MobileLLM-Pro is a 1-billion-parameter efficient device-side language model launched by Meta, optimized for mobile devices, supporting a context length of 128k and providing high-quality inference capabilities. This model is trained using knowledge distillation technology, outperforming models of the same scale in multiple benchmark tests, and supporting nearly lossless 4-bit quantization.

Natural Language ProcessingTransformersEnglish

A 1-billion-parameter instruction fine-tuning model based on the Meta Llama 3.2 architecture, specifically optimized for text generation tasks. This model uses the PEFT (Parameter-Efficient Fine-Tuning) technique to enhance instruction following and dialogue capabilities while maintaining the capabilities of the base model.

Natural Language ProcessingPeft

MobileLLM-ParetoQ is an extremely low-bit large language model quantization framework optimized for mobile devices, supporting 1-bit, 1.58-bit, 2-bit, 3-bit, and 4-bit quantization settings, significantly reducing resource consumption while maintaining high performance.

Natural Language ProcessingTransformersEnglish

Vintern-1B-v3_5 is a 1-billion-parameter vision-language model supporting image-text generation tasks.

MultimodalGguf

OLMo-2-0425-1B-Instruct is a 1-billion-parameter instruction-finetuned language model developed by AllenAI, focused on text generation tasks.

Natural Language ProcessingGguf

GLM-Z1-9B-0414 is a bilingual text generation model supporting both Chinese and English, utilizing the GGUF format and suitable for various quantization levels, from BF16 to ultra-low-bit quantization (1-2 bits).

Natural Language ProcessingGgufMultiple Languages

This is a 1-billion-parameter vision Transformer model trained on 2 billion web images through DINOv2 self-supervised learning, capable of learning visual representations without language supervision.

Computer VisionTransformers

A 1-billion-parameter Vision Transformer model trained via masked autoencoder self-supervised learning on 2 billion web images, capable of learning visual representations without language supervision.

Computer VisionTransformers

Ultra-low-bit quantization (1-2 bit) model using IQ-DynamicGate technology, suitable for memory-constrained devices and efficient inference scenarios

Natural Language ProcessingGgufEnglish

Llama-3.2-3B-Instruct GGUF is a 3B-parameter large language model released by Meta, utilizing IQ-DynamicGate technology for ultra-low-bit quantization (1-2 bits), optimizing inference performance while maintaining memory efficiency.

Natural Language ProcessingGgufMultiple Languages

phi-2 is a text generation model employing IQ-DynamicGate ultra-low bit quantization (1-2 bits), suitable for natural language processing and code generation tasks.

Natural Language ProcessingGgufMultiple Languages

CSM (Conversational Speech Model) is a 1-billion-parameter speech generation model developed by Sesame, capable of generating RVQ audio encoding from text and audio inputs.

Audio ProcessingTransformersEnglish

Ultra-low-bit quantization (1-2 bits) language model using IQ-DynamicGate technology, suitable for memory-constrained environments

Natural Language ProcessingGguf

The first open-source, native 1-bit large language model developed by Microsoft Research, with a parameter scale of 2 billion, trained on a corpus of 4 trillion tokens.

Natural Language ProcessingGgufEnglish

The first open-source 2-billion-parameter native 1-bit large language model developed by Microsoft Research, trained on 4 trillion tokens, demonstrating that native 1-bit large language models can significantly improve computational efficiency while maintaining performance comparable to full-precision open-source models of the same scale.

Natural Language ProcessingTransformersEnglish

An open-source native 1-bit large language model developed by Microsoft Research, with 2 billion parameters trained on a 4 trillion token corpus, significantly improving computational efficiency.

Natural Language ProcessingTransformersEnglish

Ultra-low-bit quantization (1-2 bits) model using IQ-DynamicGate technology, suitable for memory-constrained devices and edge computing scenarios

Natural Language ProcessingGgufEnglish

Ultra-low-bit quantization (1-2 bit) large language model using IQ-DynamicGate technology, supporting multilingual text generation tasks

Natural Language ProcessingGgufEnglish

An ultra-low bit quantized model optimized based on the Llama-3-8B architecture, utilizing IQ-DynamicGate technology for adaptive 1-2 bit precision quantization, suitable for memory-constrained environments.

Natural Language ProcessingGgufEnglish

An ultra-low-bit (1-2 bit) quantized model based on Llama-3.1-70B, utilizing IQ-DynamicGate technology for adaptive precision quantization, enhancing accuracy while maintaining memory efficiency.

Natural Language ProcessingGgufMultiple Languages