coderian/axiom-python-1.5B
coderian/axiom-python-1.5B is a 1.5 billion parameter causal language model, fine-tuned from Qwen/Qwen2.5-1.5B, specifically optimized for Python programming and code generation. It was trained using LoRA + SFT on CodeAlpaca_20K and PythonCodeInstruct_18K datasets, making it proficient in generating Python code and handling code-focused instructions in both English and Turkish. This model excels at single-line and medium-complexity Python functions, offering a specialized solution for developers needing efficient code assistance.
Loading preview...
Axiom Python 1.5B Overview
Axiom Python 1.5B is a specialized 1.5 billion parameter causal language model, fine-tuned from the Qwen/Qwen2.5-1.5B base model. Its primary focus is on generating Python code and assisting with Python programming tasks. The model was trained using LoRA (Low-Rank Adaptation) and Supervised Fine-Tuning (SFT) on a combination of the CodeAlpaca_20K and PythonCodeInstruct_18K datasets, enhancing its proficiency in Python.
Key Capabilities
- Python Code Generation: Optimized for creating Python functions and code snippets.
- Instruction Following: Responds effectively to Python-related coding instructions.
- Multilingual Support: Handles code-focused tasks in both English and Turkish.
- Efficient Performance: A compact 1.5B parameter model suitable for various deployment scenarios.
Good For
- Generating single-line and medium-complexity Python functions.
- Developers seeking a lightweight model for Python code assistance.
- Applications requiring code completion or generation based on specific instructions.
- Use cases where stable and deterministic code output is preferred (by lowering
temperature).
Limitations
- May struggle with very complex or lengthy code generation tasks due to its size.
- Performance is limited to Python; it is not designed for other programming languages.
- Consistency may degrade in very long contexts, as training data had a maximum length of 256 tokens.