the training data for this model?
#1
by
AIR-hl
- opened
hi! Can I know the dataset for training this model? It's better if i can know the hyperparams too. Thk u!
Hi, this ckpt is trained with RLHFlow SFT v1 data + dartmath subset. We use initial learning rate 2e-5 & 2 epochs.
Due to the IP issue, you may consider https://huggingface.co/datasets/RLHFlow/RLHFlow-SFT-Dataset-ver2 as alternative.