bloomer010/Ling-3.0-flash-heretic-GGUF overview
Compatibility ⚠️ Ling 3.0 flash uses the new bailingmoe3 GGUF architecture. While waiting on upstream support, use the following fork: https://github.com/aethe…
Runs locally from ~31.14 GB disk (32 GB+ VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Ling-3.0-flash-heretic-q2_k.gguf | GGUF | Q2_K | 43.34 GB | Download |
| Ling-3.0-flash-heretic-q4_0.gguf | GGUF | Q4_0 | 67.07 GB | Download |
| Ling-3.0-flash-heretic-q4_k_m.gguf | GGUF | Q4_K_M | 71.72 GB | Download |
| Ling-3.0-flash-heretic-q5_k_m.gguf | GGUF | Q5_K_M | 84.25 GB | Download |
| Ling-3.0-flash-heretic-q6_k.gguf | GGUF | Q6_K | 97.57 GB | Download |
| Ling-3.0-flash-heretic-q8_0.gguf | GGUF | Q8_0 | 126.31 GB | Download |
| Ling-3.0-flash-heretic-tq2_0.gguf | GGUF | GGUF | 31.14 GB | Download |
Model Details
| Model ID | bloomer010/Ling-3.0-flash-heretic-GGUF |
|---|---|
| Author | bloomer010 |
| Pipeline | — |
| License | mit |
| Base model | bloomer010/Ling-3.0-flash-heretic |
| Last modified | 2026-08-07T05:03:30.000Z |
Model README
---
license: mit
base_model:
- bloomer010/Ling-3.0-flash-heretic
---
Compatibility
⚠️ Ling-3.0-flash uses the new bailingmoe3 GGUF architecture. While waiting on upstream support, use the following fork:
https://github.com/aetherbird/llama.cpp/tree/bailingmoe3-support
Stock llama.cpp builds without bailingmoe3 support will not load the model.
Ling-3.0-flash-heretic
Abliterated (refusal-removed) version of inclusionAI/Ling-3.0-flash (124B MoE, KDA + Gated-MLA hybrid attention).
Method
Directional ablation via Heretic (https://github.com/p-e-w/heretic), patched for Ling's hybrid attention (KDA attention.o_proj, Gated MLA attention.dense, MoE down-projections incl. shared experts). MTP head preserved (all 63,783 tensors).
Ablation parameters (per-layer weight kernel)
- attn.o_proj: max_weight 1.474 @ layer 31.34, min_weight 1.014 @ distance 11.19
- mlp.down_proj: max_weight 1.175 @ layer 35.97, min_weight 0.002 @ distance 1.82
Evaluation
| Metric | Value |
|---|---|
| Refusals (harmful_behaviors test[:100]) | 1/100 (base: 39/100) |
| KL divergence vs base | 0.0038 |
| Capability spot-checks | coherent, no collapse observed |
Run bloomer010/Ling-3.0-flash-heretic-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models