No official torrent yet β requesting it tells the flock what to host next. You cannot seed a catalog-only page.
Cataloged from empero-ai/Qwable-9B-Claude-Fable-5-GGUF
All models by empero-ai β[Image blocked: Qwable-9B-Claude-Fable-5]
Developed by Empero
GGUF quantizations of empero-ai/Qwable-9B-Claude-Fable-5
for llama.cpp, Ollama, LM Studio, and other GGUF runtimes. This repo
ships a vision projector (mmproj), so the model runs as a full multimodal (image + text) assistant β
not just text.
Qwable-9B-Claude-Fable-5 is a full-parameter fine-tune of Qwen3.5-9B on agentic coding and reasoning traces distilled from Claude Fable 5 and a GPT-5.5 terminal agent. For full training details and the complete evaluation, see the base model card.
Early release. Strong coding and agentic behavior out of the box; a full benchmark suite is underway and will be published. See Provenance & licensing.
Text-only use needs just a quant. For image understanding, download both a text quant and the mmproj.
Or via a Modelfile (pulls in the vision projector for image support):
temp 1.0, precise coding temp 0.6; top_p 0.95, top_k 20, min_p 0. Use repeat_penalty 1.05 (a small bump from Qwen's default 1.0) to avoid rare
non-terminating reasoning loops, and allow generous -n / max_new_tokens.<think>...</think> block before the final answer β
parse and strip that span for end users.mmproj) but was inherited,
not specifically tuned.mmproj), converted and quantized with llama.cpp.The evaluation below was measured on the unquantized fine-tune. Quantized variants are very close at Q8_0/Q6_K and degrade gradually toward Q4_K_M β expect a small quality drop at the lower quants.
Training quality was tracked via held-out validation loss / token-accuracy on a 100-example split (80% Fable / 20% terminal), plus a qualitative generation review:
No overfitting: held-out loss decreased then plateaued (~0.71) through epoch 2 β it never rose even as
train loss fell to ~0.64. In a 34-prompt qualitative review, roughly 27/34 responses were clean and
correct, strongest on coding and terminal/agentic tasks β current tooling (ss over netstat,
git-filter-repo, Argon2id) with security-aware judgment (rotating a leaked key first, constant-time
comparison). Full transcripts: sample_generations.md [blocked].
<think> block; strip it for end users and allow generous
output length. Use repeat_penaltyβ1.05 for consistently crisp completions.Converted from the fine-tuned weights with llama.cpp convert_hf_to_gguf.py, then quantized with
llama-quantize. The BF16 GGUF is the conversion base; the K-quants are derived from it. The mmproj is the
base Qwen3.5-VL vision encoder (unchanged by fine-tuning). All files were verified to load and generate
in llama.cpp β text (code, reasoning) and image understanding both confirmed.
Weights are released under Apache-2.0, inherited from the Qwen3.5-9B base. The fine-tuning data comes from generated traces of Claude Fable 5 and GPT-5.5 (via the linked public datasets). Because those traces originate from third-party assistants, the providers' terms may apply to downstream training and distillation β if you plan to build on this model commercially, confirm your use aligns with those terms. Shared with the community for research and experimentation, as-is.
If this model helped you, consider supporting the project:
bc1qx6zepu6sfkvshgdmc4ewu6pk6rpadvpgffpp7vltc1qv2mefzps2vtjcpwfx8xxdrpplrcvltswm68r7x42Dbm5xg5Nq26fdyzfEU7KBnAJfhi7Cvz5J2ex5CzHXkfKuNEJzYCcmJ1GTbgjFZ5MBx72sdG1G9239Cd6rsZfv4QeDkYJYGlint-Research/Fable-5-traces,
Roman1111111/gpt5.5-terminalllama-cli -m Qwable-9B-Claude-Fable-5-Q4_K_M.gguf --jinja \ -p "Write a Python function that merges two sorted lists." \ --temp 0.6 --top-p 0.95 --top-k 20 --repeat-penalty 1.05 -n 2048llama-mtmd-cli -m Qwable-9B-Claude-Fable-5-Q4_K_M.gguf \ --mmproj mmproj-Qwable-9B-Claude-Fable-5-f16.gguf \ --image photo.jpg -p "Describe this image." \ --temp 0.6 --top-p 0.95 --top-k 20 -n 512ollama run hf.co/empero-ai/Qwable-9B-Claude-Fable-5-GGUF:Q4_K_MFROM ./Qwable-9B-Claude-Fable-5-Q4_K_M.ggufFROM ./mmproj-Qwable-9B-Claude-Fable-5-f16.ggufPARAMETER temperature 0.6PARAMETER top_p 0.95PARAMETER top_k 20PARAMETER repeat_penalty 1.05