borong-finetune-qwen-2.5

This is a fine-tuned version of Qwen/Qwen2.5-14B-Instruct.

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ehsan404/borong-finetune-qwen-2.5")
model = AutoModelForCausalLM.from_pretrained("ehsan404/borong-finetune-qwen-2.5")

# Generate response
inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=256)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Model Details

• Base Model: Qwen/Qwen2.5-14B-Instruct
• Fine-tuning: LoRA with rank 8
• Training: Custom dataset
• Precision: bfloat16