TensorRT-LLMs/tensorrt_llm/_torch/speculative
Perkz Zheng 35c5e4f1c5
feat: add CGA reduction fmha kernels on Blackwell. (#3763)
* update cubins

Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>

* add trtllm-gen kernels for eagle3 and also kernels with cga-reduction

Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>

* address the comments

Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>

---------

Signed-off-by: Perkz Zheng <67892460+PerkzZheng@users.noreply.github.com>
2025-04-29 10:43:54 +08:00
..
__init__.py Add initial EAGLE-3 implementation (#3035) 2025-03-29 22:31:24 +08:00
eagle3.py [fix] Fix flashinfer + speculation issues (#3686) 2025-04-28 14:34:22 -04:00
interface.py feat: add CGA reduction fmha kernels on Blackwell. (#3763) 2025-04-29 10:43:54 +08:00
mtp.py feat: return logits in PyTorch flow (#3221) 2025-04-24 16:56:03 -07:00
utils.py Support CUDA graphs for EAGLE3 (#3176) 2025-04-17 04:53:50 +08:00