NOSIBLE/prediction-v1.1-base
TEXT GENERATIONConcurrency Cost:1Model Size:0.8BQuant:BF16Ctx Length:32kPublished:Nov 25, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

NOSIBLE/prediction-v1.1-base is a 0.8 billion parameter classification model fine-tuned from Qwen3-0.6B, specifically designed to identify whether short text snippets contain predictions. Trained on 100k real-world financial search results from Nosible Search Feeds, it excels at accurately distinguishing predictive statements from descriptive or historical text in noisy, unstructured financial data. This lightweight model is optimized for scalable prediction mining and robust performance in financial contexts, outperforming larger LLMs in its specialized task.

Loading preview...