Parable-Nanbeige4.2-3B-Claude-Fable-5-GGUF

Part of the Parable series: small local LLMs fine-tuned on genuine agent traces. This is the reasoning-style chat variant of Nanbeige4.2-3B (to our knowledge the first published fine-tune of this base), tuned on real Claude Fable 5 agent transcripts so its step-by-step reasoning voice carries into local use.

Full-precision weights: Parable-Nanbeige4.2-3B-Claude-Fable-5

Files

File	Quant	Size
Parable-Nanbeige4.2-3B-Claude-Fable-5-Q4_K_M.gguf	Q4_K_M	2.6 GB	recommended default
Parable-Nanbeige4.2-3B-Claude-Fable-5-Q5_K_M.gguf	Q5_K_M	3.0 GB
Parable-Nanbeige4.2-3B-Claude-Fable-5-Q6_K.gguf	Q6_K	3.4 GB
Parable-Nanbeige4.2-3B-Claude-Fable-5-Q8_0.gguf	Q8_0	4.4 GB
Parable-Nanbeige4.2-3B-Claude-Fable-5-F16.gguf	F16	8.4 GB	for re-quantizing

Usage

Nanbeige support landed in llama.cpp master on 2026-07-27, so you currently need llama.cpp built from master. As of 2026-07-29 no packaged release has it yet: Homebrew's bottle (build 10150) still fails with unknown model architecture: 'nanbeige' (verified), and Ollama / LM Studio bundle older builds too. Build from source until the next releases land:

git clone https://github.com/ggml-org/llama.cpp && cd llama.cpp
cmake -B build -DCMAKE_BUILD_TYPE=Release && cmake --build build -j

llama-cli -m Parable-Nanbeige4.2-3B-Claude-Fable-5-Q4_K_M.gguf \
  -ngl 99 -c 4096 \
  -p "Write a python function that reverses a string." --temp 0.6

Runtime status (2026-07-29, stated plainly): the merged weights generate correctly (verified via transformers on GPU; sample outputs in the artifacts repo under verify/), and these GGUFs converted and quantized cleanly with num_loops=2 in the header. However, in our tests CPU-only inference of this architecture hangs with current llama.cpp master builds; the architecture is days old and its CPU path appears immature upstream. Prefer GPU offload (-ngl 99), or use the full-precision repo with transformers until upstream stabilizes. We will update this note as llama.cpp matures.

Output begins with a <think>...</think> reasoning block, then the answer. If you are building on top of this model, parse and strip the think block before showing text to end users.

The architecture runs its 22-layer stack twice (num_loops: 2, effective depth 44). This is encoded in the GGUF and handled automatically; expect roughly 2x the per-token compute of a typical 22-layer model of this width.

Model details

• Base: Nanbeige/Nanbeige4.2-3B (4.17B params, 3B non-embedding, Apache-2.0)
• Method: QLoRA (nf4, r=16, alpha=32) on the standard 7-module target set, all 22 layers (154 adapted modules)
• Data: 845 genuine Claude Fable 5 agent-session traces (chat variant: prose answers, <think> reasoning preserved) + 362 replay rows (OpenCoder educational_instruct + tulu-3 mix) to limit forgetting; 98% of rows retain their full assistant span at the 3,072-token training window
• Schedule: 380 optimizer steps, effective batch 8, completion-only loss masking, cosine LR
• Held-out trace loss: 2.333 (base) → 2.133 (tuned), same 225-trace held-out set never used in training

Evaluation

The number above is a held-out language-modeling loss on real agent traces: it measures how much better the model fits the Fable-5 reasoning distribution, and it is the honest headline for what this fine-tune does.

We do not claim benchmark gains over the base model. Nanbeige4.2-3B's model card reports strong agentic results (SWE-bench Verified 63.6, Terminal-Bench 2.0 44.1, vendor self-reported); this variant trains the reasoning voice, not tool-calling, and prose-only SFT should be expected to trade a little benchmark sharpness for style fidelity. If you need maximum benchmark performance on agentic harnesses, use the base; if you want its capability with a transparent, well-structured reasoning trace in local use, use this.

Limitations

• The base model's own benchmark figures are vendor self-reported and were not independently reproduced here.
• Training ran on a 4-bit quantized base (QLoRA); the F16 merge cannot exceed 4-bit-base quality.
• Requires llama.cpp built from master (2026-07-27+). No packaged release (brew, Ollama, LM Studio) loads this architecture yet; they will as their bundled llama.cpp catches up.

Quantization

Quantized with llama.cpp llama-quantize from the F16 merge, num_loops=2 verified present in the GGUF header at conversion time.

Provenance & licensing

Fine-tuned from Nanbeige/Nanbeige4.2-3B (Apache-2.0). Training data: Glint-Research/Fable-5-traces (AGPL-3.0) and Roman1111111/gpt5.5-terminal (MIT). Because those traces originate from third-party assistants, the providers' terms may apply to downstream training and distillation. If you plan to build on this model commercially, confirm your use aligns with those terms.

Citation

The recipe, evaluation methodology and failure analysis behind this model are documented in the tech report:

Aglawe, A. (2026). Agent-Trace Fine-Tuning of Small Language Models under Constrained Compute. Zenodo. doi:10.5281/zenodo.21676407

@misc{aglawe2026agenttrace,
  author    = {Aglawe, Ankit},
  title     = {Agent-Trace Fine-Tuning of Small Language Models under Constrained Compute},
  year      = {2026},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.21676407},
  url       = {https://doi.org/10.5281/zenodo.21676407}
}

Acknowledgements

The Nanbeige team for the base model; Glint-Research and Roman1111111 for the trace datasets; empero-ai for the recipe this series iterates on.