deepseekmirror/DeepSeek-Coder
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepSeek-Coder.git synced 2025-02-23 14:19:09 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity
108 Commits 4 Branches 0 Tags 18 MiB
78d0fd332a
Commit Graph

3 Commits

Author SHA1 Message Date
Rahul Dubey
78d0fd332a
Update finetune_deepseekcoder.py 
Using torch.float16 or torch.cuda.amp can significantly reduce memory usage and speed up training by performing computations with lower precision.
 2025-01-29 12:38:12 +05:30
DejianYang
3fd8db86c3
Update license in finetune_deepseekcoder.py 2023-11-14 18:15:08 +08:00
Yang Dejian
4f0b860d30 add deepspeed finetune 2023-11-09 22:46:45 +08:00