Closed ywen666 closed 1 year ago
Yes SantaCoder had some issues with fp16
so it's better to either use bf16
or fp32
with it
Yes SantaCoder had some issues with
fp16
so it's better to either usebf16
orfp32
with it
I encountered the same error while evaluating Humaneval on meta-llama/Llama-2-7b-chat-hf using fp16 precision. However, I was able to complete the evaluation for the CodeLlama series model and meta-llama/Llama-2-7b-hf without issues. Could you please explain the reasons behind it and the impact of changing the precision setting from fp16 to fp32 on the pass@k score? Thank you very much for your assistance!
Just wondering if we need to use fp32 for evaluation of santacoder? I tried fp16 evaluation because I fine-tuned santacoder on the stack-dedup python dataset for 1000 steps with fp16 precision. But when I ran fp16 evaluation on humaneval, it leads to the following error (for both
--model=bigcode/santacoder
and--model=myfp16_finetuned_santacoder
),The error went away if I use
--precision=fp32
, leading to 37.19% pass@100 on humaneval which is kinda close to the number reported in the paper. This is the command I used to run fp16 evaluation on humaneval.