feat: data-parallel block tuning via --data_parallel #2351
avtc
marked this pull request as draft 60 days ago
avtc
force pushed
from
849e2782
to
230317f4
60 days ago
avtc
marked this pull request as ready for review 60 days ago
avtc
force pushed
from
230317f4
to
b0533cc2
59 days ago
wenhuach21
added this to the 0.16.0 milestone 58 days ago
avtc
force pushed
from
b0533cc2
to
837728c5
56 days ago
avtc
marked this pull request as draft 55 days ago
avtc
force pushed
from
7ca13fda
to
837728c5
54 days ago
chensuyue
removed this from to the 0.16.0 milestone 47 days ago
avtc
marked this pull request as ready for review 46 days ago
avtc
marked this pull request as draft 45 days ago
avtc
force pushed
from
837728c5
to
7e7a0a4a
43 days ago
avtc
marked this pull request as ready for review 43 days ago
wenhuach21
added this to the 0.17.0 milestone 27 days ago
avtc
marked this pull request as draft 19 days ago
avtc
force pushed
from
7e7a0a4a
to
1faaa833
18 days ago
avtc
marked this pull request as ready for review 18 days ago
feat: data-parallel block tuning via --parallel_quantization
6db22dfa
fix: masked tune loss is normalized once, by the valid-token count
5dc253f9
avtc
force pushed
from
2dd31329
to
5dc253f9
18 days ago
refactor: rename --parallel_quantization to --data_parallel
627f09b4
avtc
changed the title feat: data-parallel block tuning via --parallel_quantization feat: data-parallel block tuning via --data_parallel 14 days ago
fix: address Copilot review notes on the sub-batch draw path
9d852fca
avtc
marked this pull request as draft 14 days ago
avtc
marked this pull request as ready for review 14 days ago
avtc
marked this pull request as draft 13 days ago
avtc
marked this pull request as ready for review 13 days ago
Merge upstream/main into feature/ddp-parallel-tuning
8c525dd7
Merge branch 'main' into feature/ddp-parallel-tuning
17992cf8
Merge branch 'main' into feature/ddp-parallel-tuning
98e9ee64
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub