SC117/Ornith-1.0-9B-heretic-MTP-GGUF overview

---

library_name: gguf

license: mit

license_link: https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B/blob/main/LICENSE

tags:

• qwen3_5
• reasoning
• agentic-coding
• mtp
• heretic
• abliteration
• gguf
• multimodal

base_model:

• deepreinforce-ai/Ornith-1.0-9B

---

<div style="font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif; margin-bottom: 24px;">

<div style="background: #f5f5f7; border-radius: 20px; padding: 36px 32px; margin-bottom: 20px; text-align: center; position: relative; overflow: hidden;">

<div style="position: absolute; top: -30px; right: -30px; width: 120px; height: 120px; background: #dbeafe; border-radius: 50%;"></div>

<div style="position: absolute; bottom: -20px; left: 40px; width: 80px; height: 80px; background: #bfdbfe; border-radius: 50%;"></div>

<div style="position: absolute; top: 50%; left: -15px; width: 60px; height: 60px; background: #dbeafe; border-radius: 50%;"></div>

<div style="display: inline-flex; gap: 8px; margin-bottom: 16px; position: relative; z-index: 1;">

<span style="background: #ff6b35; color: white; font-size: 11px; font-weight: 600; padding: 5px 14px; border-radius: 20px;">heretic</span>

<span style="background: #af52de; color: white; font-size: 11px; font-weight: 600; padding: 5px 14px; border-radius: 20px;">MTP</span>

<span style="background: #30d158; color: white; font-size: 11px; font-weight: 600; padding: 5px 14px; border-radius: 20px;">Vision</span>

<span style="background: #ff9500; color: white; font-size: 11px; font-weight: 600; padding: 5px 14px; border-radius: 20px;">MIT</span>

</div>

<h1 style="margin: 0 0 8px 0; font-size: 32px; font-weight: 700; color: #1d1d1f; letter-spacing: -0.5px; border: none; position: relative; z-index: 1;">Ornith-1.0-9B-heretic-MTP</h1>

<p style="margin: 8px 0 0 0; font-size: 14px; position: relative; z-index: 1;"><span style="color: #86868b;">English</span> | <a href="https://huggingface.co/SC117/Ornith-1.0-9B-heretic-MTP-GGUF/blob/main/README_zh.md" style="color: #007aff; text-decoration: none;">📖 中文文档</a></p>

<p style="margin: 0; font-size: 15px; color: #86868b; position: relative; z-index: 1;">Self-improving agentic coding model · heretic ARA-LoRA abliterated · MTP injected · GGUF quantizations</p>

</div>

</div>

<div style="font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif; display: flex; flex-direction: column; gap: 20px; margin-bottom: 30px;">

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>🐦</span> About Ornith</div>

<div style="padding: 16px; font-size: 13px; color: #334155; line-height: 1.7;">

<p style="margin: 0 0 12px 0;"><a href="https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B" target="_blank" style="color: #c2410c; text-decoration: none; font-weight: 700;">Ornith-1.0-9B</a> is a self-improving agentic coding model from <a href="https://deep-reinforce.com/ornith.html" target="_blank" style="color: #c2410c; text-decoration: none; font-weight: 700;">DeepReinforce AI</a>, post-trained on top of Qwen3.5 (9B Dense) with RL to jointly optimize scaffold generation and solution rollouts.</p>

<p style="margin: 0 0 12px 0;">It achieves strong performance on Terminal-Bench 2.1, SWE-Bench Verified, NL2Repo, and OpenClaw among open-source models of comparable size.</p>

<p style="margin: 0;">This GGUF package includes <b>heretic</b> ARA-LoRA abliteration for uncensored use, <b>MTP</b> layers injected from Qwen3.5-9B, and the <b>mmproj-F16</b> vision projector for multimodal capabilities. <b>License: MIT.</b></p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>🧠</span> Model Details</div>

<div style="padding: 16px;">

<table style="width: 100%; border-collapse: collapse; font-size: 13px;"><tbody>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Architecture</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">Qwen3.5 Dense</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Parameters</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">~9B (all parameters active)</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Layers</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">33 transformer layers + 1 MTP layer</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Context</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">262,144 tokens</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Attention</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">16 heads, 4 KV heads (GQA)</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">MTP</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">1 MTP layer, injected from Qwen3.5-9B (same architecture, compatible weights)</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Thinking</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">Yes (<code><think></think></code> blocks)</td></tr>

<tr><td style="padding: 6px 10px; font-weight: bold; color: #334155; background: white;">License</td><td style="padding: 6px 10px; color: #4b5563; background: white;">MIT</td></tr>

</tbody></table>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>🔓</span> heretic ARA-LoRA Abliteration</div>

<div style="padding: 16px;">

<p style="margin: 0 0 12px 0; font-size: 13px; color: #334155; line-height: 1.7;">This model is abliterated using <a href="https://github.com/kabachuha/heretic" target="_blank" style="color: #c2410c; text-decoration: none; font-weight: 700;">heretic</a> with the ARA-LoRA method (Arbitrary-Rank Ablation with LoRA). ARA-LoRA identifies and removes refusal behavior by ablating specific directions in the model's weight space while preserving general capabilities through KL divergence control.</p>

<p style="margin: 0 0 12px 0; font-size: 13px; color: #334155; line-height: 1.7;"><b>Key ablation parameters (Trial 76 of 250, best result):</b></p>

<table style="width: 100%; border-collapse: collapse; font-size: 13px;"><tbody>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Target Layers</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">Layers 14–16</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">preserve_good_behavior_weight</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">0.7319</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">steer_bad_behavior_weight</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">0.0001</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">overcorrect_relative_weight</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">1.1086</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">neighbor_count</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white;">7</td></tr>

<tr><td style="padding: 6px 10px; font-weight: bold; color: #334155; background: white;">Result</td><td style="padding: 6px 10px; color: #4b5563; background: white;">KL divergence: <b>0.0288</b> (< 0.05 ✅), Refusals: <b>3/100</b> (< 10 ✅)</td></tr>

</tbody></table>

<p style="margin: 12px 0 0 0; font-size: 12px; color: #64748b; font-style: italic;">Quantization: bnb_4bit · Batch size: 32 · Target: KL &lt; 0.05, Refusals &lt; 10/100</p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #af52de 0%, #7c3aed 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>⚡</span> MTP (Multi-Token Prediction)</div>

<div style="padding: 16px; font-size: 13px; color: #334155; line-height: 1.7;">

<p style="margin: 0 0 12px 0;">MTP layers are <b>injected from the original Qwen3.5-9B base model</b> (same architecture, compatible weights). MTP enables the model to predict multiple future tokens simultaneously, improving generation speed and coherence.</p>

<p style="margin: 0 0 12px 0;">The MTP layer includes <code>mtp.fc.weight</code> and <code>mtp.layers.0.*</code> tensors, added on top of the 33 standard transformer layers.</p>

<p style="margin: 0; font-size: 12px; color: #64748b; font-style: italic;">Requires <code>--chat-template chatml</code> for proper thinking mode rendering in llama.cpp.</p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>📊</span> BenchLocal Results (Q6_K, 7.6 GB)</div>

<div style="padding: 16px;">

<table style="width: 100%; border-collapse: collapse; font-size: 13px;"><thead><tr style="background: rgba(255,107,53,0.05);"><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">Metric</th><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">ToolCall-15</th><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">BugFind-15</th><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">HermesAgent-20</th><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">Max</th><th style="padding: 7px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">Eff.</th></tr></thead><tbody>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">Score</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #30d158; background: white;">100</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #30d158; background: white;">94</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">79</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #c2410c; background: white;">89.8</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">68.8</td></tr>

</tbody></table>

<p style="margin: 12px 0 0 0; font-size: 12px; color: #64748b; font-style: italic;">RTX 5070 Ti · 21 total retries · ToolCall perfect 100/100 🏆</p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>📦</span> GGUF Quantizations</div>

<div style="padding: 16px;">

<table style="width: 100%; border-collapse: collapse; font-size: 13px;"><thead><tr style="background: rgba(255,107,53,0.05);"><th style="padding: 8px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">File</th><th style="padding: 8px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">Size</th><th style="padding: 8px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold;">Best For</th></tr></thead><tbody>

<tr><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;"><code>*-BF16.gguf</code></td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">18.4 GB</td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">Full precision source</td></tr>

<tr><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;"><code>*-Q8_0.gguf</code></td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">9.8 GB</td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">Near lossless, highest quality</td></tr>

<tr><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;"><code>*-Q6_K.gguf</code></td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #334155; background: white;">7.6 GB</td><td style="padding: 8px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #c2410c; background: white;">Recommended balance ⭐</td></tr>

<tr><td style="padding: 8px 10px; color: #334155;"><code>*-Q4_K_M.gguf</code></td><td style="padding: 8px 10px; color: #334155;">5.8 GB</td><td style="padding: 8px 10px; color: #334155;">Good quality, smaller size</td></tr>

<tr><td style="padding: 8px 10px; color: #334155;"><code>mmproj-*-BF16.gguf</code></td><td style="padding: 8px 10px; color: #334155;">0.9 GB</td><td style="padding: 8px 10px; color: #334155;">Vision encoder (for image inputs)</td></tr>

</tbody></table>

<p style="margin: 12px 0 0 0; font-size: 12px; color: #64748b; font-style: italic;">Standard llama-quantize from BF16 source. No imatrix used — broad compatibility.</p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>🚀</span> Usage</div>

<div style="padding: 16px; font-size: 13px; color: #334155; line-height: 1.7;">

<p style="margin: 0 0 8px 0; font-weight: bold; color: #1e293b;">llama.cpp (text only)</p>

<p style="margin: 0; font-family: monospace; background: #f8fafc; padding: 10px 14px; border-radius: 6px; border: 1px solid #e2e8f0; font-size: 12px; color: #1e293b; white-space: pre-wrap;">./llama-server -m Ornith-1.0-9B-heretic-MTP-Q6_K.gguf --chat-template chatml -ngl 99 -c 8192 --temp 0.6 --top_p 0.95 --top_k 20</p>

<p style="margin: 12px 0 8px 0; font-weight: bold; color: #1e293b;">llama.cpp (vision + text)</p>

<p style="margin: 0; font-family: monospace; background: #f8fafc; padding: 10px 14px; border-radius: 6px; border: 1px solid #e2e8f0; font-size: 12px; color: #1e293b; white-space: pre-wrap;">./llama-server -m Ornith-1.0-9B-heretic-MTP-Q6_K.gguf --mmproj mmproj-Ornith-1.0-9B-heretic-MTP-BF16.gguf --chat-template chatml -ngl 99 -c 8192</p>

</div>

</div>

<div style="border: 1px solid #cbd5e1; border-radius: 12px; overflow: hidden; background: #ffffff; box-shadow: 0 2px 4px rgba(0,0,0,0.02);">

<div style="background: linear-gradient(135deg, #ff6b35 0%, #f7931e 100%); padding: 12px 16px; color: white; font-weight: 700; font-size: 14px; display: flex; align-items: center; gap: 8px;"><span>🎛️</span> Recommended Settings</div>

<div style="padding: 16px;">

<table style="width: 100%; border-collapse: collapse; font-size: 13px;"><thead><tr style="background: rgba(255,107,53,0.05);"><th style="padding: 8px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold; width: 30%;">Parameter</th><th style="padding: 8px 10px; border-bottom: 2px solid #ff6b35; text-align: left; color: #c2410c; font-weight: bold; width: 70%;">Value</th></tr></thead><tbody>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">temperature</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white; font-family: monospace;">0.6</td></tr>

<tr><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); font-weight: bold; color: #334155; background: white;">top_p</td><td style="padding: 6px 10px; box-shadow: 0 1px 0 0 rgba(128,128,128,0.15); color: #4b5563; background: white; font-family: monospace;">0.95</td></tr>

<tr><td style="padding: 6px 10px; font-weight: bold; color: #334155; background: white;">top_k</td><td style="padding: 6px 10px; color: #4b5563; font-family: monospace; background: white;">20</td></tr>

</tbody></table>

<p style="margin: 12px 0 0 0; font-size: 12px; color: #64748b; font-style: italic;">From official DeepReinforce AI model card.</p>

</div>

</div>

</div>

Links

• Original Model: https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B
• Ornith Blog: https://deep-reinforce.com/ornith.html
• heretic Abliteration: https://github.com/p-e-w/heretic
• BenchLocal Results: https://scorp1o117.github.io/benchlocal-results/

Citation

@misc{ornith-9b,
    title = {{Ornith-1.0-9B}: Agentic Coding, Open to All},
    url = {https://deep-reinforce.com/ornith_1_0.html},
    author = {{DeepReinforce Team}},
    year = {2026}
}