 AlpinDale
|
c2d77b1822
chore: logging refactor (#302)
|
hai 10 meses |
|
AlpinDale
|
9810daa699
feat: INT8 KV Cache (#298)
|
hai 10 meses |
|
AlpinDale
|
ac82b67f75
feat: naive context shift and various QoL changes (#289)
|
hai 10 meses |
|
AlpinDale
|
4d04ade9ef
feat: fine-grained seeds (#279)
|
hai 11 meses |
|
AlpinDale
|
697c06c4f5
fix: LoRA support for mixtral (#276)
|
hai 11 meses |
|
AlpinDale
|
4b80b42362
fix: memory leaks due to nccl cuda graphs (#275)
|
hai 11 meses |
|
AlpinDale
|
ea0f57b233
feat: allow further support for non-cuda devices (#247)
|
hai 11 meses |
|
AlpinDale
|
1a94ccf3cf
fix: prefix cache fail with lora (#239)
|
hai 11 meses |
|
AlpinDale
|
c3a221eb02
feat: GGUF, QuIP#, and Marlin support (#228)
|
hai 1 ano |
|
AlpinDale
|
31c95011a6
feat: FP8 E5M2 KV Cache (#226)
|
hai 1 ano |
|
AlpinDale
|
641bb0f6e9
feat: add custom allreduce kernels (#224)
|
hai 1 ano |
|
AlpinDale
|
c0aac15421
feat: S-LoRA support (#222)
|
hai 1 ano |
|
AlpinDale
|
8fa608aeb7
feat: replace Ray with NCCL for control plane comms (#221)
|
hai 1 ano |
|
AlpinDale
|
d54791aaa8
feat: reduce sampler overhead by making it less blocking (#198)
|
hai 1 ano |
|
AlpinDale
|
7d91e9e0f2
feat: CUDA graphs (#172)
|
hai 1 ano |
 g4rg
|
2aab3da9bd
chore: fix Python 3.8+ compatibility (#170)
|
hai 1 ano |
|
AlpinDale
|
ae57df0f44
fix: sliding window for mistral/mixtral (#163)
|
hai 1 ano |
|
AlpinDale
|
653da510d1
chore: rewrite InputMetadata (#143)
|
hai 1 ano |
