onnxruntime
save_checkpoint, load_checkpoint and aggregate_checkpoints
#6136
Merged

save_checkpoint, load_checkpoint and aggregate_checkpoints #6136

baijumeswani merged 12 commits into master from bmeswani/checkpoint
baijumeswani
baijumeswani baijumeswani requested a review from BowenBao BowenBao 5 years ago
baijumeswani baijumeswani requested a review from liqunfu liqunfu 5 years ago
baijumeswani baijumeswani requested a review from spandantiwari spandantiwari 5 years ago
baijumeswani baijumeswani requested a review from thiagocrepaldi thiagocrepaldi 5 years ago
baijumeswani baijumeswani force pushed from 926d18bc to 3f5ddd03 5 years ago
ashbhandare
ashbhandare commented on 2020-12-16
ashbhandare
ashbhandare commented on 2020-12-16
baijumeswani baijumeswani force pushed 5 years ago
baijumeswani baijumeswani force pushed to 7efc8ce3 5 years ago
thiagocrepaldi
thiagocrepaldi commented on 2020-12-16
baijumeswani remove binascii dependence
10e8bb6f
baijumeswani use default values without relying on custom defaults, pass only the …
9e1c1e37
baijumeswani use optimizer states
d296dbfd
baijumeswani save_checkpoint and load_checkpoint implementations
c8cf5269
baijumeswani checkpoint aggregation logic
71e5a6ae
baijumeswani unit tests for save_checkpoint, load_checkpoint and aggregate_checkpo…
42926ea1
baijumeswani optimizer name check, include_optimizer_states for save_checkpoint
b1f8df82
baijumeswani change name from fp32 to full_precision
d196426c
baijumeswani function name change for zero helpers, docstring typo fix
cb26971f
baijumeswani baijumeswani force pushed to cb26971f 5 years ago
ashbhandare
ashbhandare commented on 2020-12-17
ashbhandare
ashbhandare commented on 2020-12-17
ashbhandare
ashbhandare commented on 2020-12-17
baijumeswani baijumeswani force pushed to f1ecc59b 5 years ago
baijumeswani remove redundant assert, helper function for string literals
4bce5f2a
baijumeswani baijumeswani force pushed from f1ecc59b to 4bce5f2a 5 years ago
ashbhandare
ashbhandare commented on 2020-12-17
baijumeswani
baijumeswani commented on 2020-12-17
baijumeswani baijumeswani force pushed 5 years ago
baijumeswani load index 0 non aggregation checkpoint file, add trainer options val…
7b5307c0
baijumeswani baijumeswani force pushed to 7b5307c0 5 years ago
ashbhandare
ashbhandare dismissed these changes on 2020-12-17
thiagocrepaldi
thiagocrepaldi commented on 2020-12-18
baijumeswani add comments for using byte string .decode function
583bc201
baijumeswani baijumeswani dismissed their stale review via 583bc201 5 years ago
thiagocrepaldi
thiagocrepaldi approved these changes on 2020-12-18
baijumeswani
baijumeswani baijumeswani merged adc20710 into master 5 years ago
baijumeswani baijumeswani deleted the bmeswani/checkpoint branch 5 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone