pszemraj/pythia-6.9b-HC3 - Каталог нейросетей
Генерация текста

pszemraj/pythia-6.9b-HC3

Добавлено:
pszemraj/pythia-6.9b-HC3

This model is a fine-tuned version of EleutherAI/pythia-6.9b-deduped on the pszemraj/HC3-textgen-qa dataset. It achieves the following results on the evaluation set: — Loss: 1.2372 — Accuracy: 0.6769 — perplexity: 3.446 Text generation model trained on the HC3 text data of human questions + chatGPT answers. Install necessary packages for inference (unless you have a big boi GPU) The defautl GenerationConfig uses contrastive search with topk=4 and penaltyalpha=0.6. For more information on inference and parameters to use, see the transformers docs. — Intended use: research/exploration into comparing RLHF tuning vs. «guided»/specific tuning on «quality» datasets/responses of «what the human would want as answer anyway» — This is not trained/fine-tuned with RLHF and therefore will not be as helpful/generalizable/safe as chatGPT (outside of the fact that this model is ~30x smaller)

Модальности:
Генерация текста


Задача: Генерация текста
Автор: pszemraj
Теги: gpt_neox, generated_from_trainer, HC3, chatGPT, assistant, text-generation-inference
Лайков: 2  |  Загрузок: 26

Открыть на HuggingFace →

Описание основано на материалах HuggingFace. Перевод выполнен автоматически.