README.md

---
tags:
- vision
- image-to-text
- pruning
- llava
base_model: llava-hf/llava-1.5-7b-hf
---

# llava-glu-30pct

This is a pruned version of [LLaVA-1.5-7b](https://huggingface.co/llava-hf/llava-1.5-7b-hf).

## Pruning Details
- **Method**: GLU Pruning
- **Sparsity**: 30%

This model was pruned to improve efficiency while maintaining performance.

## Usage

Since this model was pruned structurally, the architecture remains compatible with the standard `LlavaForConditionalGeneration` class. However, you should use the processor from the base model to ensure correct input preprocessing.

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

model_id = "CrystalRaindropsFall/llava-glu-30pct"
base_model_id = "llava-hf/llava-1.5-7b-hf"

# 1. Load the processor from the base model
processor = AutoProcessor.from_pretrained(base_model_id)

# 2. Load the pruned model
model = LlavaForConditionalGeneration.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto"
)

# Example inference
from PIL import Image
import requests

url = "https://github.com/haotian-liu/LLaVA/blob/1a91fc274d7c35a9b50b3cb29c4247ae5837ce39/images/llava_logo.png?raw=true"
image = Image.open(requests.get(url, stream=True).raw)
prompt = "USER: <image>\nWhat is shown in this image?\nASSISTANT:"

inputs = processor(images=image, text=prompt, return_tensors="pt").to(model.device, model.dtype)

output = model.generate(**inputs, max_new_tokens=100, do_sample=False)
print(processor.decode(output[0], skip_special_tokens=True))
```