ermiaazarkhalili/Qwen3.8-2B-Function-Calling-xLAM-Unsloth

Model Overview

This model, ermiaazarkhalili/Qwen3.8-2B-Function-Calling-xLAM-Unsloth, is a LoRA fine-tune of the empero-ai/Qwen3.8-2B base model, developed by ermiaazarkhalili. It utilizes the Qwen3_5ForConditionalGeneration architecture with 2.3 billion parameters and was fine-tuned using Unsloth and TRL.

Key Capabilities

• Function Calling: Supervised fine-tuned on the Salesforce/xlam-function-calling-60k dataset, specializing in generating structured outputs for function calls.
• Efficient Fine-tuning: Leverages LoRA (rank 64, alpha 64) with QLoRA 4-bit base precision for efficient adaptation.
• Optimized for Integration: Designed to interpret user requests and translate them into executable function calls, facilitating interaction with external APIs and tools.

Limitations and Considerations

• No Benchmark Evaluation: The model has not undergone downstream benchmark evaluation; only training loss observations are available.
• Inherited Biases: It inherits the biases, knowledge cutoff, and potential failure modes of its base model.
• Specialized Training: Fine-tuned exclusively on a single instruction-following dataset, meaning its performance outside this distribution is untested.
• Merged Adapters: The LoRA adapters are merged into the base weights, preventing detachment from this specific fine-tune.

Good For

• Tool Use Applications: Ideal for scenarios where an LLM needs to interact with external functions or APIs based on user prompts.
• Structured Output Generation: Suitable for tasks requiring the model to produce structured, machine-readable outputs for automation.
• Resource-Constrained Environments: Its 2.3B parameter size and efficient fine-tuning method make it viable for deployment in environments with limited computational resources.