# Autoregressive generation

Ankh3 is a protein language model based on the T5 architecture. It is pre - trained by jointly optimizing masked language modeling and sequence completion tasks, and is suitable for protein feature extraction and sequence analysis.

This is a multimodal model fine-tuned based on Microsoft's Kosmos-2 model, specifically designed for recognizing Pokemon names on Pokemon cards.

A 7B-parameter large language model released by Salesforce AI Research, supporting 8K long sequence input, open-sourced under Apache-2.0 license

BioMedLM 2.7B is a specialized 2.7-billion-parameter language model trained on biomedical texts, demonstrating outstanding performance in biomedical NLP tasks.

A speech-to-text (S2T) model for automatic speech recognition (ASR)

A speech-to-text (S2T) model for automatic speech recognition (ASR), based on a sequence-to-sequence transformer architecture

s2t-small-librispeech-asr is a speech-to-text (S2T) model for automatic speech recognition (ASR), based on a sequence-to-sequence transformer architecture.

A speech-to-text (S2T) model for automatic speech recognition (ASR), based on a sequence-to-sequence transformer architecture

An end-to-end sequence-to-sequence transformer model for automatic speech recognition (ASR), trained on the LibriSpeech dataset

REBEL is a BART-based sequence-to-sequence model for end-to-end relation extraction, supporting over 200 different relation types.