.gitignore
.pre-commit-config.yaml
AGENTS.md
CLAUDE.md
CODE_OF_CONDUCT.md
CONTRIBUTING.md
GEMINI.md
LICENSE
README.md
README_zh.md
mkdocs.yml
pyproject.toml
setup.cfg
.claude/skills/add-new-pipeline/SKILL.md
.claude/skills/framework-analysis/SKILL.md
.claude/skills/optimize-pipeline/SKILL.md
.claude/skills/profile-pipeline/SKILL.md
.github/PULL_REQUEST_TEMPLATE.md
.github/ISSUE_TEMPLATE/bug_report.yml
.github/ISSUE_TEMPLATE/config.yml
.github/ISSUE_TEMPLATE/feature_request.yml
.github/ISSUE_TEMPLATE/roadmap.yml
.github/workflows/docs.yml
.github/workflows/lint.yml
.github/workflows/release-pypi.yml
.github/workflows/release-testpypi.yml
.github/workflows/test.yml
assets/telefuser_logo.png
benchmarks/feature_cache/README.md
benchmarks/feature_cache/analyze_residual_taylor_error.py
benchmarks/feature_cache/wan21_1_3b_ada_taylor_cache.py
benchmarks/kernel/bench_rmsnorm.py
benchmarks/kernel/bench_rotary.py
benchmarks/kernel/bench_scale_shift.py
benchmarks/kernel/run_benchmarks.sh
docker/.gitkeep
docs/requirements.txt
docs/en/adding_new_example.md
docs/en/adding_new_model.md
docs/en/adding_new_stage.md
docs/en/attention.md
docs/en/configuration.md
docs/en/feature_cache.md
docs/en/hash_config_management.md
docs/en/index.md
docs/en/latent_cache.md
docs/en/logging.md
docs/en/metrics.md
docs/en/model_loading.md
docs/en/offload.md
docs/en/ops.md
docs/en/parallel.md
docs/en/profiler.md
docs/en/release_pypi.md
docs/en/service.md
docs/en/service_metadata.md
docs/en/stream_server.md
docs/en/testing.md
docs/en/torch_compile_compatibility.md
docs/stylesheets/extra.css
docs/zh/adding_new_example.md
docs/zh/adding_new_model.md
docs/zh/adding_new_stage.md
docs/zh/adf_scoring_guide.md
docs/zh/attention.md
docs/zh/configuration.md
docs/zh/feature_cache.md
docs/zh/hash_config_management.md
docs/zh/index.md
docs/zh/latent_cache.md
docs/zh/logging.md
docs/zh/metrics.md
docs/zh/model_loading.md
docs/zh/offload.md
docs/zh/ops.md
docs/zh/parallel.md
docs/zh/profiler.md
docs/zh/release_pypi.md
docs/zh/service.md
docs/zh/service_metadata.md
docs/zh/stream_server.md
docs/zh/testing.md
docs/zh/torch_compile_compatibility.md
examples/README.md
examples/example_config.yaml
examples/run_examples.py
examples/data/1.png
examples/data/1.wav
examples/data/101235-video-720_0.png
examples/data/dag.mp4
examples/data/edit2511input.png
examples/flashvsr/README.md
examples/flashvsr/flashvsr_stream.py
examples/flux2_klein/README.md
examples/flux2_klein/flux2_klein_text_to_image_h100.py
examples/flux2_klein/flux2_klein_text_to_image_official.py
examples/hunyuan_video/README.md
examples/hunyuan_video/hunyuan_video_i2v.py
examples/hunyuan_video/hunyuan_video_i2v_cache_calibrate.py
examples/hunyuan_video/hunyuan_video_t2v.py
examples/hunyuan_video/hunyuan_video_t2v_cache_calibrate.py
examples/lingbot/stream_lingbot_world_fast.py
examples/liveact/liveact_s2v_h100.py
examples/longcat_video/README.md
examples/longcat_video/longcat_image_to_video.py
examples/longcat_video/longcat_text_to_video.py
examples/longcat_video/longcat_text_to_video_refine.py
examples/longcat_video/longcat_video_continue.py
examples/longcat_video/longcat_video_unify.py
examples/ltx_video/README.md
examples/ltx_video/ltx23_22b_image_to_video_two_stage_h100.py
examples/qwen_image/README.md
examples/qwen_image/qwen_image_cache_calibrate.py
examples/qwen_image/qwen_image_edit_plus_cache_calibrate.py
examples/qwen_image/qwen_image_edit_plus_h100.py
examples/qwen_image/qwen_image_edit_plus_official.py
examples/qwen_image/qwen_image_t2i_h100.py
examples/qwen_image/qwen_image_t2i_lightning_fp8_h100.py
examples/qwen_image/qwen_image_t2i_lora_h100.py
examples/qwen_image/qwen_image_t2i_official.py
examples/stream_server/stream_arrow_overlay.py
examples/stream_server/stream_video_replay.py
examples/stream_server/webrtc_arrow_overlay_demo.py
examples/stream_server/webrtc_bidirectional_demo.py
examples/stream_server/webrtc_client_demo.py
examples/wan_video/README.md
examples/wan_video/async_wan22_14b_image_to_video_distill_h100.py
examples/wan_video/wan21_14b_image_to_video_h100.py
examples/wan_video/wan21_14b_image_to_video_lora_h100.py
examples/wan_video/wan21_14b_text_to_video_h100.py
examples/wan_video/wan21_1_3b_text_to_video_ada_taylor_cache.py
examples/wan_video/wan21_1_3b_text_to_video_cache_calibrate.py
examples/wan_video/wan21_1_3b_text_to_video_h100.py
examples/wan_video/wan21_1_3b_text_to_video_hf.py
examples/wan_video/wan21_1_3b_text_to_video_radial.py
examples/wan_video/wan22_14b_first_last_frame_to_video_h100.py
examples/wan_video/wan22_14b_image_to_video_cache_calibrate.py
examples/wan_video/wan22_14b_image_to_video_distill_fp8_h100.py
examples/wan_video/wan22_14b_image_to_video_distill_h100.py
examples/wan_video/wan22_14b_image_to_video_h100.py
examples/wan_video/wan22_14b_image_to_video_h100_ray.py
examples/wan_video/wan22_14b_image_to_video_lora_h100.py
examples/wan_video/wan22_14b_image_to_video_mix_h100.py
examples/wan_video/wan22_14b_text_to_video_h100.py
examples/wan_video/wan22_14b_text_to_video_service.py
examples/wan_video/wan22_i2v_5b.py
examples/wan_video/wan22_t2v_5b.py
examples/z_image/README.md
examples/z_image/z_image_turbo_official.py
examples/z_image/z_image_turbo_t2i_h100.py
scripts/build_latent_dataset.py
scripts/build_telefuser_dist.sh
scripts/publish_telefuser_pypi.sh
scripts/run_ci_tests.sh
telefuser/__init__.py
telefuser/_logo.py
telefuser/_version.py
telefuser/service_types.py
telefuser.egg-info/PKG-INFO
telefuser.egg-info/SOURCES.txt
telefuser.egg-info/dependency_links.txt
telefuser.egg-info/entry_points.txt
telefuser.egg-info/requires.txt
telefuser.egg-info/scm_file_list.json
telefuser.egg-info/scm_version.json
telefuser.egg-info/top_level.txt
telefuser/cache/__init__.py
telefuser/cache/kv_cache.py
telefuser/cache_mem/__init__.py
telefuser/cache_mem/cache_types.py
telefuser/cache_mem/config.py
telefuser/cache_mem/connection.py
telefuser/cache_mem/encoders.py
telefuser/cache_mem/latent_cache.py
telefuser/cache_mem/log_monitor.py
telefuser/cache_mem/metadata.py
telefuser/cache_mem/strategies.py
telefuser/cache_mem/encoding/__init__.py
telefuser/cache_mem/encoding/interfaces.py
telefuser/cache_mem/src/__init__.py
telefuser/cache_mem/src/models/__init__.py
telefuser/cache_mem/src/models/qwen3_vl_embedding.py
telefuser/cache_mem/src/models/qwen3_vl_reranker.py
telefuser/cache_mem/state/__init__.py
telefuser/cache_mem/state/interfaces.py
telefuser/cache_mem/storage/__init__.py
telefuser/cache_mem/storage/fluxon.py
telefuser/cache_mem/storage/interfaces.py
telefuser/cache_mem/storage/local_file.py
telefuser/cache_mem/storage/memory.py
telefuser/cache_mem/vector_store/__init__.py
telefuser/cache_mem/vector_store/faiss.py
telefuser/cache_mem/vector_store/interfaces.py
telefuser/cache_mem/vector_store/qdrant.py
telefuser/client/__init__.py
telefuser/client/tf_client.py
telefuser/client/openai/__init__.py
telefuser/client/openai/client.py
telefuser/client/openai/images.py
telefuser/client/openai/videos.py
telefuser/core/__init__.py
telefuser/core/base_model.py
telefuser/core/base_pipeline.py
telefuser/core/base_stage.py
telefuser/core/config.py
telefuser/core/config_serializer.py
telefuser/core/model_registry.py
telefuser/core/module_manager.py
telefuser/distributed/__init__.py
telefuser/distributed/device_mesh.py
telefuser/distributed/fsdp.py
telefuser/distributed/parallel_shard.py
telefuser/distributed/pp_comm.py
telefuser/distributed/ring.py
telefuser/distributed/tp_parallelize.py
telefuser/distributed/ulysses_comm.py
telefuser/entrypoints/__init__.py
telefuser/entrypoints/cli/main.py
telefuser/feature_cache/__init__.py
telefuser/feature_cache/base.py
telefuser/feature_cache/ada_taylor_cache/__init__.py
telefuser/feature_cache/ada_taylor_cache/ada_taylor_cache.py
telefuser/feature_cache/ada_taylor_cache/params/HunyuanVideo15-I2V-480P.json
telefuser/feature_cache/ada_taylor_cache/params/HunyuanVideo15-T2V-480P.json
telefuser/feature_cache/ada_taylor_cache/params/Qwen-Image-2512.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_1-FL2V-14B-720P.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_1-I2V-14B-480P.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_1-I2V-14B-720P.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_1-T2V-14B.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_1-T2V-1_3B.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_2-FL2V-A14B.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_2-I2V-A14B-Camera.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_2-I2V-A14B.json
telefuser/feature_cache/ada_taylor_cache/params/Wan2_2-T2V-A14B.json
telefuser/kernel/__init__.py
telefuser/kernel/triton/__init__.py
telefuser/kernel/triton/merge_attn_states.py
telefuser/kernel/triton/norm.py
telefuser/kernel/triton/quant.py
telefuser/kernel/triton/quant_per_block.py
telefuser/kernel/triton/rotary.py
telefuser/kernel/triton/scale_shift.py
telefuser/kernel/triton/sparse_int8_attn.py
telefuser/metrics/__init__.py
telefuser/metrics/collector.py
telefuser/metrics/config.py
telefuser/metrics/exporters.py
telefuser/metrics/registry.py
telefuser/metrics/service_metrics.py
telefuser/metrics/stage_metrics.py
telefuser/models/TCDecoder.py
telefuser/models/__init__.py
telefuser/models/flashvsr_dit.py
telefuser/models/flux2_dit.py
telefuser/models/hunyuan_video_byt5.py
telefuser/models/hunyuan_video_dit.py
telefuser/models/hunyuan_video_image_encoder.py
telefuser/models/hunyuan_video_text_encoder.py
telefuser/models/hunyuan_video_upsampler.py
telefuser/models/hunyuan_video_vae.py
telefuser/models/lingbot_world_fast_dit.py
telefuser/models/liveact_dit.py
telefuser/models/longcat_video_dit.py
telefuser/models/ltx_audio_vae.py
telefuser/models/ltx_dit.py
telefuser/models/ltx_gemma_text_encoder.py
telefuser/models/ltx_upsampler.py
telefuser/models/ltx_video_vae.py
telefuser/models/qwen_image_dit.py
telefuser/models/qwen_image_text_encoder.py
telefuser/models/qwen_image_vae.py
telefuser/models/realesrgan.py
telefuser/models/rift_hdv3.py
telefuser/models/t5_tokenizer.py
telefuser/models/video_projector.py
telefuser/models/wan22_video_vae.py
telefuser/models/wan_video_dit.py
telefuser/models/wan_video_image_encoder.py
telefuser/models/wan_video_text_encoder.py
telefuser/models/wan_video_vae.py
telefuser/models/wav2vec2.py
telefuser/models/xlm_roberta.py
telefuser/models/z_image_dit.py
telefuser/models/z_image_text_encoder.py
telefuser/offload/__init__.py
telefuser/offload/async_offload.py
telefuser/offload/model_offload.py
telefuser/offload/sequential_offload.py
telefuser/ops/__init__.py
telefuser/ops/activations.py
telefuser/ops/base.py
telefuser/ops/custom_op.py
telefuser/ops/ffn.py
telefuser/ops/fp8_gemm.py
telefuser/ops/normalization.py
telefuser/ops/quantized_linear.py
telefuser/ops/rotary.py
telefuser/ops/attention/__init__.py
telefuser/ops/attention/attention_impl.py
telefuser/ops/attention/backends.py
telefuser/ops/attention/bsa.py
telefuser/ops/attention/local_sparse_attn.py
telefuser/ops/attention/sparse_patterns.py
telefuser/ops/attention/sparse_sage.py
telefuser/orchestrator/__init__.py
telefuser/orchestrator/artifact_save_stage.py
telefuser/orchestrator/pipeline_orchestrator.py
telefuser/orchestrator/stage_wrapper.py
telefuser/pipelines/__init__.py
telefuser/pipelines/common/realesrgan_upscale.py
telefuser/pipelines/common/rift_vfi.py
telefuser/pipelines/flashvsr/__init__.py
telefuser/pipelines/flashvsr/dit_denoising.py
telefuser/pipelines/flashvsr/flashvsr_stream.py
telefuser/pipelines/flashvsr/vae.py
telefuser/pipelines/flux2_klein/__init__.py
telefuser/pipelines/flux2_klein/dit_denoising.py
telefuser/pipelines/flux2_klein/pipeline.py
telefuser/pipelines/flux2_klein/text_encoding.py
telefuser/pipelines/flux2_klein/vae.py
telefuser/pipelines/hunyuan_video_1_5/__init__.py
telefuser/pipelines/hunyuan_video_1_5/dit_denoising.py
telefuser/pipelines/hunyuan_video_1_5/image_encoding.py
telefuser/pipelines/hunyuan_video_1_5/pipeline.py
telefuser/pipelines/hunyuan_video_1_5/sr_dit_denoising.py
telefuser/pipelines/hunyuan_video_1_5/text_encoding.py
telefuser/pipelines/hunyuan_video_1_5/upsampler.py
telefuser/pipelines/hunyuan_video_1_5/vae.py
telefuser/pipelines/lingbot_world_fast/__init__.py
telefuser/pipelines/lingbot_world_fast/control.py
telefuser/pipelines/lingbot_world_fast/denoising.py
telefuser/pipelines/lingbot_world_fast/pipeline.py
telefuser/pipelines/lingbot_world_fast/service.py
telefuser/pipelines/lingbot_world_fast/session.py
telefuser/pipelines/liveact/__init__.py
telefuser/pipelines/liveact/audio_encoding.py
telefuser/pipelines/liveact/denoising.py
telefuser/pipelines/liveact/pipeline.py
telefuser/pipelines/longcat_video/__init__.py
telefuser/pipelines/longcat_video/dit_denoising.py
telefuser/pipelines/longcat_video/longcat_video.py
telefuser/pipelines/longcat_video/refine_denoise.py
telefuser/pipelines/longcat_video/text_encoding.py
telefuser/pipelines/ltx_video/__init__.py
telefuser/pipelines/ltx_video/dit_denoising.py
telefuser/pipelines/ltx_video/gemma_text_encoding.py
telefuser/pipelines/ltx_video/ltx23_video.py
telefuser/pipelines/ltx_video/upsampler.py
telefuser/pipelines/ltx_video/vae.py
telefuser/pipelines/qwen_image/__init__.py
telefuser/pipelines/qwen_image/dit_denoising.py
telefuser/pipelines/qwen_image/qwen_image.py
telefuser/pipelines/qwen_image/qwen_image_edit.py
telefuser/pipelines/qwen_image/text_encoding.py
telefuser/pipelines/qwen_image/vae.py
telefuser/pipelines/wan_video/__init__.py
telefuser/pipelines/wan_video/async_wan22_video.py
telefuser/pipelines/wan_video/clip_encoding.py
telefuser/pipelines/wan_video/latent_data_utils.py
telefuser/pipelines/wan_video/moe_dit_denoising.py
telefuser/pipelines/wan_video/single_dit_denoising.py
telefuser/pipelines/wan_video/text_encoding.py
telefuser/pipelines/wan_video/ti2v_denoising.py
telefuser/pipelines/wan_video/vae.py
telefuser/pipelines/wan_video/wan21_video.py
telefuser/pipelines/wan_video/wan22_ti2v.py
telefuser/pipelines/wan_video/wan22_video.py
telefuser/pipelines/z_image/__init__.py
telefuser/pipelines/z_image/dit_denoising.py
telefuser/pipelines/z_image/text_encoding.py
telefuser/pipelines/z_image/vae.py
telefuser/pipelines/z_image/z_image.py
telefuser/platforms/__init__.py
telefuser/platforms/cpu.py
telefuser/platforms/cuda.py
telefuser/platforms/interface.py
telefuser/platforms/npu.py
telefuser/platforms/rocm.py
telefuser/schedulers/__init__.py
telefuser/schedulers/flow_match.py
telefuser/schedulers/flow_match_discrete.py
telefuser/schedulers/lcm.py
telefuser/schedulers/unipc.py
telefuser/service/__init__.py
telefuser/service/main.py
telefuser/service/api/__init__.py
telefuser/service/api/api_server.py
telefuser/service/api/middleware.py
telefuser/service/api/schema.py
telefuser/service/api/stream_schema.py
telefuser/service/api/task_contract_runtime.py
telefuser/service/api/utils.py
telefuser/service/api/openai/__init__.py
telefuser/service/api/openai/adapter.py
telefuser/service/api/openai/image_routes.py
telefuser/service/api/openai/protocol.py
telefuser/service/api/openai/video_routes.py
telefuser/service/api/routers/__init__.py
telefuser/service/api/routers/files.py
telefuser/service/api/routers/service.py
telefuser/service/api/routers/stream.py
telefuser/service/api/routers/tasks.py
telefuser/service/api/routers/webrtc.py
telefuser/service/cache/__init__.py
telefuser/service/cache/cache_factory.py
telefuser/service/cache/cache_service.py
telefuser/service/core/__init__.py
telefuser/service/core/config.py
telefuser/service/core/container.py
telefuser/service/core/contract_templates.py
telefuser/service/core/file_service.py
telefuser/service/core/pipeline_contract.py
telefuser/service/core/pipeline_loader.py
telefuser/service/core/pipeline_pool.py
telefuser/service/core/pipeline_runner.py
telefuser/service/core/pipeline_service.py
telefuser/service/core/replica_worker.py
telefuser/service/core/stream_pipeline_service.py
telefuser/service/core/task_manager.py
telefuser/service/core/task_processor.py
telefuser/service/core/task_service.py
telefuser/service/media/__init__.py
telefuser/service/media/media_base.py
telefuser/service/security/__init__.py
telefuser/service/security/security_validator.py
telefuser/service/webrtc/__init__.py
telefuser/service/webrtc/chunk_router.py
telefuser/service/webrtc/session_manager.py
telefuser/service/webrtc/track.py
telefuser/utils/__init__.py
telefuser/utils/audio.py
telefuser/utils/func.py
telefuser/utils/hf_model_analyzer.py
telefuser/utils/hf_model_utils.py
telefuser/utils/hf_utils.py
telefuser/utils/logging.py
telefuser/utils/lora_loader.py
telefuser/utils/lora_network.py
telefuser/utils/memory_snapshot.py
telefuser/utils/model_weight.py
telefuser/utils/profiler.py
telefuser/utils/stage_bench_harness.py
telefuser/utils/system.py
telefuser/utils/torch_compile.py
telefuser/utils/utils.py
telefuser/utils/video.py
telefuser/worker/__init__.py
telefuser/worker/native_worker.py
telefuser/worker/parallel_worker.py
telefuser/worker/ray_worker.py
tests/__init__.py
tests/conftest.py
tests/integration/conftest.py
tests/integration/test_pp_forward_consistency.py
tests/integration/test_service_api.py
tests/integration/test_stream_api.py
tests/integration/test_webrtc_api.py
tests/server/README.md
tests/server/__init__.py
tests/server/conftest.py
tests/server/run_integration_test.py
tests/server/run_test_server.py
tests/server/test_middleware.py
tests/server/test_openai_api.py
tests/server/test_openai_client.py
tests/server/test_server.py
tests/server/client/__init__.py
tests/server/client/test_client.py
tests/server/fixtures/__init__.py
tests/server/pipeline/__init__.py
tests/server/pipeline/fake_t2v_pipeline.py
tests/unit/__init__.py
tests/unit/test_metrics.py
tests/unit/cache_mem/__init__.py
tests/unit/cache_mem/test_concurrency.py
tests/unit/cache_mem/test_metadata.py
tests/unit/cache_mem/test_storage.py
tests/unit/cache_mem/test_types_and_config.py
tests/unit/core/test_base_pipeline.py
tests/unit/core/test_config.py
tests/unit/core/test_config_serializer.py
tests/unit/distributed/__init__.py
tests/unit/distributed/test_device_mesh.py
tests/unit/distributed/test_parallel_shard.py
tests/unit/distributed/test_pp_comm.py
tests/unit/distributed/test_ulysses_comm.py
tests/unit/feature_cache/__init__.py
tests/unit/feature_cache/test_feature_cache.py
tests/unit/kernel/__init__.py
tests/unit/kernel/test_rmsnorm.py
tests/unit/kernel/test_rotary.py
tests/unit/kernel/test_scale_shift.py
tests/unit/models/__init__.py
tests/unit/offload/__init__.py
tests/unit/offload/test_async_offload.py
tests/unit/offload/test_sequential_offload.py
tests/unit/openai/__init__.py
tests/unit/openai/test_adapter.py
tests/unit/openai/test_client.py
tests/unit/openai/test_image_routes.py
tests/unit/openai/test_integration_server.py
tests/unit/openai/test_protocol.py
tests/unit/openai/test_video_routes.py
tests/unit/ops/test_activations.py
tests/unit/ops/test_long_context_attention.py
tests/unit/ops/test_normalization.py
tests/unit/ops/test_parallel_shard_attention.py
tests/unit/pipelines/__init__.py
tests/unit/pipelines/lingbot_world_fast/test_control_alignment.py
tests/unit/pipelines/wan_video/__init__.py
tests/unit/pipelines/wan_video/test_latent_data_utils.py
tests/unit/quantize/__init__.py
tests/unit/quantize/test_quantized_linear.py
tests/unit/schedulers/test_flow_match.py
tests/unit/service/__init__.py
tests/unit/service/test_config.py
tests/unit/service/test_pipeline_contract.py
tests/unit/service/test_pipeline_pool.py
tests/unit/service/test_schema.py
tests/unit/service/test_task_contract_runtime.py
tests/unit/service/test_task_routes.py
tests/unit/service/test_task_runtime.py
tests/unit/utils/test_profiler_flags.py
tests/unit/utils/test_utils.py
tests/unit/worker/test_parallel_worker.py
tf-kernel/.clang-format
tf-kernel/.gitignore
tf-kernel/.pre-commit-config.yaml
tf-kernel/AGENTS.md
tf-kernel/CLAUDE.md
tf-kernel/CMakeLists.txt
tf-kernel/CODE_OF_CONDUCT.md
tf-kernel/CONTRIBUTING.md
tf-kernel/Dockerfile
tf-kernel/GEMINI.md
tf-kernel/LICENSE
tf-kernel/Makefile
tf-kernel/README.md
tf-kernel/README_zh.md
tf-kernel/THIRDPARTYNOTICES.txt
tf-kernel/analyze_whl_kernel_sizes.py
tf-kernel/build.sh
tf-kernel/kernel-runner-setup.sh
tf-kernel/pyproject.toml
tf-kernel/pytest.ini
tf-kernel/rename_wheels.sh
tf-kernel/.github/PULL_REQUEST_TEMPLATE.md
tf-kernel/.github/ISSUE_TEMPLATE/bug_report.yml
tf-kernel/.github/ISSUE_TEMPLATE/feature_request.yml
tf-kernel/.github/workflows/docs.yml
tf-kernel/.github/workflows/lint.yml
tf-kernel/benchmark/bench_activation.py
tf-kernel/benchmark/bench_fp4_gemm.py
tf-kernel/benchmark/bench_fp8_gemm.py
tf-kernel/benchmark/bench_int8_gemm.py
tf-kernel/benchmark/bench_nvfp4_scaled_gemm.py
tf-kernel/benchmark/bench_per_tensor_quant_fp8.py
tf-kernel/benchmark/bench_rmsnorm.py
tf-kernel/benchmark/bench_sageattn2.py
tf-kernel/cmake/detect_gpu.cmake
tf-kernel/cmake/utils.cmake
tf-kernel/csrc/common_extension.cc
tf-kernel/csrc/block_sparse_attn/block_sparse_attn_api.h
tf-kernel/csrc/block_sparse_attn/flash_api.cpp
tf-kernel/csrc/block_sparse_attn/src/alibi.h
tf-kernel/csrc/block_sparse_attn/src/block_info.h
tf-kernel/csrc/block_sparse_attn/src/flash.h
tf-kernel/csrc/block_sparse_attn/src/flash_blockmask.h
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim128_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim128_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim128_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim128_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim32_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim32_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim32_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim32_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim64_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim64_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim64_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_block_hdim64_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_kernel.h
tf-kernel/csrc/block_sparse_attn/src/flash_bwd_launch_template.h
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim128_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim128_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim128_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim128_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim32_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim32_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim32_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim32_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim64_bf16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim64_bf16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim64_fp16_causal_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_block_hdim64_fp16_sm80.cu
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_kernel.h
tf-kernel/csrc/block_sparse_attn/src/flash_fwd_launch_template.h
tf-kernel/csrc/block_sparse_attn/src/hardware_info.h
tf-kernel/csrc/block_sparse_attn/src/kernel_traits.h
tf-kernel/csrc/block_sparse_attn/src/kernel_traits_sm90.h
tf-kernel/csrc/block_sparse_attn/src/namespace_config.h
tf-kernel/csrc/block_sparse_attn/src/philox.cuh
tf-kernel/csrc/block_sparse_attn/src/softmax.h
tf-kernel/csrc/block_sparse_attn/src/static_switch.h
tf-kernel/csrc/block_sparse_attn/src/utils.h
tf-kernel/csrc/cutlass_extensions/common.hpp
tf-kernel/csrc/cutlass_extensions/detail/collective/mixed_input_utils.hpp
tf-kernel/csrc/cutlass_extensions/epilogue/epilogue_per_row_per_col_scale.h
tf-kernel/csrc/cutlass_extensions/gemm/cutlass_gemm_caller.cuh
tf-kernel/csrc/cutlass_extensions/gemm/dispatch_policy.hpp
tf-kernel/csrc/cutlass_extensions/gemm/fp8_blockwise_gemm_sm90_dispatch.cuh
tf-kernel/csrc/cutlass_extensions/gemm/gemm_universal_base_compat.h
tf-kernel/csrc/cutlass_extensions/gemm/gemm_with_epilogue_visitor.h
tf-kernel/csrc/cutlass_extensions/gemm/collective/collective_builder_mixed_input.hpp
tf-kernel/csrc/cutlass_extensions/gemm/collective/collective_mma_array_mixed_input.hpp
tf-kernel/csrc/cutlass_extensions/gemm/collective/sm90_mma_array_tma_gmma_rs_warpspecialized_mixed_input_.hpp
tf-kernel/csrc/cutlass_extensions/gemm/collective/builders/sm90_gmma_builder_mixed_input.inl
tf-kernel/csrc/elementwise/activation.cu
tf-kernel/csrc/elementwise/cast.cu
tf-kernel/csrc/elementwise/concat_mla.cu
tf-kernel/csrc/elementwise/copy.cu
tf-kernel/csrc/elementwise/fused_add_rms_norm_kernel.cu
tf-kernel/csrc/elementwise/pos_enc.cu
tf-kernel/csrc/elementwise/pos_enc.cuh
tf-kernel/csrc/elementwise/rope.cu
tf-kernel/csrc/elementwise/topk.cu
tf-kernel/csrc/elementwise/utils.cuh
tf-kernel/csrc/expert_specialization/es_fp8_blockwise.cu
tf-kernel/csrc/expert_specialization/es_fp8_blockwise_functor.cuh
tf-kernel/csrc/expert_specialization/es_fp8_blockwise_launcher.cuh
tf-kernel/csrc/expert_specialization/es_fp8_blockwise_traits.cuh
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled.cu
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled_functor.cuh
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled_group_quant.cu
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled_group_quant.cuh
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled_launcher.cuh
tf-kernel/csrc/expert_specialization/es_sm100_mxfp8_blockscaled_traits.cuh
tf-kernel/csrc/gemm/bmm_fp8.cu
tf-kernel/csrc/gemm/fp8_blockwise_gemm_kernel.cu
tf-kernel/csrc/gemm/fp8_gemm_kernel.cu
tf-kernel/csrc/gemm/int8_gemm_kernel.cu
tf-kernel/csrc/gemm/math.hpp
tf-kernel/csrc/gemm/nvfp4_expert_quant.cu
tf-kernel/csrc/gemm/nvfp4_quant.cuh
tf-kernel/csrc/gemm/nvfp4_quant_entry.cu
tf-kernel/csrc/gemm/nvfp4_quant_kernels.cu
tf-kernel/csrc/gemm/nvfp4_scaled_mm_entry.cu
tf-kernel/csrc/gemm/nvfp4_scaled_mm_kernels.cu
tf-kernel/csrc/gemm/per_tensor_quant_fp8.cu
tf-kernel/csrc/gemm/per_token_group_quant_8bit.cu
tf-kernel/csrc/gemm/per_token_group_quant_8bit_v2.cu
tf-kernel/csrc/gemm/per_token_quant_fp8.cu
tf-kernel/csrc/memory/store.cu
tf-kernel/csrc/memory/weak_ref_tensor.cpp
tf-kernel/csrc/sageattn2/cp_async.cuh
tf-kernel/csrc/sageattn2/dispatch_utils.h
tf-kernel/csrc/sageattn2/math.cuh
tf-kernel/csrc/sageattn2/mma.cuh
tf-kernel/csrc/sageattn2/numeric_conversion.cuh
tf-kernel/csrc/sageattn2/permuted_smem.cuh
tf-kernel/csrc/sageattn2/reduction_utils.cuh
tf-kernel/csrc/sageattn2/utils.cuh
tf-kernel/csrc/sageattn2/wgmma.cuh
tf-kernel/csrc/sageattn2/fused/fused.cu
tf-kernel/csrc/sageattn2/fused/fused.h
tf-kernel/csrc/sageattn2/qattn/attn_cuda_sm80.h
tf-kernel/csrc/sageattn2/qattn/attn_cuda_sm89.h
tf-kernel/csrc/sageattn2/qattn/attn_cuda_sm90.h
tf-kernel/csrc/sageattn2/qattn/attn_utils.cuh
tf-kernel/csrc/sageattn2/qattn/qk_int_sv_f16_cuda_sm80.cu
tf-kernel/csrc/sageattn2/qattn/qk_int_sv_f8_cuda_sm89.cuh
tf-kernel/csrc/sageattn2/qattn/qk_int_sv_f8_cuda_sm90.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f16_attn_inst_buf.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f16_fuse_v_scale_attn_inst_buf.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f32_attn.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f32_attn_inst_buf.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f32_fuse_v_scale_attn.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f32_fuse_v_scale_attn_inst_buf.cu
tf-kernel/csrc/sageattn2/qattn/sm89_qk_int8_sv_f8_accum_f32_fuse_v_scale_fuse_v_mean_attn.cu
tf-kernel/csrc/sageattn3/__init__.py
tf-kernel/csrc/sageattn3/api.h
tf-kernel/csrc/sageattn3/api.py
tf-kernel/csrc/sageattn3/setup.py
tf-kernel/csrc/sageattn3/blackwell/api.cu
tf-kernel/csrc/sageattn3/blackwell/block_info.h
tf-kernel/csrc/sageattn3/blackwell/blockscaled_layout.h
tf-kernel/csrc/sageattn3/blackwell/cute_extension.h
tf-kernel/csrc/sageattn3/blackwell/epilogue_tma_ws.h
tf-kernel/csrc/sageattn3/blackwell/kernel_traits.h
tf-kernel/csrc/sageattn3/blackwell/kernel_ws.h
tf-kernel/csrc/sageattn3/blackwell/launch.h
tf-kernel/csrc/sageattn3/blackwell/mainloop_tma_ws.h
tf-kernel/csrc/sageattn3/blackwell/named_barrier.h
tf-kernel/csrc/sageattn3/blackwell/params.h
tf-kernel/csrc/sageattn3/blackwell/softmax_fused.h
tf-kernel/csrc/sageattn3/blackwell/static_switch.h
tf-kernel/csrc/sageattn3/blackwell/tile_scheduler.h
tf-kernel/csrc/sageattn3/blackwell/utils.h
tf-kernel/csrc/sageattn3/quantization/cuda_utils.h
tf-kernel/csrc/sageattn3/quantization/fp4_quantization_4d.cu
tf-kernel/docs/Makefile
tf-kernel/docs/source/conf.py
tf-kernel/docs/source/contributing.rst
tf-kernel/docs/source/development.rst
tf-kernel/docs/source/index.rst
tf-kernel/docs/source/installation.rst
tf-kernel/docs/source/quickstart.rst
tf-kernel/docs/source/api/attention.rst
tf-kernel/docs/source/api/elementwise.rst
tf-kernel/docs/source/api/gemm.rst
tf-kernel/docs/source/api/index.rst
tf-kernel/include/pytorch_extension_utils_rocm.h
tf-kernel/include/scalar_type.hpp
tf-kernel/include/tf_kernel_ops.h
tf-kernel/include/tf_kernel_torch_shim.h
tf-kernel/include/utils.h
tf-kernel/scripts/check_wheel_symbols.py
tf-kernel/tests/conftest.py
tf-kernel/tests/test_activation.py
tf-kernel/tests/test_block_sparse_attn.py
tf-kernel/tests/test_block_sparse_attn_simple.py
tf-kernel/tests/test_bmm_fp8.py
tf-kernel/tests/test_copy.py
tf-kernel/tests/test_fp4_gemm.py
tf-kernel/tests/test_fp4_quantize.py
tf-kernel/tests/test_fp8_blockwise_gemm.py
tf-kernel/tests/test_fp8_gemm.py
tf-kernel/tests/test_int8_gemm.py
tf-kernel/tests/test_norm.py
tf-kernel/tests/test_sageattn2.py
tf-kernel/tests/test_sageattn3.py
tf-kernel/tests/test_torch_defaults_reset.py
tf-kernel/tests/utils.py
tf-kernel/tf_kernel/__init__.py
tf-kernel/tf_kernel/block_sparse_attn.py
tf-kernel/tf_kernel/elementwise.py
tf-kernel/tf_kernel/gemm.py
tf-kernel/tf_kernel/load_utils.py
tf-kernel/tf_kernel/memory.py
tf-kernel/tf_kernel/sageattn2.py
tf-kernel/tf_kernel/sageattn3.py
tf-kernel/tf_kernel/sampling.py
tf-kernel/tf_kernel/scalar_type.py
tf-kernel/tf_kernel/test_utils.py
tf-kernel/tf_kernel/utils.py
tf-kernel/tf_kernel/testing/__init__.py
tf-kernel/tf_kernel/testing/rotary_embedding.py
tf-kernel/tf_kernel/triton/__init__.py
tf-kernel/tf_kernel/triton/attn_qk_int8_block_varlen.py
tf-kernel/tf_kernel/triton/attn_qk_int8_per_block.py
tf-kernel/tf_kernel/triton/attn_qk_int8_per_block_causal.py
tf-kernel/tf_kernel/triton/attn_qk_int8_per_block_causal_varlen.py
tf-kernel/tf_kernel/triton/quant_per_block.py
tf-kernel/tf_kernel/triton/quant_per_block_varlen.py
tf-kernel/tf_kernel/triton/quant_per_thread.py
tools/convert/README.md
tools/convert/converter.py
tools/convert/lora_extractor.py
tools/convert/lora_loader.py
tools/convert/quant.py
tools/convert/register.py
tools/deploy/README.md
tools/deploy/docker_monitor.py
tools/deploy/multi_device_communication_test.py
tools/deploy/show_stat.py
tools/viewer/README.md
tools/viewer/weight_viewer.py
webui/__init__.py
webui/stream_app.py