Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,79 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
inference: false
|
6 |
---
|
7 |
+
|
8 |
+
<br><br>
|
9 |
+
|
10 |
+
<p align="center">
|
11 |
+
<img src="https://github.com/jina-ai/finetuner/blob/main/docs/_static/finetuner-logo-ani.svg?raw=true" alt="Finetuner logo: Finetuner helps you to create experiments in order to improve embeddings on search tasks. It accompanies you to deliver the last mile of performance-tuning for neural search applications." width="150px">
|
12 |
+
</p>
|
13 |
+
|
14 |
+
|
15 |
+
<p align="center">
|
16 |
+
<b>The text embedding suit trained by Jina AI, Finetuner team.</b>
|
17 |
+
</p>
|
18 |
+
|
19 |
+
|
20 |
+
## Intented Usage & Model Info
|
21 |
+
|
22 |
+
`jina-embedding-l-en-v1` is a language model that has been trained using Jina AI's Linnaeus-Clean dataset.
|
23 |
+
This dataset consists of 380 million pairs of sentences, which include both query-document pairs.
|
24 |
+
These pairs were obtained from various domains and were carefully selected through a thorough cleaning process.
|
25 |
+
The Linnaeus-Full dataset, from which the Linnaeus-Clean dataset is derived, originally contained 1.6 billion sentence pairs.
|
26 |
+
|
27 |
+
The model has a range of use cases, including information retrieval, semantic textual similarity, text reranking, and more.
|
28 |
+
|
29 |
+
With a size of 330 million parameters,
|
30 |
+
the model enables single-gpu inference while delivering better performance than our small and base model.
|
31 |
+
Additionally, we provide the following options:
|
32 |
+
|
33 |
+
- `jina-embedding-s-en-v1`: 35 million parameters **(you are here)**.
|
34 |
+
- `jina-embedding-b-en-v1`: 110 million parameters.
|
35 |
+
- `jina-embedding-l-en-v1`: 330 million parameters.
|
36 |
+
- `jina-embedding-1b-en-v1`: 1.2 billion parameters, 10* bert-base size (soon).
|
37 |
+
- `jina-embedding-6b-en-v1`: 6 billion parameters 30* bert-base size(soon).
|
38 |
+
|
39 |
+
## Data & Parameters
|
40 |
+
|
41 |
+
More info will be released together with the technique report.
|
42 |
+
|
43 |
+
## Metrics
|
44 |
+
|
45 |
+
We compared the model against `all-minilm-l6-v2`/`all-mpnet-base-v2` from sbert and `text-embeddings-ada-002` from OpenAI:
|
46 |
+
|
47 |
+
|Name|param |context|
|
48 |
+
|------------------------------|-----|------|
|
49 |
+
|all-minilm-l6-v2|33m |128|
|
50 |
+
|all-mpnet-base-v2 |110m |128|
|
51 |
+
|ada-embedding-002|Unknown/API based |8192|
|
52 |
+
|jina-embedding-s-en-v1|35m |512|
|
53 |
+
|jina-embedding-b-en-v1|110m |512|
|
54 |
+
|jina-embedding-l-en-v1|330m |512|
|
55 |
+
|
56 |
+
|
57 |
+
|Name|STS12|STS13|STS14|STS15|STS16|STS17|TRECOVID|Quora|SciFact|
|
58 |
+
|------------------------------|-----|-----|-----|-----|-----|-----|--------|-----|-----|
|
59 |
+
|all-minilm-l6-v2|0.724|0.806|0.756|0.854|0.79 |0.876|0.473 |0.876|0.645 |
|
60 |
+
|all-mpnet--base-v2|0.726|0.835|0.78 |0.857|0.8 |0.906|0.513 |0.875|0.656 |
|
61 |
+
|ada-embedding-002|0.698|0.833|0.761|0.861|0.86 |0.903|0.685 |0.876|0.726 |
|
62 |
+
|jina-embedding-s-en-v1|0.738|0.781|0.732|0.833|0.785|0.859|0.471 |0.852|0.567 |
|
63 |
+
|jina-embedding-b-en-v1|0.736|0.804|0.745|0.844|0.793|0.873|0.481 |0.87|0.616 |
|
64 |
+
|jina-embedding-l-en-v1|0.735|0.829|0.759|0.844|0.8|0.888|0.465 |0.876|0.645 |
|
65 |
+
|
66 |
+
For more tasks and metrics, please checkout [MTEB](https://huggingface.co/spaces/mteb/leaderboard) benchmark.
|
67 |
+
|
68 |
+
## Usage [WIP]
|
69 |
+
|
70 |
+
```python
|
71 |
+
!pip install finetuner[text]
|
72 |
+
import finetuner
|
73 |
+
model = finetuner.get_model('jinaai/jina-embedding-l-en-v1')
|
74 |
+
embeddings = model.encode(['sentence 1', 'sentence 2'])
|
75 |
+
```
|
76 |
+
|
77 |
+
## Fine-tuning [WIP]
|
78 |
+
|
79 |
+
Please consider [Finetuner](https://github.com/jina-ai/finetuner).
|