File size: 3,721 Bytes
80c6598
 
 
0297db8
 
 
 
3a858ea
 
 
db7ffe4
80c6598
 
1fe5c53
80c6598
83b9189
80c6598
 
 
 
 
90cd162
80c6598
fc9ff17
80c6598
817d582
 
3081d81
 
 
 
 
 
 
 
 
7923606
3081d81
 
817d582
 
 
 
3081d81
817d582
 
 
 
 
a1ac49a
817d582
 
 
 
 
 
 
 
 
 
34ec5c6
 
 
 
 
817d582
34ec5c6
 
 
 
817d582
34ec5c6
817d582
 
 
 
 
34ec5c6
02714ae
817d582
34ec5c6
 
02714ae
817d582
34ec5c6
 
817d582
14fb892
 
 
 
1de78e8
 
a454b7c
 
 
 
a22168a
 
 
 
 
db7ffe4
7c03122
a22168a
 
 
 
 
 
fe21202
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
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
---
language:
- en
tags:
- information retrieval
- embedding model
- visual information retrieval
metrics:
- recall
pipeline_tag: feature-extraction
license: apache-2.0
---

# OCR-free Visual Document Embedding Model as Your Personal Librarian

The model only takes images as document-side inputs and produce vectors representing document pages. `minicpm-visual-embedding-v0` is trained with over 200k query-visual document pairs, including textual document, visual document, arxiv figures, industry documents, textbooks, ebooks, etc. The performance of `minicpm-visual-embedding-v0` is on a par with our ablation text embedding model on text-oriented documents, and an advantages on visually-intensive documents.

![Memex Archtechture](images/memex.png)

# News

- 2024-06-27: 🚀 We released our first visual embedding model checkpoint minicpm-visual-embedding-v0 on [huggingface](https://huggingface.co/RhapsodyAI/minicpm-visual-embedding-v0).

- 2024-05-08: 🌍 We [open-sourced](https://github.com/RhapsodyAILab/minicpm-visual-embedding-v0) our training code (full-parameter tuning with GradCache and DeepSpeed, supports large batch size across multiple GPUs with zero-stage1) and eval code. 

# Get started

Pip install all dependencies:

```
Pillow==10.1.0
timm==0.9.10
torch==2.1.2
torchvision==0.16.2
transformers==4.36.0
sentencepiece==0.1.99
numpy==1.26.0
```

First you are suggested to git clone this huggingface repo or download repo with `huggingface_cli`.

```bash
git lfs install
git clone https://huggingface.co/RhapsodyAI/minicpm-visual-embedding-v0
```

or 

```bash
huggingface-cli download --resume-download RhapsodyAI/minicpm-visual-embedding-v0 --local-dir minicpm-visual-embedding-v0 --local-dir-use-symlinks False
```

```python
from transformers import AutoModel
from transformers import AutoTokenizer
from PIL import Image
import torch

device = 'cuda:0'

# Load model, be sure to substitute `model_path` by your model path 
model_path = '/local/path/to/model'
tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
model = AutoModel.from_pretrained(model_path, trust_remote_code=True)
model.to(device)

# Load image to PIL.Image object
image_1 = Image.open('/local/path/to/images/memex.png').convert('RGB')
image_2 = Image.open('/local/path/to/images/us2020.png').convert('RGB')
image_3 = Image.open('/local/path/to/images/hard_negative.png').convert('RGB')

# User query
query_instruction = 'Represent this query for retrieving relavant document: '
query = 'Who was elected as president of United States in 2020?'
query_full = query_instruction + query

# Embed image documents
with torch.no_grad():
    p_reps = model(text=['', '', ''], image=[image_1, image_2, image_3], tokenizer=tokenizer).reps

# Embed text queries
with torch.no_grad():
    q_reps = model(text=[query_full], image=[None], tokenizer=tokenizer).reps # [B, s, d]

# Calculate similarities
scores = torch.matmul(q_reps, p_reps.T)
print(scores)
```

# Limitations

- This checkpoint is an alpha version, and may not be strong in your tasks, for bad case, please create an issue to let us know, many thanks!

- The modeling script `modeling_minicpmv` on `huggingface` is not standard yet, the inference code could be further improved.

- The inference speed is low, because vision encoder uses `timm`, which does not yet support `flash-attn`.


# Citation

If you find our work useful, please consider cite us:

```bibtex
@misc{RhapsodyEmbedding2024,
  author = {RhapsodyAI},
  title = {OCR-free Visual Document Embedding Model as Your Personal Librarian},
  year = {2024},
  howpublished = {\url{https://huggingface.co/RhapsodyAI/minicpm-visual-embedding-v0}},
  note = {Accessed: 2024-06-28}
}
```