Orkhan

/

llama-2-7b-absa

like 13

Hi,

Thanks for providing this model.

I wanted to try it on a dataset that I have and I noticed that for some prompts it would not generate the complete output and that would cause and error and break the loop.

For example for the prompt: "Friendly environment and community"

The raw_output is : [{'generated_text': '### Human: Friendly environment and community.### Assistant: ## Aspect detected: environment,'}]

The rest of the dict is not generated and due to this I get the error list index out of range when trying to get the opinions, sentiments and phrases in process_output.

It is giving error with this statement. I tried the prompt again with the one you have in your colab and readme and it gave the correct and expected output.

Do you know why the model would not generate the full output for the above prompt?

Thanks