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

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Apr 25, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The ermiaazarkhalili/Qwen3-8B-Function-Calling-xLAM-Unsloth model is a fine-tuned 8 billion parameter Qwen3 model, developed by ermiaazarkhalili, specifically optimized for function calling tasks. It leverages Unsloth for efficient training, resulting in 2x faster fine-tuning and 60% less VRAM usage. Trained on the Salesforce/xlam-function-calling-60k dataset, this model excels at interpreting natural language queries and generating structured function calls within a 2,048 token context window.

Loading preview...

Overview

This model, developed by ermiaazarkhalili, is a fine-tuned version of the 8 billion parameter Qwen3 model, specifically optimized for function calling. It utilizes the Unsloth framework for efficient training, enabling 2x faster training and 60% less VRAM consumption compared to standard methods.

Key Capabilities

  • Function Calling: Specialized in converting natural language requests into structured function calls.
  • Efficient Training: Fine-tuned using Unsloth with QLoRA (4-bit) on an NVIDIA H100 GPU, achieving a final training loss of 0.2186 in under 4 hours.
  • Dataset: Trained on the Salesforce/xlam-function-calling-60k dataset, comprising 60,000 examples with XML-tagged queries, tool definitions, and structured answers.
  • Context Length: Supports a context window of 2,048 tokens.
  • Quantized Versions: Available in GGUF formats (Q4_K_M, Q5_K_M, Q8_0) for CPU and edge inference, compatible with Ollama and llama.cpp.

Good For

  • Applications requiring robust and efficient function calling capabilities.
  • Developers looking for a Qwen3-based model optimized for tool use and API interaction.
  • Scenarios where efficient resource utilization during fine-tuning is critical.