T-LLaMA is a Tibetan large language model trained on the LLaMA2-7B model, constructed with a corpus containing 2.2 billion Tibetan characters, demonstrating excellent performance in text classification, generation, and summarization tasks.
Large Language Model
Transformers Other