DachengLi1 / LongChat

Official repository for LongChat and LongEval
Apache License 2.0
504 stars 29 forks source link

Hi, using xformers monkey patch training llama2 got loss explosion #35

Open lucasjinreal opened 1 year ago

lucasjinreal commented 1 year ago

By using xformers to train llama2, the loss are explosion, do u know why? On V100 only