jastorj/couchmind-v5.7.6.1_qwen_stage_1-cw-6K-16bit

TEXT GENERATIONConcurrent Unit Cost:1Model Size:7.6BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jun 14, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

jastorj/couchmind-v5.7.6.1_qwen_stage_1-cw-6K-16bit is a 7.6 billion parameter Qwen2.5-Coder-7B-Instruct model fine-tuned by jastorj for text-to-SQL generation. It specializes in converting natural language questions into syntactically valid Couchbase SQL++ queries, leveraging the NL2SQL++ v5.7.6.1_qwen_stage_1 dataset with code-with-thought reasoning. This model is optimized for precise SQL++ query generation based on provided database schemas and natural language input.

Loading preview...

Model Overview

This model, jastorj/couchmind-v5.7.6.1_qwen_stage_1-cw-6K-16bit, is a specialized fine-tune of the Qwen/Qwen2.5-Coder-7B-Instruct base model. Developed by jastorj, its primary function is text-to-SQL generation, specifically for Couchbase SQL++ queries.

Key Capabilities

  • NL2SQL++ Conversion: Translates natural language questions into syntactically correct Couchbase SQL++ queries.
  • Schema-Aware Generation: Utilizes provided database schemas (including bucket, scope, and collection names) to ensure accurate query generation.
  • Code-with-Thought Reasoning: Fine-tuned on the NL2SQL++ v5.7.6.1_qwen_stage_1 dataset, which incorporates code-with-thought reasoning to enhance query logic.
  • LoRA Fine-tuning: Employs Low-Rank Adaptation (LoRA) with Unsloth for efficient fine-tuning.
  • 16-bit Quantization: Features 16-bit merged weights for optimized performance.

Training Details

The model was trained on 25,323 examples from the NL2SQL++ v5.7.6.1_qwen_stage_1 dataset. Key training parameters include a learning rate of 1e-05, 3 epochs, and a maximum sequence length of 6000 tokens, with eval_exec_accuracy as the metric for saving the best model.

Ideal Use Cases

This model is particularly well-suited for applications requiring automated generation of Couchbase SQL++ queries from natural language, such as:

  • Database interaction tools.
  • Business intelligence dashboards with natural language querying capabilities.
  • Developer assistance for Couchbase SQL++ query construction.