Closed lambdaofgod closed 2 years ago
I've uploaded the weights in half precision to the HF Hub repo, with instructions on how to load them in the model card: https://huggingface.co/facebook/incoder-6B.
I've also updated the example_usage.py script here to use these half-precision weights by default.
Please let me know if you still run into issues, and thanks for your interest!
I'm having troubles with loading 6B model even on RTX3090 - is there a possibility you would share the model weights in half precision format?
Anyway, thank you for your work, the model looks great.