MasonMac/CodeX-Thinking-Gemma-4-31B-IT

本数据集的所有提示词均源自Modotte/CodeX-2M-Thinking，该原始数据集包含每个提示的多个轨迹，而本数据集每个提示仅提供一个轨迹。生成过程使用了HuggingFace上的nvidia/Gemma-4-31B-IT-NVFP4模型，该模型配置了混合BF16/FP8权重，NVIDIA还设置了FP8键值缓存；基准测试显示，在代码生成任务中，其性能与BF16精度相似。生成过程中未使用系统提示词。All prompts were taken from Modotte/CodeX-2M-Thinking, which contains multiple traces per prompt whereas this dataset only provides one trace per prompt. Generations were with https://huggingface.co/nvidia/Gemma-4-31B-IT-NVFP4 (a mix of BF16/FP8 weights that NVIDIA configured with FP8 KV cache; benchmarks show performs similarly to BF16 for coding). No system prompt was used.