JingweiNi/test_math_no_prm800k_Qwen3-8B_texts_uhead_prm

该数据集包含多个字段，其中包括文本字段如问题、答案和回复，以及数值字段如最大序列概率、平均token熵和困惑度。还有与模型相关的特征，如JingweiNi/uhead_claim_Qwen3-8B_fixed_prm_layer1_dim512_head16_e5_lr5e-4_pos3等。数据集的结构中包含saved_stats字段，其中有多种类型的字段，包括列表和null类型。数据集分为训练集，包含23296个示例，文件大小为595399880字节。数据集的下载大小为47061449字节。

The dataset includes various fields such as text fields like question, answer, and reply, numerical features such as MaximumSequenceProbability, MeanTokenEntropy, and Perplexity, and model-related features like JingweiNi/uhead_claim_Qwen3-8B_fixed_prm_layer1_dim512_head16_e5_lr5e-4_pos3. The dataset structure contains the saved_stats field with various types including lists and null types. The dataset is split into a training set with 23296 examples and a file size of 595399880 bytes. The download size of the dataset is 47061449 bytes.