AlpinDale
|
5f851e45e5
ruff
|
10 months ago |
AlpinDale
|
41f5af0426
add python nccl wrapper, remove cupy
|
10 months ago |
AlpinDale
|
56af2e0e85
KVCache type in llava
|
10 months ago |
AlpinDale
|
73890c29e2
ipv6 fix
|
10 months ago |
AlpinDale
|
3f5ce50c19
add stop_reason
|
10 months ago |
AlpinDale
|
2efee6bcc6
optimize logprob ranks
|
10 months ago |
AlpinDale
|
278a6478f4
yapf
|
10 months ago |
AlpinDale
|
7b9c08afae
vision model support
|
10 months ago |
AlpinDale
|
9d673dc5bd
don't output two stop strings in api
|
10 months ago |
AlpinDale
|
777b6f6d51
add logprob ranks
|
10 months ago |
AlpinDale
|
0c4ead5e9f
min_tokens
|
10 months ago |
AlpinDale
|
0f1399c135
feat: attention refactor part 2
|
10 months ago |
AlpinDale
|
d1786645a3
fix formatting
|
10 months ago |
AlpinDale
|
c8a91b0b96
rope: get_device() -> device
|
10 months ago |
AlpinDale
|
0299dd41f0
fix query shape in moe models
|
10 months ago |
AlpinDale
|
c97fc0c701
fix tied embeddings in falcon
|
10 months ago |
AlpinDale
|
609710b940
LockFile -> SoftLockFile
|
10 months ago |
AlpinDale
|
eed70dff76
improve detokenization performance; improve logprobs
|
10 months ago |
AlpinDale
|
ac0595574b
fix logprobs serializer warnings
|
10 months ago |
AlpinDale
|
b738554558
add reorder scheduler policy
|
10 months ago |
AlpinDale
|
1ba9ff78cd
add scheduler delay factor
|
10 months ago |
AlpinDale
|
29eaded422
fix and re-enable custom all-reduce
|
10 months ago |
AlpinDale
|
2319b411ce
refactor: neuron support
|
10 months ago |
AlpinDale
|
3b19fa02ea
do not remove duplicate params for qwen2
|
10 months ago |
AlpinDale
|
e72165fcc6
fix quants for gemma
|
10 months ago |
AlpinDale
|
c9cb00c2a1
add warning for mismatch in vocab size
|
10 months ago |
AlpinDale
|
fbf169de06
fix pydantic serializer warning
|
10 months ago |
AlpinDale
|
b8725d0ea1
fix query shape in logits processor
|
10 months ago |
AlpinDale
|
4415f0d1f1
conflicts with _is_neuron()
|
10 months ago |
AlpinDale
|
ace9bcd53f
fix gptq for cohere
|
10 months ago |