wisenut-nlp-team/data_recipe

---
configs:
- config_name: v1_1k
data_files:
- split: train
path: v1/1k/data_recipe_v1_1k.json
- config_name: v1_5k
data_files:
- split: train
path: v1/5k/data_recipe_v1_5k.json
- config_name: v1_10k
data_files:
- split: train
path: v1/10k/data_recipe_v1_10k.json
- config_name: v2_10k
data_files:
- split: train
path: v2/10k/data_recipe_v2_10k.json
- config_name: v3_10k
data_files:
- split: train
path: v3/10k/data_recipe_v3_10k.json
---
**V1**
- 영어(8k)
- [ShareGPT](https://huggingface.co/datasets/shibing624/sharegpt_gpt4) : 3.24k
- [Claued 3 Opus](https://huggingface.co/datasets/QuietImpostor/Sao10K-Claude-3-Opus-Instruct-15K-ShareGPT) : 1.76k
- [Lima](https://huggingface.co/datasets/llamafactory/lima) : 1k
- [Slimorca](https://huggingface.co/datasets/Open-Orca/slimorca-deduped-cleaned-corrected) : 2k
- 한국어 (2k)
- [Alpca-GPT4](https://huggingface.co/datasets/FreedomIntelligence/alpaca-gpt4-korean) : 1k
- [SMR](https://huggingface.co/datasets/wisenut-nlp-team/llama_ko_smr) : 0.5k
- [MMLU](https://huggingface.co/datasets/HAERAE-HUB/KMMLU) : 0.5k

**V2**
- 영어(4k)
- [ShareGPT](https://huggingface.co/datasets/shibing624/sharegpt_gpt4) : 1k
- [Lima](https://huggingface.co/datasets/llamafactory/lima) : 1k
- [Slimorca](https://huggingface.co/datasets/Open-Orca/slimorca-deduped-cleaned-corrected) : 1k
- [Math](https://huggingface.co/datasets/EleutherAI/hendrycks_math) : 1k
- 한국어 (6k)
- [ShareGPT](https://huggingface.co/datasets/shibing624/sharegpt_gpt4) : 1k
- [Multi-turn](https://huggingface.co/datasets/maywell/koVast) : 1k
- [Alpca-GPT4](https://huggingface.co/datasets/FreedomIntelligence/alpaca-gpt4-korean) : 2k
- [SMR](https://huggingface.co/datasets/wisenut-nlp-team/llama_ko_smr) : 1k
- [MMLU](https://huggingface.co/datasets/ccw7463/Ko_MMLU_ver0.3) : 1k

**V3**
- 영어(7k)
- [ShareGPT](https://huggingface.co/datasets/shibing624/sharegpt_gpt4) : 3k
- [Claued 3 Opus](https://huggingface.co/datasets/QuietImpostor/Sao10K-Claude-3-Opus-Instruct-15K-ShareGPT) : 1k
- [Lima](https://huggingface.co/datasets/llamafactory/lima) : 1k
- [Slimorca](https://huggingface.co/datasets/Open-Orca/slimorca-deduped-cleaned-corrected) : 2k
- 한국어 (3k)
- [ShareGPT](https://huggingface.co/datasets/shibing624/sharegpt_gpt4) : 1k
- [Multi-turn](https://huggingface.co/datasets/maywell/koVast) : 1k
- [MMLU](https://huggingface.co/datasets/ccw7463/Ko_MMLU_ver0.3) : 1k