AlpinDale b3f6eeb1d2 vlm: increase the default `max_num_batched_tokens` for multimodal models (#973) 2 هفته پیش
..
__init__.py 04b53d2db5 chore: add initializer files 1 سال پیش
cache_engine.py bf88c8567e feat: mamba model support (#674) 4 ماه پیش
cpu_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
cpu_worker.py 22a4cd4595 core: fix spec decode metrics and envs circular import (#889) 3 هفته پیش
embedding_model_runner.py 89a2c6dee1 chore: refactor `MultiModalConfig` initialization and profiling (#745) 3 ماه پیش
enc_dec_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
model_runner_base.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
multi_step_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
multi_step_worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
neuron_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
neuron_worker.py 0c6d90dade neuron: add support for tensor parallelism (#923) 3 هفته پیش
openvino_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
openvino_worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
tpu_model_runner.py 5bec8fbb1b tpu: add support for async postprocessing (#968) 2 هفته پیش
tpu_worker.py ea59784f59 tpu: remove torch._dynamo.reset() (#952) 2 هفته پیش
utils.py b3f6eeb1d2 vlm: increase the default `max_num_batched_tokens` for multimodal models (#973) 2 هفته پیش
worker.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
worker_base.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
xpu_model_runner.py 0dfa6b60ec core: support logprobs with multi-step scheduling (#963) 2 هفته پیش
xpu_worker.py 15cb8d5c26 xpu: support pipeline parallel (#932) 2 هفته پیش