This is a fine-tuned Whisper speech recognition model based on the unsloth/whisper-large-v3-turbo architecture. It uses the Unsloth and Huggingface TRL libraries to achieve 2x training acceleration, significantly improving training efficiency.

Audio ProcessingTransformersEnglish

dParallel is a learnable parallel decoding method for large language models that achieves fast sampling by exploiting the intrinsic parallelism of the model. This method can significantly reduce the decoding steps while maintaining the model performance, achieving 8.5 - 10.5 times acceleration in benchmarks such as GSM8K and MBPP.

Natural Language ProcessingTransformers

dParallel is a learnable parallel decoding method for large language models. It achieves fast sampling by exploring the intrinsic parallelism of the model, significantly reducing the decoding steps while maintaining performance, and achieving 8.5 - 10.5 times acceleration in benchmark tests such as GSM8K and MBPP.

Natural Language ProcessingTransformers

Qwen3-Next-80B-A3B-Instruct is the latest generation of large language model developed by the Tongyi Qianwen team at Alibaba. It adopts an innovative hybrid attention mechanism and a highly sparse mixture-of-experts architecture. While maintaining a total of 80B parameters, only 3B parameters are activated, achieving efficient context modeling and inference acceleration. It natively supports a context length of 262K and can be extended to 1M tokens.

Natural Language ProcessingTransformers

The NVIDIA Qwen3-14B FP4 model is a quantized version of Alibaba's Qwen3-14B model, optimized using the FP4 data type and enabling efficient inference through TensorRT-LLM. This model is specifically designed for NVIDIA GPU acceleration systems and is suitable for various AI application scenarios such as AI Agent systems, chatbots, and RAG systems. It supports both commercial and non-commercial use globally.

Natural Language ProcessingSafetensors

The NVIDIA Qwen3-14B FP4 model is a quantized version of the Alibaba Cloud Qwen3-14B model. It adopts an optimized Transformer architecture and is an autoregressive language model. This model uses the TensorRT Model Optimizer for quantization, quantizing weights and activations to the FP4 data type, enabling efficient inference on NVIDIA GPU acceleration systems.

Natural Language ProcessingSafetensors

Qwen3-Next-80B-A3B-Instruct is the base model of the Qwen3-Next series. Through technologies such as the innovative hybrid attention mechanism and highly sparse mixture-of-experts layers, it achieves efficient modeling and reasoning acceleration in long text processing and demonstrates excellent performance in multiple benchmark tests.

Natural Language ProcessingTransformers

MiniCPM4.1-8B-GGUF is a GGUF quantized format version of the MiniCPM4.1-8B model. It is an efficient large language model designed for edge devices, with 8 billion parameters, supporting integrated thinking, and can achieve more than 5 times the generation acceleration on typical edge chips.

Natural Language ProcessingTransformersMultiple Languages

This is the Qwen3-8B model quantized by the PyTorch team using torchao, adopting int4 weight-only quantization and the AWQ algorithm. This model can reduce 53% of GPU memory usage and achieve 1.34x acceleration on the H100 GPU. It is specifically calibrated and optimized for the mmlu_abstract_algebra task.

Natural Language ProcessingTransformersEnglish

This is an optimized version fine-tuned based on the Qwen2.5-7B-Instruct model using Unsloth technology, achieving 2x training acceleration and focusing on text generation tasks.

Natural Language ProcessingTransformersEnglish

Jet - Nemotron - 4B is an efficient hybrid architecture language model launched by NVIDIA. It is built based on two core innovations: post - neural architecture search and the JetBlock linear attention module. In terms of performance, it surpasses open - source models such as Qwen3, Qwen2.5, Gemma3, and Llama3.2. At the same time, it achieves a maximum of 53.6 times acceleration in generation throughput on the H100 GPU.

Natural Language ProcessingTransformersEnglish

Jet-Nemotron is a new family of hybrid architecture language models that surpasses state-of-the-art open-source full-attention language models such as Qwen3, Qwen2.5, Gemma3, and Llama3.2, while achieving significant efficiency improvements - with up to 53.6x acceleration in generation throughput on H100 GPUs.

Natural Language ProcessingTransformersEnglish

This is a version of the Google Gemma3n model that has been efficiently fine-tuned using the Unsloth framework. It achieves 2x training acceleration through 4-bit quantization technology and the TRL library, while maintaining the original model's strong text understanding and generation capabilities.

Natural Language ProcessingTransformersEnglish

This model is a dialogue turn detection model fine-tuned based on Qwen3-0.6B. It uses the Unsloth and Huggingface TRL libraries to accelerate training, achieving an accuracy of 96.22% in the dialogue turn detection task.

Natural Language ProcessingTransformersEnglish

MiniCPM4 is an efficient large language model designed specifically for edge devices. While maintaining optimal performance at the same scale, it achieves extreme efficiency improvements, enabling over 5x generation acceleration on typical edge chips.

Natural Language ProcessingTransformersMultiple Languages

Hume-System2 is the pre-trained weights of System 2 for a dual-system Vision-Language-Action (VLA) model, used to accelerate the training of System 2 and provide support for relevant research and applications in the field of robotics.

MultimodalTransformersEnglish

This is the FP8 static quantization version of OpenGVLab/InternVL3-38B, optimized for high-performance inference using vLLM. It achieves approximately 2x acceleration on vision-language tasks with minimal accuracy loss.

MultimodalSafetensorsMultiple Languages

PARD is a high-performance speculative decoding method that can convert autoregressive draft models into parallel draft models at low cost, significantly accelerating the inference of large language models.

Natural Language ProcessingTransformers

MiniCPM-S-1B-sft is a 1B-parameter language model optimized with activation sparsity techniques, achieving high-sparsity inference acceleration through the ProSparse method while maintaining performance comparable to the original model.

Natural Language ProcessingTransformersMultiple Languages

Introducing an attention gating (AttnGates) weight adapter for the QwQ-32B model to accelerate long-context computation through dynamic block-level sparsity

Natural Language ProcessingTransformers