 AlpinDale
|
766ea79b89
vlm: fix feature size calculation for llava-next models (#1079)
|
7 hours ago |
|
AlpinDale
|
7b6501bd05
tests: refactor model tests (#1078)
|
7 hours ago |
|
AlpinDale
|
f6df92bde0
fix: unexpected kwarg for the legacy API server (#1076)
|
15 hours ago |
|
AlpinDale
|
814c850d89
fix: validate `n` in the sampling params (#1075)
|
15 hours ago |
|
AlpinDale
|
6212072245
api: support LoRA lineage and base model metadata management (#1072)
|
3 days ago |
|
AlpinDale
|
d9d287a288
rocm: enable multi-step scheduling for rocm (#1071)
|
3 days ago |
|
AlpinDale
|
ec17b6c4d0
fix: Phi3.5 Mini and MoE LoRA inference (#1070)
|
3 days ago |
|
AlpinDale
|
acc0c727c8
vlm: add support for molmo vision model (#1069)
|
3 days ago |
|
AlpinDale
|
7dd001ec2d
build: guard against changes in cuda library name (#1068)
|
3 days ago |
|
AlpinDale
|
ca7028d5ca
sampler: simplify logits resort in _apply_top_k_top_p (#1067)
|
3 days ago |
|
AlpinDale
|
61aed092a5
rocm: add support for FP8 KV cache in the custom paged attention kkernels (#1066)
|
3 days ago |
|
AlpinDale
|
12b0059b47
api: enable MQAphroditeEngine for embedding models (#1065)
|
3 days ago |
|
AlpinDale
|
314fa7f7d9
fix: encoder-decoder models for beam search (#1064)
|
4 days ago |
|
AlpinDale
|
34cf9b74f0
api: non-zero exit code if MQ engine startup fails (#1063)
|
4 days ago |
|
AlpinDale
|
92cee435e2
rocm: add more quants, fix _scaled_mm call (#1062)
|
4 days ago |
 AlpinDale
|
b12d5c0507
readme: add attribution to Ruliad
|
4 days ago |
|
AlpinDale
|
f81e7d7010
distributed: bind only to 127.0.0.1 for local-only usage (#1061)
|
5 days ago |
|
AlpinDale
|
58aff3771d
core: support prompt logprobs in multi-step (#1060)
|
5 days ago |
|
AlpinDale
|
9a7d5514c4
feat: introduce MQAphroditeEngine (#1056)
|
6 days ago |
|
AlpinDale
|
0b5588de5c
fix: add missing logit index increment in sampling metadata prep (#1059)
|
6 days ago |
|
AlpinDale
|
525edc1283
build: fix compilation for causal_conv1d_fwd kernel signature (#1057)
|
1 week ago |
|
AlpinDale
|
9bdf8d5bfa
mamba: enable continuous batching for mamba kernels (#1055)
|
1 week ago |
|
AlpinDale
|
11f49b5341
fix: granite logit scale in logit computation (#1054)
|
1 week ago |
|
AlpinDale
|
1264e0b5d8
api: add mistral function calling format to all models loaded with "mistral" format (#1053)
|
1 week ago |
|
AlpinDale
|
b3f9ab3b72
quant: add tensor parallel support for bitsandbytes (#1052)
|
1 week ago |
|
AlpinDale
|
a985143768
core: add cuda graph support for encoder-decoder models (#1051)
|
1 week ago |
|
AlpinDale
|
239a8cae25
torch.compile: register all-reduce operations as custom ops (#1050)
|
1 week ago |
|
AlpinDale
|
4593a3b306
chore: remove dead code from triton sampling kernels (#1049)
|
1 week ago |
|
AlpinDale
|
8976805f90
kernel: asymmetric AQ AZP quantization kernels (#1048)
|
1 week ago |
|
AlpinDale
|
638c08d9dc
fix: clean shutdown issues (#1047)
|
1 week ago |
