kpham-sgl bc4aaab6a1 [Spec][Ngram] 2/N: Rename branch length to max trie depth (#21181) ...

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-22 23:35:25 -07:00

..

cpp_ngram

[Spec][Ngram] 2/N: Rename branch length to max trie depth (#21181)

2026-03-22 23:35:25 -07:00

base_spec_worker.py

Fix flush cache API for spec v2 (#11918)

2025-10-21 23:01:16 +08:00

draft_utils.py

Support using different attention backend for draft decoding. (#14843)

2025-12-19 08:49:11 +08:00

eagle_draft_cuda_graph_runner.py

Feature/support longcat flash lite (#17838)

2026-03-09 23:00:11 +08:00

eagle_draft_extend_cuda_graph_runner.py

Feature/support longcat flash lite (#17838)

2026-03-09 23:00:11 +08:00

eagle_info_v2.py

[AMD] Support speculative decoding v2 for aiter backend on ROCm/HIP (#17450)

2026-03-11 17:01:01 -07:00

eagle_info.py

perf: avoid unnecessary gpu-cpu sync in eagle_info (#20266)

2026-03-19 19:37:29 -07:00

eagle_utils.py

[Feature] Spec-Overlap supporting DP-ATTN; PD-Disaggregation; npugraph mode (#12443)

2025-11-15 21:51:07 +08:00

eagle_worker_v2.py

[AMD] Support speculative decoding v2 for aiter backend on ROCm/HIP (#17450)

2026-03-11 17:01:01 -07:00

eagle_worker.py

[Bugfix] Fix MTP prefill cuda graph logging (#20279)

2026-03-17 16:36:52 -07:00

multi_layer_eagle_draft_extend_cuda_graph_runner.py

Feature/support longcat flash lite (#17838)

2026-03-09 23:00:11 +08:00

multi_layer_eagle_utils.py

Adjust wrong mtp meaning introduce by mimo (#15632)

2025-12-23 02:06:46 +08:00

multi_layer_eagle_worker_v2.py

[Core] Replace server_args mutation hack with explicit MemoryPoolConfig for draft worker init (#20183)

2026-03-09 11:45:54 -07:00

multi_layer_eagle_worker.py

[Bugfix] Fix MTP prefill cuda graph logging (#20279)

2026-03-17 16:36:52 -07:00

ngram_info.py

Add spec_accept_histogram request statistic (#18332)

2026-02-12 21:09:21 +08:00

ngram_worker.py

[Spec][Ngram] 2/N: Rename branch length to max trie depth (#21181)

2026-03-22 23:35:25 -07:00

spec_info.py

Tiny simplify draft worker init. (#16446)

2026-01-09 11:20:29 +08:00

spec_utils.py

fix: add ModelScope cache lookup and speculative path support (#20098)

2026-03-07 22:23:16 -08:00

standalone_worker_v2.py

[Core] Replace server_args mutation hack with explicit MemoryPoolConfig for draft worker init (#20183)

2026-03-09 11:45:54 -07:00

standalone_worker.py

[Core] Replace server_args mutation hack with explicit MemoryPoolConfig for draft worker init (#20183)

2026-03-09 11:45:54 -07:00