Home

Awesome

🔥 Models 🔥

Welcome to the Models repository! Here, you'll find a diverse collection of deep learning models and examples constructed using the Burn deep learning framework.

Collection of Official Models

ModelDescriptionRepository Link
LlamaLlama 3 and TinyLlama large language models.llama-burn
MobileNetV2A CNN model targeted at mobile devices.mobilenetv2-burn
SqueezeNetA small CNN-based model for image classification.squeezenet-burn
ResNetA CNN based on residual blocks with skip connections.resnet-burn
RoBERTaA robustly optimized BERT pretraining approach.bert-burn
YOLOXA single-stage object detector based on the YOLO series.yolox-burn

Community Contributions

Explore the curated list of models developed by the community ♥.

ModelDescriptionRepository Link
Llama 2LLMs by Meta AI, ranging from 7 billion to 70 billion parameters.Gadersd/llama2-burn
WhisperA general-purpose speech recognition model by OpenAI.Gadersd/whisper-burn
Stable Diffusion v1.4An image generation model developed by Stability AI.Gadersd/stable-diffusion-burn
kord (music note predictor)A music theory model that can detect notes in short audio clips.twitchax/kord
Whisper-LiveA fork of Gadersd/whisper-burn which has been updated for Burn 13 and provides live transcriptionsudomonikers/whisper-burn
Inception V3A CNN used for calculating FID scores.varonroy/inception-v3-burn
CRAFTA CNN for character-region aware text detectionwingertge/craft-burn

License Information

Models implemented in this repository are distributed under the terms of both the MIT license and the Apache License (Version 2.0). See LICENSE-APACHE and LICENSE-MIT for complete details.

Please note that opening a pull request signals your agreement with these licensing terms. If you copy or adapt material from other resources or codebases, ensure that you include the original license information in the NOTICES.md file under the corresponding model directory.

Community models linked in this repository may fall under different licenses, so please consult the respective repositories for specific license information.