MidTool/MidTool-Mix
Viewer • Updated • 11.2M • 586 • 4
Note Mid-training data mixture
Note Mid-training from Qwen3-4B-Base
Note Mid-training + SFT + RL from Qwen3-4B-Base
Note Mid-training from Qwen3-8B-Base
Note Mid-training + SFT + RL from Qwen3-8B-Base
Note Model used in data filtering for pdf data
Note Model used in data filtering for web data