mosaicml/llm-foundry

Pull Requests Commits

v-chen_data committed 2 years ago

c90b9b31

v-chen_data committed 2 years ago

2596dd65

v-chen_data committed 2 years ago

e9e55241

v-chen_data committed 2 years ago

59632a4d

v-chen_data committed 2 years ago

f01dc6dd

v-chen_data committed 2 years ago

b6d4a52e

update train to reflect script

v-chen_data committed 2 years ago

9a5fcd36

Merge branch 'main' into train-cli

vinchenzo-db committed 2 years ago

Verified ddad1fcf

v-chen_data committed 2 years ago

8b712441

v-chen_data committed 2 years ago

37f08293

deepcopy config (#1336)

mvpatel2000 committed 2 years ago

Verified f0535dcb

Add a config arg to just save an hf checkpoint (#1335)

dakinggg committed 2 years ago

Verified a5425655

Adding a child class of hf's rotary embedding to make hf generate work on multiple gpus. (#1334)

ShashankMosaicML committed 2 years ago

Verified 73f267ca

Update registry.py (#1333)

mvpatel2000 committed 2 years ago

Verified 1641b217

Currently multi-gpu generate does not work with hf.generate for hf checkpoints. This PR fixes that. (#1332)

ShashankMosaicML committed 2 years ago

Verified 199c3b94

Bump onnx from 1.14.0 to 1.16.1 (#1331)

dependabot[bot] committed 2 years ago

Verified fe0f25c5

Bump onnxruntime from 1.15.1 to 1.18.1 (#1329)

dependabot[bot] committed 2 years ago

Verified 546a86fd

Bump einops from 0.7.0 to 0.8.0 (#1328)

dependabot[bot] committed 2 years ago

Verified 162d4a7f

Update transformers requirement from <4.41,>=4.40 to >=4.42.3,<4.43 (#1327)

dependabot[bot] committed 2 years ago

Verified b7c097f2

Error if metadata matches existing keys (#1313)

dakinggg committed 2 years ago

Verified 1993e7fc

Relax hf hub pin (#1314)

dakinggg committed 2 years ago

Verified 62165de2

Bump version (#1326)

b-chu committed 2 years ago

Verified cece2b55

update hf token env var name (#1321)

dakinggg committed 2 years ago

Verified 73be9d33

Add dependabot (#1322)

dakinggg committed 2 years ago

Verified 80ba2b97

Registry docs update (#1323)

dakinggg committed 2 years ago

Verified 277e09fb

fixing sequence_id =-1 bug, adding tests (#1324)

ShashankMosaicML committed 2 years ago

Verified f398db10

Extra serverless (#1320)

XiaohanZhangCMU committed 2 years ago

Verified 61178447

Add optional logging of text output to EvalOutputLogging (#1283)

sjawhar committed 2 years ago

Verified 68c26256

Allows interweaving of arbitrary kinds of 'attention' layers, like sliding window, reuse prev layer kv cache etc. (#1299)

ShashankMosaicML committed 2 years ago

Verified 8604bba9

Refactor hf checkpointer (#1318)

irenedea committed 2 years ago

Verified 88511f78

Newer Older