jinaai/code_exercises
Viewer
•
Updated
•
1.47M
•
315
•
32
'Textbook Qualitiy' Datasets
High quality, education focused data sets for replicating the training process used on Microsoft Phi.
Note Synthetic, ChatGPT 3.5
nampdn-ai/tiny-textbooks
Viewer • Updated • 420k • 97 • 147Note falcon-refinedweb and minipile (cc-by-nc-sa-4.0)
nampdn-ai/tiny-webtext
Viewer • Updated • 2.32M • 50 • 32Note Web, from RefinedWeb for Falcon (MIT)
nampdn-ai/tiny-codes
Viewer • Updated • 1.63M • 174 • 232Note Synthetic, Unnamed Open Source Model (MIT)
teknium/openhermes
Viewer • Updated • 243k • 299 • 201Note Synthetic, GPT 4
LDJnr/Puffin
Viewer • Updated • 3k • 255 • 91Note Synthetic, GPT-4