Megatron-DeepSpeed
795fedbb - args should be required, don't create another latest file

Commit

3 years ago

args should be required, don't create another latest file

References

#239 - Reshape deepspeed checkpoint

#289 - No-ZeRO reshaping

#292 - a branch combining layer-norm-auto-sync and ds_ckpt_reshape

Author

stas00

stas00

Parents

Loading