Hi, I have created a model first from basic, curie, with 490 prompts, to create an anagram, provided 80 % short either names or words and around 20% with phrases. But after I had the model, testing it, it doesn’t even create anagrams, it does create some good, existing, real word or words but not e…

Check out this thread to make sure you’re configuring the fine-tuning parameters effectively: [image] Fine-Tuning In a Nutshell with a Single Line JSONL File and n_epochs Documentation Those who frequent this community easily notice that developers struggle to get good fin…

Let me read it, I only followed the guide from openai. Many thanks!

[image] whiteadi: Or what can be the reason it doesn’t understands the “simple” fact of using same letters but rearranged? The models predict the next word/token. The word/token is passed to the large language model not as a sequence of letters but as a vector, so the large language model doe…

so what are you saying is to check some of the new 4 models? I have checked the previous suggestion and used 16 epochs and one thing it did good that for each existing prompt, that was trained with if I give that one will give the good , trained with, answer, but for new names, words, will still no…

[image] whiteadi: did good that for each existing prompt, that was trained with if I give that one will give the good , trained with, answer, but for new names, words, will still not do anagrams Sounds like overfitting maybe - ie repeating something verbatim from the fine-tuning dataset. Wha…

It’s going to be very hard to fine-tune a model for Anagrams. openai/evals/blob/main/evals/registry/evals/anagrams.yaml anagrams: id: anagrams.test.v1 metrics: - accuracy anagrams.test.v1: args: few_shot_jsonl: anagrams/fewshot.jsonl num_few_shot: 5 samples_jsonl…

Thanks for the guess :slight_smile: I have 490 prompts similar to these: {"prompt":"Adrian Albu\n\n===\n\n","completion":" a bad urinal \n"} {"prompt":"A Decimal Point\n\n===\n\n","completion":" i'm a dot in place \n"} {"prompt":"A domesticated animal\n\n===\n\n","completion":" docile, as a man ta…

OOOO, many many thanks!! Will combine the data! Brb

Curious why fine tuned model doesn't work too ...fine

API

wfhbrian March 22, 2023, 1:16pm 2

Check out this thread to make sure you’re configuring the fine-tuning parameters effectively:

Topic		Replies	Views
Got awful results after fine-tuning API	11	3383	December 1, 2022
Struggling with poor performance on fine-tuned davinci model API	14	2937	February 7, 2023
Should prompts be unique for fine-tuning? Prompting	8	1879	September 18, 2022
Fine tuning very very poor results API fine-tuning , api	16	3130	July 11, 2023
Fine tune model problem API	6	874	January 10, 2023

Curious why fine tuned model doesn't work too ...fine

Related topics