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). @ bf88c8567e43522fbc91238417b06914b860cce4

AlpinDale bf88c8567e feat: mamba model support (#674)		пре 6 месеци
..
__init__.py	9d81716bfd [v0.5.3] Release Candidate (#388)	пре 10 месеци
abstract.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
blocksparse_attn.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
flash_attn.py	7df7b8ca53 optimization: reduce end-to-end overhead from python obj allocation (#666)	пре 6 месеци
flashinfer.py	67ee885293 fix: flashinfer outputs (#657)	пре 6 месеци
ipex_attn.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
openvino.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
pallas.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
placeholder_attn.py	bf88c8567e feat: mamba model support (#674)	пре 6 месеци
rocm_flash_attn.py	e200775863 feat: enable using fp8 kv and prefix caching with chunked prefill (#668)	пре 6 месеци
torch_sdpa.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	пре 6 месеци
utils.py	7df7b8ca53 optimization: reduce end-to-end overhead from python obj allocation (#666)	пре 6 месеци
xformers.py	e200775863 feat: enable using fp8 kv and prefix caching with chunked prefill (#668)	пре 6 месеци