mosaicml/llm-foundry

Pull Requests Commits

mattyding committed 269 days ago

0ae048cb

mattyding committed 269 days ago

ee8e274c

mattyding committed 269 days ago

5120abdc

i love debugging

mattyding committed 269 days ago

dda4ec7a

mattyding committed 269 days ago

084a2e48

mattyding committed 269 days ago

01c953b7

print out dtype for debugging

mattyding committed 286 days ago

2d7dbde1

mattyding committed 286 days ago

d87886ec

first try specifying bytes type

mattyding committed 286 days ago

d10e89e8

mattyding committed 286 days ago

f47093f8

convert ndarray to bytes

mattyding committed 286 days ago

512a165c

mattyding committed 286 days ago

92c58861

mattyding committed 286 days ago

f8640ea0

mattyding committed 287 days ago

38652da7

Merge branch 'main' into matt/split-mds-script

mattyding committed 287 days ago

5b918c42

mattyding committed 287 days ago

654d084f

Change RMSNorm to use PyTorch native implementation (#1809)

josejg committed 290 days ago

Verified 7993aebe

Allow subselecting the appropriate config for llama4 (#1815)

dakinggg committed 293 days ago

Verified 5c47350d

Upgrade yapf version (#1814)

dakinggg committed 293 days ago

Verified 502bf051

Update foundry version in MCLI yamls (#1813)

dakinggg committed 293 days ago

Verified ea773086

Adding temperature tuning in attention (#1793)

ShashankMosaicML committed 298 days ago

Verified 67875360

Bump foundry version to 0.21.0.dev0 (#1812)

dakinggg committed 299 days ago

Verified faa7f913

Adding support for nope positional encoding in block overrides. (#1794)

ShashankMosaicML committed 299 days ago

Verified 1a8776cf

Remove some old deprecated code/comments (#1811)

dakinggg committed 299 days ago

Verified 86741ead

Update grouped-gemm version (#1810)

dakinggg committed 299 days ago

Verified f91d8f69

Update datasets requirement from <3.4,>=3.3.2 to >=3.3.2,<3.6 (#1807)

dependabot[bot] committed 300 days ago

Verified 77f89674

Update setuptools requirement from <80.0.0 to <81.0.0 (#1803)

dependabot[bot] committed 300 days ago

Verified 71cd3a6b

added attn_implementation for hf_base.py (#1801)

ethantang-db committed 305 days ago

Verified 272dbd6c

Load model only on global rank 0 for mixed init (#1795)

dakinggg committed 306 days ago

Verified 0f6d9ba6

extended hf_checkpointer for any additional content saving (#1792)

ethantang-db committed 306 days ago

Verified ec9de523

Older