dbogunowicz

**This PR**: - Creates a common directory "vision" that contains: schema blueprint for all vision pipelines, utils shared between vision pipelines, utils for coco and imagenet dataset, utils for vision...

Create deployment folder for transformers integration

1

Testing: First I set up the sparsezoo model (picked a stub and downloaded it). I remove deployment directory: ```bash (sparsezoo_venv) damian@lambdaquad:~/.cache/sparsezoo/9768075d-2cf0-4bc7-98c3-fb6441a2ec36$ ls deployment model.md model.onnx recipe sample_inputs.tar.gz sample_outputs.tar.gz training (sparsezoo_venv)...

Create deployment folder for IC integration

1

## Testing ```bash python export.py --dataset imagenette --dataset-path /home/damian/imagenette --checkpoint-path pytorch_vision/mobilenet_v1_imagenette__03/training/checkpoint-best.pth ``` results in ```bash Model id is set to _imagenette__16 already downloaded imagenette ImagenetteSize.s160 2022-08-11 06:54:56 __main__ INFO exporting...

[SparseZoo v2 Bridge] Save transformers training artifacts to ModelDirectory directory

2

Allows creating the `ModelDirectory` local directory from the direct output of the transformers training procedure. To test: 1. You need to install locally the sparsezoo's `feature/ci-cd-refactor/validations` branch 2. Then run:...

Allows creating the `ModelDirectory` local directory from the direct output of the yolo training procedure. To test: 1. You need to install locally the sparsezoo's `feature/ci-cd-refactor/validations` branch 2. Then run:...

Implement label smoothing for CrossEntropy Loss

2

CrossEntropy losses (Categorical) now can take label smoothing are as parameter. In theory, ready to be reviewed, but for clarity, I will open this PR once the PyTorch 1.10.2 upgrade...

## Testing: ### with the variable set ```bash SPARSEZOO_MODELS_PATH=/home/damian/all_your_base_are_belong_to_us ``` ```python import sparsezoo import os stub = "zoo:cv/classification/mobilenet_v1-1.0/pytorch/sparseml/imagenet/pruned_quant-moderate" model = sparsezoo.Model(stub) print(os.getenv("SPARSEZOO_MODELS_PATH", "")) print(model._path) >> /home/damian/all_your_base_are_belong_to_us >> home/damian/all_your_base_are_belong_to_us/4ce8e2fe-29a1-4027-bead-408d28702643 ``` ###...

A simple guide to using the new sparsezoo. I would like this document to end up as a part of our user-facing docs. Let's discuss what properties this would need...