File size: 2,762 Bytes
11aad8f cdf0b51 11aad8f cdf0b51 11aad8f cdf0b51 11aad8f cdf0b51 a2e0bb7 cdf0b51 bcc0173 cdf0b51 646963e cdf0b51 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 |
---
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 [**LlaMa2-7b**](https://huggingface.co/meta-llama/Llama-2-7b) which is fine-tuned on the [**CodeSearchNet dataset**](https://github.com/github/CodeSearchNet) by using the method [**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
# You can load the LlaMa2-CodeGen model on google colab.
### 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))
``` |