Downloads · 30 days
49
7% of all-time downloads
Lufel6848/Ministral-3-8B-Reasoning-2512-GGUF
Ministral-3-8B-Reasoning-2512-GGUF is a machine learning model from Lufel6848. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Community-made GGUF conversions and quantizations of Ministral-3-8B-Reasoning-2512, intended for local text inference with GGUF-compatible software.
Downloads · 30 days
49
7% of all-time downloads
All-time downloads
735
Public
Repo size
44.2 GB
Likes
0
Public
Click a slice to open those files.
.gguf44.2 GB · 100%
From the Hugging Face model README
Community-made GGUF conversions and quantizations of Ministral-3-8B-Reasoning-2512, intended for local text inference with GGUF-compatible software.
The original model was converted from its Hugging Face / SafeTensors distribution to GGUF using tools provided by llama.cpp.
This is an unofficial community conversion.
The Ministral-3-8B-Reasoning-2512 model, architecture, and original weights were developed and released by Mistral AI. This repository provides converted and quantized GGUF files derived from the original model.
mistralai/Ministral-3-8B-Reasoning-2512Refer to the original model repository for the authoritative model card, capabilities, limitations, recommended settings, usage information, and license terms.
Ministral-3-8B-Reasoning-2512 is the reasoning post-trained variant of Ministral 3 8B, designed for tasks involving reasoning, mathematics, coding, STEM, and other workloads that benefit from multi-step reasoning.
The original Ministral-3-8B-Reasoning-2512 model includes native vision capabilities.
However, this repository provides only the GGUF conversion and quantizations of the language model. A multimodal projector (mmproj) is not included.
As a result, the GGUF files distributed in this repository are intended for text-only inference.
The absence of an mmproj file does not affect normal text generation, reasoning, coding, or other text-based use cases.
Users requiring the original model's vision capabilities should refer to the upstream model and compatible multimodal inference implementations.
This repository provides the original BF16 GGUF conversion alongside several quantized variants:
| Format / Quantization | Description |
|---|---|
BF16 | GGUF conversion retaining BF16 weight precision. Largest file and highest memory requirement among the provided variants. |
Q4_K_M | Lower storage and memory requirements. Suitable as a general-purpose local inference option. |
Q5_K_M | Balanced option with additional weight precision compared with Q4_K_M. |
Q6_K | Higher-precision quantization for systems with more available memory. |
Q8_0 | High-precision quantization with substantially larger memory and storage requirements. |
Actual memory consumption may be higher than the GGUF file size and depends on factors such as context length, KV cache configuration, inference backend, GPU offloading, and runtime settings.
The files in this repository were produced using a workflow based on llama.cpp:
Ministral-3-8B-Reasoning-2512
│
│ SafeTensors
▼
convert_hf_to_gguf.py
│
▼
BF16 GGUF
│
│ llama-quantize
▼
┌────────┬────────┬───────┬──────┐
│Q4_K_M │Q5_K_M │ Q6_K │ Q8_0 │
└────────┴────────┴───────┴──────┘
No additional training or fine-tuning is performed as part of this conversion process.
Quantization changes the numerical representation of the model weights to reduce storage and memory requirements and may affect model quality.
The multimodal projector and vision components are not included in the conversion files distributed by this repository.
These GGUF files are intended for text inference using applications and inference engines with compatible GGUF support, particularly llama.cpp and software built around it.
Example with llama.cpp:
llama-cli \
-m Ministral-3-8B-Reasoning-2512-Q5_K_M.gguf \
-p "Solve this problem step by step: If x² - 5x + 6 = 0, what are the possible values of x?"
Runtime parameters should be adjusted according to your hardware, available memory, desired context length, and inference backend.
Because this is a reasoning-oriented model, generation behavior may also depend on the chat template, system prompt, and reasoning support implemented by the inference runtime.
The upstream Ministral-3-8B-Reasoning-2512 model is multimodal and includes vision capabilities.
These capabilities are not provided by the GGUF files in this repository, as no multimodal projector (mmproj) is included.
The files distributed here should therefore be treated as text-only GGUF variants.
GGUF compatibility depends on the version of llama.cpp and its support for the underlying Mistral 3 language model architecture.
Because both llama.cpp and GGUF continue to evolve, older inference engines may not correctly load files produced by newer versions.
If you encounter GGUF compatibility problems, first test with a recent version of llama.cpp or your preferred GGUF-compatible runtime.
This repository does not provide or guarantee multimodal GGUF compatibility.
The conversion process follows the Hugging Face / SafeTensors → GGUF workflow provided by llama.cpp.
The general process consists of:
mistralai/Ministral-3-8B-Reasoning-2512 SafeTensors model;convert_hf_to_gguf.py;llama-quantize;Q4_K_M, Q5_K_M, Q6_K, and Q8_0 variants.This workflow covers the language-model GGUF files distributed in this repository and does not include generation or distribution of a multimodal projector.
Conversion and quantization behavior may vary between llama.cpp revisions as model architecture support and GGUF tooling evolve.
The original Ministral-3-8B-Reasoning-2512 model, architecture, and model weights were developed and released by Mistral AI.
This repository would not exist without their work and release of the original model.
All credit for the original model belongs to its respective authors and contributors.
GGUF conversion and quantization are performed using tools from the open-source llama.cpp project.
This workflow relies on tooling including:
convert_hf_to_gguf.pyllama-quantizeProject:
https://github.com/ggml-org/llama.cpp
Credit belongs to the llama.cpp maintainers and contributors for the conversion, quantization, GGUF, and local inference tooling used by this workflow.
The original Ministral-3-8B-Reasoning-2512 model is distributed under the Apache License 2.0.
These files are converted and quantized derivatives of the original model weights and retain the applicable licensing terms of the original model.
Please review the original Ministral-3-8B-Reasoning-2512 repository and its license before using or redistributing these files.
This repository is an unofficial community conversion and is not affiliated with, endorsed by, or maintained by Mistral AI, Vast.ai, or the llama.cpp project.
Vast.ai was used as the environment in which the conversion workflow was tested. Its use does not imply affiliation, endorsement, or a technical requirement to use Vast.ai.
The upstream Ministral-3-8B-Reasoning-2512 model includes multimodal capabilities, but this repository distributes only text-oriented GGUF model files and does not include the multimodal projector required for vision inference.
The purpose of this repository is to provide GGUF variants of the original openly released model for local text inference while documenting and crediting the upstream projects used to create them.