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). @ 0859dc3bc0eae025439fc17fb267a99736581f97

AlpinDale fe01e2ded8 chore: move `device` keys to a constant (#1020)		2 mēneši atpakaļ
..
adapter_commons	2f61644f6e SPMD optimizations (#824)	3 mēneši atpakaļ
assets	411ac4f405 vlm: add support for Qwen2-VL model (#1015)	2 mēneši atpakaļ
attention	de341ffb00 fix: ensure multistep lookahead allocation is compatible with cugraph max capture (#1008)	2 mēneši atpakaļ
common	a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019)	2 mēneši atpakaļ
compilation	0e5cf7f840 tpu: avoid dynamo guard eval overhead (#949)	2 mēneši atpakaļ
distributed	61103b92d4 tpu: support single and multi-host TPUs on GKE and RayServe (#970)	2 mēneši atpakaļ
endpoints	411ac4f405 vlm: add support for Qwen2-VL model (#1015)	2 mēneši atpakaļ
engine	fe01e2ded8 chore: move `device` keys to a constant (#1020)	2 mēneši atpakaļ
executor	f2b6dc3872 cpu: add support for W8A8 quantization via compressed-tensor (#1017)	2 mēneši atpakaļ
inputs	908ff753a1 fix: phi_3.5_v loading (#896)	2 mēneši atpakaļ
kv_quant	8a71788372 Add OLMoE (#772)	4 mēneši atpakaļ
lora	bf4a4d8516 fix: do not register punica with torch if using older torch (#948)	2 mēneši atpakaļ
modeling	a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019)	2 mēneši atpakaļ
multimodal	411ac4f405 vlm: add support for Qwen2-VL model (#1015)	2 mēneši atpakaļ
platforms	f2b6dc3872 cpu: add support for W8A8 quantization via compressed-tensor (#1017)	2 mēneši atpakaļ
plugins	22a4cd4595 core: fix spec decode metrics and envs circular import (#889)	2 mēneši atpakaļ
processing	f561a54a43 core: fix async postprocessor in case of preemption (#1000)	2 mēneši atpakaļ
prompt_adapter	30d02d0747 chore: remove peft as a requirement (#1006)	2 mēneši atpakaļ
quantization	f2b6dc3872 cpu: add support for W8A8 quantization via compressed-tensor (#1017)	2 mēneši atpakaļ
server	22a4cd4595 core: fix spec decode metrics and envs circular import (#889)	2 mēneši atpakaļ
spec_decode	5c3b94de45 spec decode: move ops.advane_step to flash attention backend (#1005)	2 mēneši atpakaļ
transformers_utils	411ac4f405 vlm: add support for Qwen2-VL model (#1015)	2 mēneši atpakaļ
triton_utils	f1d0b77c92 [0.6.0] Release Candidate (#481)	6 mēneši atpakaļ
worker	a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019)	2 mēneši atpakaļ
__init__.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	6 mēneši atpakaļ
_core_ext.py	f1ea7711bd core: do not compile ScalarType for torch < 2.4.0 (#938)	2 mēneši atpakaļ
_custom_ops.py	a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019)	2 mēneši atpakaļ
_ipex_ops.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	6 mēneši atpakaļ
connections.py	c6c91edab7 ci: update & overhaul test units (#769)	3 mēneši atpakaļ
constants.py	2f61644f6e SPMD optimizations (#824)	3 mēneši atpakaļ
py.typed	1c988a48b2 fix logging and add py.typed	1 gadu atpakaļ
scalar_type.py	f1d0b77c92 [0.6.0] Release Candidate (#481)	6 mēneši atpakaļ
version.py	cbd51a208a ci: bump to 0.6.5 (#964)	2 mēneši atpakaļ