I have a wild idea to improve smaller gpt3 esqe models by tuning their output with ppo a reinforcement learning paper. Originally, this was done to adjust gpt2’s performance to human preference. https://arxiv.org/pdf/1909.08593.pdf I propose to fine-tune gpt neo directly on “prompt driven” data. Mo…

Fine tuning gpt-neo via ppo

SUNM June 11, 2023, 11:16am 2

I have the same issue, have you been successful to fine-tune the gpt neo with ppo and get goo results? can I know your resource?

many many thanks

Topic		Replies	Views
Fine-tuning GPT for Improved Small Talk and Informal Conversations Beginners	0	27	November 9, 2024
Using GPT-J models for many NLP tasks Models	0	571	November 21, 2022
FineTuning Reimbursement Documents using 3.5 gpt turbo model Beginners	0	9	January 16, 2025
Multi-stage finetuning of transformer models Beginners	0	857	September 5, 2022
Analyze the fine tuning result Models	2	31	February 18, 2025