mrm8488/gpt2-imdb-neutral

A small GPT2 language model fine-tuned to produce neutral-ish movie reviews based on the IMDB dataset . The model is trained with rewards from a BERT sentiment classifier (lvwerra/gpt2-imdb) via PPO . The full experiment setup (for positive samples) in trl repo. The model was trained for 100 optimisation steps with a batch size of 128 which corresponds to 30000 training samples . It was trained with 100 optimisations steps with the training samples of 128 . A few examples of the model response to a query before and after optimisation:Query-before-query-response-response (before)Query-after-response,