Update README.md
Browse files
README.md
CHANGED
@@ -11,10 +11,14 @@ tags:
|
|
11 |
pipeline_tag: sentence-similarity
|
12 |
---
|
13 |
|
14 |
-
|
|
|
15 |
|
16 |
-
|
17 |
|
|
|
|
|
|
|
18 |
|
19 |
## Installation
|
20 |
|
@@ -24,13 +28,12 @@ pip install model2vec
|
|
24 |
```
|
25 |
|
26 |
## Usage
|
|
|
27 |
Load this model using the `from_pretrained` method:
|
28 |
```python
|
29 |
from model2vec import StaticModel
|
30 |
-
|
31 |
# Load a pretrained Model2Vec model
|
32 |
model = StaticModel.from_pretrained("takara-ai/e5-small-v2-static-distilled")
|
33 |
-
|
34 |
# Compute text embeddings
|
35 |
embeddings = model.encode(["Example sentence"])
|
36 |
```
|
@@ -38,13 +41,10 @@ embeddings = model.encode(["Example sentence"])
|
|
38 |
Alternatively, you can distill your own model using the `distill` method:
|
39 |
```python
|
40 |
from model2vec.distill import distill
|
41 |
-
|
42 |
# Choose a Sentence Transformer model
|
43 |
model_name = "BAAI/bge-base-en-v1.5"
|
44 |
-
|
45 |
# Distill the model
|
46 |
m2v_model = distill(model_name=model_name, pca_dims=256)
|
47 |
-
|
48 |
# Save the model
|
49 |
m2v_model.save_pretrained("m2v_model")
|
50 |
```
|
@@ -62,7 +62,7 @@ It works by passing a vocabulary through a sentence transformer model, then redu
|
|
62 |
- [Model2Vec Results](https://github.com/MinishLab/model2vec?tab=readme-ov-file#results)
|
63 |
- [Model2Vec Tutorials](https://github.com/MinishLab/model2vec/tree/main/tutorials)
|
64 |
|
65 |
-
##
|
66 |
|
67 |
Model2Vec was developed by the [Minish Lab](https://github.com/MinishLab) team consisting of [Stephan Tulkens](https://github.com/stephantul) and [Thomas van Dongen](https://github.com/Pringled).
|
68 |
|
@@ -76,4 +76,9 @@ Please cite the [Model2Vec repository](https://github.com/MinishLab/model2vec) i
|
|
76 |
year = {2024},
|
77 |
url = {https://github.com/MinishLab/model2vec},
|
78 |
}
|
79 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
11 |
pipeline_tag: sentence-similarity
|
12 |
---
|
13 |
|
14 |
+
<img src="https://takara.ai/images/logo-24/TakaraAi.svg" width="200" alt="Takara.ai Logo" />
|
15 |
+
From the Frontier Research Team at **Takara.ai** we present a distilled version of the e5-small-v2 model using Model2Vec technology for ultra-fast text embeddings.
|
16 |
|
17 |
+
---
|
18 |
|
19 |
+
# e5-small-v2-distilled
|
20 |
+
|
21 |
+
This [Model2Vec](https://github.com/MinishLab/model2vec) model is a distilled version of the [intfloat/e5-small-v2](https://huggingface.co/intfloat/e5-small-v2) Sentence Transformer. It uses static embeddings, allowing text embeddings to be computed orders of magnitude faster on both GPU and CPU. It is designed for applications where computational resources are limited or where real-time performance is critical.
|
22 |
|
23 |
## Installation
|
24 |
|
|
|
28 |
```
|
29 |
|
30 |
## Usage
|
31 |
+
|
32 |
Load this model using the `from_pretrained` method:
|
33 |
```python
|
34 |
from model2vec import StaticModel
|
|
|
35 |
# Load a pretrained Model2Vec model
|
36 |
model = StaticModel.from_pretrained("takara-ai/e5-small-v2-static-distilled")
|
|
|
37 |
# Compute text embeddings
|
38 |
embeddings = model.encode(["Example sentence"])
|
39 |
```
|
|
|
41 |
Alternatively, you can distill your own model using the `distill` method:
|
42 |
```python
|
43 |
from model2vec.distill import distill
|
|
|
44 |
# Choose a Sentence Transformer model
|
45 |
model_name = "BAAI/bge-base-en-v1.5"
|
|
|
46 |
# Distill the model
|
47 |
m2v_model = distill(model_name=model_name, pca_dims=256)
|
|
|
48 |
# Save the model
|
49 |
m2v_model.save_pretrained("m2v_model")
|
50 |
```
|
|
|
62 |
- [Model2Vec Results](https://github.com/MinishLab/model2vec?tab=readme-ov-file#results)
|
63 |
- [Model2Vec Tutorials](https://github.com/MinishLab/model2vec/tree/main/tutorials)
|
64 |
|
65 |
+
## Original Credits
|
66 |
|
67 |
Model2Vec was developed by the [Minish Lab](https://github.com/MinishLab) team consisting of [Stephan Tulkens](https://github.com/stephantul) and [Thomas van Dongen](https://github.com/Pringled).
|
68 |
|
|
|
76 |
year = {2024},
|
77 |
url = {https://github.com/MinishLab/model2vec},
|
78 |
}
|
79 |
+
```
|
80 |
+
|
81 |
+
---
|
82 |
+
For research inquiries and press, please reach out to [email protected]
|
83 |
+
|
84 |
+
> 人類を変革する
|