luckysong777/Qwen3-0.6B-JSON-SFT-GRPO
TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.8BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Oct 2, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold
The luckysong777/Qwen3-0.6B-JSON-SFT-GRPO is a 0.8 billion parameter Qwen3 model, finetuned from NotoriousH2/Qwen3-0.6B-JSON-SFT. This model is specifically optimized for JSON instruction following, leveraging Unsloth and Huggingface's TRL library for faster training. It is designed for tasks requiring structured JSON output, making it suitable for applications needing reliable data formatting.
Loading preview...
Model Overview
The luckysong777/Qwen3-0.6B-JSON-SFT-GRPO is a Qwen3-based language model with 0.8 billion parameters, finetuned by luckysong777. It builds upon the NotoriousH2/Qwen3-0.6B-JSON-SFT model and features a context length of 32768 tokens.
Key Capabilities
- JSON Instruction Following: This model is specifically finetuned for tasks that require generating structured JSON output based on instructions.
- Efficient Training: The finetuning process utilized Unsloth and Huggingface's TRL library, enabling a 2x faster training speed.
Good For
- Structured Data Generation: Ideal for applications where the output needs to adhere to a specific JSON schema.
- API Interaction: Can be used to generate JSON payloads or parse JSON responses for interacting with APIs.
- Data Extraction: Suitable for extracting information from text and formatting it into JSON objects.