Wentao Ye 3778673ea8

[Feat] Refactor for parallel_config in FusedMoEModularKernel (#30282) ...

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>

2025-12-15 04:21:36 +00:00

..

assets

…

attention

[CustomOp][MM] Extract MMEncoderAttention as CustomOp and replace the backend of QwenVisionAttention with it. (#30125)

2025-12-15 11:13:32 +08:00

benchmarks

[Bugfix][benchmarks] Fix input token calculation for rerank benchmark metrics (#30596)

2025-12-14 14:57:15 +00:00

compilation

[Bugfix] Fix fusion for VL models (#30244)

2025-12-14 21:22:37 +08:00

config

[Log] Skip piecewise cudagraph warn when using full cudagraph (#30657)

2025-12-15 02:49:45 +00:00

device_allocator

…

distributed

[NIXL][BUG FIX] Fix a bug for PD with host_buffer after merging 29665 (#30420)

2025-12-14 15:38:28 +00:00

engine

[Bug] Fix attention_backend arg string parsing (#30534)

2025-12-12 08:40:50 -07:00

entrypoints

[Bugfix] Drop empty tool_calls lists to keep assistant replies in chat template (#30648)

2025-12-15 04:21:12 +00:00

inputs

Improve parse_raw_prompt test cases for invalid input .v2 (#30512)

2025-12-14 11:18:41 +08:00

logging_utils

[Misc] Colorize logs (#29017)

2025-11-19 19:26:04 -05:00

lora

[Misc] Fix safetensors import for safe_open (#30300)

2025-12-09 06:50:06 +00:00

model_executor

[Feat] Refactor for parallel_config in FusedMoEModularKernel (#30282)

2025-12-15 04:21:36 +00:00

multimodal

additional protection for CVE-2025-62164 (#30649)

2025-12-15 03:07:10 +00:00

platforms

[CustomOp][MM] Extract MMEncoderAttention as CustomOp and replace the backend of QwenVisionAttention with it. (#30125)

2025-12-15 11:13:32 +08:00

plugins

[Frontend] Resettle pooling entrypoints (#29634)

2025-12-01 15:30:43 +08:00

profiler

[Logs] Optimize startup logs 4 (#29903)

2025-12-13 16:12:53 -05:00

ray

[Small] Capture AttributeError when checking ray dependency. (#29024)

2025-11-20 22:54:10 -08:00

reasoning

[Chore] Adjust tokenizer import to avoid circular imports (#30601)

2025-12-13 04:42:39 -08:00

third_party

…

tokenizers

[Bugfix] Fix deepseek_v32 tokenizer_mode (#30658)

2025-12-15 04:20:31 +00:00

transformers_utils

[Model][Quantization] Override HF defaults to GGUF ones (incl. Qwen3 MoE) (#30118)

2025-12-14 15:01:42 +00:00

triton_utils

[mypy] Enable type checking for more directories (#29674)

2025-11-28 08:39:27 -08:00

usage

…

utils

Revert "[Fix]Load kv-cache dtype from hf_quant_config.json automatically" (#30653)

2025-12-14 19:33:41 -05:00

v1

[Bugfix] Fix deepseek_v32 tokenizer_mode (#30658)

2025-12-15 04:20:31 +00:00

vllm_flash_attn

…

__init__.py

…

_aiter_ops.py

[Rocm][torch.compile] Adding layernorm + fp8 block quant and silu + fp8 block quant for Aiter (#25693)

2025-12-09 22:39:26 +00:00

_bc_linter.py

…

_custom_ops.py

[Bugfix] awq_gemm: fix argument order swap (#30364)

2025-12-14 18:15:37 +08:00

_ipex_ops.py

…

beam_search.py

…

collect_env.py

…

connections.py

…

env_override.py

…

envs.py

[Attention] Use sparse prefill kernel for fp8 kv-cache in DeepSeek-v3.2 (#27532)

2025-12-12 05:57:47 -08:00

forward_context.py

[BugFix] Fix assert batch_descriptor.num_tokens == num_tokens_padded (#30173)

2025-12-09 10:36:12 -05:00

logger.py

Make the httpx logger less annoying when Transformers v5 is installed (#30480)

2025-12-11 15:44:56 +00:00

logits_process.py

[Misc] Refactor tokenizer interface (#29693)

2025-11-29 04:02:21 -08:00

logprobs.py

…

outputs.py

…

pooling_params.py

add add_truncate_prompt_tokens in repr for PoolingParams (#29683)

2025-11-28 08:41:05 -08:00

py.typed

…

sampling_params.py

[Misc] Refactor tokenizer interface (#29693)

2025-11-29 04:02:21 -08:00

scalar_type.py

…

scripts.py

…

sequence.py

…

tasks.py

…

tracing.py

…

version.py

…