deepspeed
https://github.com/microsoft/deepspeed
Python
DeepSpeed is a deep learning optimization library that makes distributed training easy, efficient, and effective.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported9 Subscribers
Add a CodeTriage badge to deepspeed
Help out
- Issues
- [REQUEST] Does PPO requires a critic that backpropagates?
- [BUG]After using the code that supports llama inference, the result of the inference is different from the original one
- turn back on opt/bloom tests
- [BUG] deepspeed get_model_profile custom module zero MACs
- [BUG]KeyError: 'attention_mask'
- [BUG] generating with different batch size before causes RuntimeError
- [BUG] Non-deterministic model output if `replace_with_kernel_inject=True` for GPT-neo-1.3B/Bloom 1b7 and others
- [BUG] Train loss did not decrease during sft train process, help! Thx!
- [REQUEST] Is the cuda unified memory technology used in deepspeed? Where is its API?
- [BUG] ZeRO3 - Getting assert len(self.ckpt_list) > 0 while running validation code during fine tuning
- Docs
- Python not yet supported