This content originally appeared on HackerNoon and was authored by Writings, Papers and Blogs on Text Models
:::info Authors:
(1) Rafael Rafailo, Stanford University and Equal contribution; more junior authors listed earlier;
(2) Archit Sharma, Stanford University and Equal contribution; more junior authors listed earlier;
(3) Eric Mitchel, Stanford University and Equal contribution; more junior authors listed earlier;
(4) Stefano Ermon, CZ Biohub;
(5) Christopher D. Manning, Stanford University;
(6) Chelsea Finn, Stanford University.
:::
Table of Links
4 Direct Preference Optimization
7 Discussion, Acknowledgements, and References
\ A Mathematical Derivations
A.1 Deriving the Optimum of the KL-Constrained Reward Maximization Objective
A.2 Deriving the DPO Objective Under the Bradley-Terry Model
A.3 Deriving the DPO Objective Under the Plackett-Luce Model
A.4 Deriving the Gradient of the DPO Objective and A.5 Proof of Lemma 1 and 2
\ B DPO Implementation Details and Hyperparameters
\ C Further Details on the Experimental Set-Up and C.1 IMDb Sentiment Experiment and Baseline Details
C.2 GPT-4 prompts for computing summarization and dialogue win rates
\ D Additional Empirical Results
D.1 Performance of Best of N baseline for Various N and D.2 Sample Responses and GPT-4 Judgments
C Further Details on the Experimental Set-Up
In this section, we include additional details relevant to our experimental design.
C.1 IMDb Sentiment Experiment and Baseline Details
The prompts are prefixes from the IMDB dataset of length 2-8 tokens. We use the pre-trained sentiment classifier siebert/sentiment-roberta-large-english as a ground-truth reward model and gpt2-large as a base model. We use these larger models as we found the default ones to generate low-quality text and rewards to be somewhat inaccurate. We first use supervised fine-tuning on a subset of the IMDB data for 1 epoch. We then use this model to sample 4 completions for 25000 prefixes and create 6 preference pairs for each prefix using the ground-truth reward model. The RLHF reward model is initialized from the gpt2-large model and trained for 3 epochs on the preference datasets, and we take the checkpoint with the highest validation set accuracy. The “TRL” run uses the hyper-parameters in the TRL library. Our implementation uses larger batch samples of 1024 per PPO step.
\
:::info This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license.
:::
\
This content originally appeared on HackerNoon and was authored by Writings, Papers and Blogs on Text Models
Writings, Papers and Blogs on Text Models | Sciencx (2024-08-26T20:45:14+00:00) Fine-Tuning GPT-2 for IMDb Sentiment Analysis. Retrieved from https://www.scien.cx/2024/08/26/fine-tuning-gpt-2-for-imdb-sentiment-analysis/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.