Fine tuning response is cropping up

API
1

Fine-tuning using the davinchi model
I made a request to see the results.

According to my search, the number of limit tokens is 2048,
Even though the answer is significantly smaller than the number of limit tokens,
The results seem to keep cropping up.

What should I do?
Can I adjust the number of limit tokens during fine tuning?

Anyone who knows why please let me know…

스크린샷 2023-03-14 오후 12.39.08
스크린샷 2023-03-14 오후 12.39.082612×196 256 KB

2

Please post all your API parameters to get the assistance you need. Please do not post them as an image. Post them using Markdown wrapped in triple backticks, as follows:

```
#  your API code and parameters here
```

Thanks

:slight_smile:

3

@ruby_coder
i just make like this
‘’’
curl ‘https://api.openai.com/v1/fine-tunes’
-X POST
-H “Content-Type: application/json”
-H “Authorization: Bearer ************************************************”
-d ‘{ “training_file”: “file-QwIgzvlgsUmm7rGCnjGmapXC”, “model”: “davinci”}’
‘’’

4

Yes, that is correct. You must pass the fine-tuning training data file ID to the fine-tuned process.

From the OpenAI API docs, as reference only:

curl https://api.openai.com/v1/fine-tunes \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
    "training_file": "file-XGinujblHPwGLSztz8cPS8XY"
  }'

Glad you are back on track @parksj0923

:slight_smile: