Small Language Model baseline model

[Aisuko]

Here is the small language model I mentioned before, please take a look.

I fine-tuned it with ORPO on ultra-feedback dataset see https://www.kaggle.com/code/aisuko/ft-smollm-135m-instruct-on-hf-ultrafeedback. The entire training process takes 6 hours on single P100 GPU. The learning rate I used same to original model.

And test it performance of inference on CPU also GPU please.
https://huggingface.co/aisuko/ft-smollm-135M-instruct-on-hf-ultrafeedback