LFM2.5-1.2B-Base-linux-safe

LINUX-optimized | Safe pruning | 25% weights pruned

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

Performance Comparison

Category	Original	Pruned	Change
Python	0.0%	0.0%	→
Html	0.0%	0.0%	→
Trivia	40.0%	10.0%	↓ 30.0%
Math	15.0%	15.0%	→
Reasoning	0.0%	0.0%	→
Medical	20.0%	0.0%	↓ 20.0%
Linux	45.0%	20.0% ⭐	↓ 25.0%
Writing	10.0%	5.0%	↓ 5.0%

Average: 16.2% -> 6.2% (-10.0%)

Linux Retention: 44.4%

Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer

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

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	Linux
Prune Mode	Safe
Weight Reduction	25% weights pruned

License

This model inherits the license from the base model.