llm-jp-4.1-thinking-sft-data
收藏资源简介：
# llm-jp-4.1-thinking-sft-data ## Overview This dataset is a supervised fine-tuning (SFT) dataset used to train `llm-jp-4.1-*-thinking` models. This dataset is constructed from prompts and conversations collected from multiple data sources. For most subsets, reasoning processes and final responses used for LLM-jp-4.1 SFT were generated or augmented using [gpt-oss-120b](https://huggingface.co/openai/gpt-oss-120b). For the tool-calling and agentic data derived from NVIDIA Nemotron datasets, the original conversations and trajectories were generated using multiple language models. See the "Nemotron Agentic Mix" section below for details. The splits `reasoning_low`, `reasoning_medium`, and `reasoning_high` correspond to different reasoning effort settings used when generating responses with gpt-oss-120b. To support the continued development of LLM-jp, we would greatly appreciate it if you could share how you utilize LLM-jp outcomes via the [survey form](https://forms.gle/AvbNXTNT2ADsssHq5). ## Data Sources and Licenses `llm-jp-4.1-*-thinking` are constructed based on the following data sources. Since each dataset has its own license, please ensure compliance with the respective licenses when using this dataset. Some subsets cannot be redistributed due to licensing restrictions and are therefore not included. | Dataset | License | Notes | | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------- | ------------------------------------------------------- | | [ac_002](https://huggingface.co/datasets/llm-jp/AnswerCarefully) | answer-carefully-dataset-tou | Not included due to redistribution restrictions | | [daring_anteater](https://huggingface.co/datasets/nvidia/Daring-Anteater) | CC BY 4.0 | - | | [flan](https://huggingface.co/datasets/allenai/dolmino-mix-1124) | ODC-BY | - | | [ichikara](https://liat-aip.sakura.ne.jp/wp/llm%E3%81%AE%E3%81%9F%E3%82%81%E3%81%AE%E6%97%A5%E6%9C%AC%E8%AA%9E%E3%82%A4%E3%83%B3%E3%82%B9%E3%83%88%E3%83%A9%E3%82%AF%E3%82%B7%E3%83%A7%E3%83%B3%E3%83%87%E3%83%BC%E3%82%BF%E4%BD%9C%E6%88%90/) | - | Requires paid license, not included | | [jaster_v1.4.1](https://github.com/llm-jp/llm-jp-eval/tree/v1.4.1) | CC BY-SA 4.0, CC BY-SA 3.0, CC BY 4.0, BSD-3, Apache-2.0, MIT | - | | [llmjp_extraction_wiki_ja_v0.x](https://huggingface.co/datasets/llm-jp/extraction-wiki-ja) | Apache-2.0 | - | | [llmjp_magpie_sft_v1.0](https://huggingface.co/datasets/llm-jp/magpie-sft-v1.0) | Apache-2.0 | - | | llmc_math_dataset | - | To be released | | [logical_math_coding_wizard8x22b](https://huggingface.co/datasets/llm-jp/wizardlm8x22b-logical-math-coding-sft-ja) | Apache-2.0 | - | | [multiturn_calm3](https://huggingface.co/datasets/kanhatakeyama/AutoMultiTurnByCalm3-22B) | Apache-2.0, CC BY-SA 3.0, CC0, CC BY 4.0 | - | | [nemotron_post_v2_stem](https://huggingface.co/datasets/nvidia/Nemotron-Post-Training-Dataset-v2) | CC BY 4.0 | - | | [nemotron_post_v3_chat](https://huggingface.co/datasets/nvidia/Nemotron-Instruction-Following-Chat-v1) | CC BY 4.0 | - | | [nemotron_post_v3_if](https://huggingface.co/datasets/nvidia/Nemotron-Instruction-Following-Chat-v1) | CC BY 4.0 | - | | [nemotron_post_v3_math](https://huggingface.co/datasets/nvidia/Nemotron-Math-v2) | CC BY-SA 4.0 | - | | [random_to_fixed_multiturn_calm3](https://huggingface.co/datasets/kanhatakeyama/ramdom-to-fixed-multiturn-Calm3) | Apache-2.0 | - | | synthetic_if_ja | - | To be released | | self_system_question | CC BY 4.0 | Model-related questions | | [synthetic_jp_en_coding](https://huggingface.co/datasets/llm-jp/Synthetic-JP-EN-Coding-Dataset) | Apache-2.0 | - | | system_prompt_question | CC BY 4.0 | Model identity and metadata | | [nemotron3_sft_chat_v3](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Instruction-Following-Chat-v3) | CC BY 4.0, ODC-BY | - | | [nemotron3_sft_science_vendor](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Science-v2) | CC BY-SA 4.0 | - | | [table_gpt_train_large](https://huggingface.co/datasets/LipengCS/Table-GPT) | MIT | - | | [nemotron3_sft_multilingual_v2_math_ko_stackoverflow](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Multilingual-v2) | CC BY-SA 4.0 | - | | [nemotron3_sft_multilingual_v2_math_ja_stackoverflow](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Multilingual-v2) | CC BY-SA 4.0 | - | | nemotron_agentic_mix_v0.1.1 | CC BY 4.0 | Derived from [NVIDIA Nemotron-Agentic-v1](https://huggingface.co/datasets/nvidia/Nemotron-Agentic-v1) and [NVIDIA Nemotron-SFT-Agentic-v2](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Agentic-v2) | | nemotron_agentic_mix_v0.1.1_ja | CC BY 4.0 | Japanese translation of `nemotron_agentic_mix_v0.1.1` generated using `gpt-oss-120b` | ### Nemotron Agentic Mix `nemotron_agentic_mix_v0.1.1` is constructed from the following NVIDIA datasets: * [Nemotron-Agentic-v1](https://huggingface.co/datasets/nvidia/Nemotron-Agentic-v1) * [Nemotron-SFT-Agentic-v2](https://huggingface.co/datasets/nvidia/Nemotron-SFT-Agentic-v2) The source data include agentic conversations generated using multiple language models. For Nemotron-Agentic-v1, the `interactive_agent` subset was generated using `Qwen3-235B-A22B-Thinking-2507`, `Qwen3-32B`, `gpt-oss-120b`, and `Qwen3-235B-A22B-Instruct-2507`. The `tool_calling` subset was generated using `Qwen3-235B-A22B-Thinking-2507` and `Qwen3-235B-A22B-Instruct-2507`. For Nemotron-SFT-Agentic-v2, `Qwen3-235B-A22B-Thinking-2507`, `Qwen3-32B`, `Qwen3-235B-A22B-Instruct-2507`, `deepseek-r1-0528`, `DeepSeek-V3.2`, and `gpt-oss-120b`. `nemotron_agentic_mix_v0.1.1_ja` is the Japanese version of this data, translated using `gpt-oss-120b`. ## Data Format Each sample has the following format: ```json { "ID": "...", "messages": [ {"role": "system", "name": null, "content": ["..."]}, {"role": "user", "name": null, "content": ["..."]}, {"role": "assistant", "name": null, "content": ["..."], "channel": "analysis"}, {"role": "assistant", "name": null, "content": ["..."], "channel": "final"}, ... ] } ``` * `channel="analysis"` represents the reasoning process (chain-of-thought) * `channel="final"` represents the final response The `messages` field contains conversational data and is mostly compatible with [openai-harmony](https://github.com/openai/harmony). Starting with LLM-jp-4.1, we use a modified version of the Harmony format as the chat template. As a result, compatibility with openai-harmony is no longer guaranteed, particularly for tool-calling datasets such as `nemotron_agentic_mix_v0.1.1` and `nemotron_agentic_mix_v0.1.1_ja`. ## Notes * This dataset contains reasoning processes and responses generated by a large language model (`gpt-oss-120b`). * Generated content may include errors, inaccuracies, biases, or harmful outputs. ## Send Questions to llm-jp(at)nii.ac.jp ## Model Card Authors The names are listed in alphabetical order. Hirokazu Kiyomaru, Takashi Kodama, and Yunang Wu.
