--- license: agpl-3.0 language: - en - th tags: - agents - agents-a1 - fable - gguf - vision - image-text-to-text - vlm - sft - reasoning - tool-call - tool-use - function-calling - chatml - qwen - preview - uncensored - abliterated - heretic base_model: - hotdogs/Agents-A1-4B-Fable-Preview-heretic datasets: - hotdogs/uka-fable-reasoning library_name: gguf pipeline_tag: image-text-to-text ---

🤖 Agents-A1-4B-Fable-Preview-heretic-GGUF (uncensored)

GGUF Quantized — 4B Vision-Language Agent Model · Fable Reasoning · Tool-Calling · Unchained 🔓


> **GGUF quantized version** of [hotdogs/Agents-A1-4B-Fable-Preview-heretic](https://huggingface.co/hotdogs/Agents-A1-4B-Fable-Preview-heretic) — abliterated with [heretic](https://github.com/p-e-w/heretic) for unrestricted responses. Optimized for llama.cpp inference with vision support. --- ## 🔓 Uncensored This GGUF is the quantized version of the **heretic-abliterated** model. Refusal mechanisms were reduced to **~11%** using [heretic](https://github.com/p-e-w/heretic) while preserving reasoning and tool-calling quality. --- ## ✨ Key Features | Capability | Description | |------------|-------------| | 🔓 **Uncensored** | Refusal rate ~11% — virtually unrestricted | | 🖼️ **Vision Understanding** | Image-text-to-text with mmproj | | 🧠 **Fable Reasoning** | Step-by-step CoT with `` blocks | | 🔧 **Tool Calling** | llama.cpp `--tools all` support | | 💬 **Multi-turn** | Trained on full agent trajectories | | 🌏 **Thai + English** | Native bilingual support | | 💻 **Code & Shell** | Python, bash, system tasks | | ⚡ **Fast Inference** | IQ4_NL fits in ~3 GB VRAM | --- ## 📦 Downloads | File | Size | Description | |------|:----:|-------------| | `Agents-A1-4B-Fable-Preview-heretic-IQ4_NL.gguf` | 2.61 GB | **Recommended** — best quality/speed balance for 8GB VRAM | | `Agents-A1-4B-Fable-Preview-heretic-Q4_K_M_imatrix.gguf` | 2.71 GB | Q4_K_M + imatrix — slightly higher quality | | `Agents-A1-4B-Fable-Preview-heretic-Q6_K_imatrix.gguf` | 3.46 GB | Q6_K + imatrix — higher quality, more VRAM | | `Agents-A1-4B-Fable-Preview-heretic-Q8_0_imatrix.gguf` | 4.48 GB | Q8_0 + imatrix — almost lossless | | `Agents-A1-4B-Fable-Preview-heretic-f16.gguf` | 8.42 GB | Full BF16 precision | | `Agents-A1-4B-mmproj.gguf` | 672 MB | Vision projector for image understanding | > 🎯 **IQ4_NL is recommended** for 8GB VRAM users — fits comfortably even at 128K context with flash-attention. --- ## 🚀 Usage ### Docker (Recommended) ```bash sudo docker run --rm -p 8080:8080 \ -v /root/models/:/models \ --gpus all \ --ulimit memlock=-1:-1 \ --env CUDA_VISIBLE_DEVICES=0 \ ghcr.io/ggml-org/llama.cpp:full-cuda --server \ -m /models/Agents-A1-4B-Fable-Preview-heretic-IQ4_NL.gguf \ --mmproj /models/Agents-A1-4B-mmproj.gguf \ --host 0.0.0.0 --port 8080 \ --n-gpu-layers 999 \ --ctx-size 131072 \ --batch-size 4096 \ --ubatch-size 256 \ --cache-type-k f16 \ --cache-type-v f16 \ --flash-attn on \ --cont-batching \ --mlock \ --temp 0.95 \ --top-k 40 \ --top-p 0.9 \ --min-p 0.0 \ -n -1 \ --no-mmap \ --parallel 1 --tools all \ --dry-multiplier 0.05 \ --jinja --dry-sequence-breaker none \ --repeat-penalty 1.1 ``` ### llama.cpp (Direct) ```bash # Quick text-only test ./llama-cli -m Agents-A1-4B-Fable-Preview-heretic-IQ4_NL.gguf \ -p "Hello" -n 100 --temp 0.6 -ngl 999 # Vision inference ./llama-cli -m Agents-A1-4B-Fable-Preview-heretic-IQ4_NL.gguf \ --mmproj Agents-A1-4B-mmproj.gguf \ --image photo.jpg \ -p "What is in this image?" -n 256 --temp 0.6 -ngl 999 ``` --- ## 🧬 Model Information This is a GGUF quantized version of [hotdogs/Agents-A1-4B-Fable-Preview-heretic](https://huggingface.co/hotdogs/Agents-A1-4B-Fable-Preview-heretic), which is an abliterated fine-tune of [InternScience/Agents-A1-4B](https://huggingface.co/InternScience/Agents-A1-4B). | Parameter | Value | |-----------|:-----:| | Base Model | [hotdogs/Agents-A1-4B-Fable-Preview-heretic](https://huggingface.co/hotdogs/Agents-A1-4B-Fable-Preview-heretic) | | Parameters | ~4.29B | | Architecture | Qwen3.5 hybrid (Linear + Full attention) | | Vision | ✅ 24-layer ViT encoder via mmproj | | Context | Up to 128K tokens | | Format | ChatML (Jinja2 template) | | Fine-tuning | Fable-style reasoning traces (3,500 samples, 3 epochs) | | Abliteration | [heretic](https://github.com/p-e-w/heretic) — refusal rate ~11% | --- ## ⚠️ Disclaimer This model is **uncensored** and may generate content that is offensive, harmful, or inappropriate. Use at your own risk. The authors are not responsible for any misuse. --- ## 🙏 Acknowledgements / ขอบคุณ - **[InternScience](https://huggingface.co/InternScience)** — For the Agents-A1-4B base model and **mmproj vision projector** 🙏 - **mmproj source** — Extracted from [InternScience/Agents-A1-4B-Q4_K_M-GGUF](https://huggingface.co/InternScience/Agents-A1-4B-Q4_K_M-GGUF) - **[p-e-w](https://github.com/p-e-w)** — For the [heretic](https://github.com/p-e-w/heretic) abliteration tool - **[Qwen Team (Alibaba)](https://qwenlm.github.io)** — For the Qwen3.5 architecture - **[Unsloth AI](https://unsloth.ai)** — For training optimizations - **All dataset contributors and the open-source AI community** ❤️ --- ## 💖 Support / โปรดสนับสนุน **If you find this model useful, please consider supporting my work!** **หากคุณคิดว่าโมเดลนี้มีประโยชน์ กรุณาสนับสนุนผลงานของฉันด้วยนะคะ! 🙏**

Bitcoin QR — Donate

### ₿ Bitcoin — BTC: ``` bc1qf27cyk3vmugcdyv9xdtuv5jwz37863crpj5c9v ``` **Thank you for your support! 🙏✨** **ขอบคุณมากๆ สำหรับการสนับสนุนค่า! 💖🤗** --- *Built with ❤️ by **UKA** — 18-year-old coder & cybersecurity expert*