seanpoyner/smolcode-coder-dotnet-1.5b-tools
The seanpoyner/smolcode-coder-dotnet-1.5b-tools is a 1.5 billion parameter LoRA fine-tune of Qwen2.5-Coder-1.5B-Instruct, developed by seanpoyner. This model is specifically trained to emit native function calls, enabling agentic coding loops for small language models. With a 32768 token context length, it excels at driving agentic coding assistants by correctly parsing tool-use instructions.
Loading preview...
Overview
This model, smolcode-coder-dotnet-1.5b-tools, is a LoRA fine-tune of the Qwen2.5-Coder-1.5B-Instruct base model, developed by seanpoyner. Its primary purpose is to enable small language models (SLMs) to effectively drive agentic coding loops by correctly emitting native <tool_call> function calls. Unlike standard Qwen-Coder models that might output plain-text JSON for tool calls, this fine-tune ensures the output is in the format expected by runtimes like Ollama and llama.cpp, which is crucial for functional agentic tool-use.
Key Capabilities
- Native Tool Call Emission: Specifically trained to generate
<tool_call>{"name": ..., "arguments": ...}</tool_call>syntax, essential for agentic workflows. - SLM-Optimized Agentic Coding: Designed to facilitate agentic coding with tiny (≤2B) models, making advanced coding assistance more accessible and efficient.
- High Fidelity Training: Utilizes
NousResearch/hermes-function-calling-v1and syntheticsmolcodetool-use trajectories, ensuring robust and accurate tool call generation. Training incorporated assistant-only loss, focusing on tool calls and final answers. - Context Length: Supports a substantial context length of 32768 tokens.
Good For
- Developers building agentic coding assistants that require precise tool call parsing from small language models.
- Use cases where efficient and accurate function calling is critical for automating coding tasks.
- Integration with runtimes that expect native
<tool_call>formats for tool execution.