iromu/Qwen3-0.6B-tools

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.8BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 26, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The iromu/Qwen3-0.6B-tools model is a 0.8 billion parameter Qwen3-based causal language model, fine-tuned with LoRA for enhanced tool calling and agent-style interactions. It features a 32768-token context length and is specifically optimized for structured function calling and multi-step agentic workflows. This model is designed for small-footprint on-device or edge deployments where efficient tool use is critical.

Loading preview...

Model Overview

iromu/Qwen3-0.6B-tools is a 0.8 billion parameter model based on the Qwen3 architecture, specifically fine-tuned using LoRA for robust tool calling and agent-style interactions. It leverages a 32768-token context window, making it suitable for complex multi-step tasks requiring function execution.

Key Capabilities

  • Structured Tool/Function Calling: Achieves significantly improved performance in tool call emission and argument matching compared to its base model, with the BF16 version reaching 66.0% exact-args match.
  • Agent-Style Interactions: Optimized for multi-step conversational agents that require external tool use.
  • Efficient Deployment: Designed for small-footprint applications, including on-device or edge deployments, due to its compact size.

Training Details

The model was fine-tuned from Qwen/Qwen3-0.6B using NVIDIA NeMo AutoModel with LoRA (dimension 32, alpha 32, dropout 0.05) on the sft_tools split of the r0b0tlab/qwen3.8-max-glm5.2-kimi-k3-distillation dataset. Training involved 336 steps with a max sequence length of 4096 and a learning rate of 5e-5, utilizing bf16 mixed precision.

Intended Use Cases

This model is ideal for developers building applications that require:

  • Reliable and accurate function calling from natural language prompts.
  • Autonomous agents capable of interacting with external systems or APIs.
  • Deployment in resource-constrained environments where larger models are impractical.