david/aphrodite-engine: PygmalionAI's large-scale inference engine pygmalion.chat It is designed to serve as the inference endpoint for the PygmalionAI website, and to allow serving the Pygmalion models to a large number of users with blazing fast speeds (thanks to vLLM's Paged Attention). @ rejection_sampling_kernels

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