 AlpinDale
|
141672a0d4
kernels: disambiguate quantized types via a new ScalarType
|
há 5 meses atrás |
|
AlpinDale
|
6979ff658e
chore: perform allreduce in fp32 for marlin, better logging
|
há 5 meses atrás |
|
AlpinDale
|
84a9cd25c9
fix: some naming issues
|
há 5 meses atrás |
 AlpinDale
|
ba371fbbbd
feat: AWQ marlin kernels (#603)
|
há 5 meses atrás |
|
AlpinDale
|
7e9d4f3c71
chore: some more marlin cleanups
|
há 5 meses atrás |
|
AlpinDale
|
058e629f8e
chore: refactor marlin python utils
|
há 5 meses atrás |
|
AlpinDale
|
88a638d793
chore: debug logs for all available endpoints
|
há 5 meses atrás |
|
AlpinDale
|
98cb1c4cd1
feat: support fp8 via `llm-compressor`
|
há 5 meses atrás |
