Text Generation
Transformers
PyTorch
Safetensors
English
llama
text-generation-inference
Inference Endpoints

will there be a tinymistral?

#3
by LaferriereJC - opened

that would be nice.

I'm really looking for a 3b model, but I'll settle for 1.1b

You would need the Mistral dataset.

Otherwise someone can do a 'distillation'. Let Mistral blabber on, and train on that.

There might be better pruning methods though.

Sign up or log in to comment