- 18 Aug, 2020 1 commit
-
-
Jeff Rasley authored
* turn off multi-node launch if only 1 node
-
- 14 Aug, 2020 1 commit
-
-
Jeff Rasley authored
-
- 13 Aug, 2020 2 commits
-
-
Jeff Rasley authored
-
Jeff Rasley authored
* update fan out flag for pdsh
-
- 12 Aug, 2020 2 commits
-
-
Jeff Rasley authored
-
Shaden Smith authored
-
- 10 Aug, 2020 1 commit
-
-
Jeff Rasley authored
* add fix and tests for get_lr from lr_scheduler before training starts
-
- 08 Aug, 2020 1 commit
-
-
Shaden Smith authored
-
- 07 Aug, 2020 2 commits
-
-
Jeff Rasley authored
Add webinar on-demand links and update readme
-
Shaden Smith authored
The parenthesis alter the evaluation of the assert() and it will always evaluate to True.
-
- 01 Aug, 2020 1 commit
-
-
Emmanuel Kahembwe authored
mpu object is bound to the class instance.. the if statement uses `self.mpu' but just `mpu` is called in the following lines.. This raises a NameError
-
- 28 Jul, 2020 2 commits
-
-
Jeff Rasley authored
* fix nv_peer_mem version in dockerfile * fix security issue, remove pillow dependency (this is only needed for cifar example which has its own requirements.txt)
-
Emmanuel Kahembwe authored
-
- 27 Jul, 2020 1 commit
-
-
Jeff Rasley authored
-
- 25 Jul, 2020 1 commit
-
-
Shaden Smith authored
-
- 24 Jul, 2020 2 commits
-
-
Jeff Rasley authored
-
Jeff Rasley authored
-
- 23 Jul, 2020 1 commit
-
-
Jeff Rasley authored
* updates to amp to support grad clip and grad accumulation * zero grad using optimizer if in amp mode
-
- 22 Jul, 2020 2 commits
-
-
Olatunji Ruwase authored
* Avoid deadlock for unsynchronized non-zero checkpointing * Fix formatting issues Co-authored-by:Jeff Rasley <jerasley@microsoft.com>
-
Jeff Rasley authored
-
- 21 Jul, 2020 1 commit
-
-
Arash Ashari authored
-
- 15 Jul, 2020 2 commits
-
-
Jeff Rasley authored
* empty grad fix * add unit tests for empty grad
-
Olatunji Ruwase authored
-
- 14 Jul, 2020 1 commit
-
-
Olatunji Ruwase authored
* Support saving and loading ZeRO checkpoints on different data parallelism degree. * Fix formatting * Support checkpoint with varying GPU count in ZeRO stage 1 * Fix formatting * Formatting fixes * Update model tests * Remove pprint * Minor fix * Fix formatting * Update model tests Co-authored-by:Jeff Rasley <jerasley@microsoft.com>
-
- 13 Jul, 2020 1 commit
-
-
Jeff Rasley authored
* add amp docs
-
- 11 Jul, 2020 1 commit
-
-
Jeff Rasley authored
* add amp support for deepspeed (non-ZeRO) * tests for amp mode
-
- 06 Jul, 2020 1 commit
-
-
Olatunji Ruwase authored
* Load non-DeepSpeed checkpoints into ZeRO optimizer * Handle parameters smaller than DP * Formatting fixes * Handle empty partitions * Fix perf bug Co-authored-by:Jeff Rasley <jerasley@microsoft.com>
-
- 29 Jun, 2020 1 commit
-
-
Jeff Rasley authored
-
- 25 Jun, 2020 1 commit
-
-
Jeff Rasley authored
-
- 24 Jun, 2020 1 commit
-
-
Conglong Li authored
* syntax/typo fix * add README for documentation * fix links * update navigation * typo fix * docs readme fix
-
- 23 Jun, 2020 1 commit
-
-
Olatunji Ruwase authored
* Load non-DeepSpeed checkpoints into ZeRO optimizer * Handle parameters smaller than DP * Formatting fixes
-
- 20 Jun, 2020 4 commits
-
-
Shaden Smith authored
-
Samyam Rajbhandari authored
* Removing handle_overflow debugging code in deepspeed_utils.py * Removing handle_overflow debugging code in deepspeed_zero_optimizer.py Removing unnecessary overflow handle code. Not sure why it was there in the first place.
-
Shaden Smith authored
This reverts commit 54c0267e.
-
Tunji Ruwase authored
-
- 17 Jun, 2020 3 commits
-
-
Shaden Smith authored
-
Jeff Rasley authored
-
Shaden Smith authored
-
- 16 Jun, 2020 1 commit
-
-
RezaYazdaniAminabadi authored
* add the fine-tuning results * updating tutorial and blog-post * updated the tutorials and links
-
- 12 Jun, 2020 1 commit
-
-
Chunyang Wen authored
-