AlpinDale
|
4e71bd1d12
feat: add PagedAttention V2 kernels (#76)
|
1 year ago |
50h100a
|
d0eadd4dbd
Added `min_tokens` and reimplemented `ignore_eos` using a new logit processor (#70)
|
1 year ago |
AlpinDale
|
04a27c6aeb
fix: revert mirostat v2 (#79)
|
1 year ago |
AlpinDale
|
9c353a0e02
fix: unnecessary import
|
1 year ago |
AlpinDale
|
ce5e2332ea
fix: launch AWQ kernels on the current CUDAStream (#75)
|
1 year ago |
AlpinDale
|
561773dec8
fix: hopefully fixes github actions
|
1 year ago |
AlpinDale
|
2c1d6a8cf2
fix: fast tokenizer latency reduction
|
1 year ago |
AlpinDale
|
28db67fd78
fix: mistral support
|
1 year ago |
AlpinDale
|
d8f04f29a9
readme: update sampling params
|
1 year ago |
Stefan Gligorijevic
|
5acc27adeb
chore: fix parameter validation on ooba endpoint
|
1 year ago |
Stefan Gligorijevic
|
93daff0384
chore: delete leftover debug prints
|
1 year ago |
Stefan Gligorijevic
|
5dbd262033
feat: Mirostat v2 (#69)
|
1 year ago |
AlpinDale
|
f393dc2af1
fix: broken GPTQ layer
|
1 year ago |
AlpinDale
|
3bf6197afb
fix: prompt processing delay introduced by #66 (#71)
|
1 year ago |
AlpinDale
|
9df91fe863
bump version to 0.3.6
|
1 year ago |
AlpinDale
|
2b42a1ada2
bump the version to 0.3.5
|
1 year ago |
AlpinDale
|
c55c8f7bd8
update readme
|
1 year ago |
AlpinDale
|
380206038e
fix: change the timing of logit sorting (#66)
|
1 year ago |
AlpinDale
|
bdad759503
feat: YaRN context window extension support (#67)
|
1 year ago |
AlpinDale
|
f04588203e
feat: mistral AWQ support and file blacklisting
|
1 year ago |
AlpinDale
|
7572e1dd59
overflow in AWQ GEMM kernel
|
1 year ago |
AlpinDale
|
c1fa7e8567
chore: fix datatype check (#65)
|
1 year ago |
AlpinDale
|
a6a4220fa6
feat: refactor megatron and quants (#57)
|
1 year ago |
AlpinDale
|
9a9e59b871
update readme with new instructions
|
1 year ago |
g4rg
|
16bf6b61a3
fix: requests stalling in KAI non-streaming endpoint (#46)
|
1 year ago |
LitreallyNone
|
b526a7b3bc
Update requirements.txt (#58)
|
1 year ago |
AlpinDale
|
2e70a6d5ed
chore: allow the user to specify install method (#56)
|
1 year ago |
official-elinas
|
46e472062a
chore: make NVCC work for different versions (#55)
|
1 year ago |
AlpinDale
|
6682ede3de
fix: clean up API servers
|
1 year ago |
henk717
|
0b2b62fe96
Micromamba Runtime (#54)
|
1 year ago |