cahlen/Convergent-7B
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Apr 6, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

Convergent-7B by cahlen is a 7.6 billion parameter QLoRA fine-tuned model based on Qwen2.5-7B-Instruct, specifically designed as an agentic research companion for computational mathematics. It excels in number theory, generating CUDA kernel scaffolding, and performing agentic tool calls to the bigcompute.science MCP server. This model is optimized for reasoning about mathematical findings and suggesting novel research directions in number theory.

Loading preview...