FineWeb is a pre-trained model specifically designed for text classification tasks in the financial domain. It is trained on high-quality financial text data and can accurately identify and classify the content of financial-related documents.

Natural Language ProcessingSafetensorsEnglish

A 360-million-parameter language model based on the LLaMA architecture and using MoR (Mixture of Recursions) technology, fine-tuned on the FineWeb-Edu deduplicated dataset, achieving efficient text generation capabilities through a dynamic routing mechanism and recursive KV cache.

Natural Language ProcessingTransformersEnglish

A GPT-2 model trained on the FineWeb-EDU dataset, demonstrating natural language processing capabilities

Natural Language ProcessingTransformers

This is a 774M-parameter language model based on the GPT-2 architecture, trained on 150 billion tokens from the FineWeb dataset.

Natural Language ProcessingTransformersEnglish

A lightweight FastText-based classifier for evaluating the educational value of web content, optimized for CPU processing speed

Natural Language ProcessingFasttextEnglish

This model is part of the FineWeb ablation experiment, with 1.82 billion parameters, based on the Llama architecture, trained using the FineWeb-Edu dataset, and suitable for English text completion tasks.

Natural Language ProcessingTransformersEnglish

A webpage educational value assessment classifier trained on the FineWeb dataset, used for filtering high-quality educational content

Natural Language ProcessingTransformersEnglish