NVIDIA GPT-OSS-120B Eagle3 is an optimized version based on the OpenAI gpt-oss-120b model. It adopts the Mixture of Experts (MoE) architecture, with a total of 120 billion parameters and 5 billion active parameters. This model supports both commercial and non-commercial use and is suitable for text generation tasks, especially for the development of AI Agent systems, chatbots, and other applications.

Natural Language ProcessingSafetensors

This is the GGUF format version of the GPT-OSS 20B model based on WeOpenML, which first implements the innovative MoE CPU expert offloading technology. This technology achieves a 99.9% reduction in video memory while maintaining the full generation quality, and only requires 2MB of video memory to run a 20 billion parameter mixture of experts model.

Natural Language ProcessingSafetensorsMultiple Languages

This project distills the inference ability of GPT into the Llama-3.1-8B model through an innovative two-stage training process. First, knowledge distillation and format alignment are performed through supervised fine-tuning, and then reinforcement learning is used to encourage the model to independently explore and optimize inference strategies, focusing on breakthroughs in the field of mathematical reasoning.

Natural Language ProcessingTransformersMultiple Languages

GLM-4.6-GPTQ-Int4-Int8Mix is a quantized version based on the zai-org/GLM-4.6 base model, which performs excellently in text generation tasks. This model achieves optimization of the model size through GPTQ quantization technology while maintaining good performance.

Natural Language ProcessingTransformers

WEBGEN DEVSTRAL IMAGES is an AI model focused on web page generation. It can generate single-page web pages using HTML, CSS, JS, and Tailwind technologies. This project is trained based on custom templates and uses the supervised fine-tuning method, training with a dataset generated by GPT-OSS-120B.

Natural Language ProcessingTransformersEnglish

This model is based on GPT-OSS-20B and fine-tuned using the Unsloth reinforcement learning framework. The aim is to optimize inference efficiency and reduce vulnerabilities that occur during reinforcement learning from human feedback (RLHF) training. The fine-tuning process focuses on the robustness and efficiency of alignment, ensuring that the model maintains inference depth without incurring excessive computational overhead.

Natural Language ProcessingTransformersEnglish

This is the GGUF quantized version of the Qwen/Qwen3-14B language model, with 14 billion parameters. It has deep reasoning ability, research-level accuracy, and an autonomous workflow. After conversion, it can be used in local inference frameworks such as llama.cpp, LM Studio, OpenWebUI, and GPT4All.

Natural Language ProcessingGgufMultiple Languages

This is the GGUF quantized version of the Qwen/Qwen3-0.6B language model, a compact large language model with 600 million parameters, designed for ultra-fast inference on low-resource devices. It supports frameworks such as llama.cpp, LM Studio, OpenWebUI, and GPT4All, and can be used offline anywhere for private AI.

Natural Language ProcessingGgufMultiple Languages

Jinx GPT OSS 20B MXFP4 MLX is a large language model with 20 billion parameters converted based on the MLX framework. It is optimized using mixed-precision quantization technology and is suitable for efficient inference on Apple chip devices.

Natural Language ProcessingMlx

This is a quantized version model based on AmanPriyanshu/gpt-oss-6.0b-specialized-all-pruned-moe-only-7-experts. It adopts a mixture of experts architecture and pruning technology, providing multiple quantization types to meet the needs of different scenarios. The model supports English text generation tasks.

Natural Language ProcessingTransformersMultiple Languages

GPT-OSS-20B is an open-weight model launched by OpenAI, designed for powerful reasoning, agent tasks, and diverse developer use cases. This model has 21 billion parameters (3.6 billion of which are active parameters), adopts a mixture-of-experts architecture, and supports low-latency local deployment and specific scenario applications.

Natural Language ProcessingTransformers

This is a 4.2 billion parameter multi-expert model based on the GPT architecture, which has been pruned and optimized to retain only 4 experts. It offers multiple quantization versions, is designed specifically for consumer-grade hardware, and supports efficient inference and convenient deployment.

Natural Language ProcessingGguf

GPT-OSS-20B-NVFP4 is a quantized version of the OpenAI GPT-OSS-20B model, using NVIDIA's advanced NVFP4 quantization format. While maintaining a significant improvement in memory efficiency, this model can provide a better accuracy retention rate compared to MXFP4 quantization, and is designed for high-performance inference scenarios.

Natural Language ProcessingSafetensorsEnglish

GPT-OSS-20B is an open-weight model launched by OpenAI, with 21 billion parameters (3.6 billion of which are active parameters). It is designed for low-latency, local deployment, and specific use cases. This model is trained based on the harmony response format and has powerful inference capabilities and agent task processing capabilities.

Natural Language ProcessingTransformers

This is an MLX format version converted from the OpenAI GPT-OSS-120B model, converted using the mlx-lm 0.27.0 tool, and supports efficient text generation tasks.

Natural Language ProcessingMlx

This is the MLX format conversion version of the OpenAI GPT-OSS-20B model. It is converted using the mlx-lm 0.27.0 tool and supports efficient operation on Apple Silicon devices. The model uses MXFP4-Q8 quantization technology to significantly reduce memory usage while maintaining performance.

Natural Language ProcessingMlx

This is an MLX format version converted from the OpenAI GPT-OSS-20B model, using the mlx-lm 0.27.0 tool for conversion, which supports efficient operation on Apple Silicon devices.

Natural Language ProcessingMlx

This is the GGUF quantized version of the GPT-OSS-Code-Reasoning-20B model, optimized for efficient reasoning and reducing memory requirements. This model is specifically fine-tuned for competition programming and algorithm reasoning tasks, capable of generating Python/C++ solutions and reasoning processes.

Natural Language ProcessingGgufEnglish

This is a financial analysis model fine - tuned based on GPT - OSS 20B, specifically designed for financial market analysis. It can provide technical analysis, risk assessment, trading signals, and price forecasts for various securities. Quantized in Q8_0 GGUF format, it reduces inference requirements while ensuring model quality.

Natural Language ProcessingTransformersEnglish

This is a 4.2B parameter Mixture of Experts model based on the GPT-OSS architecture. After pruning optimization, only 4 experts are retained, and it is converted to a quantized version in GGUF format. The model focuses on efficient inference and text generation tasks, using Q4_K_M quantization technology to reduce memory usage while maintaining performance.

Natural Language ProcessingGgufMultiple Languages