Pinkstack/Superthoughts-lite-v1

Text Generation • Updated 3 days ago • 1.6k • • 2

Note From the super thoughts series, trained on SFT & GRPO to reason properly.

Pinkstack/PARM-V2-QwQ-Qwen-2.5-o1-3B-GGUF

Text Generation • Updated Jan 15 • 550 • 6

Note From the PARM V2 series, trained on a mix of 4 SFT datasets.

Pinkstack/Parm-2-CoT-14B-16k-o1-QwQ

Text Generation • Updated 12 days ago • 281 • 7

Note Trained on just a QwQ distil dataset, generally liked but does not compare to superthoughts lite.