Commits · 5ce89059f8149eaf313c63e9ded4199670cd74bb · OpenDAS / text-generation-inference

12 Jun, 2023 1 commit
- feat(server): pre-allocate past key values for flash causal LM (#412) · 5ce89059
  OlivierDehaene authored Jun 12, 2023
  
  5ce89059
14 Apr, 2023 1 commit
- fea(dockerfile): better layer caching (#159) · 53ee09c0
  OlivierDehaene authored Apr 14, 2023
  
  53ee09c0
09 Apr, 2023 1 commit
- feat(docker): improve flash_attention caching (#160) · 1883d8ec
  OlivierDehaene authored Apr 09, 2023
  
  1883d8ec