Dataset · shibing624
ShareGPT GPT-4: multilingual dialogues
shibing624/sharegpt_gpt4
Curated multi-turn GPT-4 dialogues in Chinese, English, and other languages.
What's inside
The dataset contains multi-turn GPT-4 question-answer pairs selected from ShareGPT in multiple languages, including Chinese and English. The data is in JSONL format: the conversations field is a list of turns with from and value. According to metadata, languages are zh, en, gl, ko, with a size of 100K-1M.
What you can do with it
- SFT of a multilingual chat assistant
- Fine-tune a dialogue model
- Verification of multi-turn responses
- Tasks
- Text classification, Text generation
- Languages
- Chinese, English, Galician, Korean
- Size
- 100K–1M rows
- Downloads this month
- 1.5K
- Likes
- 139
Models trained on it
- mradermacher/AMD-OLMo-1B-IT-GGUF 501
- mradermacher/AMD-OLMo-1B-IT-i1-GGUF 448
- mradermacher/Qwen2.5-14B-Wernicke-SFT-GGUF 349
- HWERI/pythia-1.4b-deduped-sharegptText generation 131
- HWERI/pythia-70m-deduped-cleansharegpt-enText generation 91
- tensorblock/pythia-1.4b-deduped-sharegpt-GGUFText generation 91
- tensorblock/HWERI_pythia-70m-deduped-cleansharegpt-en-GGUF 77
- CultriX/Qwen2.5-14B-Wernicke-SFTText generation 74
How to load it
from datasets import load_dataset
ds = load_dataset("shibing624/sharegpt_gpt4")