mosaicml/llm-foundry

Pull Requests Commits

v-chen_data committed 1 year ago

ccc6c95d

v-chen_data committed 1 year ago

1d18939e

v-chen_data committed 1 year ago

7645de5e

v-chen_data committed 1 year ago

6a8e32b3

v-chen_data committed 1 year ago

4417fd50

v-chen_data committed 1 year ago

3b48af8a

v-chen_data committed 1 year ago

c3c3db24

v-chen_data committed 1 year ago

27a033f7

v-chen_data committed 1 year ago

105b3c26

v-chen_data committed 1 year ago

3fe90c7b

v-chen_data committed 1 year ago

180e6fae

v-chen_data committed 1 year ago

b7cea356

v-chen_data committed 1 year ago

198c38f5

v-chen_data committed 1 year ago

5b98bfb3

v-chen_data committed 1 year ago

766700f0

Update huggingface-hub requirement from <0.24,>=0.19.0 to >=0.19.0,<0.25 (#1379)

dependabot[bot] committed 1 year ago

Verified 78e91ef7

Update accelerate requirement from <0.33,>=0.25 to >=0.25,<0.34 (#1403)

dependabot[bot] committed 1 year ago

Verified 9417417d

Kevin/ghcr build (#1413)

KevDevSha committed 1 year ago

Verified 9a62bfd6

Update setup.py (#1411)

mvpatel2000 committed 1 year ago

Verified 396d06f4

Enable QuickGelu Function for CLIP models (#1408)

gupta-abhay committed 1 year ago

Verified 7a7f6df3

Set pretrained model name correctly, if provided, in HF Checkpointer (#1407)

snarayan21 committed 1 year ago

Verified 6f4aa8c9

Remove curriculum learning error when duration less than saved timestamp (#1406)

b-chu committed 1 year ago

Verified 6d5d016b

Add spin_dataloaders flag (#1405)

dakinggg committed 1 year ago

Verified 5c7e99be

Remove orig params false default (#1401)

dakinggg committed 1 year ago

Verified 5e07a05d

Add pip requirements directly for mlflow save (#1400)

dakinggg committed 1 year ago

Verified 799279bc

Add pre register method for mlflow (#1396)

dakinggg committed 1 year ago

Verified 0c93331f

Enable passing epsilon when building norm layers (#1399)

gupta-abhay committed 1 year ago

Verified d8c15521

Update README.md (#1398)

dakinggg committed 1 year ago

Verified 7de4969e

Add a pretrained guard for the initial meta init on global rank 0 (#1397)

irenedea committed 1 year ago

Verified bb385f65

Dtensor oom (#1395)

dakinggg committed 1 year ago

Verified e8826585

Older