Apertus is a fully open multilingual large language model developed by Swiss AI, offering two parameter scales of 7 billion and 8 billion. This model supports over 1000 languages, uses fully compliant and open training data, and its performance can rival that of closed-source models. Apertus was pre-trained on 15T tokens and adopts a phased curriculum training method, supporting a context length of up to 65,536 tokens.

Natural Language ProcessingTransformersOther

Granite-4.0-Micro-Base is a decoder-only long-context language model developed by IBM. It is trained from scratch on approximately 15 trillion tokens and adopts a four-stage training strategy. This model is designed for various text generation tasks and supports multilingual and fill-in-the-middle code completion functions.

Natural Language ProcessingTransformers

MiniCPM4.1-8B GGUF is an efficient large language model designed specifically for edge devices. It has carried out systematic innovations in four key dimensions: model architecture, training data, training algorithm, and inference system, achieving extreme efficiency improvement. It supports a context length of 65,536 tokens and a hybrid inference mode.

Natural Language ProcessingTransformersMultiple Languages

Kuaishou Keye-VL is a cutting - edge multimodal large language model developed by the Kuaishou Keye team, which performs excellently in video understanding, visual perception, and reasoning tasks. The 1.5 version reaches a new height in video understanding, image perception, and reasoning ability through innovative fast - slow video encoding strategies, LongCoT cold - start data pipelines, and reinforcement learning training strategies, and supports an extended context length of up to 128k tokens.

MultimodalSafetensorsMultiple Languages

Moonlight is a Mixture of Experts (MoE) model with a total of 16B parameters and 3B active parameters, trained based on the Muon optimizer using 5.7T tokens. This model improves the current Pareto frontier, achieving better performance with fewer training FLOPs and excelling in multiple benchmark tests.

Natural Language ProcessingTransformers

Tri-21B is a flagship large language model that achieves state-of-the-art performance with only 2T training tokens, redefining the boundary of training efficiency.

Natural Language ProcessingTransformersMultiple Languages

Qwen3-8B-Base is the latest generation of large language models in the Qwen series. Based on 36 trillion tokens of multilingual pre-training data, it optimizes the model architecture and training techniques to provide an efficient and accurate language interaction experience.

Natural Language ProcessingTransformers

This model originates from karpathy's llm.c project, converted to HuggingFace format for researching bfloat16 performance, with a training process consuming 150 billion tokens.

Natural Language ProcessingTransformers

This is a speech language model based on discrete Hubert tokens, focusing on efficient training and capable of generating speech segment continuations.

Audio ProcessingTransformers

Yuna Ai V4 is a multilingual text generation model based on the LLaMA 3.1 8B architecture, focusing on providing intelligent dialogue and knowledge Q&A experiences. After training with over 3B tokens, it has advanced features such as quantum thinking and performs outstandingly in dialogue ability and creativity.

Natural Language ProcessingGgufMultiple Languages

Qwen2.5-Aloe-Beta-7B is an open-source large medical language model that achieves state-of-the-art performance in multiple medical tasks. It is fine-tuned based on the Qwen2.5-7B architecture, and the training data covers 1.8 billion tokens of diverse medical tasks.

Natural Language ProcessingTransformersEnglish

Granite-3.0-8B-Base is a decoder-only language model developed by IBM. It adopts a two-stage training strategy and is trained on multi-domain data with 12 trillion tokens. It supports 12 languages and can handle various text generation tasks such as text summarization, classification, information extraction, and question answering.

Natural Language ProcessingTransformers

Minitron-8B-Base is a large language model obtained by pruning Nemotron-4 15B, employing distillation and continuous training methods, saving 40 times the training tokens and 1.8 times the computational cost compared to training from scratch.

Natural Language ProcessingTransformersEnglish

Fietje 2 Chat Version is an open and efficient Dutch large language model, adapted from microsoft/phi-2 and customized through training on 28B tokens of Dutch text.

Natural Language ProcessingTransformersOther

GLA-1.3B-100B is a gated linear attention transformer model with 1.3 billion parameters, trained on 100 billion tokens. This model uses a hardware-efficient training method, focuses on text generation tasks, and supports English language processing.

Natural Language ProcessingTransformersMultiple Languages

SambaLingo-Turkish-Base is a bilingual (Turkish and English) model based on Llama-2-7b pre-training, adapted for Turkish by training on 42 billion tokens from the Turkish portion of the Cultura-X dataset.

Natural Language ProcessingTransformersMultiple Languages

TinyMistral-248M-v3 is a small language model with 248M parameters, currently still in training, having processed approximately 21 billion tokens.

Natural Language ProcessingTransformersEnglish

GEITje-7B is a large open-source Dutch language model based on Mistral 7B, significantly enhancing Dutch language capabilities and local knowledge through additional training on 10 billion Dutch text tokens.

Natural Language ProcessingTransformersOther

Large open-source Dutch conversational model based on Mistral 7B, significantly improving Dutch language capabilities through additional training on 10 billion Dutch text tokens

Natural Language ProcessingTransformersOther

A large-scale open-source Dutch language model based on Mistral 7B, significantly enhancing Dutch language comprehension and local knowledge coverage through incremental training on 10 billion Dutch text tokens.

Natural Language ProcessingTransformersOther