File size: 3,260 Bytes
a3a7806
 
 
 
 
 
 
 
 
 
 
ca4fb56
a3a7806
 
 
e1c449a
befc36c
641f724
a3a7806
641f724
a3a7806
f2df3cf
a3a7806
 
 
 
 
 
 
 
 
 
 
 
a65ec98
 
506f9f2
 
a65ec98
a3a7806
 
 
 
 
 
 
 
 
 
 
 
 
a65ec98
8f4c803
a3a7806
 
 
 
a65ec98
a3a7806
 
 
 
 
 
 
 
 
 
a65ec98
a3a7806
 
 
a65ec98
a3a7806
 
 
 
 
 
45417f7
 
 
a3a7806
 
 
 
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
---
inference: false
license: apache-2.0
---

# Model Card

<p align="center">
  <img src="./icon.png" alt="Logo" width="350">
</p>

๐Ÿ“– [Technical report](https://arxiv.org/abs/2402.11530) | ๐Ÿ  [Code](https://github.com/BAAI-DCAI/Bunny) | ๐Ÿฐ [Demo](http://bunny.baai.ac.cn) | ๐Ÿค— [GGUF](https://huggingface.co/BAAI/Bunny-Llama-3-8B-V-gguf)

This is Bunny-Llama-3-8B-V.

We also provide v1.1 version accepting high-resolution images up to 1152x1152. ๐Ÿค— [v1.1](https://huggingface.co/BAAI/Bunny-v1_1-Llama-3-8B-V)

Bunny is a family of lightweight but powerful multimodal models. It offers multiple plug-and-play vision encoders, like EVA-CLIP, SigLIP and language backbones, including Llama-3-8B, Phi-1.5, StableLM-2, Qwen1.5, MiniCPM and Phi-2. To compensate for the decrease in model size, we construct more informative training data by curated selection from a broader data source.

We provide Bunny-Llama-3-8B-V, which is built upon [SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) and [Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct). More details about this model can be found in [GitHub](https://github.com/BAAI-DCAI/Bunny).

![comparison](comparison.png)


# Quickstart

Here we show a code snippet to show you how to use the model with transformers.

Before running the snippet, you need to install the following dependencies:

```shell
pip install torch transformers accelerate pillow
```

If the CUDA memory is enough, it would be faster to execute this snippet by setting `CUDA_VISIBLE_DEVICES=0`.

Users especially those in Chinese mainland may want to refer to a HuggingFace [mirror site](https://hf-mirror.com).


```python
import torch
import transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
from PIL import Image
import warnings

# disable some warnings
transformers.logging.set_verbosity_error()
transformers.logging.disable_progress_bar()
warnings.filterwarnings('ignore')

# set device
device = 'cuda'  # or cpu
torch.set_default_device(device)

# create model
model = AutoModelForCausalLM.from_pretrained(
    'BAAI/Bunny-Llama-3-8B-V',
    torch_dtype=torch.float16, # float32 for cpu
    device_map='auto',
    trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(
    'BAAI/Bunny-Llama-3-8B-V',
    trust_remote_code=True)

# text prompt
prompt = 'Why is the image funny?'
text = f"A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: <image>\n{prompt} ASSISTANT:"
text_chunks = [tokenizer(chunk).input_ids for chunk in text.split('<image>')]
input_ids = torch.tensor(text_chunks[0] + [-200] + text_chunks[1][1:], dtype=torch.long).unsqueeze(0).to(device)

# image, sample images can be found in images folder
image = Image.open('example_2.png')
image_tensor = model.process_images([image], model.config).to(dtype=model.dtype, device=device)

# generate
output_ids = model.generate(
    input_ids,
    images=image_tensor,
    max_new_tokens=100,
    use_cache=True,
    repetition_penalty=1.0 # increase this to avoid chattering
)[0]

print(tokenizer.decode(output_ids[input_ids.shape[1]:], skip_special_tokens=True).strip())
```