Cross-link: DistilQwen collection spotlight — 2026-03-29
Browse files
README.md
CHANGED
|
@@ -218,3 +218,19 @@ HuggingFace: [reaperdoesntknow](https://huggingface.co/reaperdoesntknow)
|
|
| 218 |
|
| 219 |
|
| 220 |
*Last updated: 2026-03-28 12:56 UTC*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
|
| 219 |
|
| 220 |
*Last updated: 2026-03-28 12:56 UTC*
|
| 221 |
+
|
| 222 |
+
<!-- CIX-CROSSLINK-START -->
|
| 223 |
+
|
| 224 |
+
---
|
| 225 |
+
|
| 226 |
+
## From the Convergent Intelligence Portfolio
|
| 227 |
+
|
| 228 |
+
**[DistilQwen Collection](https://huggingface.co/collections/reaperdoesntknow/distilqwen-69bf40ec669117e3f069ef1c)** — Proof-weighted distillation from Qwen3-30B-A3B → 1.7B and 0.6B. Three teacher variants (Instruct, Thinking, Coder), nine models, 2,788 combined downloads. Structure beats scale.
|
| 229 |
+
|
| 230 |
+
Top model: [Qwen3-1.7B-Coder-Distilled-SFT](https://huggingface.co/reaperdoesntknow/Qwen3-1.7B-Coder-Distilled-SFT) — 508 downloads
|
| 231 |
+
|
| 232 |
+
Full methodology: [Structure Over Scale (DOI: 10.57967/hf/8165)](https://doi.org/10.57967/hf/8165)
|
| 233 |
+
|
| 234 |
+
*Convergent Intelligence LLC: Research Division*
|
| 235 |
+
|
| 236 |
+
<!-- CIX-CROSSLINK-END -->
|