DistilQwen
Proof-weighted distillation, Qwen3-30B to 1.7B/0.6B. Three teachers: Instruct, Thinking, Coder. The core method series. DOI 10.57967/hf/8165
Text Generation • 2B • Updated • 2.57k • 2Note 30B teacher, 1.7B student. Proof-weighted KD at 2.25× on reasoning.
reaperdoesntknow/Qwen3-1.7B-Distilled-30B-A3B-SFT-GGUF
Text Generation • 2B • Updated • 3.35kNote Edge deployment of the full Instruct pipeline. Apache 2.0.
reaperdoesntknow/Qwen3-1.7B-Distilled-30B-A3B-SFT
2B • Updated • 360Note Source model for the SFT-GGUF quantizations.
reaperdoesntknow/Qwen3-0.6B-Distilled-30B-A3B
Text Generation • 0.8B • Updated • 2.38kNote 0.6B student. Proves the methodology works at extreme scales.
reaperdoesntknow/Qwen3-0.6B-Distilled-30B-A3B-Thinking-SFT
Text Generation • 0.8B • Updated • 2.74k • 2Note The smallest model with Thinking teacher signal. 0.8B params.
reaperdoesntknow/Qwen3-0.6B-Distilled-30B-A3B-Thinking-SFT-GGUF
Text Generation • 0.8B • Updated • 2.58kNote mradermacher also auto-quantized this one — 420+ shadow downloads.
reaperdoesntknow/Qwen3-1.7B-Coder-Distilled-SFT
Text Generation • 2B • Updated • 2.91k • 1Note Coder teacher. Structured decomposition → STEM derivation.
reaperdoesntknow/Qwen3-1.7B-Coder-Distilled-SFT-GGUF
Text Generation • 2B • Updated • 3.58k • 4Note Pair with Thinking variant for comparative analysis.
reaperdoesntknow/DistilQwen3-1.7B-uncensored
Text Generation • 2B • Updated • 2kNote Foundation for research applications requiring unfiltered output.
reaperdoesntknow/TopologicalQwen
Text Generation • 2B • Updated • 3.12k • 1Note TKD flagship. BV decomposition → jump detection → curriculum.
reaperdoesntknow/DiStil-Qwen3-1.7B-uncensored
2B • Updated • 374 • 2Note DISC-informed distillation. Uncensored. Research-focused.
reaperdoesntknow/Disctil-Qwen3-1.7B
Text Generation • 2B • Updated • 1.94k • 1Note Structural refinement via DISC operator before TKD stage.
reaperdoesntknow/DistilQwen3-1.7B-uncensored-GGUF
2B • Updated • 2.53k • 4Note Community validated — third-party quantizations exist.
reaperdoesntknow/Qwen3-1.7B-Thinking-Distil
Text Generation • 2B • Updated • 2.69k • 2Note The most popular model. Thinking teacher = richest signal.
reaperdoesntknow/LFM2.5-1.2B-Distilled-SFT
Text Generation • 1B • Updated • 1.82kNote Proves TKD works across architecture families, not just within Qwen.
reaperdoesntknow/Discrepancy_Calculus
UpdatedNote Continuous Thought Dynamics — mathematical backbone of DualMind.