Here is a collective list of instruction dataset used for Neural Chat fine-tuning. The total number of instruction samples and tokens are about 1.1M and 326M respectively.
Type
Language
Dataset
Number
HC3
en
HC3
24K
dolly
en
databricks-dolly-15k
15K
alpaca-zh
zh
tigerbot-alpaca-zh-0.5m
500K
alpaca-en
en
TigerResearch/tigerbot-alpaca-en-50k50K
math
en
tigerbot-gsm-8k-en
8K
general
en
tigerbot-stackexchange-qa-en-0.5m
500K
The collective dataset has been validated on… See the full description on the dataset page:
https://huggingface.co/datasets/Intel/neural-chat-dataset-v1-1.
