Can anyone please share the code to reuse the fine-tuned model and to use it.I got some code from internet, but unfortunately it is not working correctly
Fine tuning is done so for specific purpose. You fine tune the model so it gives you more accurate answers to your specific needs, it is not just more accurate or less accurate, it depends on the use-case.
That being said, I am just guessing what your questions means.
Can you please share any codes to reuse the fine-tuned GPT4 model. without accuracy score, precision and F1 score , how to know our fine-tuned model’s performance?
If you are indeed fine-tuning and have access to GPT-4 fine-tune via an OpenAI account, you can perform mostly without code in the new GUI.
This will show a graph of training loss and validation loss after the job is done.
Validation loss is a metric calculated by a validation file, containing examples similar to those you train on, but held-out to check the quality.
Since you are mostly saying things that aren’t fine-tune terms, I suggest you read a tutorial.
1 Like
Yeah, i completed the finetuning using this GUI