4 people 3a64faa1f2 Fix disagg PP MTP for GLM-5.2 (#39378) ...

Co-authored-by: Julien Lin <jullin@nvidia.com>
Co-authored-by: YAMY1234 <74099316+YAMY1234@users.noreply.github.com>
Co-authored-by: Yangmin Li <yangminl@nvidia.com>

2026-09-19 13:55:07 -07:00

..

model_runner_components

Fix disagg PP MTP for GLM-5.2 (#39378)

2026-09-19 13:55:07 -07:00

runner

[Logprob] Serve input-logprob temporaries from CUDA-graph-pool dead space (#40038)

2026-09-19 12:03:52 -07:00

runner_backend

[Runtime] Add decode CUDA graph hooks for eager logits processing (#40222)

2026-09-18 23:27:03 -07:00

runner_backend_utils

GLM-5.3-Flash support (#36507)

2026-09-06 02:27:59 -07:00

runner_utils

[Logprob] Borrow graph-pool memory for input logprob logits construction (#40007)

2026-09-18 13:18:01 -07:00

cpu_graph_runner.py

[CPU] Add fp8_per_tensor_scaled_mm_cpu kernel (#32618)

2026-09-16 12:31:05 +08:00

cuda_graph_buffer_registry.py

Pipeline parallelism x speculative decoding (EAGLE/MTP) compatibility (#30775)

2026-09-16 21:41:54 -07:00

cuda_graph_config.py

dsv4.1: remaining model and runtime integration (#38798)

2026-09-18 02:55:30 -07:00

encoder_swa_replay.py

dsv4.1: remaining model and runtime integration (#38798)

2026-09-18 02:55:30 -07:00

forward_batch_deepseek_mha_mixin.py

[AMD] Optimize Kimi-K3 Triton MLA prefill on gfx950 (#35770)

2026-09-04 17:44:30 -07:00

forward_batch_info.py

[Runtime] Add decode CUDA graph hooks for eager logits processing (#40222)

2026-09-18 23:27:03 -07:00

forward_context.py

[Refactor] Cuda Graph Runner/Backend Refactor (#23906)

2026-06-09 21:36:57 -07:00

graph_memory_usage.py

[Observability] Add startup, memory, and hybrid SWA diagnostics (#33375)

2026-08-04 12:50:09 -07:00

graph_shared_output.py

[Runtime] Add decode CUDA graph hooks for eager logits processing (#40222)

2026-09-18 23:27:03 -07:00

hook_manager.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

input_buffers.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

mindspore_runner.py

[Config] Round 6.4: the runtime reads the bags, not the record (#38049)

2026-09-06 21:41:46 -07:00

model_runner.py

Fix disagg PP MTP for GLM-5.2 (#39378)

2026-09-19 13:55:07 -07:00

pool_configurator.py

feat(kv-cache): support SM100 NVFP4 GenMHA and speculative decoding (#36340)

2026-09-18 14:50:46 -07:00

step_span_utils.py

Profiling Enhancements [2/3]: detailed execution step annotations (#24911)

2026-08-18 01:09:07 -07:00