Training Script
#1
by NDugar - opened
Hello, Thanks for the great model and detailed training details, I am however still confused as each dataset seems to have separate losses attached to them, Would you be kind enough to share the training script you used ?
Thanks.
Thank you for your question! I followed https://www.sbert.net/docs/sentence_transformer/training_overview.html closely for training this model and you can just swap base model with keeeeenw/MicroLlama and start training.
It has been a while so I don't remember if this was the final version of the training code but this was definitely one of the versions I used and you can use this as reference:
https://huggingface.co/keeeeenw/MicroLlama-text-embedding/blob/main/train.py