 AlpinDale
|
50c2434267
move megatron to a top-level directory
|
il y a 9 mois |
 AlpinDale
|
4d33ce60da
feat: Triton flash attention backend for ROCm (#407)
|
il y a 9 mois |
|
AlpinDale
|
41beab5dc1
add exllamav2 tensor paralell, fused MoE for GPTQ/AWQ
|
il y a 9 mois |
|
AlpinDale
|
0f1399c135
feat: attention refactor part 2
|
il y a 9 mois |
|
AlpinDale
|
d1786645a3
fix formatting
|
il y a 9 mois |
|
AlpinDale
|
688d56993a
add logit scale for command-r
|
il y a 9 mois |
|
AlpinDale
|
f1ea36a445
add some imports
|
il y a 9 mois |
|
AlpinDale
|
f8dfac6372
chore: attention refactor and upstream sync apr01 (#365)
|
il y a 9 mois |
|
AlpinDale
|
da223153c6
feat&fix: cohere support and missing GPU blocks (#333)
|
il y a 10 mois |
|
AlpinDale
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
il y a 10 mois |
|
AlpinDale
|
e31c6f0b45
feat: refactor modeling logic and support more models (#274)
|
il y a 11 mois |
|
AlpinDale
|
842912d022
feat: on-the-fly gguf conversion (#250)
|
il y a 11 mois |
|
AlpinDale
|
c3a221eb02
feat: GGUF, QuIP#, and Marlin support (#228)
|
il y a 1 an |
|
AlpinDale
|
8fa608aeb7
feat: replace Ray with NCCL for control plane comms (#221)
|
il y a 1 an |
|
AlpinDale
|
b9b295d74e
chore: backlogs 1 (#191)
|
il y a 1 an |
|
AlpinDale
|
7d91e9e0f2
feat: CUDA graphs (#172)
|
il y a 1 an |
|
AlpinDale
|
653da510d1
chore: rewrite InputMetadata (#143)
|
il y a 1 an |
|
AlpinDale
|
8b2bbbd98b
chore: attention rewrite + models (#135)
|
il y a 1 an |
|
AlpinDale
|
0d51eac374
feat: awq for all models (#124)
|
il y a 1 an |
|
AlpinDale
|
e7b6a2d5a0
chore: tensor parallel refactors part 2 (#116)
|
il y a 1 an |
|
AlpinDale
|
74604eb252
fix: pylint complaints (#91)
|
il y a 1 an |
|
AlpinDale
|
efc6f7fbec
chore: reformats (#90)
|
il y a 1 an |
|
AlpinDale
|
a6a4220fa6
feat: refactor megatron and quants (#57)
|
il y a 1 an |
|
AlpinDale
|
0495c50a3e
GPTQ+exllama support (#21)
|
il y a 1 an |
|
AlpinDale
|
75c27d3e65
massive overhaul
|
il y a 1 an |
|
AlpinDale
|
6b9561ef07
adapt TGI incremental detokenization
|
il y a 1 an |
|
AlpinDale
|
45f6d9f923
initial refactor commit
|
il y a 1 an |
|
AlpinDale
|
06e71fc492
feat: add GPT-NeoX support for testing purposes
|
il y a 1 an |
