mosaicml/llm-foundry

Pull Requests Commits

changes for cosine restarts with wamup

gupta-abhay committed 251 days ago

0c923c2d

Fix monolithic checkpointing against composer main (#1836)

dakinggg committed 252 days ago

Verified a51c96cb

Bump Composer 0.31.0 (#1835)

bowenyang008 committed 253 days ago

Verified 429eb847

Configurable shard size (#1833)

dakinggg committed 253 days ago

Verified 9cf23209

Bump coverage[toml] from 7.8.0 to 7.8.2 (#1830)

dependabot[bot] committed 254 days ago

Verified 83b47adb

Update ci-testing version to latest (#1827)

dakinggg committed 261 days ago

Verified 358b23aa

Delete useless print("here") (#1826)

Omar Zoloev committed 261 days ago

Verified 6b6ec053

Bump docformatter for python3.12 and change blank_line_before_module_docstring = false (#1825)

sashaDoubov committed 262 days ago

Verified f0018598

Bump onnx from 1.17.0 to 1.18.0 (#1823)

dependabot[bot] committed 262 days ago

Verified 8aad84d5

Update accelerate requirement from <1.7,>=0.25 to >=0.25,<1.8 (#1824)

dependabot[bot] committed 262 days ago

Verified f83434f6

Fix Dtensor initialization (#1820)

bowenyang008 committed 265 days ago

Verified d39debbe

Deprecate inference API wrappers (#1821)

dakinggg committed 266 days ago

Verified 8c865f4d

Update huggingface-hub[hf_xet] requirement from <0.31,>=0.30.0 to >=0.30.0,<0.32 (#1818)

dependabot[bot] committed 269 days ago

Verified dedcfe3b

Bump onnxruntime from 1.19.2 to 1.22.0 (#1819)

dependabot[bot] committed 269 days ago

Verified 31527818

Update datasets requirement from <3.6,>=3.3.2 to >=3.3.2,<3.7 (#1817)

dependabot[bot] committed 269 days ago

Verified 96427baf

Change RMSNorm to use PyTorch native implementation (#1809)

josejg committed 273 days ago

Verified 7993aebe

Allow subselecting the appropriate config for llama4 (#1815)

dakinggg committed 276 days ago

Verified 5c47350d

Upgrade yapf version (#1814)

dakinggg committed 276 days ago

Verified 502bf051

Update foundry version in MCLI yamls (#1813)

dakinggg committed 276 days ago

Verified ea773086

Adding temperature tuning in attention (#1793)

ShashankMosaicML committed 281 days ago

Verified 67875360

Bump foundry version to 0.21.0.dev0 (#1812)

dakinggg committed 282 days ago

Verified faa7f913

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

ShashankMosaicML committed 282 days ago

Verified 1a8776cf

Remove some old deprecated code/comments (#1811)

dakinggg committed 282 days ago

Verified 86741ead

Update grouped-gemm version (#1810)

dakinggg committed 282 days ago

Verified f91d8f69

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

dependabot[bot] committed 283 days ago

Verified 77f89674

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

dependabot[bot] committed 283 days ago

Verified 71cd3a6b

added attn_implementation for hf_base.py (#1801)

ethantang-db committed 288 days ago

Verified 272dbd6c

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

dakinggg committed 289 days ago

Verified 0f6d9ba6

extended hf_checkpointer for any additional content saving (#1792)

ethantang-db committed 289 days ago

Verified ec9de523

Update accelerate requirement from <1.4,>=0.25 to >=0.25,<1.7 (#1799)

dependabot[bot] committed 290 days ago

Verified a1fce9e7

Older