AlpinDale 5c3b94de45 spec decode: move ops.advane_step to flash attention backend (#1005) 2 주 전
..
__init__.py 9d81716bfd [v0.5.3] Release Candidate (#388) 8 달 전
batch_expansion.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
draft_model_runner.py 5c3b94de45 spec decode: move ops.advane_step to flash attention backend (#1005) 2 주 전
interfaces.py f1d0b77c92 [0.6.0] Release Candidate (#481) 4 달 전
medusa_worker.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
metrics.py 2f61644f6e SPMD optimizations (#824) 1 개월 전
mlp_speculator_worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 주 전
multi_step_worker.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
ngram_worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 주 전
proposer_worker_base.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
smaller_tp_proposer_worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 주 전
spec_decode_worker.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
target_model_runner.py 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 주 전
top1_proposer.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 주 전
util.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 주 전