Rain Jiang 0ffd0a3995 Nsa trtllm mla sparse fp8 support with Deepseek v3.2 NVFP4 (#18389)

2026-02-16 09:29:54 +08:00

..

cpu_graph_runner.py

feat(SpecEagleV2): add standalone_worker_v2 (#12625)

2025-12-30 17:55:04 +08:00

cuda_graph_runner.py

refactor context parallel state (#17213)

2026-02-13 23:18:17 +08:00

forward_batch_deepseek_mha_mixin.py

[2/n] deepseek_v2.py Refactor: Migrate MHA forward method in deepseek_v2.py (#16817)

2026-01-17 09:36:25 +08:00

forward_batch_info.py

refactor context parallel state (#17213)

2026-02-13 23:18:17 +08:00

hook_manager.py

Rename: --hooks to --forward-hooks (#13994)

2025-11-26 22:26:28 +08:00

input_buffers.py

[Qwen3-next] support mamba radix cache for overlap scheduler (#14792)

2025-12-14 18:54:16 -08:00

mindspore_runner.py

[NPU]mindspore model support moe (#15363)

2026-02-02 17:52:49 +08:00

model_runner_kv_cache_mixin.py

Nsa trtllm mla sparse fp8 support with Deepseek v3.2 NVFP4 (#18389)

2026-02-16 09:29:54 +08:00

model_runner.py

Model: Support IBM Granite (Dense/Mamba + MoE) (#18040)

2026-02-15 11:24:41 +08:00

piecewise_cuda_graph_runner.py

[PCG] GPT OSS Triton Kernel Support (#18405)

2026-02-12 09:23:55 +08:00