---

base_model:

• prithivMLmods/Q3.5-9B-DS-v4-Flash-v2.0

tags:

• text-generation-inference
• llama-cpp
• deepseek-v4
• flash
• math
• reasoning
• abliterated

license: apache-2.0

language:

• en

pipeline_tag: image-text-to-text

library_name: transformers

---

Q3.5-9B-DS-v4-Flash-v2.0-GGUF

> Q3.5-9B-DS-v4-Flash-v2.0 is a reasoning-capable 9B-parameter language model built on top of Qwen/Qwen3.5-9B. The model was trained through a multi-stage training pipeline using approximately 3K long-context DeepSeek V4 Flash reasoning traces, along with additional high-quality reasoning traces, to improve long-form reasoning, mathematical problem solving, scientific analysis, and instruction-following capabilities.

> [!NOTE]

> This model is an experimental release and may generate unexpected behaviors or reasoning artifacts in certain scenarios.

Model Files

File Name | Quant Type | File Size | File Link |

|-----------|------------|-----------|-----------|

| Q3.5-9B-DS-v4-Flash-v2.0.BF16.gguf | BF16 | 17.9 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.F16.gguf | F16 | 17.9 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q3_K_L.gguf | Q3_K_L | 4.93 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q3_K_M.gguf | Q3_K_M | 4.62 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q3_K_S.gguf | Q3_K_S | 4.26 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q4_K_M.gguf | Q4_K_M | 5.63 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q4_K_S.gguf | Q4_K_S | 5.35 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q5_K_M.gguf | Q5_K_M | 6.47 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q5_K_S.gguf | Q5_K_S | 6.31 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.Q8_0.gguf | Q8_0 | 9.53 GB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.mmproj-bf16.gguf | mmproj-bf16 | 922 MB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.mmproj-f16.gguf | mmproj-f16 | 922 MB | Download |

| Q3.5-9B-DS-v4-Flash-v2.0.mmproj-q8_0.gguf | mmproj-q8_0 | 624 MB | Download |

llama.cpp

LLM inference in C/C++ — https://github.com/ggml-org/llama.cpp