GitLab

* uneql rank.

* trim.

* enable passing in number of heads for each rank.

* simplify.

* simplify.

* cleanup.

* fix col parallel.

* fix bug with row parallel.

* fit out proj.

* refac.

* fix sharding logic.

* refac sharding.

* refac.

* support multiple of.

* make fn reuseable.

* fix bug in dimensions.

* scaffold.

* test uneven heads.

* fix test by adding barrier.

* refac.

* reuse code.

* clean up.

* prelim.

* add hf convertion fn.

* mlp.

* change name.

* fix bug.

* inverse permute.

* change comment.

* revert style changes.

* fix.

* add doc.

* revert.

* enable load safe.

* fix safe load.

* fix import.

* fix typing-related lints.

* fix ckpt loading logic.

* make single gpu work.

* test with parallel.

* ckpt format.

* enable pretrained state dict.

* remove unused imports.

* remove unused.

* mark idea related.

* Bump version to 2.0.2
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Update version in Dockerfile
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Add RNG state to kernel launch params
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Save seed and offset for backward
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Single thread write to global mem
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* compute_dq_dk_dv_1colblock get seed and offset from launch params
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* compute_dq_dk_dv_1rowblock get seed and offset from launch params
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Change forward c++ APIs to save RNG state for backward
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Change backward c++ APIs to set RNG state for bprop launcher
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Bug fixes
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Python side API changes
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Bug fix; only save seeds instead of full offset
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Account for 3D grid size
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

This resolves RuntimeErrors after running evaluation in inference mode:

```
  File "/home/proger/.local/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/proger/.local/lib/python3.10/site-packages/flash_attn/modules/mha.py", line 492, in forward
    qkv = self.rotary_emb(qkv)
  File "/home/proger/.local/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/proger/.local/lib/python3.10/site-packages/flash_attn/layers/rotary.py", line 229, in forward
    return apply_rotary_emb_qkv_(
  File "/home/proger/.local/lib/python3.10/site-packages/torch/autograd/function.py", line 506, in apply
    return super().apply(*args, **kwargs)  # type: ignore[misc]
RuntimeError: Inference tensors cannot be saved for backward. To work around you can make a clone to get a normal tensor and use it in autograd.
```