LFM2.5-1.2B-Base-math-aggressive

🎯 MATH-optimized | 📦 Aggressive pruning | ⚡ 20% weights pruned

This model is a aggressively pruned version of LiquidAI/LFM2.5-1.2B-Base.

Performance Comparison

Category	Original	Pruned	Change
Python	0.0%	0.0%	→
Html	16.7%	25.0%	↑ 8.3%
Trivia	91.7%	83.3%	↓ 8.3%
Math	75.0%	83.3% ⭐	↑ 8.3%
Reasoning	41.7%	58.3%	↑ 16.7%
Medical	66.7%	50.0%	↓ 16.7%
Linux	16.7%	8.3%	↓ 8.3%
Writing	33.3%	41.7%	↑ 8.3%

Average: 42.7% → 43.8% (+1.0%)

Math Retention: 111.1%

Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("CompactAI/LFM2.5-1.2B-Base-math-aggressive")
tokenizer = AutoTokenizer.from_pretrained("CompactAI/LFM2.5-1.2B-Base-math-aggressive")

inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Technical Details

Property	Value
Base Model	LiquidAI/LFM2.5-1.2B-Base
Specialization	Math
Prune Mode	Aggressive
Weight Reduction	20% weights pruned

License

This model inherits the license from the base model.