Skip to content

Relicense to Apache-2.0 (dual-license; external-contributor files remain MIT) - #408

Merged
Anerudhan merged 1 commit into
NVIDIA:developfrom
Anerudhan:relicense-apache-2.0
Jul 30, 2026
Merged

Relicense to Apache-2.0 (dual-license; external-contributor files remain MIT)#408
Anerudhan merged 1 commit into
NVIDIA:developfrom
Anerudhan:relicense-apache-2.0

Conversation

@Anerudhan

@Anerudhan Anerudhan commented Jul 17, 2026

Copy link
Copy Markdown
Collaborator

Summary

Relicenses NVIDIA-authored code in cudnn-frontend from MIT to the Apache License 2.0, using the exact NVIDIA OSS SPDX header. Files that still carry contributions from external contributors whose permission has not been established remain under MIT. Rebased onto current develop (b950af1).

License assignment

Every source file carries an SPDX SPDX-License-Identifier: tag.

License Files Which
Apache-2.0 541 NVIDIA-authored, plus files cleared by consent or employment
MIT (consent pending) 50 Surviving lines from external contributors who have not yet consented
MIT (third-party-derived) 29 FlashAttention / QuACK-derived files carrying external authors' copyright

Per-file mapping — including the commit that introduced each surviving external line — is in LICENSING.md.

How the split is computed

Blame-based, not touch-based. A file stays MIT only if a not-yet-cleared external contributor's lines survive in current develop. Files whose external lines were fully overwritten by NVIDIA, or whose contributors are cleared, go to Apache-2.0.

Licensing files

  • LICENSE.txt (Apache-2.0), LICENSE-MIT.txt (MIT subset)
  • LICENSING.md — manifest: every MIT file, its pending contributor, and the introducing commit; plus the consent/employment roster
  • THIRD_PARTY_LICENSES.txt — nlohmann/json, FlashAttention, CUTLASS, Megatron-LM, QuACK, labml.ai, dlpack, pybind11, Catch2
  • NOTICE, pyproject.toml (license = "Apache-2.0 AND MIT"), README.md

Verification

  • Every source file has exactly one SPDX tag; none unlabeled
  • All Python compiles (py_compile); git diff -U0 confirms every source change is comment/header-only — compile-neutral by construction

🤖 Generated with Claude Code

@coderabbitai

coderabbitai Bot commented Jul 17, 2026

Copy link
Copy Markdown
Contributor

Important

Review skipped

Too many files!

This PR contains 621 files, which is 321 over the limit of 300.

To get a review, narrow the scope:
• coderabbit review --committed # exclude uncommitted changes
• coderabbit review --dir # limit to a subdirectory
• coderabbit review --base # compare against a closer base

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: f0e9742c-6cd9-409c-ad38-0626dc4dda39

📥 Commits

Reviewing files that changed from the base of the PR and between b950af1 and 0c08141.

⛔ Files ignored due to path filters (6)
  • include/cudnn_frontend/generated/rms_norm_silu/sm100/ln_fwd_silu_kernel.h is excluded by !**/generated/**
  • include/cudnn_frontend/generated/rms_norm_silu/sm100/ln_headers.h is excluded by !**/generated/**
  • include/cudnn_frontend/generated/sdpa/sm100/prefill/full_seqlens/d128_fprop_kernel.h is excluded by !**/generated/**
  • include/cudnn_frontend/generated/sdpa/sm100/prefill/full_seqlens/d64_fprop_kernel.h is excluded by !**/generated/**
  • include/cudnn_frontend/generated/sdpa/sm90/prefill/full_seqlens/d128_fprop_kernel.h is excluded by !**/generated/**
  • include/cudnn_frontend/generated/sdpa/sm90/prefill/full_seqlens/d64_fprop_kernel.h is excluded by !**/generated/**
📒 Files selected for processing (621)
  • CMakeLists.txt
  • LICENSE-MIT.txt
  • LICENSE.txt
  • LICENSING.md
  • NOTICE
  • README.md
  • THIRD_PARTY_LICENSES.txt
  • benchmark/bench_moe.py
  • benchmark/csa/bench_csa_compressor.py
  • benchmark/csa/gate_csa_compressor_r128.py
  • benchmark/csa/reg_probe_csa_compressor_r128.py
  • benchmark/cutedsl_fusion_kernels/cutedsl_fusion_benchmarks.py
  • benchmark/dsa/benchmark_dsa_sparse_attention_backward.py
  • benchmark/norms/Dockerfile
  • benchmark/norms/__init__.py
  • benchmark/norms/benchmark_single_norm.py
  • benchmark/norms/charts.py
  • benchmark/norms/config_types.py
  • benchmark/norms/configs/__init__.py
  • benchmark/norms/configs/all_models.py
  • benchmark/norms/configs/deepseek_v3.py
  • benchmark/norms/configs/gpt3_175b.py
  • benchmark/norms/configs/llama31_405b.py
  • benchmark/norms/configs/llama3_70b.py
  • benchmark/norms/configs/llama3_8b.py
  • benchmark/norms/configs/llama4_e16.py
  • benchmark/norms/configs/mixtral_8x22b.py
  • benchmark/norms/configs/mixtral_8x7b.py
  • benchmark/norms/configs/nemotronh_56b.py
  • benchmark/norms/configs/qwen3_235b.py
  • benchmark/norms/configs/qwen3_30b.py
  • benchmark/norms/runner.py
  • benchmark/sdpa_benchmark_training/Dockerfile
  • benchmark/sdpa_benchmark_training/__init__.py
  • benchmark/sdpa_benchmark_training/bench_ar_dit_peak.py
  • benchmark/sdpa_benchmark_training/benchmark_single_sdpa.py
  • benchmark/sdpa_benchmark_training/charts.py
  • benchmark/sdpa_benchmark_training/config_types.py
  • benchmark/sdpa_benchmark_training/configs/__init__.py
  • benchmark/sdpa_benchmark_training/configs/auto_regressive_dit.py
  • benchmark/sdpa_benchmark_training/configs/dsv3.py
  • benchmark/sdpa_benchmark_training/configs/gpt_oss.py
  • benchmark/sdpa_benchmark_training/configs/kimiK26.py
  • benchmark/sdpa_benchmark_training/configs/llama.py
  • benchmark/sdpa_benchmark_training/configs/ltx2.py
  • benchmark/sdpa_benchmark_training/configs/qwen35.py
  • benchmark/sdpa_benchmark_training/configs/wan22.py
  • benchmark/sdpa_benchmark_training/runner.py
  • cmake/cuDNN.cmake
  • cudnn_frontend-config.cmake.in
  • include/cudnn_backend_base.h
  • include/cudnn_frontend.h
  • include/cudnn_frontend/backend/backend_descriptor.h
  • include/cudnn_frontend/backend/device_properties.h
  • include/cudnn_frontend/backend/execution_helpers.h
  • include/cudnn_frontend/backend/kernel_cache.h
  • include/cudnn_frontend/backend/plan_helpers.h
  • include/cudnn_frontend/context.h
  • include/cudnn_frontend/cudnn_interface.h
  • include/cudnn_frontend/experimental/attention_utils.h
  • include/cudnn_frontend/experimental/nvrtc_shim.h
  • include/cudnn_frontend/experimental/oss_engine_interface.h
  • include/cudnn_frontend/experimental/sm100_rms_norm_silu_engine.h
  • include/cudnn_frontend/experimental/sm100_rms_norm_silu_knobs.h
  • include/cudnn_frontend/experimental/sm100_sdpa_prefill_engine.h
  • include/cudnn_frontend/experimental/sm90_sdpa_prefill_engine.h
  • include/cudnn_frontend/graph_helpers.h
  • include/cudnn_frontend/graph_interface.h
  • include/cudnn_frontend/graph_properties.h
  • include/cudnn_frontend/knobs.h
  • include/cudnn_frontend/node/adaptive_layernorm.h
  • include/cudnn_frontend/node/batchnorm.h
  • include/cudnn_frontend/node/batchnorm_inference.h
  • include/cudnn_frontend/node/block_scale_dequantize.h
  • include/cudnn_frontend/node/block_scale_quantize.h
  • include/cudnn_frontend/node/bn_finalize.h
  • include/cudnn_frontend/node/concatenate.h
  • include/cudnn_frontend/node/conv_dgrad.h
  • include/cudnn_frontend/node/conv_fprop.h
  • include/cudnn_frontend/node/conv_wgrad.h
  • include/cudnn_frontend/node/dbn.h
  • include/cudnn_frontend/node/dbn_weight.h
  • include/cudnn_frontend/node/diagonal_band_mask.h
  • include/cudnn_frontend/node/dln.h
  • include/cudnn_frontend/node/genstats.h
  • include/cudnn_frontend/node/instancenorm.h
  • include/cudnn_frontend/node/layernorm.h
  • include/cudnn_frontend/node/matmul.h
  • include/cudnn_frontend/node/matmul_fp8.h
  • include/cudnn_frontend/node/moe_grouped_matmul.h
  • include/cudnn_frontend/node/moe_grouped_matmul_bwd.h
  • include/cudnn_frontend/node/paged_cache_load.h
  • include/cudnn_frontend/node/pointwise.h
  • include/cudnn_frontend/node/reduction.h
  • include/cudnn_frontend/node/resample.h
  • include/cudnn_frontend/node/reshape.h
  • include/cudnn_frontend/node/rmsnorm.h
  • include/cudnn_frontend/node/rng.h
  • include/cudnn_frontend/node/rope.h
  • include/cudnn_frontend/node/rope_backward.h
  • include/cudnn_frontend/node/scaled_dot_product_flash_attention.h
  • include/cudnn_frontend/node/sdpa_fp8_bwd.h
  • include/cudnn_frontend/node/sdpa_support_surface.h
  • include/cudnn_frontend/node/slice.h
  • include/cudnn_frontend/node/softmax.h
  • include/cudnn_frontend/node/transpose.h
  • include/cudnn_frontend/node_interface.h
  • include/cudnn_frontend/plans.h
  • include/cudnn_frontend/utils/attn_score_modifiers.h
  • include/cudnn_frontend/utils/serialize.h
  • include/cudnn_frontend_ConvDesc.h
  • include/cudnn_frontend_Engine.h
  • include/cudnn_frontend_EngineConfig.h
  • include/cudnn_frontend_EngineConfigGenerator.h
  • include/cudnn_frontend_EngineFallbackList.h
  • include/cudnn_frontend_Errata.h
  • include/cudnn_frontend_ExecutionPlan.h
  • include/cudnn_frontend_ExecutionPlanCache.h
  • include/cudnn_frontend_Filters.h
  • include/cudnn_frontend_Heuristics.h
  • include/cudnn_frontend_Logging.h
  • include/cudnn_frontend_MatMulDesc.h
  • include/cudnn_frontend_Operation.h
  • include/cudnn_frontend_OperationGraph.h
  • include/cudnn_frontend_PointWiseDesc.h
  • include/cudnn_frontend_ReductionDesc.h
  • include/cudnn_frontend_Reorder_Tensor.h
  • include/cudnn_frontend_Resample.h
  • include/cudnn_frontend_Rng.h
  • include/cudnn_frontend_Tensor.h
  • include/cudnn_frontend_VariantPack.h
  • include/cudnn_frontend_find_plan.h
  • include/cudnn_frontend_get_plan.h
  • include/cudnn_frontend_shim.h
  • include/cudnn_frontend_utils.h
  • include/cudnn_frontend_version.h
  • pyproject.toml
  • python/CMakeLists.txt
  • python/cudnn/__init__.py
  • python/cudnn/_experimental_warnings.py
  • python/cudnn/_pygraph.py
  • python/cudnn/api_base.py
  • python/cudnn/block_sparse_attention/__init__.py
  • python/cudnn/block_sparse_attention/_interface.py
  • python/cudnn/block_sparse_attention/api.py
  • python/cudnn/block_sparse_attention/csrc/__init__.py
  • python/cudnn/block_sparse_attention/csrc/bwd/__init__.py
  • python/cudnn/block_sparse_attention/csrc/bwd/bsa_bwd_postprocess.py
  • python/cudnn/block_sparse_attention/csrc/bwd/bsa_bwd_prepost.py
  • python/cudnn/block_sparse_attention/csrc/bwd/bsa_bwd_preprocess.py
  • python/cudnn/block_sparse_attention/csrc/bwd/bucketed_k2q_csr.py
  • python/cudnn/block_sparse_attention/csrc/bwd/sm100_blk128/__init__.py
  • python/cudnn/block_sparse_attention/csrc/bwd/sm100_blk128/bsa_bwd_sm100.py
  • python/cudnn/block_sparse_attention/csrc/bwd/sm100_blk64/__init__.py
  • python/cudnn/block_sparse_attention/csrc/bwd/sm100_blk64/bsa_bwd_sm100.py
  • python/cudnn/block_sparse_attention/csrc/bwd/sm90_blk64/bsa_bwd_sm90.py
  • python/cudnn/block_sparse_attention/csrc/fwd/__init__.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm100_blk128/__init__.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm100_blk128/bsa_fwd_sm100.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm100_blk64/bsa_fwd_combine.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm100_blk64/bsa_fwd_helpers.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm100_blk64/bsa_fwd_sm100.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm120_blk64/bsa_fwd_sm120.py
  • python/cudnn/block_sparse_attention/csrc/fwd/sm90_blk64/bsa_fwd_sm90.py
  • python/cudnn/block_sparse_attention/csrc/utils/__init__.py
  • python/cudnn/block_sparse_attention/csrc/utils/batched_static_scheduler.py
  • python/cudnn/block_sparse_attention/csrc/utils/block_info.py
  • python/cudnn/block_sparse_attention/csrc/utils/block_sparse_tile_scheduler.py
  • python/cudnn/block_sparse_attention/csrc/utils/copy_utils.py
  • python/cudnn/block_sparse_attention/csrc/utils/cute_dsl_utils.py
  • python/cudnn/block_sparse_attention/csrc/utils/kernel_utils.py
  • python/cudnn/block_sparse_attention/csrc/utils/layout_utils.py
  • python/cudnn/block_sparse_attention/csrc/utils/mma_sm100_desc.py
  • python/cudnn/block_sparse_attention/csrc/utils/named_barrier.py
  • python/cudnn/block_sparse_attention/csrc/utils/pack_gqa.py
  • python/cudnn/block_sparse_attention/csrc/utils/pipeline.py
  • python/cudnn/block_sparse_attention/csrc/utils/seqlen_info.py
  • python/cudnn/block_sparse_attention/csrc/utils/sm90_utils.py
  • python/cudnn/block_sparse_attention/csrc/utils/softmax.py
  • python/cudnn/block_sparse_attention/csrc/utils/tcgen05_mma_helpers.py
  • python/cudnn/block_sparse_attention/csrc/utils/tile_scheduler.py
  • python/cudnn/collect_env.py
  • python/cudnn/csa/__init__.py
  • python/cudnn/csa/compressor/__init__.py
  • python/cudnn/csa/compressor/api.py
  • python/cudnn/csa/compressor/compressor_sm100.py
  • python/cudnn/csa/compressor/compressor_sm100_r128.py
  • python/cudnn/datatypes.py
  • python/cudnn/deepseek_sparse_attention/__init__.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/__init__.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/api.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/dense_indexer_backward_sm100.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/dense_indexer_backward_sm90.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/indexer_backward_sm100.py
  • python/cudnn/deepseek_sparse_attention/indexer_backward/indexer_backward_sm90.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/__init__.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/_interface.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/_interface_sm90.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/api.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/indexer_fwd_sm100.py
  • python/cudnn/deepseek_sparse_attention/indexer_forward/indexer_fwd_sm90.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/__init__.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/api.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/block_scan.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/compactify.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/indexer_top_k_decode_varlen.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/indexer_top_k_varlen_util.py
  • python/cudnn/deepseek_sparse_attention/indexer_top_k/local_to_global_dsl.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/__init__.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/_interface_sm100.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/_interface_sm90.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/api.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/dense_score_recompute_sm100.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/dense_score_recompute_sm90.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/pack_gqa.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/sparse_score_recompute_sm100.py
  • python/cudnn/deepseek_sparse_attention/score_recompute/sparse_score_recompute_sm90.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/__init__.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/_interface_sm100.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/_interface_sm90.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/api.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/dsa_bwd_sm100.py
  • python/cudnn/deepseek_sparse_attention/sparse_attention_backward/dsa_bwd_sm90.py
  • python/cudnn/deepseek_sparse_attention/utils/__init__.py
  • python/cudnn/deepseek_sparse_attention/utils/compiler.py
  • python/cudnn/deepseek_sparse_attention/utils/copy.py
  • python/cudnn/deepseek_sparse_attention/utils/runtime.py
  • python/cudnn/deepseek_sparse_attention/utils/seqlen.py
  • python/cudnn/deepseek_sparse_attention/utils/sm100/__init__.py
  • python/cudnn/deepseek_sparse_attention/utils/sm100/gemm.py
  • python/cudnn/deepseek_sparse_attention/utils/sm100/mma_desc.py
  • python/cudnn/deepseek_sparse_attention/utils/sm90/__init__.py
  • python/cudnn/deepseek_sparse_attention/utils/sm90/bwd_barriers.py
  • python/cudnn/deepseek_sparse_attention/utils/sm90/bwd_tile_scheduler.py
  • python/cudnn/deepseek_sparse_attention/utils/sm90/mma.py
  • python/cudnn/deepseek_sparse_attention/utils/sm90/primitives.py
  • python/cudnn/deepseek_sparse_attention/utils/tensor_conversion.py
  • python/cudnn/discrete_grouped_gemm/__init__.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_dswiglu/__init__.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_dswiglu/api.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_dswiglu/discrete_B_blockscaled_grouped_gemm_dglu_dbias.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_swiglu/__init__.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_swiglu/api.py
  • python/cudnn/discrete_grouped_gemm/discrete_grouped_gemm_swiglu/discrete_B_blockscaled_grouped_gemm_glu_bias.py
  • python/cudnn/discrete_grouped_gemm/discrete_kernel_utils.py
  • python/cudnn/discrete_grouped_gemm/moe_persistent_scheduler.py
  • python/cudnn/discrete_grouped_gemm/moe_sched_extension.py
  • python/cudnn/discrete_grouped_gemm/moe_utils.py
  • python/cudnn/engines/__init__.py
  • python/cudnn/engines/base.py
  • python/cudnn/engines/engine_ids.py
  • python/cudnn/engines/reference_matmul_engine.py
  • python/cudnn/engines/router.py
  • python/cudnn/experimental/__init__.py
  • python/cudnn/experimental/ops/__init__.py
  • python/cudnn/experimental/ops/moe_grouped_matmul.py
  • python/cudnn/experimental/ops/sdpa.py
  • python/cudnn/gemm_amax/__init__.py
  • python/cudnn/gemm_amax/api.py
  • python/cudnn/gemm_amax/dense_blockscaled_gemm_persistent_amax.py
  • python/cudnn/gemm_dsrelu/__init__.py
  • python/cudnn/gemm_dsrelu/api.py
  • python/cudnn/gemm_dsrelu/dense_blockscaled_gemm_persistent_dsrelu_quant.py
  • python/cudnn/gemm_proj_rope_mxfp8/__init__.py
  • python/cudnn/gemm_proj_rope_mxfp8/api.py
  • python/cudnn/gemm_proj_rope_mxfp8/gemm_proj_rope_mxfp8_bf16in.py
  • python/cudnn/gemm_proj_rope_mxfp8/gemm_proj_rope_mxfp8_mxfp8in.py
  • python/cudnn/gemm_srelu/__init__.py
  • python/cudnn/gemm_srelu/api.py
  • python/cudnn/gemm_srelu/dense_blockscaled_gemm_persistent_srelu_quant.py
  • python/cudnn/gemm_swiglu/__init__.py
  • python/cudnn/gemm_swiglu/api.py
  • python/cudnn/gemm_swiglu/dense_blockscaled_gemm_persistent_swiglu_interleaved_quant.py
  • python/cudnn/gemm_swiglu/dense_gemm_persistent_swiglu.py
  • python/cudnn/graph.py
  • python/cudnn/graph_types.py
  • python/cudnn/grouped_gemm/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/_bf16_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/_blockscaled_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/moe_blockscaled_grouped_gemm_dglu_dbias.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/moe_blockscaled_grouped_gemm_dglu_rubin.py
  • python/cudnn/grouped_gemm/grouped_gemm_dglu/moe_grouped_gemm_dglu_dbias.py
  • python/cudnn/grouped_gemm/grouped_gemm_dsrelu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_dsrelu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_dsrelu/moe_blockscaled_grouped_gemm_dsrelu_quant.py
  • python/cudnn/grouped_gemm/grouped_gemm_dswiglu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_dswiglu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_dswiglu/grouped_gemm_dswiglu_quant.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/_bf16_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/_blockscaled_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/moe_blockscaled_grouped_gemm_glu_bias.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/moe_blockscaled_grouped_gemm_glu_rubin.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu/moe_grouped_gemm_glu_bias.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu_hadamard/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu_hadamard/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu_hadamard/hadamard_utils.py
  • python/cudnn/grouped_gemm/grouped_gemm_glu_hadamard/moe_blockscaled_grouped_gemm_glu_hadamard.py
  • python/cudnn/grouped_gemm/grouped_gemm_quant/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_quant/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_quant/grouped_gemm_quant.py
  • python/cudnn/grouped_gemm/grouped_gemm_quant/moe_blockscaled_grouped_gemm_quant_rubin.py
  • python/cudnn/grouped_gemm/grouped_gemm_srelu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_srelu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_srelu/moe_blockscaled_grouped_gemm_srelu_quant.py
  • python/cudnn/grouped_gemm/grouped_gemm_swiglu/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_swiglu/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_swiglu/grouped_gemm_swiglu_quant.py
  • python/cudnn/grouped_gemm/grouped_gemm_unfused/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_unfused/_bf16_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_unfused/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_unfused/moe_grouped_gemm.py
  • python/cudnn/grouped_gemm/grouped_gemm_utils.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/__init__.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/_bf16_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/_blockscaled_api.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/api.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/moe_blockscaled_grouped_gemm_wgrad.py
  • python/cudnn/grouped_gemm/grouped_gemm_wgrad/moe_grouped_gemm_wgrad.py
  • python/cudnn/grouped_gemm/moe_kernel_helpers.py
  • python/cudnn/grouped_gemm/moe_persistent_scheduler.py
  • python/cudnn/grouped_gemm/moe_sched_extension.py
  • python/cudnn/grouped_gemm/moe_utils.py
  • python/cudnn/grouped_gemm/utils.py
  • python/cudnn/native_sparse_attention/__init__.py
  • python/cudnn/native_sparse_attention/compression/__init__.py
  • python/cudnn/native_sparse_attention/compression/api.py
  • python/cudnn/native_sparse_attention/compression/fmha.py
  • python/cudnn/native_sparse_attention/compression/fmha_helpers.py
  • python/cudnn/native_sparse_attention/selection/NSA_select_attn_fwd_hmma.py
  • python/cudnn/native_sparse_attention/selection/__init__.py
  • python/cudnn/native_sparse_attention/selection/api.py
  • python/cudnn/native_sparse_attention/sliding_window_attention/__init__.py
  • python/cudnn/native_sparse_attention/sliding_window_attention/api.py
  • python/cudnn/native_sparse_attention/top_k/__init__.py
  • python/cudnn/native_sparse_attention/top_k/api.py
  • python/cudnn/native_sparse_attention/top_k/nsa_top_k_reduction_fwd.py
  • python/cudnn/native_sparse_attention/utils.py
  • python/cudnn/nodes.py
  • python/cudnn/ops/__init__.py
  • python/cudnn/ops/causal_conv1d.py
  • python/cudnn/rmsnorm_rht_amax/__init__.py
  • python/cudnn/rmsnorm_rht_amax/api.py
  • python/cudnn/rmsnorm_rht_amax/kernel.py
  • python/cudnn/sdpa/__init__.py
  • python/cudnn/sdpa/bwd/__init__.py
  • python/cudnn/sdpa/bwd/api.py
  • python/cudnn/sdpa/bwd/fmha_backward_sm100_2kernel.py
  • python/cudnn/sdpa/bwd/fmha_dkdv_d256_sm100.py
  • python/cudnn/sdpa/bwd/fmha_dq_d256_sm100.py
  • python/cudnn/sdpa/fmha_utils.py
  • python/cudnn/sdpa/fwd/__init__.py
  • python/cudnn/sdpa/fwd/api.py
  • python/cudnn/sdpa/fwd/fmha_forward_sm100_d256.py
  • python/cudnn/sdpa/utils.py
  • python/cudnn/wrapper.py
  • python/cudnn/yarn.py
  • python/properties.cpp
  • python/pycudnn.cpp
  • python/pygraph/norm.cpp
  • python/pygraph/pointwise.cpp
  • python/pygraph/pygraph.cpp
  • python/pygraph/pygraph.h
  • python/pygraph/sdpa.cpp
  • samples/CMakeLists.txt
  • samples/cpp/CMakeLists.txt
  • samples/cpp/causal_conv1d/b2b_causal_conv1d.cpp
  • samples/cpp/causal_conv1d/causal_conv1d.cpp
  • samples/cpp/causal_conv1d/causal_conv1d_nwh.cpp
  • samples/cpp/convolution/conv_dynamic_shape_benchmark.cpp
  • samples/cpp/convolution/dgrads.cpp
  • samples/cpp/convolution/fp8_fprop.cpp
  • samples/cpp/convolution/fprop.cpp
  • samples/cpp/convolution/int8_fprop.cpp
  • samples/cpp/convolution/wgrads.cpp
  • samples/cpp/matmul/blackwell_nvfp4_mxfp8_block_scale_matmul.cpp
  • samples/cpp/matmul/complex_fp32_matmul.cpp
  • samples/cpp/matmul/fp8_matmul.cpp
  • samples/cpp/matmul/general_block_scale_matmul.cpp
  • samples/cpp/matmul/int8_matmul.cpp
  • samples/cpp/matmul/matmuls.cpp
  • samples/cpp/matmul/mixed_matmul.cpp
  • samples/cpp/membound/boolean_fusion.cpp
  • samples/cpp/membound/concat.cpp
  • samples/cpp/membound/membound_fusion.cpp
  • samples/cpp/membound/reshape.cpp
  • samples/cpp/membound/slice.cpp
  • samples/cpp/membound/transpose.cpp
  • samples/cpp/misc/autotuning.cpp
  • samples/cpp/misc/compile_time_constant_example.cpp
  • samples/cpp/misc/cudagraphs.cpp
  • samples/cpp/misc/custom_plan.cpp
  • samples/cpp/misc/deviceless_aot_compilation.cpp
  • samples/cpp/misc/parallel_compilation.cpp
  • samples/cpp/misc/pointwise.cpp
  • samples/cpp/misc/resample.cpp
  • samples/cpp/misc/serialization.cpp
  • samples/cpp/misc/slice.cpp
  • samples/cpp/misc/sm_carveout.cpp
  • samples/cpp/moe_grouped_matmul/moe_grouped_matmul.cpp
  • samples/cpp/norm/adaptive_layernorm.cpp
  • samples/cpp/norm/batchnorm.cpp
  • samples/cpp/norm/layernorm.cpp
  • samples/cpp/norm/layernorm_bitmask_relu.cpp
  • samples/cpp/norm/norm_block_scale.cpp
  • samples/cpp/norm/norm_zero_centered_gamma.cpp
  • samples/cpp/norm/rmsnorm.cpp
  • samples/cpp/sdpa/fp16_benchmark.cpp
  • samples/cpp/sdpa/fp16_bwd.cpp
  • samples/cpp/sdpa/fp16_bwd_with_cudagraphs.cpp
  • samples/cpp/sdpa/fp16_bwd_with_flexible_graphs.cpp
  • samples/cpp/sdpa/fp16_bwd_with_sink_token.cpp
  • samples/cpp/sdpa/fp16_cached.cpp
  • samples/cpp/sdpa/fp16_dynamic_shapes.cpp
  • samples/cpp/sdpa/fp16_fwd.cpp
  • samples/cpp/sdpa/fp16_fwd_paged_decode_and_prefill.cpp
  • samples/cpp/sdpa/fp16_fwd_with_block_mask.cpp
  • samples/cpp/sdpa/fp16_fwd_with_cu_seq_len.cpp
  • samples/cpp/sdpa/fp16_fwd_with_cudagraphs.cpp
  • samples/cpp/sdpa/fp16_fwd_with_custom_dropout.cpp
  • samples/cpp/sdpa/fp16_fwd_with_flexible_graphs.cpp
  • samples/cpp/sdpa/fp16_fwd_with_max_and_sum_exp.cpp
  • samples/cpp/sdpa/fp16_fwd_with_paged_caches.cpp
  • samples/cpp/sdpa/fp16_fwd_with_sink_token.cpp
  • samples/cpp/sdpa/fp8_bwd.cpp
  • samples/cpp/sdpa/fp8_bwd_bottom_right_causal_mask.cpp
  • samples/cpp/sdpa/fp8_bwd_with_current_scaling.cpp
  • samples/cpp/sdpa/fp8_fwd.cpp
  • samples/cpp/sdpa/fp8_fwd_bottom_right_causal_mask.cpp
  • samples/cpp/sdpa/fp8_fwd_current_scaling.cpp
  • samples/cpp/sdpa/mxfp8_bwd.cpp
  • samples/cpp/sdpa/mxfp8_fwd.cpp
  • samples/cpp/sdpa/prefill_oss_engine.cpp
  • samples/cpp/utils/helpers.h
  • samples/legacy_samples/CMakeLists.txt
  • samples/legacy_samples/conv_sample.cpp
  • samples/legacy_samples/conv_sample.h
  • samples/legacy_samples/cpu_references.h
  • samples/legacy_samples/fp16_dev.cu
  • samples/legacy_samples/fp16_emu.cpp
  • samples/legacy_samples/fp8_sample.cpp
  • samples/legacy_samples/fp8_sample.h
  • samples/legacy_samples/fusion_sample.cpp
  • samples/legacy_samples/fusion_sample.h
  • samples/legacy_samples/helpers.cpp
  • samples/legacy_samples/norm_samples.cpp
  • samples/legacy_samples/norm_samples.h
  • samples/legacy_samples/test_list.cpp
  • samples/legacy_samples/utils/error_util.h
  • samples/legacy_samples/utils/fp16_dev.h
  • samples/legacy_samples/utils/fp16_emu.h
  • samples/legacy_samples/utils/helpers.h
  • samples/llama/100_download_weight.py
  • samples/llama/101_hf_llama_tieout.py
  • samples/llama/102_torch_llama_tieout.py
  • samples/llama/103_cudnn_llama_tieout.py
  • samples/llama/104_torch_llama_nvtx.py
  • samples/llama/105_cudnn_llama_nvtx.py
  • samples/llama/decode_nvtx_profile.py
  • samples/llm_coverage/test_gqa_b+h+s+d.py
  • samples/llm_coverage/test_linear+swish_1+bs+d.py
  • samples/llm_coverage/test_linear_1+bs+d.py
  • samples/llm_coverage/test_linear_b+s+d.py
  • samples/llm_coverage/test_linear_b+s+d_fp32compute.py
  • samples/llm_coverage/test_rmsnorm_b+s+d.py
  • samples/llm_coverage/test_rmsnorm_bs+d.py
  • samples/llm_coverage/test_rope_b+s+h+d.py
  • samples/llm_coverage/test_swiglu_1+bs+d.py
  • samples/llm_coverage/test_swiglu_layer_1+bs+d.py
  • setup.py
  • test/CMakeLists.txt
  • test/cpp/CMakeLists.txt
  • test/cpp/get_engine_and_knobs.cpp
  • test/cpp/pointwise_tests.cpp
  • test/cpp/serialize.cpp
  • test/cpp/tensor.cpp
  • test/cpp/validate.cpp
  • test/cpp/version.cpp
  • test/python/conftest.py
  • test/python/fe_api/bsa/__init__.py
  • test/python/fe_api/bsa/bsa_reference.py
  • test/python/fe_api/bsa/bsa_utils.py
  • test/python/fe_api/bsa/test_BSA_attention_backward.py
  • test/python/fe_api/bsa/test_BSA_attention_forward.py
  • test/python/fe_api/csa/test_CSA_compressor.py
  • test/python/fe_api/dsa/dsa_reference.py
  • test/python/fe_api/dsa/dsa_utils.py
  • test/python/fe_api/dsa/test_DSA_dense_indexer_backward.py
  • test/python/fe_api/dsa/test_DSA_dense_score_recompute.py
  • test/python/fe_api/dsa/test_DSA_indexer_backward.py
  • test/python/fe_api/dsa/test_DSA_indexer_forward.py
  • test/python/fe_api/dsa/test_DSA_indexer_top_k.py
  • test/python/fe_api/dsa/test_DSA_runtime.py
  • test/python/fe_api/dsa/test_DSA_sparse_attention_backward.py
  • test/python/fe_api/dsa/test_DSA_sparse_score_recompute.py
  • test/python/fe_api/gemm/test_gemm_amax.py
  • test/python/fe_api/gemm/test_gemm_amax_utils.py
  • test/python/fe_api/gemm/test_gemm_dsrelu.py
  • test/python/fe_api/gemm/test_gemm_dsrelu_utils.py
  • test/python/fe_api/gemm/test_gemm_proj_rope_mxfp8.py
  • test/python/fe_api/gemm/test_gemm_proj_rope_mxfp8_utils.py
  • test/python/fe_api/gemm/test_gemm_srelu.py
  • test/python/fe_api/gemm/test_gemm_srelu_utils.py
  • test/python/fe_api/gemm/test_gemm_swiglu.py
  • test/python/fe_api/gemm/test_gemm_swiglu_utils.py
  • test/python/fe_api/grouped_gemm/test_discrete_grouped_gemm_dswiglu.py
  • test/python/fe_api/grouped_gemm/test_discrete_grouped_gemm_dswiglu_utils.py
  • test/python/fe_api/grouped_gemm/test_discrete_grouped_gemm_swiglu.py
  • test/python/fe_api/grouped_gemm/test_discrete_grouped_gemm_swiglu_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_dglu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_dsrelu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_dsrelu_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_dswiglu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_dswiglu_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_glu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_glu_hadamard.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_quant.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_quant_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_srelu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_srelu_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_swiglu.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_swiglu_utils.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_wgrad.py
  • test/python/fe_api/grouped_gemm/test_grouped_gemm_wgrad_utils.py
  • test/python/fe_api/norm/test_rmsnorm_rht_amax.py
  • test/python/fe_api/nsa/nsa_reference.py
  • test/python/fe_api/nsa/nsa_utils.py
  • test/python/fe_api/nsa/test_NSA_compression_attention.py
  • test/python/fe_api/nsa/test_NSA_selection_attention.py
  • test/python/fe_api/nsa/test_NSA_swa.py
  • test/python/fe_api/nsa/test_NSA_topk_reduction.py
  • test/python/fe_api/sdpa/test_sdpa_bwd.py
  • test/python/fe_api/sdpa/test_sdpa_bwd_utils.py
  • test/python/fe_api/sdpa/test_sdpa_fwd.py
  • test/python/fe_api/sdpa/test_sdpa_fwd_utils.py
  • test/python/fe_api/test_api_base_logging.py
  • test/python/fe_api/test_fe_api_utils.py
  • test/python/fe_api/test_grouped_gemm_bf16.py
  • test/python/fe_api/test_grouped_gemm_bf16_utils.py
  • test/python/fe_api/test_grouped_gemm_dglu_bf16_utils.py
  • test/python/fe_api/test_grouped_gemm_glu_bf16_utils.py
  • test/python/fe_api/test_grouped_gemm_wgrad_bf16_utils.py
  • test/python/fe_api/test_rubin_kernel_dispatch.py
  • test/python/sdpa/blocked.py
  • test/python/sdpa/fp16.py
  • test/python/sdpa/fp16_ref.py
  • test/python/sdpa/fp8.py
  • test/python/sdpa/fp8_ref.py
  • test/python/sdpa/helpers.py
  • test/python/sdpa/mxfp8.py
  • test/python/sdpa/mxfp8_ref.py
  • test/python/sdpa/random_config.py
  • test/python/test_api_signature_parity.py
  • test/python/test_apply_rope.py
  • test/python/test_batchnorm.py
  • test/python/test_block_scale_quantize.py
  • test/python/test_block_scale_quantize_dynamic_shape.py
  • test/python/test_collect_env.py
  • test/python/test_conv_bias.py
  • test/python/test_conv_fprop.py
  • test/python/test_conv_fuzzer.py
  • test/python/test_conv_genstats.py
  • test/python/test_conv_reduction.py
  • test/python/test_cudnn_sdpa_op.py
  • test/python/test_deviceless_aot_compilation.py
  • test/python/test_engine_router.py
  • test/python/test_flexible_sdpa.py
  • test/python/test_flexible_sdpa_bprop.py
  • test/python/test_graph_native.py
  • test/python/test_instancenorm.py
  • test/python/test_kernel_cache.py
  • test/python/test_layernorm.py
  • test/python/test_low_precision_matmul.py
  • test/python/test_matmul_bias_relu.py
  • test/python/test_matmul_fuzzer.py
  • test/python/test_mhas.py
  • test/python/test_mhas_v2.py
  • test/python/test_moe_grouped_matmul.py
  • test/python/test_moe_grouped_matmul_op.py
  • test/python/test_native_backend_lowering.py
  • test/python/test_norm_fuzzer.py
  • test/python/test_oss_rope.py
  • test/python/test_rmsnorm.py
  • test/python/test_sdpa_chunked_prefill.py
  • test/python/test_sdpa_custom_features.py
  • test/python/test_sdpa_edge_cases.py
  • test/python/test_sdpa_fp32_rejected.py
  • test/python/test_sdpa_thd.py
  • test/python/test_sdpa_with_caching.py
  • test/python/test_silu_and_mul.py
  • test/python/test_slice.py
  • test/python/test_sm100_prefill_oss_engine.py
  • test/python/test_sm100_rms_norm_silu_graph_api.py
  • test/python/test_sm90_prefill_oss_engine.py
  • test/python/test_utils.py
  • test/python/test_wgrads.py
  • test/python/test_yarn_rope.py
  • tools/cudnn_repro/cudnn_repro/__init__.py
  • tools/cudnn_repro/cudnn_repro/__main__.py
  • tools/cudnn_repro/cudnn_repro/log_parser.py
  • tools/cudnn_repro/cudnn_repro/operations.py
  • tools/cudnn_repro/cudnn_repro/repro_command.py
  • tools/cudnn_repro/cudnn_repro/sdpa_bwd.py
  • tools/cudnn_repro/cudnn_repro/sdpa_fp8_bwd.py
  • tools/cudnn_repro/cudnn_repro/sdpa_fp8_fwd.py
  • tools/cudnn_repro/cudnn_repro/sdpa_fwd.py
  • tools/cudnn_repro/cudnn_repro/utils.py
  • tools/cudnn_repro/tests/__init__.py
  • tools/cudnn_repro/tests/helpers.py
  • tools/cudnn_repro/tests/test_cudnn_repro_bwd.py
  • tools/cudnn_repro/tests/test_cudnn_repro_cli.py
  • tools/cudnn_repro/tests/test_cudnn_repro_closed_loop.py
  • tools/cudnn_repro/tests/test_cudnn_repro_fp8.py
  • tools/cudnn_repro/tests/test_cudnn_repro_fp8_closed_loop.py
  • tools/cudnn_repro/tests/test_cudnn_repro_log_parser.py
  • tools/cudnn_repro/tests/test_cudnn_repro_mxfp8_closed_loop.py
  • tools/cudnn_repro/tests/test_cudnn_repro_schema.py
  • tools/cudnn_repro/tests/test_cudnn_repro_utils.py

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

Comment @coderabbitai help to get the list of available commands.

@Anerudhan Anerudhan mentioned this pull request Jul 23, 2026
18 tasks
@Anerudhan
Anerudhan force-pushed the relicense-apache-2.0 branch from 158ed02 to 9f5a3cb Compare July 23, 2026 19:18
@Anerudhan Anerudhan changed the title Relicense from MIT to Apache 2.0 Relicense to Apache-2.0 (dual-license; external-contributor files remain MIT) Jul 23, 2026
@Anerudhan
Anerudhan force-pushed the relicense-apache-2.0 branch 3 times, most recently from c85e9df to 1fdbcc4 Compare July 30, 2026 06:08
…MIT)

Rebased onto current develop (b950af1). Relicenses NVIDIA-authored code from
MIT to Apache-2.0 using the exact NVIDIA OSS SPDX header, keeping under MIT
only files that still carry contributions from external contributors whose
permission has not been established.

License assignment (per-file SPDX tag on every source file):
- Apache-2.0: 541 files
- MIT (50): surviving lines from external contributors, consent pending
- MIT (29): FlashAttention/QuACK-derived files carrying external authors'
  copyright

Cleared, freeing 16 files to Apache-2.0:
- Written consent on issue NVIDIA#431 (10): take-cheeze, fallintoplace, zianglih,
  JackRao123, zkyue, Hyaloid, haowen-han, junaire, szluyu99, dimitar-asenov.
- NVIDIA employment (2), commits under personal email addresses:
  HollowMan6, and hxbai (Hongxiao Bai, PR NVIDIA#410) -- frees
  .../indexer_top_k/indexer_top_k_varlen_util.py and
  test/python/fe_api/dsa/test_DSA_indexer_top_k.py.
Files touched by both a cleared and a still-pending contributor remain MIT.

PR NVIDIA#427 (CSA fused Compressor kernels ported from Megatron-LM, author @zkyue
who consented): its 13 files had no license header at all; they now carry the
NVIDIA Apache-2.0 SPDX header. Megatron-LM added to THIRD_PARTY_LICENSES.txt
as an NVIDIA Apache-2.0 provenance note.

Still pending (7): ConnorBaker, DrDirk, EmilienM, jyknight, sbcd90, valgur,
and co-author Benjamin Leff.

Every change is comment/header-only; all Python compiles.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@Anerudhan
Anerudhan force-pushed the relicense-apache-2.0 branch from 1fdbcc4 to 0c08141 Compare July 30, 2026 06:13
@Anerudhan Anerudhan self-assigned this Jul 30, 2026
@Anerudhan Anerudhan added cat-feature Requests for new functionality, APIs, examples, or behavior improvements. mod-frontend cuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers. orig-nv-eng Reported or requested by NVIDIA engineering. labels Jul 30, 2026
@Anerudhan Anerudhan added this to the Frontend 1.27.0 milestone Jul 30, 2026
@Anerudhan
Anerudhan marked this pull request as ready for review July 30, 2026 17:42
@Anerudhan
Anerudhan requested a review from vedaanta July 30, 2026 17:47
@Anerudhan
Anerudhan merged commit 57dc579 into NVIDIA:develop Jul 30, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cat-feature Requests for new functionality, APIs, examples, or behavior improvements. mod-frontend cuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers. orig-nv-eng Reported or requested by NVIDIA engineering.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants