Lianmin Zheng 8e66fbecee Improve DP attention (#4390) ...

Co-authored-by: dhou-xai <dhou@x.ai>
Co-authored-by: SangBin Cho <rkooo567@gmail.com>

2025-03-13 08:23:56 -07:00

..

attention

model: Support Janus-pro (#3203)

2025-03-12 11:02:11 -07:00

moe

[moe] fix: correct the cache size in the last chunk (#3679)

2025-03-12 22:22:13 -07:00

quantization

Revert "[XPU][CPU] Enable the native path of DeepSeek" (#4367)

2025-03-12 23:45:52 -07:00

activation.py

fix: support gelu_new activation function in gpt2 (#3712)

2025-03-04 04:09:52 -08:00

dp_attention.py

Improve DP attention (#4390)

2025-03-13 08:23:56 -07:00

layernorm.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

linear.py

Apply sgl w8a8 fp8 kernel (#3148)

2025-03-09 00:03:32 -08:00

logits_processor.py

Improve DP attention (#4390)

2025-03-13 08:23:56 -07:00

parameter.py

Apply sgl w8a8 fp8 kernel (#3148)

2025-03-09 00:03:32 -08:00

pooler.py

Rename InputMetadata -> ForwardBatch (#1543)

2024-09-30 02:41:11 -07:00

radix_attention.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

rotary_embedding.py

Revert "[XPU][CPU] Enable the native path of DeepSeek" (#4367)

2025-03-12 23:45:52 -07:00

sampler.py

Move output processing logic from scheduler.py into a separate file (#4354)

2025-03-12 16:21:49 -07:00

torchao_utils.py

Support loading of larger models with on-the-fly quantization (#3061)

2025-01-22 21:33:17 -08:00

vocab_parallel_embedding.py

Clean up fp8 support (#4230)

2025-03-09 21:46:35 -07:00