GitLab

* fix condition checks related to FA head_dim
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* force q,k,v contiguous when RoPE is in use
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* fix lint
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* Expand FA version
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

---------
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Refactor sharding.py for the further custom_partitioning migration
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating both FWD and BWD of LayerNorm/RMSNorm from xmap to custom_partitioning.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating both FWD and BWD of all kinds of softmax from xmap to custom_partitioning.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Fix the wrong order of parameters to LN/RMSN bwd in ln_mlp_fp8.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* WAR to LN/RMSN_fp8 before migrating to CP.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Fix the wrong order of parameters of bwd of LN/RMSN_fp8.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Following review feedback to modify
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Force the hidden dim in Norm ops to no sharding and add warning msg.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Reuse fwd_rule in VJP functions
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating both FWD and BWD of self-fused-attn from xmap to custom_partitioning.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating both FWD and BWD of cross-fused-attn from xmap to custom_partitioning.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* add gelu and dgelu.
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Reuse fwd_rule in VJP functions for attentions
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Apply native FP8 Dtypes to fp8.py
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating cast_and_transpose from xmap to custom_partitioning
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Migrating transpose from xmap to custom_partitioning
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Apply XLA pattern match to perform FP8 GEMM.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* migrate layernorm_fp8 to custom_partitioning.
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Unify code style
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Extend supported of Transpose with FP8
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Implementing layernorm_fp8_dot based on migrated custom calls.
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Renaming variables and publish NVTE_FP8_COLLECTION_NAME
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Replace Q/DQ custom calls with native XLA implementations
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* migrate gelu_fp to custom_partitioning.
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Miner fix
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Support custom calls with mutli-dims
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Support gerneral dot indices in _fp8_dot_impl
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Implementing layernrom_geglu_fp8_mlp
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Remove GEMM custom calls
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Remove xmap related code
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Fix typo and add query-function to FP8MetaPackage
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Fix some bugs of custom calls
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Fix CT's bugs
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Update UTs/eaxmaples to adapt to the API changes.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Unify kernel initilization in MLP.
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Modifing with code review's feedback
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Update README and Add deprecating warning to *ShardingType
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>

* Canonicalize the dtype
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Adding assertion for non-supported batch dims.
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Adding doc/examples to _multidim_transpose
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Set FP8 meta as WeightHParamsCollection.OVERWRITE_WITH_GRADIENT in Praxis modules.
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Set FP8 meta as WeightHParamsCollection.OVERWRITE_WITH_GRADIENT in Praxis modules.
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Apply dtype-based rtol/atol to UTs
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Deprecate QKV_INTERLEAVED enum
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Skip test_distributed_custom_ops.py
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Fix the wrong sharding of bias in SelfAttn
Signed-off-by: Ming Huang <mingh@nvidia.com>

* WAR to fix the wrong cu_seqlen of MHA when DP/FSDP enabled
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Adding distributed ops unit-tests
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Adding license to test_distributed_*
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Follow review feedback to modify
Signed-off-by: Ming Huang <mingh@nvidia.com>

* Use total bytes involved in collective ops as criteria.
Signed-off-by: Ming Huang <mingh@nvidia.com>

---------
Signed-off-by: Ming Huang <mingh@nvidia.com>
Signed-off-by: Ming-Xu Huang <mingh@nvidia.com>
Co-authored-by: Donglin Yang <dongliny@nvidia.com>

* Update README.rst - Installation section

Added pip install instructions and cleaned up pre-reqs and FlashAttention-2 section
Signed-off-by: Santosh Bhavani <santosh@semantic.md>

* Update README.rst
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Santosh Bhavani <santosh@semantic.md>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

Improve PyTorch memory usage
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

[C/JAX] Support more mask types for the arbitrary seqlen kernels and minor changes of JAX bias (#469)

* Move bias to float32
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Enable varlen
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Increase neg infinity abs values
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Enable varlen tests
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Remove unnecessary code
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Fix lint
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Support variable sequence length after cuDNN 8.9.6
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Use unique_ptr instead of shared_ptr
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Add a new mask type: PADDING_CAUSAL_MASK
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Support flash padding mask after 8.9.6
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Enhance the Max512 handling for causal masking and add the related tests
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Update the fused attn support lists
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Remove padding_aware from the caching
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Fix libtransformer.so issue
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Reduce the pad ratio tests
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Fix a bug with cuDNN 8.9.5
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Release backend resource after the module level unit test
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Clean the jax live arrays before running the unit tests
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Fix too-few-public-methods lint
Signed-off-by: Reese Wang <rewang@nvidia.com>

---------
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Make user buffer name configurable
Signed-off-by: Sangkug Lym <slym@nvidia.com>

* Apply suggestions from code review
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Fix duplicate argument
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

---------
Signed-off-by: Sangkug Lym <slym@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Returning an empty tensor of param dtype for wgrad
Signed-off-by: Selvaraj Anandaraj <selvaraja@computelab-frontend-4-ub22.nvidia.com>

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

---------
Signed-off-by: Selvaraj Anandaraj <selvaraja@computelab-frontend-4-ub22.nvidia.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Selvaraj Anandaraj <selvaraja@computelab-frontend-4-ub22.nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Deprecate QKV_INTERLEAVED use in JAX
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Deprecate QKV_INTERLEAVED use in Paddle
Signed-off-by: Reese Wang <rewang@nvidia.com>

* Enhance qkv enum mappings
Signed-off-by: rewang <rewang@nvidia.com>

* Fix LD_LIBRARY_PATH issue
Signed-off-by: rewang <rewang@nvidia.com>

* Arbitrary seqlen kernels only support self attention currently
Signed-off-by: rewang <rewang@nvidia.com>

---------
Signed-off-by: Reese Wang <rewang@nvidia.com>
Signed-off-by: rewang <rewang@nvidia.com>

* Use FP8 tolerances in JAX FP8 tests
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Programmatically compute expected floating point error
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Loosen tolerance for MNIST test
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Tim Moon <tmoon@nvidia.com>

Disable large test cases for PyTorch Transformer layer
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fixed minor bug with DistributedConfigsHelper prematurely crashing the test for insufficient GPUs before @pytest.skip condition.
Signed-off-by: Alp Dener <adener@nvidia.com>

* Update tests/jax/distributed_configs_helper.py
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Debug PyTest errors when running on single-GPU system
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Alp Dener <adener@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>

fix bwd error with FA v2
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

[JAX] Regression tests for custom ops sharding with both xmap and custom_partitioning.

Coverage:
- layernorm: fwd/grad, zero_centered_gamma, DP, TP_COL, DP_TP_COL
- rmsnorm: fwd/grad, DP, TP_COL, DP_TP_COL
- softmax: fwd/grad, SCALED, SCALED_MASKED, SCALED_UPPER_TRIANG_MASKED, DP, TP_COL, TP_ROW, DP_TP_COL, DP_TP_ROW
- self_fused_attn: fwd/grad, NO_BIAS, PRE_SCALE_BIAS, POST_SCALE_BIAS, NO_MASK, CAUSAL_MASK, PADDING_MASK, DP, TP_COL, DP_TP_COL
- cross_fused_attn: fwd/grad, NO_BIAS, NO_MASK, PADDING_MASK, DP, TP_COL, DP_TP_COL
Signed-off-by: Alp Dener <adener@nvidia.com>

* Experimental FP8 tensor
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Sudhakar Singh <sudhakars@nvidia.com>
Co-authored-by: Przemyslaw Tredak <ptrendx@gmail.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Add fp8 tensor to ci test
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* review comments and tests
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

* Default to FP8 usage
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

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

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

* Fix transpose caching
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Debug transpose caching

Handle case where transpose cache is updated externally.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Rename FP8GlobalStateManager.with_fp8_parameters
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* remove Float8Tensor from import API
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Avoid caching FP8 transposes if not required
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix import error in FP8 tensor tests
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix tranpose caching and checkpointing bug
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Improve caching and fix distopt case
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Update transformer_engine/pytorch/float8_tensor.py
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Remove recursive logic
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fix cache reset bug
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Store FP8 attributes in dict

Easier for multiple tensors to share, e.g. detached tensors.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Make sure scale_inv is 1D tensor
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Make sure scale_inv is 1D tensor
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fixes and detach recipe
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Set default fp8 data type
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Sudhakar Singh <sudhakars@nvidia.com>
Co-authored-by: Przemyslaw Tredak <ptrendx@gmail.com>

Inherit the existing XLA_FLAGS
Signed-off-by: Reese Wang <rewang@nvidia.com>

Signed-off-by: Frederic Bastien <fbastien@nvidia.com>

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

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

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

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

* Do not include logging macros in installed C headers
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug logging macros
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug C++ tests

Use Google style for header includes.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Update CUDA driver macros

Incorporating changes from #389.
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Jan Bielak <jbielak@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Use core error checking macros in PyTorch extensions

Hack to get around macro redefinition warning.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix missing arg when getting CUDA driver error string
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Reuse logging header in frameworks
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Jan Bielak <jbielak@nvidia.com>

Signed-off-by: niboshi <niboshi000@gmail.com>

* initial test fix
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Drop eval for selective checkpointing tests
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Remove redundant recompute for FA
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* CI fix; Decouple fused attention and numerics tests
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

fixed incorrect of extend_fsdp_sharding_meta() in cross_fused_attn()
Signed-off-by: Alp Dener <adener@nvidia.com>

canonicalize the dtype for the better user experience
Signed-off-by: Reese Wang <rewang@nvidia.com>

Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Ability to check cuDNN version from Python
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Modify the fused attention test to not use the CUDNN_VERSION env
variable which is specific to NGC containers
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

---------
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

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

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

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

Signed-off-by: Tim Moon <tmoon@nvidia.com>

Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add class for RNG state tracker.
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fix docs for checkpoint
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

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

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

* Debug PyTorch and Paddle tests on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Only run Paddle layer tests with cuDNN fMHA on supported archs
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug PyTorch fMHA tests
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Reduce JAX FP8 GEMM sizes

Avoid split-k kernels on Ada.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Disable JAX fused self-attention test on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Run supported fused attention tests on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Run supported fused attention JAX tests on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Enable Paddle fused attention on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Update reference scale calculation in TensorFlow test
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Restore backend support to reference FP8 attention impl in PyT test

Review suggestion from @cyanguwa
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix merge conflicts
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug Paddle tests on Ada
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Loosen tolerances for Paddle attention tests
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Assume causal mask implies equal seqlens in Paddle attention tests
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Tim Moon <tmoon@nvidia.com>

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

* rename set_context_parallel_running to set_context_parallel_group
Signed-off-by: xren <xren@nvidia.com>

* bug fix
Signed-off-by: xren <xren@nvidia.com>

---------
Signed-off-by: xren <xren@nvidia.com>

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

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

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

* add support for h2d/2hd in 8.9.6
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* cull unit tests in fused_attn.py and add skipif for layout tests
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* add workopt=1 flag for dpa tests
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* update support table for arbi_seqlen backend
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* fix rotary position embedding and add unit tests accordingly
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* further cut down unit tests for CI efficiency
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* fix lint
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

* remove einops dependency
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

---------
Signed-off-by: Charlene Yang <8636796+cyanguwa@users.noreply.github.com>

Signed-off-by: Tim Moon <tmoon@nvidia.com>

Signed-off-by: Tim Moon <tmoon@nvidia.com>

Signed-off-by: Przemek Tredak <ptredak@nvidia.com>