=*DT1c=yu5?2D4^spXN#6)7qkJr?pmbfL9txZgeoihQlQc&a&h
zy2?C