Improve README
Browse files- README.md +177 -19
- Sentiment.svg +578 -0
README.md
CHANGED
@@ -1,34 +1,192 @@
|
|
1 |
---
|
2 |
-
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
tags:
|
5 |
-
-
|
6 |
-
|
7 |
-
- name: multilingual-e5-small-aligned-sentiment-20241214-new
|
8 |
-
results: []
|
9 |
---
|
10 |
|
11 |
-
|
12 |
-
should probably proofread and complete it, then remove this comment. -->
|
13 |
|
14 |
-
|
15 |
|
16 |
-
|
17 |
-
It achieves the following results on the evaluation set:
|
18 |
-
- Loss: 0.1455
|
19 |
-
- Mse: 0.1455
|
20 |
|
21 |
-
|
|
|
|
|
22 |
|
23 |
-
|
24 |
|
25 |
-
|
|
|
|
|
|
|
26 |
|
27 |
-
|
|
|
|
|
28 |
|
29 |
-
##
|
30 |
|
31 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
## Training procedure
|
34 |
|
|
|
1 |
---
|
2 |
+
license: mit
|
3 |
+
language:
|
4 |
+
- multilingual
|
5 |
+
- af
|
6 |
+
- am
|
7 |
+
- ar
|
8 |
+
- as
|
9 |
+
- az
|
10 |
+
- be
|
11 |
+
- bg
|
12 |
+
- bn
|
13 |
+
- br
|
14 |
+
- bs
|
15 |
+
- ca
|
16 |
+
- cs
|
17 |
+
- cy
|
18 |
+
- da
|
19 |
+
- de
|
20 |
+
- el
|
21 |
+
- en
|
22 |
+
- eo
|
23 |
+
- es
|
24 |
+
- et
|
25 |
+
- eu
|
26 |
+
- fa
|
27 |
+
- fi
|
28 |
+
- fr
|
29 |
+
- fy
|
30 |
+
- ga
|
31 |
+
- gd
|
32 |
+
- gl
|
33 |
+
- gu
|
34 |
+
- ha
|
35 |
+
- he
|
36 |
+
- hi
|
37 |
+
- hr
|
38 |
+
- hu
|
39 |
+
- hy
|
40 |
+
- id
|
41 |
+
- is
|
42 |
+
- it
|
43 |
+
- ja
|
44 |
+
- jv
|
45 |
+
- ka
|
46 |
+
- kk
|
47 |
+
- km
|
48 |
+
- kn
|
49 |
+
- ko
|
50 |
+
- ku
|
51 |
+
- ky
|
52 |
+
- la
|
53 |
+
- lo
|
54 |
+
- lt
|
55 |
+
- lv
|
56 |
+
- mg
|
57 |
+
- mk
|
58 |
+
- ml
|
59 |
+
- mn
|
60 |
+
- mr
|
61 |
+
- ms
|
62 |
+
- my
|
63 |
+
- ne
|
64 |
+
- nl
|
65 |
+
- 'no'
|
66 |
+
- om
|
67 |
+
- or
|
68 |
+
- pa
|
69 |
+
- pl
|
70 |
+
- ps
|
71 |
+
- pt
|
72 |
+
- ro
|
73 |
+
- ru
|
74 |
+
- sa
|
75 |
+
- sd
|
76 |
+
- si
|
77 |
+
- sk
|
78 |
+
- sl
|
79 |
+
- so
|
80 |
+
- sq
|
81 |
+
- sr
|
82 |
+
- su
|
83 |
+
- sv
|
84 |
+
- sw
|
85 |
+
- ta
|
86 |
+
- te
|
87 |
+
- th
|
88 |
+
- tl
|
89 |
+
- tr
|
90 |
+
- ug
|
91 |
+
- uk
|
92 |
+
- ur
|
93 |
+
- uz
|
94 |
+
- vi
|
95 |
+
- xh
|
96 |
+
- yi
|
97 |
+
- zh
|
98 |
+
datasets:
|
99 |
+
- agentlans/en-translations
|
100 |
+
base_model:
|
101 |
+
- agentlans/multilingual-e5-small-aligned
|
102 |
+
pipeline_tag: text-classification
|
103 |
tags:
|
104 |
+
- multilingual
|
105 |
+
- sentiment-assessment
|
|
|
|
|
106 |
---
|
107 |
|
108 |
+
# multilingual-e5-small-aligned-sentiment
|
|
|
109 |
|
110 |
+
This model is a fine-tuned version of [agentlans/multilingual-e5-small-aligned](https://huggingface.co/agentlans/multilingual-e5-small-aligned) designed for assessing text sentiment across multiple languages.
|
111 |
|
112 |
+
## Key Features
|
|
|
|
|
|
|
113 |
|
114 |
+
- Multilingual support
|
115 |
+
- Sentiment assessment for text
|
116 |
+
- Based on E5 small model architecture
|
117 |
|
118 |
+
## Intended Uses & Limitations
|
119 |
|
120 |
+
This model is intended for:
|
121 |
+
- Assessing the sentiment of multilingual text
|
122 |
+
- Filtering multilingual content
|
123 |
+
- Comparative analysis of corpus text sentiment across different languages
|
124 |
|
125 |
+
Limitations:
|
126 |
+
- Performance may vary for languages not well-represented in the training data
|
127 |
+
- Should not be used as the sole criterion for sentiment assessment
|
128 |
|
129 |
+
## Usage Example
|
130 |
|
131 |
+
```python
|
132 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
133 |
+
import torch
|
134 |
+
|
135 |
+
model_name = "agentlans/multilingual-e5-small-aligned-sentiment"
|
136 |
+
|
137 |
+
# Initialize tokenizer and model
|
138 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
139 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
140 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
141 |
+
model = model.to(device)
|
142 |
+
|
143 |
+
def sentiment(text):
|
144 |
+
"""Assess the sentiment of the input text."""
|
145 |
+
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
|
146 |
+
with torch.no_grad():
|
147 |
+
logits = model(**inputs).logits.squeeze().cpu()
|
148 |
+
return logits.tolist()
|
149 |
+
|
150 |
+
# Example usage
|
151 |
+
score = sentiment("Your text here.")
|
152 |
+
print(f"Sentiment score: {score}")
|
153 |
+
```
|
154 |
+
|
155 |
+
## Performance Results
|
156 |
+
|
157 |
+
The model was evaluated on a diverse set of multilingual text samples:
|
158 |
+
|
159 |
+
- 10 English text samples of varying sentiment were translated into Arabic, Chinese, French, Russian, and Spanish.
|
160 |
+
- The model demonstrated consistent sentiment assessment across different languages for the same text.
|
161 |
+
|
162 |
+
<details>
|
163 |
+
<summary>Click here for the 10 original texts and their translations.</summary>
|
164 |
+
|
165 |
+
| **Text** | **English** | **French** | **Spanish** | **Chinese** | **Russian** | **Arabic** |
|
166 |
+
|---|---|---|---|---|---|---|
|
167 |
+
| A | Nothing seems to go right, and I'm constantly frustrated. | Rien ne semble aller bien et je suis constamment frustré. | Nada parece ir bien y me siento constantemente frustrado. | 一切似乎都不顺利,我总是感到很沮丧。 | Кажется, все идет не так, как надо, и я постоянно расстроен. | يبدو أن لا شيء يسير على ما يرام، وأنا أشعر بالإحباط باستمرار. |
|
168 |
+
| B | Everything is falling apart, and I can't see any way out. | Tout s’effondre et je ne vois aucune issue. | Todo se está desmoronando y no veo ninguna salida. | 一切都崩溃了,我看不到任何出路。 | Все рушится, и я не вижу выхода. | كل شيء ينهار، ولا أستطيع أن أرى أي مخرج. |
|
169 |
+
| C | I feel completely overwhelmed by the challenges I face. | Je me sens complètement dépassé par les défis auxquels je suis confronté. | Me siento completamente abrumado por los desafíos que enfrento. | 我感觉自己完全被所面临的挑战压垮了。 | Я чувствую себя совершенно подавленным из-за проблем, с которыми мне приходится сталкиваться. | أشعر بأنني غارق تمامًا في ��لتحديات التي أواجهها. |
|
170 |
+
| D | There are some minor improvements, but overall, things are still tough. | Il y a quelques améliorations mineures, mais dans l’ensemble, les choses restent difficiles. | Hay algunas mejoras menores, pero en general las cosas siguen siendo difíciles. | 虽然有一些小的改进,但是总的来说,事情仍然很艰难。 | Есть некоторые незначительные улучшения, но в целом ситуация по-прежнему сложная. | هناك بعض التحسينات الطفيفة، ولكن بشكل عام، لا تزال الأمور صعبة. |
|
171 |
+
| E | I can see a glimmer of hope amidst the difficulties I encounter. | Je vois une lueur d’espoir au milieu des difficultés que je rencontre. | Puedo ver un rayo de esperanza en medio de las dificultades que encuentro. | 我在遇到的困难中看到了一线希望。 | Среди трудностей, с которыми я сталкиваюсь, я вижу проблеск надежды. | أستطيع أن أرى بصيص أمل وسط الصعوبات التي أواجهها. |
|
172 |
+
| F | Things are starting to look up, and I'm cautiously optimistic. | Les choses commencent à s’améliorer et je suis prudemment optimiste. | Las cosas están empezando a mejorar y me siento cautelosamente optimista. | 事情开始好转,我持谨慎乐观的态度。 | Ситуация начинает улучшаться, и я настроен осторожно и оптимистично. | بدأت الأمور تتجه نحو التحسن، وأنا متفائل بحذر. |
|
173 |
+
| G | I'm feeling more positive about my situation than I have in a while. | Je me sens plus positif à propos de ma situation que je ne l’ai été depuis un certain temps. | Me siento más positivo sobre mi situación que en mucho tiempo. | 我对自己处境的感觉比以前更加乐观了。 | Я чувствую себя более позитивно относительно своей ситуации, чем когда-либо за последнее время. | أشعر بإيجابية أكبر تجاه وضعي مقارنة بأي وقت مضى. |
|
174 |
+
| H | There are many good things happening, and I appreciate them. | Il se passe beaucoup de bonnes choses et je les apprécie. | Están sucediendo muchas cosas buenas y las aprecio. | 有很多好事发生,我对此表示感谢。 | Происходит много хорошего, и я это ценю. | هناك الكثير من الأشياء الجيدة التي تحدث، وأنا أقدرها. |
|
175 |
+
| I | Every day brings new joy and possibilities; I feel truly blessed. | Chaque jour apporte de nouvelles joies et possibilités ; je me sens vraiment béni. | Cada día trae nueva alegría y posibilidades; me siento verdaderamente bendecida. | 每天都有新的快乐和可能性;我感到非常幸福。 | Каждый день приносит новую радость и возможности; я чувствую себя по-настоящему благословенной. | كل يوم يجلب فرحة وإمكانيات جديدة؛ أشعر بأنني محظوظة حقًا. |
|
176 |
+
| J | Life is full of opportunities, and I'm excited about the future. | La vie est pleine d’opportunités et je suis enthousiaste quant à l’avenir. | La vida está llena de oportunidades y estoy entusiasmado por el futuro. | 生活充满机遇,我对未来充满兴奋。 | Жизнь полна возможностей, и я с нетерпением жду будущего. | الحياة مليئة بالفرص، وأنا متحمس للمستقبل. |
|
177 |
+
|
178 |
+
</details>
|
179 |
+
|
180 |
+
<img src="Sentiment.svg" alt="Scatterplot of predicted sentiment scores grouped by text sample and language" width="100%"/>
|
181 |
+
|
182 |
+
## Training Data
|
183 |
+
|
184 |
+
The model was trained on the [Multilingual Parallel Sentences dataset](https://huggingface.co/datasets/agentlans/en-translations), which includes:
|
185 |
+
|
186 |
+
- Parallel sentences in English and various other languages
|
187 |
+
- Semantic similarity scores calculated using LaBSE
|
188 |
+
- Additional sentiment metrics
|
189 |
+
- Sources: JW300, Europarl, TED Talks, OPUS-100, Tatoeba, Global Voices, and News Commentary
|
190 |
|
191 |
## Training procedure
|
192 |
|
Sentiment.svg
ADDED