Parameter	Value
Base model	`Qwen/Qwen2.5-0.5B`
Method	SFT + LoRA
LoRA r / alpha	32 / 16
LoRA targets	all-linear
Epochs	1
Learning rate	0.0002
Batch size	8 × 4 (grad accum)
Training examples	1024
Precision	bf16

Answer Format

The model is trained to end responses with:

The answer is: {number}.

System Prompt

You are a helpful math assistant. Solve the problem step by step, then give your final answer as a single number on the last line in exact format
The answer is: {number}.

tripathysagar/Qwen2.5-0.5B-GSM8K-SFT

Qwen2.5-0.5B-GSM8K-SFT

Training Details

Answer Format

System Prompt

tripathysagar/
Qwen2.5-0.5B-GSM8K-SFT