README.md

# LoRA Adapter Model

This is a LoRA adapter model fine-tuned on llava-hf/llava-1.5-7b-hf.

## Model Details
- Base Model: llava-hf/llava-1.5-7b-hf
- Training Parameters:
  - Learning Rate: 1e-4
  - Batch Size: 16
  - Training Steps: 58

## Usage

```python
from transformers import LlavaForConditionalGeneration, AutoProcessor
from peft import PeftModel
import torch

# Load base model
base_model = LlavaForConditionalGeneration.from_pretrained(
    "llava-hf/llava-1.5-7b-hf",
    revision='a272c74',
    torch_dtype=torch.float16,
    device_map="auto"
)
tokenizer = AutoProcessor.from_pretrained("llava-hf/llava-1.5-7b-hf", revision='a272c74')

# Load LoRA adapter
model = PeftModel.from_pretrained(
    base_model,
    "Dipto084/RepLLaVA4",
    torch_dtype=torch.float16,
    device_map="auto"
)
```