xtools-at/DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic
text-generation model by xtools-at
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
Click here for the 9B version. Also check out DeepThought, my science finetune based on this model: xtools-at/Qwen3.5-4B-DeepThought-STEM 🧬 The 9B Training Recipe, Compressed into a 4B Student 1 · Teacher GenerationDeepSeek-V4-Pro in Max Effect mode produces detailed mathematics and STEM solutions. 2 · Data PreparationApproximately 250K samples are cleaned and formatted for supervised reasoning transfer. 3 · Unsloth SFTQwen3.5-4B is fine-tuned through an Unsloth LoRA pipeline and merged into BF16. 4 · Deployment BuildAn MTP-enabled Q8 GGUF build is used for the reported llama.cpp evaluations. The MTP-Q8 build was evaluated twice on the complete 1,319-question GSM8K test split. Across 2,638 sampled answers, the model produced 2,421 exact-answer matches, yielding a two-run pass@1 mean of 91.7741%. The two independent run scores were 91.6603% and 91.8878%, a spread of only 0.23 percentage points. 📈 Compact-model resultThe 4B student is 2.73 points behind the published 9B score while remaining 2.47 points above…
Read the full model card ↗ · Preview checked 2026-11-14T09:42:08.805Z
Inside the original model card — Document outline
- This is a decensored version of Jackrong/DeepSeek-V4-Pro-Qwen3.5-4B, made using Heretic v2.0.0
- Arbitrary-Rank Ablation (ARA) parameters
- Performance
- 🤝 Collaboration & Training Support
- 🔬 1. Distillation & Training Data
- 📊 2. Benchmark Results
- 2.1 GSM8K
- 2.2 4B vs. 9B
- 2.3 MMLU-Pro: Math, Physics & Chemistry
- ⚙️ 3. Deployment Profile
- 🎯 4. Recommended Uses
- ⚠️ 5. Limitations
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Documentation belongs to its respective authors. Reported project/model license: apache-2.0. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
xtools-at/DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic is a text-generation model repository published by xtools-at on Hugging Face. The source reports the transformers library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- text-generation
- Library
- transformers
- Recent downloads (30 days)
- 12,595
- Cumulative likes
- 77
- Hugging Face trending score
- 47
- Reported safetensors parameters
- 4,539,265,536
- Architecture
- Qwen3_5ForConditionalGeneration
- License
- apache-2.0
- Access
- Not gated by Hugging Face
- Created
- 2026-11-02T10:29:00.000Z
- Last modified
- 2026-11-12T21:00:42.000Z
Compatibility and lineage
Reported languages: en, zh, ko, ja, es, ru
Reported base models: Jackrong/DeepSeek-V4-Pro-Qwen3.5-4B
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-11-14T06:56:53.534Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.