VOOZH about

URL: https://huggingface.co/shivash/enhanced-hybrid-transformer-768d-trained/discussions/1

⇱ shivash/enhanced-hybrid-transformer-768d-trained · Add trained multi-dataset model

Add trained multi-dataset model

#1
by ziadrone - opened

Trained on GSM8K and AQuA-RAT datasets with 3 epochs. Final loss: 3.8256

shivash changed pull request status to closed

· Sign up or log in to comment