DeepSpeed
Several fixes to unblock CI
#3047
Merged

Several fixes to unblock CI #3047

jeffra merged 34 commits into master from loadams/update-megatron
loadams
loadams Update a few workflows to use torch 1.13.1 for now
c20f881e
loadams Pin torch version on two other workflows
4d94c0b3
loadams Accelerate works when reverting to 1.13.1, checking if we use latest …
107e25a7
loadams Update --torch-ver in the torch-latest-v100
be294c6e
loadams Update fixes for pytorch lightning test
2efdb9f2
loadams Fix formatting
e6494a31
loadams Bump CUDA version to 11.7 for torch latest
6c18aa63
loadams Update Megatron version to latest
62749630
loadams Fix outdated flag
291edb59
jeffra switch to testing MDS main branch
05ac2fab
loadams Switch name from GPT2Model to GPTModel
34bc4b1d
loadams Merge branch 'loadams/update-megatron' of https://github.com/microsof…
12e12ba2
loadams Formatting issue
0daed47c
loadams Merge branch 'master' into loadams/update-megatron
c3350885
jeffra jeffra changed the title Update megatron version Several fixes to unblock CI 2 years ago
jeffra jeffra marked this pull request as ready for review 2 years ago
jeffra jeffra requested a review from jeffra jeffra 2 years ago
jeffra jeffra requested a review from mrwyattii mrwyattii 2 years ago
jeffra jeffra requested a review from tjruwase tjruwase 2 years ago
loadams Pin to previous accelerate version
cec22fb0
loadams Merge branch 'loadams/update-megatron' of https://github.com/microsof…
7f75fc3f
loadams Merge convlits
df47f3a5
loadams Revert "Merge convlits"
8f94bec6
loadams Add comment clarifying pinning pytorch-lightning and add microbatchsize
ec257b51
loadams formatting
0f676a63
mrwyattii add skip for megatron tests
eb88ce5e
loadams Remove changes for micro batch
ee8c168c
mrwyattii skip more tests with latest torch
ba2ee86f
loadams Revert megatron-lm import change when changing version
9cf95a60
loadams Update nv-accelerate workflow to use latest since they added cpuadam …
d1394e79
loadams Test with older transformers version
364ab044
loadams Remove fix to previous git commit as it fails with torch > 2.
f16bdcf4
loadams Revert "Test with older transformers version"
09ef3603
loadams revert change updating flag for new megatron-lm
b7c902a4
loadams Formatting
1c0b1046
loadams revert changes to megatron_model file
71c4f3f6
jeffra Merge branch 'master' into loadams/update-megatron
72c5315a
jeffra
jeffra approved these changes on 2023-03-21
jeffra skip pp tests on p40, add pytest defaults
7fd99ea2
jeffra fix typos
0f744d31
jeffra jeffra merged 4e068623 into master 2 years ago
jeffra jeffra deleted the loadams/update-megatron branch 2 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone