MT-NLG - Are we ever getting access to the 530 B parameters trained model?

Just following up on the Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, the World’s Largest and Most Powerful Generative Language Model

The framework is awesome, however it would be awesome to get the full model at NGC. A tremendous Head Start.

We recommend you to raise this query in TRITON Inference Server Github instance issues section.


How can I move it into there? SHould I delete this thread or edit and it will move there ?


Sorry above one is not related to Triton either.
This forum talks more about updates and issues related to the TensorRT.
We recommend you to please reach out to the same post you mentioned.

Thank you.