A 64-layer frankenmerge of two of Jackrong's incredible Qwen3.5-9B finetunes, stacking all 32 layers from each to create an ~18B parameter model, then healed with a 1000-step QLoRA fine-tune to smooth the layer boundary.
This was a fun experiment! A lot of people have been asking for something between Jackrong's 27B and 9B models — something that runs well on 12–16 GB GPUs. This frankenmerge is an attempt at filling that gap, and the results are surprisingly good.
Heal Fine-Tune — It Works
The raw frankenmerge had a known issue: garbled code output. Because two separately-trained models were stacked at layer 32, structured output (code blocks, HTML, bracket matching) would occasionally come out malformed or hallucinated.
We ran a 1000-step QLoRA heal fine-tune using Jackrong's own training data to let gradients flow across the layer boundary — and the results are significant:
HTML generation is now clean and production-quality. We tested a complex single-page weather dashboard (navbar, dark mode toggle, 5-day forecast grid, responsive sidebar, CSS variables, JavaScript) — the model produced 14,500+ chars of valid HTML/CSS/JS with perfectly balanced CSS braces, perfectly balanced JS parentheses, no garbled text, and a complete </html> closure.
Programming benchmark improved: 11/15 (raw) -> 12/15 (healed), recovering the longest_substring sliding-window algorithm test (8/8 cases passing)
Overall score improved: 39/44 -> 40/44 (90.9%), still comfortably beating Qwen 3.6 MoE (38/44) at less than half the VRAM
Loss dropped 39% during training (1.02 -> 0.62), confirming the layer boundary was a real source of error that training could address
The healed GGUF (Qwopus-GLM-18B-Healed-Q4_K_M.gguf) is the only version in this repo. If you're interested in the raw unhealed merge for research purposes, reach out.
This is still an experimental model — it may have quirks or issues. If you run into anything weird, or if you make something cool with it, reach out on X: @KyleHessling1
Benchmark Results
We ran a 44-test capability suite covering basic generation, reasoning, tool calling, agentic workflows, structured output, context handling, multilingual, programming, and performance.
The healed merge outperforms the brand new Qwen 3.6-35B-A3B MoE (Q4_K_M, 22 GB) despite being significantly smaller (Q4_K_M, 9.2 GB):
Category
Qwopus 9B (source)
Qwopus-GLM-18B (healed)
Qwen 3.6-35B MoE
Basic
6/6
6/6
5/6
Reasoning
4/4
4/4
4/4
Tool Calling
6/6
6/6
6/6
Agentic
4/4
4/4
4/4
Structured Output
2/2
2/2
2/2
Context
2/3
2/3
2/3
Multilingual
2/2
2/2
2/2
Programming
13/15
12/15
12/15
Performance
2/2
2/2
1/2
TOTAL
41/44 (93.2%)
40/44 (90.9%)
38/44 (86.4%)
Throughput
126.0 tok/s
66.0 tok/s
174.2 tok/s
GGUF Size
5.3 GB
9.2 GB
22 GB
Key Takeaways
40/44 tests passed (90.9% healed) — beats Qwen 3.6 MoE's 38/44 (86.4%) at less than half the VRAM
Heal training recovered programming capability: 11/15 raw -> 12/15 healed (matching Qwen 3.6 MoE)
Highest Chinese output density of any model tested: 129-138 CJK characters
~66 tok/s with low throughput variance — stable inference
Fits in 12 GB VRAM at Q4_K_M — runs on consumer GPUs like RTX 3060/4070
Heal Fine-Tune Details
The raw frankenmerge had code formatting issues (garbled code blocks, missing brackets). We ran a 1000-step QLoRA heal fine-tune using Jackrong's training data to smooth the layer-32 boundary:
Data: Blend of Jackrong/Qwen3.5-reasoning-700x (70%), Jackrong/Competitive-Programming-python-blend (15%), Jackrong/MultiReason-ChatAlpaca (15%)
Training: 1000 steps, batch 8, lr 2e-5 cosine, ~14 hours on RTX 5090
Loss: 1.02 -> 0.62 (39% reduction)
Result: Recovered 1 programming test, HTML/CSS output is now clean and production-quality
Where It Falls Short
Three programming tests still fail on the healed version: one function naming issue, one missing JS paren, and one that doesn't produce a code block for pytest generation. These are residual formatting artifacts from the merge.
Frontend Code Generation — Stress Test Results
We put the healed model through a rigorous frontend stress test: 6 increasingly complex HTML/CSS/JS generation tasks, each requiring thousands of tokens of structurally valid code output. The results speak for themselves:
Test
What We Asked For
Checks Passed
Output Size
Weather Dashboard
Responsive dashboard, CSS vars, dark mode toggle, 5-day forecast grid
9/9
14.5K chars
E-Commerce Product Page
Image gallery, color swatches, quantity selector, tabbed content, sticky mobile bar
SVG bar chart with tooltips, SVG donut chart, sortable data table, collapsible sidebar, dark theme
13/13
22.3K chars
Multi-Step Registration
3-step form wizard, real-time validation, password strength meter, state dropdown, animated transitions, success modal
12/12
23.3K chars
Snake Game
Canvas game loop, arrow key controls, collision detection, localStorage high score, increasing difficulty
11/12
11.2K chars
62/63 total checks passed (98.4%)
Every single output had:
Perfectly balanced CSS braces (zero imbalance across all 6 files)
Perfectly balanced JS parentheses (zero imbalance across all 6 files)
Zero garbled or hallucinated text
Working JavaScript — dark mode toggles, IntersectionObserver animations, SVG chart rendering, form validation, canvas game loops
The only miss: the Snake game had a minor closing tag typo (html> instead of </html>) at the very end.
This is remarkable for a frankenmerge of two 9B models with only 1000 steps of QLoRA healing. The model is producing production-quality frontend code — not just syntactically valid HTML, but sophisticated interactive applications with modern CSS (Grid, Flexbox, custom properties, keyframe animations) and non-trivial JavaScript (IntersectionObserver, requestAnimationFrame game loops, real-time form validation, SVG chart generation).
All 6 sample HTML files are included in the samples/ directory of this repo — download them and open in a browser to see for yourself.
All credit for the source models goes to Jackrong, who created both of these excellent finetunes. I just stacked the layers — the quality comes from his work.
Jackrong/Qwopus3.5-9B-v3.5
A reasoning-enhanced finetune of Qwen3.5-9B trained with ~2x more SFT data than v3, focused on structured reasoning, tool-augmented workflows, and multi-step agentic tasks.
Key insight from the v3.5 design: "Scaling high-quality SFT data may further enhance the generalization ability of large language models." Reasoning SFT helps models better utilize existing knowledge and activate latent knowledge through structured reasoning, rather than simply memorizing long Chain-of-Thought outputs.
A distilled variant of Qwen3.5-9B trained on high-quality reasoning data from a GLM-5.1 teacher model (~700x scale of Qwen3.5-reasoning-700x). Focused on structured reasoning ability, instruction-following consistency, and problem decomposition.
The model learns a structured reasoning scaffold: understand the task, break down the problem, reason step by step, then construct the final answer.
Training data:
Primary: Jackrong/GLM-5.1-Reasoning-1M-Cleaned (cleaned from Kassadin88/GLM-5.1-1000000x)
By stacking two differently-distilled reasoning models, this merge combines:
Qwopus v3.5's strengths in agentic tool use, code generation, and token-efficient reasoning (Opus-style training)
GLM-5.1 Distill's strengths in structured problem decomposition, instruction adherence, and chain-of-thought organization (GLM-style reasoning scaffold)
The hypothesis: deeper networks with diverse reasoning training produce more robust, capable models — and the benchmark results suggest it works, at least for the capabilities we tested.
Merge Details
Method: Passthrough frankenmerge (layer stacking)
Tool: Custom script (mergekit did not support Qwen3.5's hybrid linear/full attention architecture)
Embeddings / LM Head / Visual / MTP: From Qwopus3.5-9B-v3.5
The full BF16 safetensors are not included in this GGUF repo. If you need them for further fine-tuning or experimentation, reach out on X.
Limitations
Experimental frankenmerge — no additional training was done. Layer boundaries between the two source models may cause occasional coherence or formatting issues.
Code formatting — the model sometimes garbles fenced code blocks (returns code without proper markdown fencing). The reasoning is usually correct even when the formatting isn't.
Not exhaustively tested — this was a fun weekend project. There may be edge cases we haven't found yet.
Hallucination risk — as with all autoregressive LLMs, outputs may contain factual errors.
Supported Research
Both source models reference: Ren et al., 2026 — Rethinking Generalization in Reasoning SFT (arXiv:2604.06628)
Key findings: reasoning SFT generalizes when sufficiently trained; high-quality long-CoT data enables cross-domain transfer; stronger models learn reasoning structure, not just longer outputs.
Acknowledgements
Jackrong — the real MVP. Both source models, training pipelines, datasets, and documentation are his work. This merge exists because his finetunes are so good that even stacking them naively produces something surprisingly capable.
