Orpheus-bangla-GGUF Open-source Model - Freely Achieve Bangla Text-to-Speech Tasks

Orpheus Bangla GGUF

Developed by mradermacher

This is the static quantized version of the asif00/orpheus-bangla-tts model, supporting Bengali text-to-speech tasks.

Speech Synthesis OtherOpen Source License:Apache-2.0 #Bengali TTS #Lightweight Speech Synthesis #Multiple Quantization Versions

Downloads 416

Release Time : 3/28/2025

Model Overview

This model is a Bengali text-to-speech (TTS) model based on the transformers architecture, offering multiple quantization versions to accommodate different hardware requirements.

Model Features

Multiple Quantization Versions

Provides 12 different quantization versions from Q2_K to f16 to meet various performance and quality needs.

Static Quantization Optimization

Reduces model size and improves inference efficiency through static quantization techniques.

Bengali Language Support

A text-to-speech model specifically optimized for the Bengali language.

Model Capabilities

Bengali Text-to-Speech

Supports Multiple Quantization Inference

Use Cases

Speech Synthesis

Bengali Voice Assistant

Provides voice interaction functionality for Bengali-speaking users.

Audiobook Generation

Converts Bengali text into natural speech.

🚀 Orpheus Bangla TTS Quantized Model

This project provides static quantizations of the Orpheus Bangla TTS model, offering various quantization types for different usage scenarios.

🚀 Quick Start

If you're new to using GGUF files, refer to TheBloke's READMEs for detailed instructions, including how to concatenate multi - part files.

📚 Documentation

About

This is a static quantization of the Orpheus Bangla TTS model. At present, weighted/imatrix quants seem unavailable to me. If they don't appear about a week after the static ones, I probably haven't planned for them. You can request them by opening a Community Discussion.

Provided Quants

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

Link	Type	Size/GB	Notes
GGUF	Q2_K	1.5
GGUF	Q3_K_S	1.7
GGUF	Q3_K_M	1.9	lower quality
GGUF	Q3_K_L	2.0
GGUF	IQ4_XS	2.0
GGUF	Q4_K_S	2.1	fast, recommended
GGUF	Q4_K_M	2.2	fast, recommended
GGUF	Q5_K_S	2.4
GGUF	Q5_K_M	2.5
GGUF	Q6_K	2.8	very good quality
GGUF	Q8_0	3.6	fast, best quality
GGUF	f16	6.7	16 bpw, overkill

Here is a useful 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

For answers to your questions or to request quantization of other models, see here.

📄 License

This project is licensed under the Apache - 2.0 license.

🙏 Thanks

I'm grateful to my company, nethype GmbH, for allowing me to use its servers and upgrading my workstation, enabling me to do this work in my free time. I also thank @nicoboss for giving me access to his private supercomputer, which allows me to provide many more high - quality imatrix quants.

Property	Details
Base Model	asif00/orpheus-bangla-tts
Datasets	SUST - CSE - Speech/banspeech
Language	bn
Library Name	transformers
License	apache - 2.0
Quantized By	mradermacher
Tags	transformers, llama, text - to - speech

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