Update README.md (#1)
Browse files- Update README.md (86a4ad9735478db389f930bfe9fb7837b284ec09)
README.md
CHANGED
@@ -22,35 +22,98 @@ tags:
|
|
22 |
- transformers.js
|
23 |
language:
|
24 |
- multilingual
|
|
|
|
|
25 |
- ar
|
|
|
|
|
|
|
|
|
26 |
- bn
|
|
|
|
|
|
|
|
|
|
|
27 |
- da
|
28 |
- de
|
29 |
- el
|
30 |
- en
|
|
|
31 |
- es
|
|
|
|
|
|
|
32 |
- fi
|
33 |
- fr
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
- hi
|
|
|
|
|
|
|
35 |
- id
|
|
|
36 |
- it
|
37 |
- ja
|
|
|
38 |
- ka
|
|
|
|
|
|
|
39 |
- ko
|
|
|
|
|
|
|
|
|
|
|
40 |
- lv
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
- nl
|
42 |
- no
|
|
|
|
|
|
|
43 |
- pl
|
|
|
44 |
- pt
|
45 |
- ro
|
46 |
- ru
|
|
|
|
|
|
|
47 |
- sk
|
|
|
|
|
|
|
|
|
|
|
48 |
- sv
|
|
|
|
|
|
|
49 |
- th
|
|
|
50 |
- tr
|
|
|
51 |
- uk
|
52 |
- ur
|
|
|
53 |
- vi
|
|
|
|
|
54 |
- zh
|
55 |
inference: false
|
56 |
---
|
@@ -70,15 +133,19 @@ inference: false
|
|
70 |
<b>Jina CLIP: your CLIP model is also your text retriever!</b>
|
71 |
</p>
|
72 |
|
|
|
|
|
|
|
|
|
73 |
|
74 |
## Intended Usage & Model Info
|
75 |
|
76 |
`jina-clip-v2` is a state-of-the-art **multilingual and multimodal (text-image) embedding model**.
|
77 |
|
78 |
`jina-clip-v2` is a successor to the [`jina-clip-v1`](https://huggingface.co/jinaai/jina-clip-v1) model and brings new features and capabilities, such as:
|
79 |
-
* *support for multiple languages* - the text tower now supports
|
80 |
-
* *embedding truncation on both image and text vectors* - both towers are trained using [Matryoshka Representation Learning](https://arxiv.org/abs/2205.13147) which enables slicing the output vectors and in as a result computation and storage costs as well
|
81 |
-
* *visual document retrieval performance boost* - with an image resolution of
|
82 |
|
83 |
Similar to our predecessor model, `jina-clip-v2` bridges the gap between text-to-text and cross-modal retrieval. Via a single vector space, `jina-clip-v2` offers state-of-the-art performance on both tasks.
|
84 |
This dual capability makes it an excellent tool for multimodal retrieval-augmented generation (MuRAG) applications, enabling seamless text-to-text and text-to-image searches within a single model.
|
@@ -210,38 +277,4 @@ If you find `jina-clip-v2` useful in your research, please cite the following pa
|
|
210 |
Year = {2024},
|
211 |
Eprint = {arXiv:2405.20204},
|
212 |
}
|
213 |
-
```
|
214 |
-
|
215 |
-
## FAQ
|
216 |
-
|
217 |
-
### I encounter this problem, what should I do?
|
218 |
-
|
219 |
-
```
|
220 |
-
ValueError: The model class you are passing has a `config_class` attribute that is not consistent with the config class you passed (model has <class 'transformers_modules.jinaai.jina-clip-implementation.7f069e2d54d609ef1ad2eb578c7bf07b5a51de41.configuration_clip.JinaCLIPConfig'> and you passed <class 'transformers_modules.jinaai.jina-clip-implementation.7f069e2d54d609ef1ad2eb578c7bf07b5a51de41.configuration_cli.JinaCLIPConfig'>. Fix one of those so they match!
|
221 |
-
```
|
222 |
-
|
223 |
-
There was a bug in Transformers library between 4.40.x to 4.41.1. You can update transformers to >4.41.2 or <=4.40.0
|
224 |
-
|
225 |
-
### Given one query, how can I merge its text-text and text-image cosine similarity?
|
226 |
-
|
227 |
-
Our emperical study shows that text-text cosine similarity is normally larger than text-image cosine similarity!
|
228 |
-
If you want to merge two scores, we recommended 2 ways:
|
229 |
-
|
230 |
-
1. weighted average of text-text sim and text-image sim:
|
231 |
-
|
232 |
-
```python
|
233 |
-
combined_scores = sim(text, text) + lambda * sim(text, image) # optimal lambda depends on your dataset, but in general lambda=2 can be a good choice.
|
234 |
-
```
|
235 |
-
|
236 |
-
2. apply z-score normalization before merging scores:
|
237 |
-
|
238 |
-
```python
|
239 |
-
# pseudo code
|
240 |
-
query_document_mean = np.mean(cos_sim_text_texts)
|
241 |
-
query_document_std = np.std(cos_sim_text_texts)
|
242 |
-
text_image_mean = np.mean(cos_sim_text_images)
|
243 |
-
text_image_std = np.std(cos_sim_text_images)
|
244 |
-
|
245 |
-
query_document_sim_normalized = (cos_sim_query_documents - query_document_mean) / query_document_std
|
246 |
-
text_image_sim_normalized = (cos_sim_text_images - text_image_mean) / text_image_std
|
247 |
-
```
|
|
|
22 |
- transformers.js
|
23 |
language:
|
24 |
- multilingual
|
25 |
+
- af
|
26 |
+
- am
|
27 |
- ar
|
28 |
+
- as
|
29 |
+
- az
|
30 |
+
- be
|
31 |
+
- bg
|
32 |
- bn
|
33 |
+
- br
|
34 |
+
- bs
|
35 |
+
- ca
|
36 |
+
- cs
|
37 |
+
- cy
|
38 |
- da
|
39 |
- de
|
40 |
- el
|
41 |
- en
|
42 |
+
- eo
|
43 |
- es
|
44 |
+
- et
|
45 |
+
- eu
|
46 |
+
- fa
|
47 |
- fi
|
48 |
- fr
|
49 |
+
- fy
|
50 |
+
- ga
|
51 |
+
- gd
|
52 |
+
- gl
|
53 |
+
- gu
|
54 |
+
- ha
|
55 |
+
- he
|
56 |
- hi
|
57 |
+
- hr
|
58 |
+
- hu
|
59 |
+
- hy
|
60 |
- id
|
61 |
+
- is
|
62 |
- it
|
63 |
- ja
|
64 |
+
- jv
|
65 |
- ka
|
66 |
+
- kk
|
67 |
+
- km
|
68 |
+
- kn
|
69 |
- ko
|
70 |
+
- ku
|
71 |
+
- ky
|
72 |
+
- la
|
73 |
+
- lo
|
74 |
+
- lt
|
75 |
- lv
|
76 |
+
- mg
|
77 |
+
- mk
|
78 |
+
- ml
|
79 |
+
- mn
|
80 |
+
- mr
|
81 |
+
- ms
|
82 |
+
- my
|
83 |
+
- ne
|
84 |
- nl
|
85 |
- no
|
86 |
+
- om
|
87 |
+
- or
|
88 |
+
- pa
|
89 |
- pl
|
90 |
+
- ps
|
91 |
- pt
|
92 |
- ro
|
93 |
- ru
|
94 |
+
- sa
|
95 |
+
- sd
|
96 |
+
- si
|
97 |
- sk
|
98 |
+
- sl
|
99 |
+
- so
|
100 |
+
- sq
|
101 |
+
- sr
|
102 |
+
- su
|
103 |
- sv
|
104 |
+
- sw
|
105 |
+
- ta
|
106 |
+
- te
|
107 |
- th
|
108 |
+
- tl
|
109 |
- tr
|
110 |
+
- ug
|
111 |
- uk
|
112 |
- ur
|
113 |
+
- uz
|
114 |
- vi
|
115 |
+
- xh
|
116 |
+
- yi
|
117 |
- zh
|
118 |
inference: false
|
119 |
---
|
|
|
133 |
<b>Jina CLIP: your CLIP model is also your text retriever!</b>
|
134 |
</p>
|
135 |
|
136 |
+
## Quick Start
|
137 |
+
|
138 |
+
[Blog](https://jina.ai/news/jina-embeddings-v3-a-frontier-multilingual-embedding-model/#parameter-dimensions) | [Azure](https://azuremarketplace.microsoft.com/en-us/marketplace/apps/jinaai.jina-clip-v2) | [AWS SageMaker](https://aws.amazon.com/marketplace/pp/prodview-kdi3xkt62lo32) | [API](https://jina.ai/embeddings)
|
139 |
+
|
140 |
|
141 |
## Intended Usage & Model Info
|
142 |
|
143 |
`jina-clip-v2` is a state-of-the-art **multilingual and multimodal (text-image) embedding model**.
|
144 |
|
145 |
`jina-clip-v2` is a successor to the [`jina-clip-v1`](https://huggingface.co/jinaai/jina-clip-v1) model and brings new features and capabilities, such as:
|
146 |
+
* *support for multiple languages* - the text tower now supports 100 languages with tuning focus on **Arabic, Bengali, Chinese, Danish, Dutch, English, Finnish, French, Georgian, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Latvian, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Thai, Turkish, Ukrainian, Urdu,** and **Vietnamese.**
|
147 |
+
* *embedding truncation on both image and text vectors* - both towers are trained using [Matryoshka Representation Learning](https://arxiv.org/abs/2205.13147) which enables slicing the output vectors and in as a result computation and storage costs as well.
|
148 |
+
* *visual document retrieval performance boost* - with an image resolution of 512 (compared to 224 on `jina-clip-v1`) the image tower can now capture finer visual details. This feature along with a more diverse training set enable the model to perform much better on visual document retrieval tasks. This enable `jina-clip-v2` as a strong encoder for future vLLM based retriever.
|
149 |
|
150 |
Similar to our predecessor model, `jina-clip-v2` bridges the gap between text-to-text and cross-modal retrieval. Via a single vector space, `jina-clip-v2` offers state-of-the-art performance on both tasks.
|
151 |
This dual capability makes it an excellent tool for multimodal retrieval-augmented generation (MuRAG) applications, enabling seamless text-to-text and text-to-image searches within a single model.
|
|
|
277 |
Year = {2024},
|
278 |
Eprint = {arXiv:2405.20204},
|
279 |
}
|
280 |
+
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|