Qwen3-0.6B-seo-finetuned

Qwen3-0.6B SEO Fine-tuned

Fine-tuned version of Qwen/Qwen3-0.6B on the Global SEO Knowledge dataset.

Training Details

• Method: LoRA (r=16, alpha=32)
• Trainable params: 10M / 606M (1.67%)
• Epochs: 3
• Dataset: 2,065 SEO knowledge examples
• Final loss: 1.14
• Token accuracy: 79.8%

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("Kelnux/Qwen3-0.6B-seo-finetuned")
tokenizer = AutoTokenizer.from_pretrained("Kelnux/Qwen3-0.6B-seo-finetuned")

messages = [{"role": "user", "content": "What is Core Web Vitals in SEO?"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Categories Covered

• Technical SEO (robots.txt, sitemaps, Core Web Vitals, canonical URLs, schema markup)
• On-Page SEO (keyword density, meta tags, content optimization)
• Off-Page SEO (backlinking, domain authority, link building)
• Content SEO (E-E-A-T, content quality)