Xin Yang 1f19d8f899

[Perf] Set split_k to 1 for triton_kernels (#30528) ...

Signed-off-by: Xin Yang <xyangx@amazon.com>

2025-12-12 14:07:57 -05:00

..

layers

[Perf] Set split_k to 1 for triton_kernels (#30528)

2025-12-12 14:07:57 -05:00

model_loader

[Model][Quantization] Restore MoE + GGUF models support (incl. Qwen3 MoE) by allowing Sideload Parameters (#30116)

2025-12-09 05:30:05 +00:00

models

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

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

warmup

[BugFix] Fix AttributeError: 'MergedColumnParallelLinear' object has no attribute 'weight_scale' (#30399)

2025-12-10 07:59:23 -08:00

__init__.py

…

custom_op.py

…

parameter.py

…

utils.py

[Quantization] FP8 Weight Reloading for Quantized RL Rollout (#28480)

2025-12-09 13:54:32 -08:00