huggingface / optimum-neuron

Easy, fast and very cheap training and inference on AWS Trainium and Inferentia chips.
Apache License 2.0
195 stars 59 forks source link

LLaMa 2 training with examples/tutorials #283

Closed mmcclean-aws closed 1 month ago

mmcclean-aws commented 11 months ago

LLaMa 2 is highly requested by customers. Can we ensure we have LLaMa 2 fine-tuning working with neuronx-distributed including sample code and tutorials for the 7B, 13B and 70B models ?

rgrandhiamzn commented 10 months ago

Specifically for Llama-2-7B can we have one with TP and Zero-1. similar to the one in Neuron tutorials but we need a corresponding optimum-neuron version.

https://awsdocs-neuron.readthedocs-hosted.com/en/latest/libraries/neuronx-distributed/tutorials/training_llama2_7b.html#llama2-7b-tp-zero1-tutorial.

rgrandhiamzn commented 10 months ago

Tutorial for Llama-2-70B Training using (TP+PP)

HuggingFaceDocBuilderDev commented 6 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

HuggingFaceDocBuilderDev commented 5 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

HuggingFaceDocBuilderDev commented 4 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

HuggingFaceDocBuilderDev commented 3 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

HuggingFaceDocBuilderDev commented 2 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

HuggingFaceDocBuilderDev commented 2 months ago

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!