---
license: apache-2.0
language: c++
tags:
  - code-generation
  - codellama
  - peft
  - unit-tests
  - causal-lm
  - text-generation
base_model: codellama/CodeLlama-7b-hf
model_type: llama
pipeline_tag: text-generation
---

# 🧪 CodeLLaMA Unit Test Generator (Merged Model)

This is a merged model that combines [`codellama/CodeLlama-7b-hf`](https://huggingface.co/codellama/CodeLlama-7b-hf) with LoRA fine-tuning trained on a dataset of embedded C/C++ functions and corresponding unit tests.

It specializes in generating **comprehensive unit tests** for C/C++ code using frameworks like **GoogleTest** or **CppUTest**, focusing on:
- Edge cases
- Boundary conditions
- Exception handling
- MISRA C compliance (if applicable)

---

## 📌 Example Usage

```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model = AutoModelForCausalLM.from_pretrained("Utkarsh524/codellama_utests_full")
tokenizer = AutoTokenizer.from_pretrained("Utkarsh524/codellama_utests_full")

prompt = "<|system|>\nGenerate comprehensive unit tests for C/C++ code. Cover all edge cases, boundary conditions, and error scenarios.\nOutput Constraints:\n1. ONLY include test code (no explanations, headers, or main functions)\n2. Start directly with TEST(...)\n3. End after last test case\n4. Never include framework boilerplate\n<|user|>\nCreate tests for:\nint add(int a, int b) { return a + b; }\n<|assistant|>\n"

inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))