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). @ 9bbc75d2e3389428a587af92f60e94482898ae15

AlpinDale 9bbc75d2e3 wip		6 mesiacov pred
..
__init__.py	5884e0b904 add bitnetforcausallm support	6 mesiacov pred
arctic.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
baichuan.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
bitnet.py	9bbc75d2e3 wip	6 mesiacov pred
bloom.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
chatglm.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
commandr.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
dbrx.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
decilm.py	50b7c13db0 refactor: attention selector (#552)	6 mesiacov pred
deepseek.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
falcon.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
gemma.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
gpt2.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
gpt_bigcode.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
gpt_j.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
gpt_neox.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
internlm2.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
jais.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
llama.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred
llama_embedding.py	50b7c13db0 refactor: attention selector (#552)	6 mesiacov pred
llava.py	24a2d9c2c8 minor llava refactoring	6 mesiacov pred
minicpm.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
mixtral.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred
mixtral_quant.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred
mpt.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
olmo.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
opt.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
orion.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
phi.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
phi3_small.py	696f2cd59c add phi3_small support with blocksparse attention	6 mesiacov pred
qwen.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
qwen2.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred
qwen2_moe.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
stablelm.py	656459fd84 make fp8_e4m3 work on nvidia	6 mesiacov pred
starcoder2.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred
vlm_base.py	f970f3f3fb add base class for VLMs	6 mesiacov pred
xverse.py	ac79d115b3 add guards for prefix caching, fp8, chunked, etc	6 mesiacov pred