Support for Llama 3.3 (transformers >= 4.45.0)

[alessandro-zucca]

I tried to export Llama 3.3 70B with the latest version of neuronx-tgi (0.0.27) but it fails with error "No tokenizer found ...".
The problems should be related to the transformers/tokenizers version: Llama 3.3 requires transformers >= 4.45.0, while in optimum neuron 0.0.27 we have 4.43.x.

Is there a plan to upgrade and add support for Llama 3.3?

Thanks

[github-actions]

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.

[marcomarinodev]

Having the same issue here

[dacorvo]

The upcoming 0.0.28 version will support LLama 3.3. In the meantime, you can build your own optimum-neuron version from the main branch.

[github-actions]

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.

[github-actions]

This issue was closed because it has been stalled for 5 days with no activity.