Add new SentenceTransformer model.
Browse files- 1_Pooling/config.json +10 -0
- README.md +485 -0
- config.json +24 -0
- config_sentence_transformers.json +10 -0
- model.safetensors +3 -0
- modules.json +14 -0
- sentence_bert_config.json +4 -0
- special_tokens_map.json +51 -0
- tokenizer.json +0 -0
- tokenizer_config.json +72 -0
- vocab.txt +0 -0
1_Pooling/config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"word_embedding_dimension": 768,
|
3 |
+
"pooling_mode_cls_token": false,
|
4 |
+
"pooling_mode_mean_tokens": true,
|
5 |
+
"pooling_mode_max_tokens": false,
|
6 |
+
"pooling_mode_mean_sqrt_len_tokens": false,
|
7 |
+
"pooling_mode_weightedmean_tokens": false,
|
8 |
+
"pooling_mode_lasttoken": false,
|
9 |
+
"include_prompt": true
|
10 |
+
}
|
README.md
ADDED
@@ -0,0 +1,485 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- es
|
4 |
+
license: apache-2.0
|
5 |
+
library_name: sentence-transformers
|
6 |
+
tags:
|
7 |
+
- sentence-transformers
|
8 |
+
- sentence-similarity
|
9 |
+
- feature-extraction
|
10 |
+
- generated_from_trainer
|
11 |
+
- dataset_size:74124
|
12 |
+
- loss:MultipleNegativesRankingLoss
|
13 |
+
base_model: microsoft/mpnet-base
|
14 |
+
datasets: []
|
15 |
+
metrics:
|
16 |
+
- cosine_accuracy
|
17 |
+
- dot_accuracy
|
18 |
+
- manhattan_accuracy
|
19 |
+
- euclidean_accuracy
|
20 |
+
- max_accuracy
|
21 |
+
widget:
|
22 |
+
- source_sentence: ¿Cuál es la tensión nominal en las redes monofásicas trifilares
|
23 |
+
con el punto medio conectado a tierra?
|
24 |
+
sentences:
|
25 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 14.3.1 trata sobre: Circuitos trifásicos
|
26 |
+
son aquellos que emplean las tres fases de la energía que provee la ANDE, con
|
27 |
+
interruptor y protección adecuados en el tablero de arranque y se emplean en líneas
|
28 |
+
distribuidoras de fuerza motriz, calefacción, refrigeración y similares, comprendiendo
|
29 |
+
incluso aparatos monofásicos, sin limitaciones de carga, siempre que: a) Se realice
|
30 |
+
el equilibrio de cargas de los equipos monofásicos. b) Se atienda correctamente
|
31 |
+
a 13.4.2. c) Que ninguna de las cargas trifásicas individuales sea igual o superior
|
32 |
+
a 15 A nominales en el caso de motores, o 20 A si son equipos de calefacción o
|
33 |
+
similares. d) Se use un circuito trifásico independiente por cada motor de 15
|
34 |
+
A nominales o más. e) Se use un circuito trifásico independiente por cada equipo
|
35 |
+
de calefacción o similar de 20 A o más.. El 14.3.1 pertenece a la sección: <section>14.3</section>'
|
36 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 24.5.1 trata sobre: No podrán ser usadas
|
37 |
+
curvas de abertura inferior a 90°. En una tubería comprendida entre dos cajas,
|
38 |
+
o entre extremidades libres, o entre una caja y una extremidad libre, no se podrán
|
39 |
+
usar más de 3 curvas.. El 24.5.1 pertenece a la sección: <section>24.5</section>'
|
40 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 1.1 trata sobre: Este reglamento se
|
41 |
+
aplica a las instalaciones eléctricas de baja tensión conectadas y a ser conectadas
|
42 |
+
a las redes de la Administración Nacional de Electricidad (ANDE). Observación:
|
43 |
+
ANDE distribuye energía eléctrica en baja tensión a la frecuencia nominal de 50
|
44 |
+
Hz según los siguientes sistemas: a) Mediante redes trifásicas trifilares sin
|
45 |
+
neutro, a la tensión nominal de 220 Voltios entre fases; b) Mediante redes monofásicas
|
46 |
+
trifilares con el punto medio conectado a tierra, a la tensión nominal de 220
|
47 |
+
Voltios; c) Mediante redes trifásicas tetrafilares con neutro conectado a tierra,
|
48 |
+
a la tensión nominal de 380 Voltios entre fases y de 220 Voltios entre fase y
|
49 |
+
neutro. NOTA 1: Entiéndase por “frecuencia nominal” la citada, admitiéndose una
|
50 |
+
variación de hasta 2% en más o menos. NOTA 2: Entiéndase por “tensión nominal”
|
51 |
+
las citadas, admitiéndose una variación de hasta 2% en más o en menos. NOTA 3:
|
52 |
+
Las tolerancias indicadas en las notas precedentes se refieren a servicio normal,
|
53 |
+
pudiendo ser excedidas en situaciones anormales no permanentes. NOTA 4: La Baja
|
54 |
+
Tensión de las instalaciones servidas en Media o Alta Tensión, con transformador
|
55 |
+
de uso exclusivo, podrá ser optativa.. El 1.1 pertenece a la sección: <section>1-</section>'
|
56 |
+
- source_sentence: ¿Cuál es la altura mínima requerida en relación a patios, jardines
|
57 |
+
y paseos de exclusivo uso de peatones?
|
58 |
+
sentences:
|
59 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 36.3.2 trata sobre: Deberán ser usados
|
60 |
+
transformadores monofásicos de potencia no superior a 1 kVA, con tensión en el
|
61 |
+
secundario, en vacío, no mayor de 20 kV y corriente de corto circuito no superior
|
62 |
+
a 80 mA.. El 36.3.2 pertenece a la sección: <section>36.3</section>'
|
63 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 36.6.1 trata sobre: Las partes conductoras
|
64 |
+
de los circuitos de este tipo que operen con tensiones superiores a 1000 voltios,
|
65 |
+
deberán mantener las siguientes separaciones mínimas: a) 1,5 m en relación a
|
66 |
+
ventanas, terrazas, balcones y lugares semejantes. b) 1,5 m en relación a líneas
|
67 |
+
aéreas de luz, fuerza motriz, teléfonos y similares. c) 2,5 m de altura en relación
|
68 |
+
al piso en instalaciones interiores, no protegidas, así como también en el caso
|
69 |
+
de balcones, terrazas y lugares semejantes. d) 3,5 m de altura en relación a patios,
|
70 |
+
jardines y paseos de exclusivo uso de peatones. e) 5,5 m de altura en relación
|
71 |
+
a calles, patios y áreas de circulación de vehículos.. El 36.6.1 pertenece a la
|
72 |
+
sección: <section>36.6</section>'
|
73 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 32.5 trata sobre Vanos: y tiene las
|
74 |
+
siguientes sub-secciones: <sub-section>32.5.1</sub-section>, <sub-section>32.5.2</sub-section>'
|
75 |
+
- source_sentence: Determine quién es responsable de la conservación y mantenimiento
|
76 |
+
del material de la acometida una vez conectada la energía eléctrica.
|
77 |
+
sentences:
|
78 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 11.2.1 trata sobre: Entrada es la conexión
|
79 |
+
eléctrica entre el servicio y el equipo de medición.. El 11.2.1 pertenece a la
|
80 |
+
sección: <section>11.2</section>'
|
81 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 31.1.2 trata sobre: Las barras desnudas
|
82 |
+
podrán ser instaladas en canaletas, incluso en recintos de tránsito, siempre que
|
83 |
+
las mismas aseguren adecuada protección eléctrica y mecánica.. El 31.1.2 pertenece
|
84 |
+
a la sección: <section>31.1</section>'
|
85 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 11.4.2 trata sobre: La entrada debe
|
86 |
+
ser instalada por el electricista responsable. El material para el servicio será
|
87 |
+
adquirido por el usuario e instalado por ANDE. Una vez conectada la energía eléctrica,
|
88 |
+
todo el material de la acometida queda a cargo de la ANDE que, en consecuencia,
|
89 |
+
es la responsable de su conservación y mantenimiento.. El 11.4.2 pertenece a la
|
90 |
+
sección: <section>11.4</section>'
|
91 |
+
- source_sentence: Identifique el número de la sección que trata sobre la Medida de
|
92 |
+
la resistencia de aislación.
|
93 |
+
sentences:
|
94 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 21.2 trata sobre Medida de la resistencia
|
95 |
+
de aislación: y tiene las siguientes sub-secciones: <sub-section>21.2.1</sub-section>,
|
96 |
+
<sub-section> 21.2.2</sub-section>, <sub-section>21.2.3</sub-section>'
|
97 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 17.7 trata sobre Protección de los
|
98 |
+
conductores: y tiene las siguientes sub-secciones: <sub-section>17.7.1</sub-section>,
|
99 |
+
<sub-section>17.7.2</sub-section>, <sub-section>17.7.3</sub-section>, <sub-section>17.7.4</sub-section>'
|
100 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 21.1 trata sobre Aplicación: y tiene
|
101 |
+
las siguientes sub-secciones: <sub-section>21.1.1</sub-section>'
|
102 |
+
- source_sentence: ¿Cuál es el número del artículo que trata sobre la mínima sección
|
103 |
+
permisible para una lámpara o grupo de lámparas?
|
104 |
+
sentences:
|
105 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 12.1.1 trata sobre: El material con
|
106 |
+
que se confeccionen los soportes, dependerá del material que se emplee en los
|
107 |
+
elementos de maniobra, protección y control. Así, podrán usarse bastidores de
|
108 |
+
metal, chapas metálicas, placas de mármol o de otra substancia aislante y maderas
|
109 |
+
debidamente tratadas en dependencia de las características de los seccionadores,
|
110 |
+
portafusibles, interruptores, llaves y demás elementos componentes del tablero..
|
111 |
+
El 12.1.1 pertenece a la sección: <section>12.1</section>'
|
112 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 14.7.3 trata sobre: La mínima sección
|
113 |
+
permisible para una lámpara, o grupo de lámparas que forman un solo artefacto
|
114 |
+
de iluminación, será de 1 mm².. El 14.7.3 pertenece a la sección: <section>14.7</section>'
|
115 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 19.2.1 trata sobre: La caída de tensión
|
116 |
+
máxima permisible, es la siguiente: a) Para iluminación, en general (19.1.1),
|
117 |
+
hasta 4%. -2% en el alimentador, y -2% en el circuito (19.1.2). b) Para fuerza
|
118 |
+
motriz y/o calefacción, hasta 5%. -4% en el alimentador, y -1% en el ramal. c)
|
119 |
+
En el caso de clientes que reciban la energía a tensión diferente de las normales
|
120 |
+
de utilización (19.1.3), hasta 4%.. El 19.2.1 pertenece a la sección: <section>19.2</section>'
|
121 |
+
pipeline_tag: sentence-similarity
|
122 |
+
model-index:
|
123 |
+
- name: andegpt-embed
|
124 |
+
results:
|
125 |
+
- task:
|
126 |
+
type: triplet
|
127 |
+
name: Triplet
|
128 |
+
dataset:
|
129 |
+
name: andegpt dev
|
130 |
+
type: andegpt-dev
|
131 |
+
metrics:
|
132 |
+
- type: cosine_accuracy
|
133 |
+
value: 0.9970859640602234
|
134 |
+
name: Cosine Accuracy
|
135 |
+
- type: dot_accuracy
|
136 |
+
value: 0.0031568722680913063
|
137 |
+
name: Dot Accuracy
|
138 |
+
- type: manhattan_accuracy
|
139 |
+
value: 0.9968431277319086
|
140 |
+
name: Manhattan Accuracy
|
141 |
+
- type: euclidean_accuracy
|
142 |
+
value: 0.9970859640602234
|
143 |
+
name: Euclidean Accuracy
|
144 |
+
- type: max_accuracy
|
145 |
+
value: 0.9970859640602234
|
146 |
+
name: Max Accuracy
|
147 |
+
---
|
148 |
+
|
149 |
+
# andegpt-embed
|
150 |
+
|
151 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
|
152 |
+
|
153 |
+
## Model Details
|
154 |
+
|
155 |
+
### Model Description
|
156 |
+
- **Model Type:** Sentence Transformer
|
157 |
+
- **Base model:** [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base) <!-- at revision 6996ce1e91bd2a9c7d7f61daec37463394f73f09 -->
|
158 |
+
- **Maximum Sequence Length:** 512 tokens
|
159 |
+
- **Output Dimensionality:** 768 tokens
|
160 |
+
- **Similarity Function:** Cosine Similarity
|
161 |
+
<!-- - **Training Dataset:** Unknown -->
|
162 |
+
- **Language:** es
|
163 |
+
- **License:** apache-2.0
|
164 |
+
|
165 |
+
### Model Sources
|
166 |
+
|
167 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
168 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
169 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
170 |
+
|
171 |
+
### Full Model Architecture
|
172 |
+
|
173 |
+
```
|
174 |
+
SentenceTransformer(
|
175 |
+
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: MPNetModel
|
176 |
+
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
|
177 |
+
)
|
178 |
+
```
|
179 |
+
|
180 |
+
## Usage
|
181 |
+
|
182 |
+
### Direct Usage (Sentence Transformers)
|
183 |
+
|
184 |
+
First install the Sentence Transformers library:
|
185 |
+
|
186 |
+
```bash
|
187 |
+
pip install -U sentence-transformers
|
188 |
+
```
|
189 |
+
|
190 |
+
Then you can load this model and run inference.
|
191 |
+
```python
|
192 |
+
from sentence_transformers import SentenceTransformer
|
193 |
+
|
194 |
+
# Download from the 🤗 Hub
|
195 |
+
model = SentenceTransformer("enpaiva/embed-andegpt-280724")
|
196 |
+
# Run inference
|
197 |
+
sentences = [
|
198 |
+
'¿Cuál es el número del artículo que trata sobre la mínima sección permisible para una lámpara o grupo de lámparas?',
|
199 |
+
'Reglamento de Baja Tensión de la ANDE: El 14.7.3 trata sobre: La mínima sección permisible para una lámpara, o grupo de lámparas que forman un solo artefacto de iluminación, será de 1 mm².. El 14.7.3 pertenece a la sección: <section>14.7</section>',
|
200 |
+
'Reglamento de Baja Tensión de la ANDE: El 19.2.1 trata sobre: La caída de tensión máxima permisible, es la siguiente: a) Para iluminación, en general (19.1.1), hasta 4%. -2% en el alimentador, y -2% en el circuito (19.1.2). b) Para fuerza motriz y/o calefacción, hasta 5%. -4% en el alimentador, y -1% en el ramal. c) En el caso de clientes que reciban la energía a tensión diferente de las normales de utilización (19.1.3), hasta 4%.. El 19.2.1 pertenece a la sección: <section>19.2</section>',
|
201 |
+
]
|
202 |
+
embeddings = model.encode(sentences)
|
203 |
+
print(embeddings.shape)
|
204 |
+
# [3, 768]
|
205 |
+
|
206 |
+
# Get the similarity scores for the embeddings
|
207 |
+
similarities = model.similarity(embeddings, embeddings)
|
208 |
+
print(similarities.shape)
|
209 |
+
# [3, 3]
|
210 |
+
```
|
211 |
+
|
212 |
+
<!--
|
213 |
+
### Direct Usage (Transformers)
|
214 |
+
|
215 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
216 |
+
|
217 |
+
</details>
|
218 |
+
-->
|
219 |
+
|
220 |
+
<!--
|
221 |
+
### Downstream Usage (Sentence Transformers)
|
222 |
+
|
223 |
+
You can finetune this model on your own dataset.
|
224 |
+
|
225 |
+
<details><summary>Click to expand</summary>
|
226 |
+
|
227 |
+
</details>
|
228 |
+
-->
|
229 |
+
|
230 |
+
<!--
|
231 |
+
### Out-of-Scope Use
|
232 |
+
|
233 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
234 |
+
-->
|
235 |
+
|
236 |
+
## Evaluation
|
237 |
+
|
238 |
+
### Metrics
|
239 |
+
|
240 |
+
#### Triplet
|
241 |
+
* Dataset: `andegpt-dev`
|
242 |
+
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
|
243 |
+
|
244 |
+
| Metric | Value |
|
245 |
+
|:-------------------|:-----------|
|
246 |
+
| cosine_accuracy | 0.9971 |
|
247 |
+
| dot_accuracy | 0.0032 |
|
248 |
+
| manhattan_accuracy | 0.9968 |
|
249 |
+
| euclidean_accuracy | 0.9971 |
|
250 |
+
| **max_accuracy** | **0.9971** |
|
251 |
+
|
252 |
+
<!--
|
253 |
+
## Bias, Risks and Limitations
|
254 |
+
|
255 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
256 |
+
-->
|
257 |
+
|
258 |
+
<!--
|
259 |
+
### Recommendations
|
260 |
+
|
261 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
262 |
+
-->
|
263 |
+
|
264 |
+
## Training Details
|
265 |
+
|
266 |
+
### Training Hyperparameters
|
267 |
+
#### Non-Default Hyperparameters
|
268 |
+
|
269 |
+
- `prediction_loss_only`: False
|
270 |
+
- `learning_rate`: 2e-05
|
271 |
+
- `lr_scheduler_type`: cosine
|
272 |
+
- `log_level_replica`: passive
|
273 |
+
- `log_on_each_node`: False
|
274 |
+
- `logging_nan_inf_filter`: False
|
275 |
+
- `bf16`: True
|
276 |
+
- `batch_sampler`: no_duplicates
|
277 |
+
|
278 |
+
#### All Hyperparameters
|
279 |
+
<details><summary>Click to expand</summary>
|
280 |
+
|
281 |
+
- `overwrite_output_dir`: False
|
282 |
+
- `do_predict`: False
|
283 |
+
- `prediction_loss_only`: False
|
284 |
+
- `per_device_train_batch_size`: 8
|
285 |
+
- `per_device_eval_batch_size`: 8
|
286 |
+
- `per_gpu_train_batch_size`: None
|
287 |
+
- `per_gpu_eval_batch_size`: None
|
288 |
+
- `gradient_accumulation_steps`: 1
|
289 |
+
- `eval_accumulation_steps`: None
|
290 |
+
- `learning_rate`: 2e-05
|
291 |
+
- `weight_decay`: 0.0
|
292 |
+
- `adam_beta1`: 0.9
|
293 |
+
- `adam_beta2`: 0.999
|
294 |
+
- `adam_epsilon`: 1e-08
|
295 |
+
- `max_grad_norm`: 1.0
|
296 |
+
- `num_train_epochs`: 3
|
297 |
+
- `max_steps`: -1
|
298 |
+
- `lr_scheduler_type`: cosine
|
299 |
+
- `lr_scheduler_kwargs`: {}
|
300 |
+
- `warmup_ratio`: 0
|
301 |
+
- `warmup_steps`: 0
|
302 |
+
- `log_level`: passive
|
303 |
+
- `log_level_replica`: passive
|
304 |
+
- `log_on_each_node`: False
|
305 |
+
- `logging_nan_inf_filter`: False
|
306 |
+
- `save_safetensors`: True
|
307 |
+
- `save_on_each_node`: False
|
308 |
+
- `save_only_model`: False
|
309 |
+
- `no_cuda`: False
|
310 |
+
- `use_cpu`: False
|
311 |
+
- `use_mps_device`: False
|
312 |
+
- `seed`: 42
|
313 |
+
- `data_seed`: None
|
314 |
+
- `jit_mode_eval`: False
|
315 |
+
- `use_ipex`: False
|
316 |
+
- `bf16`: True
|
317 |
+
- `fp16`: False
|
318 |
+
- `fp16_opt_level`: O1
|
319 |
+
- `half_precision_backend`: auto
|
320 |
+
- `bf16_full_eval`: False
|
321 |
+
- `fp16_full_eval`: False
|
322 |
+
- `tf32`: None
|
323 |
+
- `local_rank`: 0
|
324 |
+
- `ddp_backend`: None
|
325 |
+
- `tpu_num_cores`: None
|
326 |
+
- `tpu_metrics_debug`: False
|
327 |
+
- `debug`: []
|
328 |
+
- `dataloader_drop_last`: False
|
329 |
+
- `dataloader_num_workers`: 0
|
330 |
+
- `dataloader_prefetch_factor`: None
|
331 |
+
- `past_index`: -1
|
332 |
+
- `disable_tqdm`: False
|
333 |
+
- `remove_unused_columns`: True
|
334 |
+
- `label_names`: None
|
335 |
+
- `load_best_model_at_end`: False
|
336 |
+
- `ignore_data_skip`: False
|
337 |
+
- `fsdp`: []
|
338 |
+
- `fsdp_min_num_params`: 0
|
339 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
340 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
341 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True}
|
342 |
+
- `deepspeed`: None
|
343 |
+
- `label_smoothing_factor`: 0.0
|
344 |
+
- `optim`: adamw_torch
|
345 |
+
- `optim_args`: None
|
346 |
+
- `adafactor`: False
|
347 |
+
- `group_by_length`: False
|
348 |
+
- `length_column_name`: length
|
349 |
+
- `ddp_find_unused_parameters`: None
|
350 |
+
- `ddp_bucket_cap_mb`: None
|
351 |
+
- `ddp_broadcast_buffers`: False
|
352 |
+
- `dataloader_pin_memory`: True
|
353 |
+
- `dataloader_persistent_workers`: False
|
354 |
+
- `skip_memory_metrics`: True
|
355 |
+
- `use_legacy_prediction_loop`: False
|
356 |
+
- `push_to_hub`: False
|
357 |
+
- `resume_from_checkpoint`: None
|
358 |
+
- `hub_model_id`: None
|
359 |
+
- `hub_strategy`: every_save
|
360 |
+
- `hub_private_repo`: False
|
361 |
+
- `hub_always_push`: False
|
362 |
+
- `gradient_checkpointing`: False
|
363 |
+
- `gradient_checkpointing_kwargs`: None
|
364 |
+
- `include_inputs_for_metrics`: False
|
365 |
+
- `fp16_backend`: auto
|
366 |
+
- `push_to_hub_model_id`: None
|
367 |
+
- `push_to_hub_organization`: None
|
368 |
+
- `mp_parameters`:
|
369 |
+
- `auto_find_batch_size`: False
|
370 |
+
- `full_determinism`: False
|
371 |
+
- `torchdynamo`: None
|
372 |
+
- `ray_scope`: last
|
373 |
+
- `ddp_timeout`: 1800
|
374 |
+
- `torch_compile`: False
|
375 |
+
- `torch_compile_backend`: None
|
376 |
+
- `torch_compile_mode`: None
|
377 |
+
- `dispatch_batches`: None
|
378 |
+
- `split_batches`: None
|
379 |
+
- `include_tokens_per_second`: False
|
380 |
+
- `include_num_input_tokens_seen`: False
|
381 |
+
- `neftune_noise_alpha`: None
|
382 |
+
- `optim_target_modules`: None
|
383 |
+
- `batch_sampler`: no_duplicates
|
384 |
+
- `multi_dataset_batch_sampler`: proportional
|
385 |
+
|
386 |
+
</details>
|
387 |
+
|
388 |
+
### Training Logs
|
389 |
+
| Epoch | Step | Training Loss | loss | andegpt-dev_max_accuracy |
|
390 |
+
|:------:|:----:|:-------------:|:------:|:------------------------:|
|
391 |
+
| 0 | 0 | - | - | 0.6136 |
|
392 |
+
| 0.0270 | 250 | 0.8269 | 0.3100 | 0.9658 |
|
393 |
+
| 0.0540 | 500 | 0.3667 | 0.2169 | 0.9721 |
|
394 |
+
| 0.0809 | 750 | 0.2305 | 0.1594 | 0.9801 |
|
395 |
+
| 0.1079 | 1000 | 0.1866 | 0.1372 | 0.9830 |
|
396 |
+
| 0.1349 | 1250 | 0.1639 | 0.1114 | 0.9859 |
|
397 |
+
| 0.1619 | 1500 | 0.1375 | 0.0983 | 0.9871 |
|
398 |
+
| 0.1889 | 1750 | 0.1082 | 0.0815 | 0.9886 |
|
399 |
+
| 0.2158 | 2000 | 0.1023 | 0.0723 | 0.9900 |
|
400 |
+
| 0.2428 | 2250 | 0.0777 | 0.0703 | 0.9905 |
|
401 |
+
| 0.2698 | 2500 | 0.0809 | 0.0656 | 0.9896 |
|
402 |
+
| 0.2968 | 2750 | 0.0639 | 0.0662 | 0.9891 |
|
403 |
+
| 0.3238 | 3000 | 0.0633 | 0.0590 | 0.9922 |
|
404 |
+
| 0.3507 | 3250 | 0.0545 | 0.0533 | 0.9930 |
|
405 |
+
| 0.3777 | 3500 | 0.0541 | 0.0458 | 0.9932 |
|
406 |
+
| 0.4047 | 3750 | 0.0475 | 0.0365 | 0.9947 |
|
407 |
+
| 0.4317 | 4000 | 0.0394 | 0.0330 | 0.9939 |
|
408 |
+
| 0.4587 | 4250 | 0.0561 | 0.0345 | 0.9939 |
|
409 |
+
| 0.4856 | 4500 | 0.0432 | 0.0327 | 0.9942 |
|
410 |
+
| 0.5126 | 4750 | 0.0417 | 0.0328 | 0.9944 |
|
411 |
+
| 0.5396 | 5000 | 0.0388 | 0.0252 | 0.9949 |
|
412 |
+
| 0.5666 | 5250 | 0.033 | 0.0284 | 0.9959 |
|
413 |
+
| 0.5936 | 5500 | 0.0243 | 0.0229 | 0.9964 |
|
414 |
+
| 0.6205 | 5750 | 0.023 | 0.0223 | 0.9959 |
|
415 |
+
| 0.6475 | 6000 | 0.0313 | 0.0209 | 0.9966 |
|
416 |
+
| 0.6745 | 6250 | 0.0285 | 0.0208 | 0.9961 |
|
417 |
+
| 0.7015 | 6500 | 0.022 | 0.0192 | 0.9961 |
|
418 |
+
| 0.7285 | 6750 | 0.0219 | 0.0235 | 0.9956 |
|
419 |
+
| 0.7555 | 7000 | 0.0258 | 0.0186 | 0.9954 |
|
420 |
+
| 0.7824 | 7250 | 0.0226 | 0.0230 | 0.9959 |
|
421 |
+
| 0.8094 | 7500 | 0.0226 | 0.0240 | 0.9961 |
|
422 |
+
| 0.8364 | 7750 | 0.0208 | 0.0173 | 0.9968 |
|
423 |
+
| 0.8634 | 8000 | 0.0147 | 0.0200 | 0.9956 |
|
424 |
+
| 0.8904 | 8250 | 0.0193 | 0.0147 | 0.9971 |
|
425 |
+
| 0.9173 | 8500 | 0.0254 | 0.0136 | 0.9968 |
|
426 |
+
| 0.9443 | 8750 | 0.0148 | 0.0132 | 0.9971 |
|
427 |
+
| 0.9713 | 9000 | 0.0174 | 0.0157 | 0.9968 |
|
428 |
+
| 0.9983 | 9250 | 0.0221 | 0.0144 | 0.9971 |
|
429 |
+
|
430 |
+
|
431 |
+
### Framework Versions
|
432 |
+
- Python: 3.11.0
|
433 |
+
- Sentence Transformers: 3.0.1
|
434 |
+
- Transformers: 4.39.3
|
435 |
+
- PyTorch: 2.2.0+cu121
|
436 |
+
- Accelerate: 0.28.0
|
437 |
+
- Datasets: 2.20.0
|
438 |
+
- Tokenizers: 0.15.2
|
439 |
+
|
440 |
+
## Citation
|
441 |
+
|
442 |
+
### BibTeX
|
443 |
+
|
444 |
+
#### Sentence Transformers
|
445 |
+
```bibtex
|
446 |
+
@inproceedings{reimers-2019-sentence-bert,
|
447 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
448 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
449 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
450 |
+
month = "11",
|
451 |
+
year = "2019",
|
452 |
+
publisher = "Association for Computational Linguistics",
|
453 |
+
url = "https://arxiv.org/abs/1908.10084",
|
454 |
+
}
|
455 |
+
```
|
456 |
+
|
457 |
+
#### MultipleNegativesRankingLoss
|
458 |
+
```bibtex
|
459 |
+
@misc{henderson2017efficient,
|
460 |
+
title={Efficient Natural Language Response Suggestion for Smart Reply},
|
461 |
+
author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
|
462 |
+
year={2017},
|
463 |
+
eprint={1705.00652},
|
464 |
+
archivePrefix={arXiv},
|
465 |
+
primaryClass={cs.CL}
|
466 |
+
}
|
467 |
+
```
|
468 |
+
|
469 |
+
<!--
|
470 |
+
## Glossary
|
471 |
+
|
472 |
+
*Clearly define terms in order to be accessible across audiences.*
|
473 |
+
-->
|
474 |
+
|
475 |
+
<!--
|
476 |
+
## Model Card Authors
|
477 |
+
|
478 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
479 |
+
-->
|
480 |
+
|
481 |
+
<!--
|
482 |
+
## Model Card Contact
|
483 |
+
|
484 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
485 |
+
-->
|
config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "./models/mpnet-base-andegpt-triplet/checkpoint-9266/",
|
3 |
+
"architectures": [
|
4 |
+
"MPNetModel"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"bos_token_id": 0,
|
8 |
+
"eos_token_id": 2,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-05,
|
15 |
+
"max_position_embeddings": 514,
|
16 |
+
"model_type": "mpnet",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 1,
|
20 |
+
"relative_attention_num_buckets": 32,
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.39.3",
|
23 |
+
"vocab_size": 30527
|
24 |
+
}
|
config_sentence_transformers.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"__version__": {
|
3 |
+
"sentence_transformers": "3.0.1",
|
4 |
+
"transformers": "4.39.3",
|
5 |
+
"pytorch": "2.2.0+cu121"
|
6 |
+
},
|
7 |
+
"prompts": {},
|
8 |
+
"default_prompt_name": null,
|
9 |
+
"similarity_fn_name": null
|
10 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d2bcb1c4e8a60e4ff675490bc56a5c27bb5ae5d4194f69f41e458c269838034
|
3 |
+
size 437967672
|
modules.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"idx": 0,
|
4 |
+
"name": "0",
|
5 |
+
"path": "",
|
6 |
+
"type": "sentence_transformers.models.Transformer"
|
7 |
+
},
|
8 |
+
{
|
9 |
+
"idx": 1,
|
10 |
+
"name": "1",
|
11 |
+
"path": "1_Pooling",
|
12 |
+
"type": "sentence_transformers.models.Pooling"
|
13 |
+
}
|
14 |
+
]
|
sentence_bert_config.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"max_seq_length": 512,
|
3 |
+
"do_lower_case": false
|
4 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": {
|
3 |
+
"content": "<s>",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"cls_token": {
|
10 |
+
"content": "<s>",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"eos_token": {
|
17 |
+
"content": "</s>",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"mask_token": {
|
24 |
+
"content": "<mask>",
|
25 |
+
"lstrip": true,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"pad_token": {
|
31 |
+
"content": "<pad>",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
},
|
37 |
+
"sep_token": {
|
38 |
+
"content": "</s>",
|
39 |
+
"lstrip": false,
|
40 |
+
"normalized": false,
|
41 |
+
"rstrip": false,
|
42 |
+
"single_word": false
|
43 |
+
},
|
44 |
+
"unk_token": {
|
45 |
+
"content": "[UNK]",
|
46 |
+
"lstrip": false,
|
47 |
+
"normalized": false,
|
48 |
+
"rstrip": false,
|
49 |
+
"single_word": false
|
50 |
+
}
|
51 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "<pad>",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "</s>",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "<unk>",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": true,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"104": {
|
36 |
+
"content": "[UNK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
},
|
43 |
+
"30526": {
|
44 |
+
"content": "<mask>",
|
45 |
+
"lstrip": true,
|
46 |
+
"normalized": false,
|
47 |
+
"rstrip": false,
|
48 |
+
"single_word": false,
|
49 |
+
"special": true
|
50 |
+
}
|
51 |
+
},
|
52 |
+
"bos_token": "<s>",
|
53 |
+
"clean_up_tokenization_spaces": true,
|
54 |
+
"cls_token": "<s>",
|
55 |
+
"do_lower_case": true,
|
56 |
+
"eos_token": "</s>",
|
57 |
+
"mask_token": "<mask>",
|
58 |
+
"max_length": 512,
|
59 |
+
"model_max_length": 512,
|
60 |
+
"pad_to_multiple_of": null,
|
61 |
+
"pad_token": "<pad>",
|
62 |
+
"pad_token_type_id": 0,
|
63 |
+
"padding_side": "right",
|
64 |
+
"sep_token": "</s>",
|
65 |
+
"stride": 0,
|
66 |
+
"strip_accents": null,
|
67 |
+
"tokenize_chinese_chars": true,
|
68 |
+
"tokenizer_class": "MPNetTokenizer",
|
69 |
+
"truncation_side": "right",
|
70 |
+
"truncation_strategy": "longest_first",
|
71 |
+
"unk_token": "[UNK]"
|
72 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|