philschmid
Internal ModelError
Hi, I tried to deploy gptj https://huggingface.co/EleutherAI/gpt-j-6B with text-generation as the task. It seemed to work but then I tried to curl the post request and got this error ModelError 400
{"error": "ModelError('An error occurred (ModelError) when calling the InvokeEndpoint operation: Received client error (400) from primary with message "{\n "code": 400,\n "type": "InternalServerException",\n "message": "\\u0027gptj\\u0027"\n}\n". See https://us-east-1.console.aws.amazon.com/cloudwatch/home?region=us-east-.... for more information.')"}%
Do you know how to fix this?
Hello @ninajlu,
did you check "See https://us-east-1.console.aws.amazon.com/cloudwatch/home?region=us-east-.... for more information.'"?
We also have a dedicated blog post on how to deploy GPT-J: https://huggingface.co/blog/gptj-sagemaker
I was getting this error when my payload wasn't converted into a 'byte' format. I solved it by doing this.
payload = json.dumps(payload, indent=2).encode('utf-8')
