Chengyi NieandChengyi Nie a572baff5e [Model Performance] Add Qwen3MoE tuned MoE configs for H200 (#35457) ...

Signed-off-by: Chengyi Nie <cnie@roblox.com>
Co-authored-by: Chengyi Nie <cnie@roblox.com>

2026-02-27 13:51:14 +08:00

..

attention

[Attn,KV-cache] Use per-head scales in the attention selector (#34281)

2026-02-24 09:02:43 -05:00

fla

[Model] Ring 2.5 (#35102)

2026-02-26 02:17:11 -08:00

fused_moe

[Model Performance] Add Qwen3MoE tuned MoE configs for H200 (#35457)

2026-02-27 13:51:14 +08:00

mamba

[Model] Ring 2.5 (#35102)

2026-02-26 02:17:11 -08:00

pooler

[Bugfix] Fix corner case of sparse embedding (#33886)

2026-02-05 02:51:22 -08:00

quantization

[Bugfix] Fix KV Scale loading for MLA Models (#35430)

2026-02-26 23:38:19 +00:00

rotary_embedding

[ROCm]: Enable customop and rope+kvcache fusion for AITER RoPE (#35180)

2026-02-25 04:36:40 +00:00

__init__.py

Change the name to vLLM (#150)

2023-06-17 03:07:40 -07:00

activation.py

[XPU][1/N] Deprecate ipex and switch to vllm-xpu-kernels for xpu platform (#33379)

2026-02-02 22:46:10 -08:00

attention_layer_base.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

batch_invariant.py

[Platform] Add current_platform.num_compute_units interface (#35042)

2026-02-24 22:22:49 -08:00

conv.py

[Doc] Add developer guide for CustomOp (#30886)

2026-01-09 16:21:11 +00:00

kda.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

layernorm.py

[Bug] correct out dtype of rms_norm_gated native path (#35369)

2026-02-27 05:19:51 +00:00

lightning_attn.py

[CI/Build][Kernel][AMD] Move extra dim to after load in _fwd_kv_parallel in lighting_attn.py (#29132)

2025-11-21 11:53:09 -05:00

linear.py

[Bugfix] [Qwen3.5]Fix Qwen3.5 FP8 quantization: tuple shard_id weight loading (#35289)

2026-02-26 18:26:15 +08:00

logits_processor.py

[Spec Decode] Reduce TP communication for speculative decoding draft token generation (#34049)

2026-02-22 14:59:16 -08:00

mla.py

[Model Bash] DeepSeek R1 BF16 Min Latency QKV A GEMM (0.5% E2E Speedup) (#34758)

2026-02-18 07:42:36 -08:00

resampler.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

sparse_attn_indexer.py

[Bugfix] Enforce DeepGEMM when using sparse_attn_indexer on CUDA (#34374)

2026-02-12 12:08:06 -08:00

utils.py

[Platform] Add current_platform.num_compute_units interface (#35042)

2026-02-24 22:22:49 -08:00

vocab_parallel_embedding.py

[Doc] Add developer guide for CustomOp (#30886)

2026-01-09 16:21:11 +00:00