This is a fine-tuned version of conceptofmind/LLongMA-2-7b intended for summarization and literary analysis of fiction stories. It contains custom modeling code to use Flash Attention 2 during inference, which provides a significant speedup, especially at longer context lengths. To enable, pass trustremotecode=True,useflashattention=True to AutoModelForCausalLM. > A nineteen-seventies pair of twins approaches Micheala and Gerty at a cafe. They ask them to write down their story of their disappearance. Gerty takes notes on their story as they tell it. The twins’ names are Daphne and Claude. The twins explain that they came from London, England and were leading a movement against fascism there. But in Waystation City, they were stuck in the limbo between timelines with no real way of returning home to continue their work. In fact, it was harder for them to remember their lives at home as time went on. They tell Gerty that they cannot stay and Micheala encourages them not to go down to the stairs where Waystation City’s ferry drivers will take them home. Daphne writes an inscription on a wall, leaving her mark as one of the disappeared, but she and Claude relent. Together they…
Модальности:
Генерация текста
Задача: Генерация текста
Автор: emozilla
Теги: llama, custom_code, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 13
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
