Qwen2.5-3B-Instruct
Collection
Collection of pruned models based on Qwen/Qwen2.5-3B-Instruct
•
6 items
•
Updated
Qwen2.5-3B-Instruct-math-medium-heavy
This model is a medium-heavy pruned version of Qwen/Qwen2.5-3B-Instruct, specialized for MATH tasks.
Pruning Details
- Base Model: Qwen/Qwen2.5-3B-Instruct
- Specialization: Math
- Prune Mode: Medium-heavy
- Method: Activation-based weight pruning
Performance Comparison
| Category | Original | Pruned |
|---|---|---|
| Python | 100.0% | 60.0% |
| HTML | 6.7% | 6.7% |
| Trivia | 100.0% | 100.0% |
| Math | 60.0% | 53.3% |
| Reasoning | N/A | N/A |
| Medical | 100.0% | 100.0% |
| Linux | 100.0% | 100.0% |
| Writing | 73.3% | 66.7% |
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("CompactAI/Qwen2.5-3B-Instruct-math-medium-heavy")
tokenizer = AutoTokenizer.from_pretrained("CompactAI/Qwen2.5-3B-Instruct-math-medium-heavy")
License
This model inherits the license from the base model.
- Downloads last month
- -
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support