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). @ control-vectors

AlpinDale fa15bad2ea chore: minor AMD fixes		hace 5 meses
..
__init__.py	9d81716bfd [v0.5.3] Release Candidate (#388)	hace 8 meses
abstract.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
blocksparse_attn.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
flash_attn.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
flashinfer.py	8adc496a2a fix: use paged attention for bloc swapping/copying in flashinfer	hace 5 meses
ipex_attn.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
openvino.py	0886c361f4 feat: OpenVINO CPU backend (#576)	hace 5 meses
pallas.py	9d7beaa5b9 chore: separate kv_scale into k_scale and v_scale	hace 5 meses
rocm_flash_attn.py	fa15bad2ea chore: minor AMD fixes	hace 5 meses
torch_sdpa.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
utils.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses
xformers.py	22305c91e9 refactor _prepare_model_input_tensor and attn metadata builder for most backends	hace 5 meses