huggingface / exporters

Export Hugging Face models to Core ML and TensorFlow Lite
Apache License 2.0
622 stars 46 forks source link

Allow max sequence length to follow the config #31

Closed pcuenca closed 1 year ago

pcuenca commented 1 year ago

This requires custom configurations, as the default still uses 128. Not fully sure this is a great idea, I need to test memory consumption in more depth.

The rationale for this change is that I'm following the steps that would be required to create a modern LM chat UI using Core ML, and see how far we can get.