VitreousCut/Tiel-Inspired-Coder-9B-GGUF

Tiel-Inspired-Coder-9B (GGUF)

Tiel-Coder's methodology, applied to Ornith-1.5-9B. No fine-tuning — just an imatrix, a chat template, and dynamic quantization.

What this is

This is Ornith-1.5-9B re-quantized with a custom imatrix and carrying the Sharp chat template inside the GGUF — the same recipe used by Tiel-Coder-35B-A3B, scaled down to the 9B.

This is not a Tiel-Coder release. It is an independent reproduction of peculiar-ragdoll's methodology (imatrix + Sharp template, no weight changes) on the smaller Ornith-1.5-9B base. The name "Tiel-Inspired-Coder" makes that clear. Full credit to peculiar-ragdoll for the approach and the Sharp template. Full credit to ornith-ai for the base model.

The imatrix

The calibration corpus is ~49M characters drawn from eaddario/imatrix-calibration (MIT): about three quarters code_medium and one quarter combined_all_large, interleaved — the same mix Tiel-Coder 35B used.

The matrix was measured on a Q8_0 of the original BF16 weights, over 3,000 chunks of 512 tokens (~1.5M tokens). Final PPL estimate: 6.1217.

The imatrix file (Ornith-1.5-9B.imatrix.gguf) ships in this repo for anyone who wants to cut additional tiers.

The Sharp template

The Sharp chat template (qwen3.8-froggeric-v22.4.1) is embedded in every GGUF. It enables reasoning by default, supports configurable reasoning effort, and adds a terseness directive that keeps answers lean.

Available quants

File	Size	Notes
Tiel-Inspired-Coder-9B-UD-IQ4_XS.gguf	5.0 GB	Smallest 4-bit
Tiel-Inspired-Coder-9B-UD-Q4_K_S.gguf	5.2 GB	Tight 4-bit
Tiel-Inspired-Coder-9B-UD-Q4_K_M.gguf	5.4 GB	Standard 4-bit
Tiel-Inspired-Coder-9B-UD-Q5_K_M.gguf	6.2 GB	5-bit
Tiel-Inspired-Coder-9B-UD-Q6_K.gguf	7.1 GB	Near-lossless
Tiel-Inspired-Coder-9B-UD-Q8_0.gguf	9.2 GB	Reference

All tiers were quantized from the BF16 source using the imatrix. The Sharp template was injected into each file's metadata.

How to run

hf download VitreousCut/Tiel-Inspired-Coder-9B-GGUF Tiel-Inspired-Coder-9B-UD-Q4_K_M.gguf --local-dir .
llama-server -m Tiel-Inspired-Coder-9B-UD-Q4_K_M.gguf -ngl 99 --jinja -c 32768

Sampling: temperature 0.2, top_p 0.95, top_k 20 for coding tasks.

Important: This model is a massive overthinker

The Sharp template enables reasoning by default. On simple problems the model may produce a few hundred tokens of reasoning before answering — fine. On complex or ambiguous problems, it can spiral into 32,000+ tokens of reasoning and never produce an answer at all, even with a 32k token budget.

This is the single biggest weakness of this build. In HumanEval testing, 3 out of 164 problems failed purely because the model spent its entire token budget thinking without emitting code — and no amount of additional budget fixed them. These are "incurable overthinkers" where the model gets stuck in a reasoning loop.

Recommended fix: use reasoning_effort: low

The Sharp template supports configurable reasoning effort. For most use cases, pass reasoning_effort: low via chat template kwargs:

llama-server -m Tiel-Inspired-Coder-9B-UD-Q4_K_M.gguf -ngl 99 --jinja -c 32768 \
  --reasoning-effort low

Or via API:

{
  "messages": [...],
  "chat_template_kwargs": {"reasoning_effort": "low"}
}

This keeps reasoning concise and focused. Reserve medium (default) or high for genuinely hard problems where you can afford a longer think.

HumanEval results

All runs on Q4_K_M, 3 parallel workers, temperature 0.2.

Primary run (reasoning_effort=medium, 16k token budget)

Metric	Value
Pass@1	93.3%
Passed	153/164
Failed (code bugs)	5
Errors (overthinking — no code produced)	6
Avg tokens generated	1,761
Time	24.6 min

Error re-run (32k token budget, 131k context)

Re-ran the 6 error problems with 32k token budget and 131k context:

Problem	Result	Tokens	Diagnosis
HumanEval/129	PASS	4,827	Just needed more room
HumanEval/134	PASS	1,880	Barely needed effort
HumanEval/160	PASS	4,354	Needed room to think
HumanEval/32	FAIL	5,152	Produced code but it hangs (infinite loop)
HumanEval/116	FAIL	32,768	Incurable overthinking — 32k of reasoning, no code
HumanEval/132	FAIL	32,768	Incurable overthinking — same pattern

Adjusted Pass@1 with re-run: 95.1% (156/164)

The 3 remaining overthinking failures are problems where the model loops endlessly in its reasoning block. No amount of token budget fixes them — this is a behavioral issue that would require fine-tuning to address.

Comparison: reasoning_effort=low

Metric	medium (16k)	low (8k)
Pass@1	93.3%	89.0%
Failed	5	14
Errors	6	4
Avg tokens	1,761	854
Time	24.6 min	12.0 min

low is 2x faster and half the tokens, but loses ~4 points of accuracy. The sweet spot for quality is medium with a big budget; the sweet spot for speed is low.

Baseline comparison

Ornith-1.5-9B base typically scores in the 72-78% range on HumanEval at FP16. The imatrix + Sharp template + reasoning budget pushed it to 93.3% (95.1% adjusted) at 4-bit quantization.

Hardware

Tested on:

• CPU: Intel i7-6700K
• GPU: NVIDIA RTX 3090 (24GB VRAM)
• RAM: 32GB

All tiers run fully offloaded on a 3090 with room for 32k+ context.

Credits

• ornith-ai — the Ornith-1.5-9B weights (MIT)
• peculiar-ragdoll — the Tiel-Coder methodology, the Sharp chat template, and the imatrix recipe
• eaddario — the calibration corpora (MIT)
• froggeric — the template lineage Sharp builds on
• llama.cpp — llama-quantize / llama-imatrix / llama-server

MIT, inheriting Ornith-1.5's license.