ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

Doc: update harness readme (#12324)

Browse source
This commit is contained in:
Chu,Youcheng 2024-11-04 14:58:54 +08:00 committed by GitHub
parent 4644cb640c
commit a01371f90b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 6 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
python/llm/dev/benchmark/harness/README.md
View file

@ -78,3 +78,8 @@ Please refer to these:
- [Security features from the Hugging Face datasets library · Issue #1135 · EleutherAI/lm-evaluation-harness (github.com)](https://github.com/EleutherAI/lm-evaluation-harness/issues/1135#issuecomment-1961928695) - [Security features from the Hugging Face datasets library · Issue #1135 · EleutherAI/lm-evaluation-harness (github.com)](https://github.com/EleutherAI/lm-evaluation-harness/issues/1135#issuecomment-1961928695)
You have to manually run `export HF_DATASETS_TRUST_REMOTE_CODE=1` to solve the problem. You have to manually run `export HF_DATASETS_TRUST_REMOTE_CODE=1` to solve the problem.
### 3.Error: xe_addons.rotary_half_inplaced(self.rotary_emb.inv_freq, position_ids,RuntimeError: unsupported dtype, only fp32 and fp16 are supported.
This error is because `ipex-llm` currently only support models with `torch_dtype` of `fp16` or `fp32`.
You can add `--model_args dtype=float16` to your command to solve this problem.