Downloads · 30 days
570
18% of all-time downloads
esatapedico/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-BUDGET-GGUF
Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-BUDGET-GGUF is a text generation model from esatapedico. Use it when you need the model to write or continue text. It is set up for gguf. The card lists the license as apache-2.0.
Two compact GGUF files of DavidAU/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored (TWIN-TURBO Fable Cold Fusion Heretic Uncensored, 709-L arc class tune on Qwen3.8-27B, 27B dense hybrid with Gated DeltaNet p…
Downloads · 30 days
570
18% of all-time downloads
All-time downloads
3.2K
Public
Repo size
88.9 GB
Likes
1
Public
Click a slice to open those files.
.gguf30.2 GB · 100%
From the Hugging Face model README
Two compact GGUF files of DavidAU/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored (TWIN-TURBO Fable Cold Fusion Heretic Uncensored, 709-L arc class tune on Qwen3.8-27B, 27B dense hybrid with Gated DeltaNet plus Gated Attention, 262,144-token native context, native vision tower) without the MTP speculative head.
These are the budget members of the esatapedico/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-GGUF family. They share the same byte-identical 448-tensor native NVFP4 backbone (all 64 layers attention plus MLP) as the MTP tiers, but the MTP draft block has been stripped and the extra tensors pinned to the smallest types. No drafter, no speculative decoding, just the model at the lowest possible footprint. Pairs with the original Qwen3.8 vision projector.
The goal is to let 16 GB VRAM users benefit from native NVFP4. These are the smallest possible ways to serve this TWIN-TURBO tune on a single 16 GB Blackwell card. There are, of course, accuracy trade-offs vs the larger MTP tiers, plus the lack of MTP. Like every compact tier of the family, the NVFP4 backbone is preserved byte identical, and here the head tensors are pinned to the smallest types.
BUDGET is the quality-leaning budget tier: Q3_K lm_head plus Q2_K token embedding. Single-16GB friendly.STARVED is the absolute floor: Q2_K lm_head and Q2_K token embedding. Smallest possible variant of the family.Both files were re-issued with the upstream multi-modal fix from the same tool-call thread (the REASON-scanner pass no longer flags every message as system content, so image and video messages render vision tokens instead of raising; the system-message image guard is retained). Keeps the 2026-09-12 toolcall2 fixes plus the string-argument guard (upstream's reg file still lacks the latter, so our template stays ahead on that point; credit apollo-mg, donboyle, DavidAU). Tensor data is byte-identical to the previous release; only the embedded template changed, so each file grew by 544 bytes and both SHA-256 below are new.
Both files were re-issued with the corrected chat template from the upstream tool-call thread (preserved tool_calls, REASON-scanner guards, string-argument guard; credit apollo-mg, donboyle, DavidAU). Tensor data is byte-identical to the previous release; only the embedded template changed, so each file grew by 544 bytes and both SHA-256 below are new. For tool calling, upstream recommends reasoning mode with temp 0.6 to 0.7; instruct-mode tool calling stays poor.
I post updates on new conversions, benchmarks, and what I'm working on over on my site. Follow along there to keep up with new releases and the work in progress. If you'd like to support more of it, you will find the options there. I do this on consumer hardware and like seeing how far it goes. More is on the way.
🌐 esatapedico.com. Updates, work-in-progress, and ways to support.
| File | Size (decimal GB) | lm_head (output.weight) | token_embd | MTP head |
|---|---|---|---|---|
Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-BUDGET.gguf | 14.72 GB | Q3_K | Q2_K | none, stripped |
Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-STARVED.gguf | 14.59 GB | Q2_K | Q2_K | none, stripped |
Both contain 1,107 tensors: the 448-tensor native NVFP4 backbone (byte-identical across the whole TWIN-TURBO family, verified per-tensor SHA-256) plus 657 F32 norms, scales and gates plus the two head tensors. qwen35.nextn_predict_layers=0 and qwen35.block_count=64 (the MTP block is gone), so they load in any recent llama.cpp build with NVFP4 support without the draft-mtp path.
The MTP family is built around a byte-identical NVFP4 backbone. The MTP draft head costs VRAM and compute on every pass. If you do not use speculative decoding, the stripped versions free that head entirely and shave the extra tensors down to the cheapest types. That is the smallest possible way to serve this TWIN-TURBO tune on a single 16 GB Blackwell card.
At a high level, the steps were:
block_count 64 and nextn_predict_layers 0.The NVFP4 tensors are native GGML type 40, preserved from the source checkpoint through GGUF conversion with no re-quantization round trip.
Vision works. The TWIN-TURBO 709-L tune leaves the original Qwen3.8 vision tower untouched. Pair either GGUF with the bundled mmproj-BF16.gguf via --mmproj (byte-identical to unsloth/Qwen3.8-27B-GGUF's projector). If you need the absolute minimum VRAM footprint on a single 16 GB card, you can also load the model without the projector (no --mmproj, text-only). That frees the projector's about 1 GB and is another lever to fit bigger context.
sm_120 support (Blackwell).mmproj-BF16.gguf.temp 1.0, top_p 0.95, top_k 20. Instruct mode temp 0.7, top_p 0.80, top_k 20. Pick per use case.These are rough sanity checks to confirm the files load and generate, not a formal benchmark. Method and hardware are noted so you can interpret them in context. Results will vary with hardware, sampling and context length.
All direct GPU runs used a diverse synthetic English payload (about 75 kB). Endpoint numbers below are end to end and include endpoint overhead, do not compare them to direct decode.
Prompt What is a black hole? via single turn sampling. Both tiers answered coherently with no repetition collapse (BUDGET distinct-5 0.9936, STARVED 0.9932).
Direct perplexity on the 75 kB diverse payload, single GPU, same chunks as the 6-tier MTP family:
| Tier | PPL (Final estimate) |
|---|---|
| BUDGET | 3.2523 +/- 0.07129 |
| STARVED | 3.4022 +/- 0.07629 |
BUDGET 3.25 lands at the top edge of the MTP family range (3.19 to 3.25 on the same payload). STARVED 3.40 sits about 0.15 above BUDGET, the expected Q2_K floor cost.
Single 16 GB class card with full GPU offload, three runs per tier:
| Tier | VRAM | pp512 tok/s | tg128 tok/s | Notes |
|---|---|---|---|---|
| BUDGET | 13.70 GiB | 2696.60 +/- 318.95 | 47.30 +/- 0.08 | fits, full GPU |
| STARVED | 13.58 GiB | 2698.86 +/- 301.13 | 47.71 +/- 0.05 | fits, full GPU |
Prefill about 2697 tok/s and decode about 47 to 48 tok/s on a single 16 GB card, flat across both tiers and in line with the MTP compact tiers. Both fit a single 5070 Ti class card with full GPU offload.
Each tier gets one large request with max_tokens=20000 and finish_reason=stop for both. These are end to end timings, not pure decode, so treat as sanity and responsiveness, not a formal benchmark.
| Tier | request_seconds | tokens per sec (20000 / s) | finish_reason | reasoning chars | content chars |
|---|---|---|---|---|---|
| BUDGET | 190.37 | 105.06 | stop | 1244 | 1185 |
| STARVED | 185.54 | 107.79 | stop | 1220 | 735 |
Both tiers ran repetition clean: content distinct-5-gram 1.0 with no adjacent dupes. BUDGET reasoning showed mild source citation repetition (0.94, max repeat 3) which per runbook is not a loop signature.
What to take away, naive framing
This is a derivative work built entirely from existing Apache 2.0 artifacts. Nothing here was trained or fine tuned. Credit belongs to:
The NVFP4 tensors are native GGML type 40, preserved from the source checkpoint through GGUF conversion with no re-quantization round trip.
Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-BUDGET.gguf (14.72 GB)Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-STARVED.gguf (14.59 GB)56bc8b7ccdc7b3eb3cd66a6faab23f967c1dce8a6097592d4f99094c691e561a Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-BUDGET.gguf
0357b80c563600bea38863ca621e3c27146365056701d2b4ec1dbae3ecc1a248 Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-STARVED.gguf
The full MTP family (VERY-LOW, LOW, MEDIUM, MID-HIGH, HIGH, VERY-HIGH, all with the speculative head baked in) lives in esatapedico/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NVFP4-GGUF. The budget tiers share its byte-identical 448-tensor NVFP4 backbone with the MTP block removed.
apache-2.0 (inherits from Qwen base and DavidAU tune). general.license = apache-2.0 is set inside every GGUF. general.name matches the tier name, qwen35.block_count 64, qwen35.context_length 262144.
Card written by AI assistance at the request of the repository author, who did the engineering. As with any AI-generated text, there may be errors. Please verify anything important (hashes, sizes, commands) against the file itself before relying on it.