prunaai/qwen3-32b 📝🔢✓ → 📝

▶️ 21.3K runs 📅 May 2025 ⚙️ Cog 0.14.7 🔗 GitHub
code-generation question-answering text-generation

About

A 2x faster qwen 3 model through pruna oss

Example Output

Prompt:

"How many helicopters can a human eat in one sitting?"

Output

A human cannot eat helicopters, as helicopters are made of metal, plastic, and other non-edible materials. Therefore, the answer is zero—a human cannot eat any helicopters in one sitting or at all. If this was meant as a humorous or hypothetical question, feel free to clarify!

Performance Metrics

2.30s Prediction Time
2.31s Total Time
All Input Parameters 
{
  "prompt": "How many helicopters can a human eat in one sitting?",
  "max_new_tokens": 2048,
  "enable_thinking": false
}
Input Parameters
prompt Type: stringDefault: How many helicopters can a human eat in one sitting?
Prompt for the model
max_new_tokens Type: integerDefault: 512
Maximum number of new tokens to generate
enable_thinking Type: booleanDefault: true
Enable thinking mode (outputs intermediate reasoning)
Output Schema

Output

Type: string

Example Execution Logs 
Number of tokens generated: 61
Version Details
Version ID
5ea670f086464b322044e36d46f65506db15334e7d777efc0e886162684d4831
Version Created
May 3, 2025
Run on Replicate →