---
tags:
- Code-Generation
- autotrain
- text-generation
- Llama2
- Pytorch
- PEFT
- QLoRA
- code
- coding
pipeline_tag: text-generation
widget:
- text: 'Write a program that add five numbers'
- text: 'Write a python code for reading multiple images'
- text: 'Write a python code for the name Ahmed to be in a reversed order'
---

# LlaMa2-CodeGen
This model is **LlaMa-2 7b** [Llama-2](https://huggingface.co/meta-llama/Llama-2-7b) fine-tuned on the **CodeSearchNet dataset** [CodeSearchNet](https://github.com/github/CodeSearchNet) by using the method **QLoRA** [QLoRA](https://github.com/artidoro/qlora) with [PEFT](https://github.com/huggingface/peft) library.

# Model Trained on Google Colab Pro Using AutoTrain, PEFT and QLoRA



### Example 
```py


import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig

peft_model_id = "AhmedSSoliman/Llama2-CodeGen-PEFT-QLoRA"
config = PeftConfig.from_pretrained(peft_model_id)
model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path, trust_remote_code=True, return_dict=True, load_in_4bit=True, device_map='auto')
tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)

# Load the Lora model
model = PeftModel.from_pretrained(model, peft_model_id)



def create_prompt(instruction):
  system = "You are a coding assistant that will help the user to resolve the following instruction:"
  instruction = "\n### Input: " + instruction
  return system + "\n" + instruction + "\n\n" + "### Response:" + "\n"

def generate(
        instruction,
        max_new_tokens=128,
        temperature=0.1,
        top_p=0.75,
        top_k=40,
        num_beams=4,
        **kwargs,
):
    prompt = create_prompt(instruction)
    print(prompt)
    inputs = tokenizer(prompt, return_tensors="pt")
    input_ids = inputs["input_ids"].to("cuda")
    attention_mask = inputs["attention_mask"].to("cuda")
    generation_config = GenerationConfig(
        temperature=temperature,
        top_p=top_p,
        top_k=top_k,
        num_beams=num_beams,
        **kwargs,
    )
    with torch.no_grad():
        generation_output = model.generate(
            input_ids=input_ids,
            attention_mask=attention_mask,
            generation_config=generation_config,
            return_dict_in_generate=True,
            output_scores=True,
            max_new_tokens=max_new_tokens,
            early_stopping=True
        )
    s = generation_output.sequences[0]
    output = tokenizer.decode(s)
    return output.split("### Response:")[1].lstrip("\n")


instruction = """
 Write a python code for the name Ahmed to be in a reversed order
"""
print(generate(instruction))
```