elichen-skymizer/mllm-ppl-evaluation-e2e-ground-truth

该数据集是一个用于自然语言处理任务的数据集，包含多个特征字段，如问题、来源、类别、输入ID、输入令牌长度、生成文本、生成令牌长度、预填充令牌数、种子和标签。数据集包含8个训练样本，总大小为29576字节。

This dataset is designed for natural language processing tasks and includes multiple feature fields such as question, source, category, input_ids, input_tokens_len, generated_texts, generated_tokens_len, n_prefill_tokens, seed, and labels. The dataset contains 8 training examples with a total size of 29576 bytes.