Spaces:
Running
on
CPU Upgrade
Repeated Model Evaluation Failures
Hi, I've submitted 'augtoma/qCammel-70-x', a few time's, each resulting in failures. I've been able to download the model from the hub and run the evaluation (with limits) myself in the lm-evaluation-harness package (same version as the one cited in the about section), is there anything I should do to troubleshoot this? Thanks.
Hi! Could you upload your model as safetensors? We'll check the logs in the meantime
Hi, it seems like we cannot load your model. There is an issue when reading the file pytorch_model-00014-of-00015.bin from your repo. I tried multiple times but the issue subsists. More precisely, it seems to be an issue with the encoding of the file: UnicodeDecodeError: 'utf-8' codec can't decode byte 0x80 in position 128: invalid start byte