OpenThinker-Agent-v1-SFT is an agent model obtained through supervised fine-tuning (SFT) based on Qwen/Qwen3-8B. It is the first-stage model of the complete training process (SFT + RL) of OpenThinker-Agent-v1, specifically optimized for agent tasks (such as terminal operations and code repair).

Natural Language ProcessingTransformersOther

This is a quantized version of TheDrummer's Snowpiercer-15B-v4 model, quantized using llama.cpp and imatrix technology. This version offers multiple quantization levels from BF16 to IQ2_S, meeting the needs of users with different hardware conditions and performance requirements. The model is calibrated and optimized using a specific dataset and supports the online repackaging function to improve performance on ARM and AVX hardware.

Natural Language ProcessingGguf

This project is a llama.cpp GGUF quantized version based on the Qwen3-VL-32B-Instruct model, offering various quantization types to meet different hardware and performance requirements. Quantization is performed using a specific dataset in combination with the imatrix option to ensure quantization quality. It supports online weight repackaging to improve the performance of ARM and AVX machines.

MultimodalGguf

This is a repaired image text-to-text model that solves the problem of inference failure of the original model in the latest Transformers version. This model is specifically designed for image recognition and text generation tasks and supports multimodal input processing.

MultimodalTransformersEnglish

A quantized version of the Tencent Hunyuan 4B Instruction Model based on llama.cpp. Through quantization technology, the model can run more efficiently in resource-constrained environments while maintaining a certain level of performance and quality. It offers multiple quantization type options and supports online repackaging optimization.

Natural Language ProcessingGguf

A quantized and repaired large language model based on Qwen3 technology, designed for efficient code generation and programming tasks. This model uses AWQ quantization technology to significantly reduce the computational resource requirements while maintaining high performance, and supports fast deployment and inference in a multi-GPU environment.

Natural Language ProcessingTransformers

Granite-3B-Code-Instruct-128K is a 3-billion parameter long context code instruction model developed by IBM Research and released under the Apache 2.0 license. This model is specifically optimized for code generation tasks, supports 128K long context processing, and performs well in code generation, interpretation, and repair.

Natural Language ProcessingTransformersOther

CodeModernBERT-Owl-3.0 is the final pre-trained version of the multilingual long context encoder model in the CodeModernBERT series, optimized for downstream code-related tasks such as code search, code summarization, error repair, and representation learning.

Natural Language ProcessingTransformersMultiple Languages

This is an experimental model that has not undergone complete training and is mainly used for research purposes. It is created through model fusion technology and requires a specific repair process to restore normal functionality.

Natural Language ProcessingTransformers

VIRES is a sketch- and text-guided video instance repainting method that supports video instance repainting, replacement, generation, and removal.

Computer VisionDiffusers

PMRF is a model for blind face image restoration, capable of effectively repairing damaged or low-quality facial images.

Computer VisionSafetensors

PAVE is a model focused on repairing and adapting video large language models, aiming to enhance the conversion capability between video and text.

MultimodalTransformers

Granite - 8B - Code - Base - 128K is a code generation model developed by IBM Research. Through a progressive training strategy, the context length is extended from 4K to 128K. It supports 116 programming languages and can handle various software engineering tasks such as code generation, interpretation, and repair.

Natural Language ProcessingTransformersOther

NatGen is a generative pre-training model that 'naturalizes' source code, focusing on code generation, translation, and defect repair tasks.

Natural Language ProcessingTransformersOther

RepVGG-A1 is a lightweight image classification model pre-trained on the ImageNette dataset, designed with a reparameterizable architecture.

Computer VisionTransformers

RepVGG-A0 is an efficient convolutional neural network based on the VGG-style architecture, achieving separation between training and inference structures through structural reparameterization techniques to enhance performance.

Computer VisionTransformers

This is a Vietnamese automatic speech recognition (ASR) repair model based on the MT5 architecture, fine-tuned for Vietnamese speech recognition tasks.

Audio ProcessingJaxOther