Qwen2-Audio-7B-Instruct-i1-GGUF Open Source Model - Supports Efficient Transcription of English Audio to Text

Home

Qwen2 Audio 7B Instruct I1 GGUF

Developed by mradermacher

Weighted/matrix quantized model of Qwen2-Audio-7B-Instruct, supporting English audio-to-text transcription tasks

Text-to-Audio

Transformers

EnglishOpen Source License:Apache-2.0 #Audio-to-text transcription #Low-resource quantization #Multi-precision options

Downloads 282

Release Time : 5/31/2025

Model Overview

This project provides quantized versions of Qwen/Qwen2-Audio-7B-Instruct, suitable for audio processing tasks in resource-constrained environments

Model Features

Multiple quantization options

Offers 20 different quantization levels from IQ1_S to Q6_K to meet various resource requirements

Resource efficient

The smallest quantized version is only 2.2GB, suitable for resource-constrained environments

Performance optimized

IQ quantized versions typically outperform non-IQ quantized versions at the same size

Model Capabilities

Audio-to-text conversion

English speech understanding

Instruction following

Use Cases

Speech processing

Speech-to-text

Convert English speech content into text

Resource-constrained environments

Edge device speech processing

Implement speech understanding functionality on devices with limited computing resources

🚀 Qwen2-Audio-7B-Instruct Quantized Model

This project provides quantized weights of Qwen2-Audio-7B-Instruct, facilitating efficient deployment and usage.

🚀 Quick Start

If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files.

✨ Features

Quantized Weights: Offers weighted/imatrix quants of https://huggingface.co/Qwen/Qwen2-Audio-7B-Instruct.
Multiple Quant Types: Provides a variety of quant types with different sizes and qualities.

📚 Documentation

About

Weighted/imatrix quants of https://huggingface.co/Qwen/Qwen2-Audio-7B-Instruct. Static quants are available at https://huggingface.co/mradermacher/Qwen2-Audio-7B-Instruct-GGUF.

Provided Quants

(sorted by size, not necessarily quality. IQ - quants are often preferable over similar sized non - IQ quants)

Link	Type	Size/GB	Notes
GGUF	i1-IQ1_S	2.2	for the desperate
GGUF	i1-IQ1_M	2.3	mostly desperate
GGUF	i1-IQ2_XXS	2.5
GGUF	i1-IQ2_XS	2.7
GGUF	i1-IQ2_S	2.9
GGUF	i1-Q2_K_S	3.0	very low quality
GGUF	i1-IQ2_M	3.0
GGUF	i1-Q2_K	3.2	IQ3_XXS probably better
GGUF	i1-IQ3_XXS	3.3	lower quality
GGUF	i1-IQ3_XS	3.5
GGUF	i1-IQ3_S	3.7	beats Q3_K*
GGUF	i1-Q3_K_S	3.7	IQ3_XS probably better
GGUF	i1-IQ3_M	3.9
GGUF	i1-Q3_K_M	4.0	IQ3_S probably better
GGUF	i1-Q3_K_L	4.3	IQ3_M probably better
GGUF	i1-IQ4_XS	4.4
GGUF	i1-IQ4_NL	4.6	prefer IQ4_XS
GGUF	i1-Q4_0	4.6	fast, low quality
GGUF	i1-Q4_K_S	4.7	optimal size/speed/quality
GGUF	i1-Q4_K_M	4.9	fast, recommended
GGUF	i1-Q4_1	5.1
GGUF	i1-Q5_K_S	5.5
GGUF	i1-Q5_K_M	5.7
GGUF	i1-Q6_K	6.5	practically like static Q6_K

Here is a handy graph by ikawrakow comparing some lower - quality quant types (lower is better):

And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9

FAQ / Model Request

See https://huggingface.co/mradermacher/model_requests for some answers to questions you might have and/or if you want some other model quantized.

Thanks

I thank my company, nethype GmbH, for letting me use its servers and providing upgrades to my workstation to enable this work in my free time. Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.

📄 License

This project is licensed under the apache - 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご