Add new SentenceTransformer model.
Browse files- 1_Pooling/config.json +10 -0
- README.md +503 -0
- config.json +26 -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 +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +57 -0
- vocab.txt +0 -0
1_Pooling/config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"word_embedding_dimension": 384,
|
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 +1,503 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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: nreimers/MiniLM-L6-H384-uncased
|
14 |
+
datasets: []
|
15 |
+
metrics:
|
16 |
+
- cosine_accuracy
|
17 |
+
- dot_accuracy
|
18 |
+
- manhattan_accuracy
|
19 |
+
- euclidean_accuracy
|
20 |
+
- max_accuracy
|
21 |
+
widget:
|
22 |
+
- source_sentence: Enumere los tres casos en los que se aplican las prescripciones
|
23 |
+
del artículo 40.1.1.
|
24 |
+
sentences:
|
25 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 39.1.1 trata sobre: Estas prescripciones
|
26 |
+
se aplican a todos los aparatos domésticos y de iluminación de cualquier tipo,
|
27 |
+
forma y tamaño, siempre que: a) La instalación eléctrica de la residencia se
|
28 |
+
halle capacitada para servirlos. b) Se respeten las demás prescripciones de este
|
29 |
+
Reglamento que le sean aplicables.. El 39.1.1 pertenece a la sección: <section>39.1</section>'
|
30 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 14.3.1 trata sobre: Circuitos trifásicos
|
31 |
+
son aquellos que emplean las tres fases de la energía que provee la ANDE, con
|
32 |
+
interruptor y protección adecuados en el tablero de arranque y se emplean en líneas
|
33 |
+
distribuidoras de fuerza motriz, calefacción, refrigeración y similares, comprendiendo
|
34 |
+
incluso aparatos monofásicos, sin limitaciones de carga, siempre que: a) Se realice
|
35 |
+
el equilibrio de cargas de los equipos monofásicos. b) Se atienda correctamente
|
36 |
+
a 13.4.2. c) Que ninguna de las cargas trifásicas individuales sea igual o superior
|
37 |
+
a 15 A nominales en el caso de motores, o 20 A si son equipos de calefacción o
|
38 |
+
similares. d) Se use un circuito trifásico independiente por cada motor de 15
|
39 |
+
A nominales o más. e) Se use un circuito trifásico independiente por cada equipo
|
40 |
+
de calefacción o similar de 20 A o más.. El 14.3.1 pertenece a la sección: <section>14.3</section>'
|
41 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 40.1.1 trata sobre: Sin perjuicio de
|
42 |
+
las demás disposiciones de este Reglamento, estas prescripciones se aplican para:
|
43 |
+
a) Conexiones entre las instalaciones fijas y los aparatos portátiles, o que deban
|
44 |
+
ser desplazadas con alguna frecuencia. b) Conexiones de las partes móviles de
|
45 |
+
aparatos y máquinas fijas. c) Conexiones de aparatos de iluminación, colgantes,
|
46 |
+
etc. con la observación del numeral 39. 40.2 Tipos de conductores:. El 40.1.1
|
47 |
+
pertenece a la sección: <section>40.1</section>'
|
48 |
+
- source_sentence: ¿Hasta dónde llega la conexión eléctrica según el punto 11.1.1
|
49 |
+
del reglamento de baja tensión de la ANDE?
|
50 |
+
sentences:
|
51 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 11.5.3 trata sobre: Los conductores,
|
52 |
+
equipo eléctrico auxiliar y la mano de obra para el servicio y entrada subterránea,
|
53 |
+
serán aportados por la ANDE, y abonados por el usuario, quedando de propiedad
|
54 |
+
de la ANDE, la que, en consecuencia, tendrá a su cargo su conservación y buen
|
55 |
+
servicio.. El 11.5.3 pertenece a la sección: <section>11.5</section>'
|
56 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 11.1.1 trata sobre: Servicio es la
|
57 |
+
conexión eléctrica desde el punto de toma de energía de la red, hasta la parte
|
58 |
+
externa de la propiedad del usuario, sobre la calle, en el punto escogido para
|
59 |
+
la entrada de energía.. El 11.1.1 pertenece a la sección: <section>11.1</section>'
|
60 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 22.2.3 trata sobre: En los sistemas
|
61 |
+
de distribución sin neutro, la conexión a tierra de equipos es obligatoria (22.1.2).
|
62 |
+
En los sistemas de distribución con neutro, la red de interconexión de tierra
|
63 |
+
de los equipos quedará substituida por el conductor neutro, que hará sus veces,
|
64 |
+
debiendo conectarse a él todas las partes que normalmente irían conectadas a la
|
65 |
+
red de conexión de tierra de equipos.. El 22.2.3 pertenece a la sección: <section>22.2</section>'
|
66 |
+
- source_sentence: ¿Cuál es el propósito de los grupos mencionados en el reglamento
|
67 |
+
de baja tensión de la ANDE?
|
68 |
+
sentences:
|
69 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 6.1 trata sobre: En todos los planos
|
70 |
+
de instalaciones eléctricas deberá usarse la simbología indicada en el Anexo N°
|
71 |
+
2.. El 6.1 pertenece a la sección: <section>6-</section>'
|
72 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 25- trata sobre Instalación en tubos
|
73 |
+
flexibles: y tiene las siguientes sub-secciones: <sub-section>25.1</sub-section>,
|
74 |
+
<sub-section>25.2</sub-section>'
|
75 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 42.3.1 trata sobre: Estos grupos se
|
76 |
+
utilizarán para proveer energía eléctrica complementaria a la que se toma de la
|
77 |
+
red de distribución de ANDE y, exclusivamente, a las instalaciones del usuario..
|
78 |
+
El 42.3.1 pertenece a la sección: <section>42.3</section>'
|
79 |
+
- source_sentence: ¿Cuántas redes independientes debe comprender la instalación de
|
80 |
+
iluminación de cines, teatros y locales semejantes?
|
81 |
+
sentences:
|
82 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 41.5.2 trata sobre: La instalación
|
83 |
+
de iluminación de cines, teatros y locales semejantes deberá comprender las siguientes
|
84 |
+
redes independientes: a) Iluminación del edificio propiamente dicho (oficinas,
|
85 |
+
pasillos, taquilla, baños, entradas y salidas, guardarropas, camarines, etc.). b)
|
86 |
+
Iluminación del escenario. c) Iluminación del local ocupado por las personas (plateas,
|
87 |
+
palcos, balcones, así como para circulación de las personas durante la realización
|
88 |
+
del programa con luces generales apagadas). d) Iluminación de emergencia.. El
|
89 |
+
41.5.2 pertenece a la sección: <section>41.5</section>'
|
90 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 33.3.2 trata sobre: Los cables, aunque
|
91 |
+
del tipo adecuado, no podrán ser enterrados en el suelo en el interior de edificios,
|
92 |
+
salvo que se trate de áreas industriales. La profundidad de instalación no deberá
|
93 |
+
ser inferior a 60 cm, debiendo llevar el cable en toda su extensión, una capa
|
94 |
+
de arena, y encima de la misma una línea continua de ladrillos simplemente apoyados
|
95 |
+
(no unidos con argamasa), cuya función fundamental es denunciar la presencia del
|
96 |
+
cable, además de brindar una pequeña protección mecánica adicional.. El 33.3.2
|
97 |
+
pertenece a la sección: <section>33.3</section>'
|
98 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 12.1.2 trata sobre: Los tableros se
|
99 |
+
instalarán en lugares secos y de fácil acceso y, si es posible, en lugares expresamente
|
100 |
+
reservados, ventilados e iluminados.. El 12.1.2 pertenece a la sección: <section>12.1</section>'
|
101 |
+
- source_sentence: ¿Cuál es el nombre del reglamento que se menciona en la información
|
102 |
+
proporcionada?
|
103 |
+
sentences:
|
104 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 22.1.2 trata sobre: Se entiende “tierra
|
105 |
+
de equipos”, la conexión a tierra de las partes metálicas de la instalación o
|
106 |
+
de los aparatos que no transportan corriente, tales como: tubos de metal, blindajes
|
107 |
+
metálicos de los cables, cajas de conexión y/o derivación, estructuras de tableros
|
108 |
+
o cuadros, cajas de interruptores, bastidores de máquinas y, en general, cualquier
|
109 |
+
parte metálica relacionada con la instalación eléctrica y no destinada a la conducción
|
110 |
+
de corriente.. El 22.1.2 pertenece a la sección: <section>22.1</section>'
|
111 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 10- trata sobre Partes de que se compone
|
112 |
+
una instalación eléctrica: y tiene las siguientes sub-secciones: <sub-section>10.1</sub-section>'
|
113 |
+
- 'Reglamento de Baja Tensión de la ANDE: El 37- trata sobre Soldadura eléctrica:
|
114 |
+
y tiene las siguientes sub-secciones: <sub-section>37.1</sub-section>, <sub-section>37.2</sub-section>,
|
115 |
+
<sub-section>37.3</sub-section>, <sub-section>37.4</sub-section>'
|
116 |
+
pipeline_tag: sentence-similarity
|
117 |
+
model-index:
|
118 |
+
- name: embed-andegpt-H384
|
119 |
+
results:
|
120 |
+
- task:
|
121 |
+
type: triplet
|
122 |
+
name: Triplet
|
123 |
+
dataset:
|
124 |
+
name: andegpt dev
|
125 |
+
type: andegpt-dev
|
126 |
+
metrics:
|
127 |
+
- type: cosine_accuracy
|
128 |
+
value: 0.998300145701797
|
129 |
+
name: Cosine Accuracy
|
130 |
+
- type: dot_accuracy
|
131 |
+
value: 0.002185526954832443
|
132 |
+
name: Dot Accuracy
|
133 |
+
- type: manhattan_accuracy
|
134 |
+
value: 0.9985429820301117
|
135 |
+
name: Manhattan Accuracy
|
136 |
+
- type: euclidean_accuracy
|
137 |
+
value: 0.998300145701797
|
138 |
+
name: Euclidean Accuracy
|
139 |
+
- type: max_accuracy
|
140 |
+
value: 0.9985429820301117
|
141 |
+
name: Max Accuracy
|
142 |
+
- task:
|
143 |
+
type: triplet
|
144 |
+
name: Triplet
|
145 |
+
dataset:
|
146 |
+
name: andegpt test
|
147 |
+
type: andegpt-test
|
148 |
+
metrics:
|
149 |
+
- type: cosine_accuracy
|
150 |
+
value: 0.9973288003885381
|
151 |
+
name: Cosine Accuracy
|
152 |
+
- type: dot_accuracy
|
153 |
+
value: 0.0024283632831471587
|
154 |
+
name: Dot Accuracy
|
155 |
+
- type: manhattan_accuracy
|
156 |
+
value: 0.9970859640602234
|
157 |
+
name: Manhattan Accuracy
|
158 |
+
- type: euclidean_accuracy
|
159 |
+
value: 0.9973288003885381
|
160 |
+
name: Euclidean Accuracy
|
161 |
+
- type: max_accuracy
|
162 |
+
value: 0.9973288003885381
|
163 |
+
name: Max Accuracy
|
164 |
+
---
|
165 |
+
|
166 |
+
# embed-andegpt-H384
|
167 |
+
|
168 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [nreimers/MiniLM-L6-H384-uncased](https://huggingface.co/nreimers/MiniLM-L6-H384-uncased). It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
|
169 |
+
|
170 |
+
## Model Details
|
171 |
+
|
172 |
+
### Model Description
|
173 |
+
- **Model Type:** Sentence Transformer
|
174 |
+
- **Base model:** [nreimers/MiniLM-L6-H384-uncased](https://huggingface.co/nreimers/MiniLM-L6-H384-uncased) <!-- at revision 3276f0fac9d818781d7a1327b3ff818fc4e643c0 -->
|
175 |
+
- **Maximum Sequence Length:** 512 tokens
|
176 |
+
- **Output Dimensionality:** 384 tokens
|
177 |
+
- **Similarity Function:** Cosine Similarity
|
178 |
+
<!-- - **Training Dataset:** Unknown -->
|
179 |
+
- **Language:** es
|
180 |
+
- **License:** apache-2.0
|
181 |
+
|
182 |
+
### Model Sources
|
183 |
+
|
184 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
185 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
186 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
187 |
+
|
188 |
+
### Full Model Architecture
|
189 |
+
|
190 |
+
```
|
191 |
+
SentenceTransformer(
|
192 |
+
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
|
193 |
+
(1): Pooling({'word_embedding_dimension': 384, '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})
|
194 |
+
)
|
195 |
+
```
|
196 |
+
|
197 |
+
## Usage
|
198 |
+
|
199 |
+
### Direct Usage (Sentence Transformers)
|
200 |
+
|
201 |
+
First install the Sentence Transformers library:
|
202 |
+
|
203 |
+
```bash
|
204 |
+
pip install -U sentence-transformers
|
205 |
+
```
|
206 |
+
|
207 |
+
Then you can load this model and run inference.
|
208 |
+
```python
|
209 |
+
from sentence_transformers import SentenceTransformer
|
210 |
+
|
211 |
+
# Download from the 🤗 Hub
|
212 |
+
model = SentenceTransformer("enpaiva/embed-andegpt-H384")
|
213 |
+
# Run inference
|
214 |
+
sentences = [
|
215 |
+
'¿Cuál es el nombre del reglamento que se menciona en la información proporcionada?',
|
216 |
+
'Reglamento de Baja Tensión de la ANDE: El 10- trata sobre Partes de que se compone una instalación eléctrica: y tiene las siguientes sub-secciones: <sub-section>10.1</sub-section>',
|
217 |
+
'Reglamento de Baja Tensión de la ANDE: El 37- trata sobre Soldadura eléctrica: y tiene las siguientes sub-secciones: <sub-section>37.1</sub-section>, <sub-section>37.2</sub-section>, <sub-section>37.3</sub-section>, <sub-section>37.4</sub-section>',
|
218 |
+
]
|
219 |
+
embeddings = model.encode(sentences)
|
220 |
+
print(embeddings.shape)
|
221 |
+
# [3, 384]
|
222 |
+
|
223 |
+
# Get the similarity scores for the embeddings
|
224 |
+
similarities = model.similarity(embeddings, embeddings)
|
225 |
+
print(similarities.shape)
|
226 |
+
# [3, 3]
|
227 |
+
```
|
228 |
+
|
229 |
+
<!--
|
230 |
+
### Direct Usage (Transformers)
|
231 |
+
|
232 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
233 |
+
|
234 |
+
</details>
|
235 |
+
-->
|
236 |
+
|
237 |
+
<!--
|
238 |
+
### Downstream Usage (Sentence Transformers)
|
239 |
+
|
240 |
+
You can finetune this model on your own dataset.
|
241 |
+
|
242 |
+
<details><summary>Click to expand</summary>
|
243 |
+
|
244 |
+
</details>
|
245 |
+
-->
|
246 |
+
|
247 |
+
<!--
|
248 |
+
### Out-of-Scope Use
|
249 |
+
|
250 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
251 |
+
-->
|
252 |
+
|
253 |
+
## Evaluation
|
254 |
+
|
255 |
+
### Metrics
|
256 |
+
|
257 |
+
#### Triplet
|
258 |
+
* Dataset: `andegpt-dev`
|
259 |
+
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
|
260 |
+
|
261 |
+
| Metric | Value |
|
262 |
+
|:-------------------|:-----------|
|
263 |
+
| cosine_accuracy | 0.9983 |
|
264 |
+
| dot_accuracy | 0.0022 |
|
265 |
+
| manhattan_accuracy | 0.9985 |
|
266 |
+
| euclidean_accuracy | 0.9983 |
|
267 |
+
| **max_accuracy** | **0.9985** |
|
268 |
+
|
269 |
+
#### Triplet
|
270 |
+
* Dataset: `andegpt-test`
|
271 |
+
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
|
272 |
+
|
273 |
+
| Metric | Value |
|
274 |
+
|:-------------------|:-----------|
|
275 |
+
| cosine_accuracy | 0.9973 |
|
276 |
+
| dot_accuracy | 0.0024 |
|
277 |
+
| manhattan_accuracy | 0.9971 |
|
278 |
+
| euclidean_accuracy | 0.9973 |
|
279 |
+
| **max_accuracy** | **0.9973** |
|
280 |
+
|
281 |
+
<!--
|
282 |
+
## Bias, Risks and Limitations
|
283 |
+
|
284 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
285 |
+
-->
|
286 |
+
|
287 |
+
<!--
|
288 |
+
### Recommendations
|
289 |
+
|
290 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
291 |
+
-->
|
292 |
+
|
293 |
+
## Training Details
|
294 |
+
|
295 |
+
### Training Hyperparameters
|
296 |
+
#### Non-Default Hyperparameters
|
297 |
+
|
298 |
+
- `prediction_loss_only`: False
|
299 |
+
- `per_device_train_batch_size`: 32
|
300 |
+
- `learning_rate`: 2e-05
|
301 |
+
- `lr_scheduler_type`: cosine
|
302 |
+
- `log_level_replica`: passive
|
303 |
+
- `log_on_each_node`: False
|
304 |
+
- `logging_nan_inf_filter`: False
|
305 |
+
- `bf16`: True
|
306 |
+
- `batch_sampler`: no_duplicates
|
307 |
+
|
308 |
+
#### All Hyperparameters
|
309 |
+
<details><summary>Click to expand</summary>
|
310 |
+
|
311 |
+
- `overwrite_output_dir`: False
|
312 |
+
- `do_predict`: False
|
313 |
+
- `prediction_loss_only`: False
|
314 |
+
- `per_device_train_batch_size`: 32
|
315 |
+
- `per_device_eval_batch_size`: 8
|
316 |
+
- `per_gpu_train_batch_size`: None
|
317 |
+
- `per_gpu_eval_batch_size`: None
|
318 |
+
- `gradient_accumulation_steps`: 1
|
319 |
+
- `eval_accumulation_steps`: None
|
320 |
+
- `learning_rate`: 2e-05
|
321 |
+
- `weight_decay`: 0.0
|
322 |
+
- `adam_beta1`: 0.9
|
323 |
+
- `adam_beta2`: 0.999
|
324 |
+
- `adam_epsilon`: 1e-08
|
325 |
+
- `max_grad_norm`: 1.0
|
326 |
+
- `num_train_epochs`: 3
|
327 |
+
- `max_steps`: -1
|
328 |
+
- `lr_scheduler_type`: cosine
|
329 |
+
- `lr_scheduler_kwargs`: {}
|
330 |
+
- `warmup_ratio`: 0
|
331 |
+
- `warmup_steps`: 0
|
332 |
+
- `log_level`: passive
|
333 |
+
- `log_level_replica`: passive
|
334 |
+
- `log_on_each_node`: False
|
335 |
+
- `logging_nan_inf_filter`: False
|
336 |
+
- `save_safetensors`: True
|
337 |
+
- `save_on_each_node`: False
|
338 |
+
- `save_only_model`: False
|
339 |
+
- `no_cuda`: False
|
340 |
+
- `use_cpu`: False
|
341 |
+
- `use_mps_device`: False
|
342 |
+
- `seed`: 42
|
343 |
+
- `data_seed`: None
|
344 |
+
- `jit_mode_eval`: False
|
345 |
+
- `use_ipex`: False
|
346 |
+
- `bf16`: True
|
347 |
+
- `fp16`: False
|
348 |
+
- `fp16_opt_level`: O1
|
349 |
+
- `half_precision_backend`: auto
|
350 |
+
- `bf16_full_eval`: False
|
351 |
+
- `fp16_full_eval`: False
|
352 |
+
- `tf32`: None
|
353 |
+
- `local_rank`: 0
|
354 |
+
- `ddp_backend`: None
|
355 |
+
- `tpu_num_cores`: None
|
356 |
+
- `tpu_metrics_debug`: False
|
357 |
+
- `debug`: []
|
358 |
+
- `dataloader_drop_last`: False
|
359 |
+
- `dataloader_num_workers`: 0
|
360 |
+
- `dataloader_prefetch_factor`: None
|
361 |
+
- `past_index`: -1
|
362 |
+
- `disable_tqdm`: False
|
363 |
+
- `remove_unused_columns`: True
|
364 |
+
- `label_names`: None
|
365 |
+
- `load_best_model_at_end`: False
|
366 |
+
- `ignore_data_skip`: False
|
367 |
+
- `fsdp`: []
|
368 |
+
- `fsdp_min_num_params`: 0
|
369 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
370 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
371 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True}
|
372 |
+
- `deepspeed`: None
|
373 |
+
- `label_smoothing_factor`: 0.0
|
374 |
+
- `optim`: adamw_torch
|
375 |
+
- `optim_args`: None
|
376 |
+
- `adafactor`: False
|
377 |
+
- `group_by_length`: False
|
378 |
+
- `length_column_name`: length
|
379 |
+
- `ddp_find_unused_parameters`: None
|
380 |
+
- `ddp_bucket_cap_mb`: None
|
381 |
+
- `ddp_broadcast_buffers`: False
|
382 |
+
- `dataloader_pin_memory`: True
|
383 |
+
- `dataloader_persistent_workers`: False
|
384 |
+
- `skip_memory_metrics`: True
|
385 |
+
- `use_legacy_prediction_loop`: False
|
386 |
+
- `push_to_hub`: False
|
387 |
+
- `resume_from_checkpoint`: None
|
388 |
+
- `hub_model_id`: None
|
389 |
+
- `hub_strategy`: every_save
|
390 |
+
- `hub_private_repo`: False
|
391 |
+
- `hub_always_push`: False
|
392 |
+
- `gradient_checkpointing`: False
|
393 |
+
- `gradient_checkpointing_kwargs`: None
|
394 |
+
- `include_inputs_for_metrics`: False
|
395 |
+
- `fp16_backend`: auto
|
396 |
+
- `push_to_hub_model_id`: None
|
397 |
+
- `push_to_hub_organization`: None
|
398 |
+
- `mp_parameters`:
|
399 |
+
- `auto_find_batch_size`: False
|
400 |
+
- `full_determinism`: False
|
401 |
+
- `torchdynamo`: None
|
402 |
+
- `ray_scope`: last
|
403 |
+
- `ddp_timeout`: 1800
|
404 |
+
- `torch_compile`: False
|
405 |
+
- `torch_compile_backend`: None
|
406 |
+
- `torch_compile_mode`: None
|
407 |
+
- `dispatch_batches`: None
|
408 |
+
- `split_batches`: None
|
409 |
+
- `include_tokens_per_second`: False
|
410 |
+
- `include_num_input_tokens_seen`: False
|
411 |
+
- `neftune_noise_alpha`: None
|
412 |
+
- `optim_target_modules`: None
|
413 |
+
- `batch_sampler`: no_duplicates
|
414 |
+
- `multi_dataset_batch_sampler`: proportional
|
415 |
+
|
416 |
+
</details>
|
417 |
+
|
418 |
+
### Training Logs
|
419 |
+
| Epoch | Step | Training Loss | loss | andegpt-dev_max_accuracy | andegpt-test_max_accuracy |
|
420 |
+
|:------:|:----:|:-------------:|:------:|:------------------------:|:-------------------------:|
|
421 |
+
| 0 | 0 | - | - | 0.5920 | - |
|
422 |
+
| 0.1079 | 250 | 2.3094 | 0.7200 | 0.9597 | - |
|
423 |
+
| 0.2158 | 500 | 0.7952 | 0.3598 | 0.9813 | - |
|
424 |
+
| 0.3237 | 750 | 0.4862 | 0.2162 | 0.9910 | - |
|
425 |
+
| 0.4316 | 1000 | 0.3304 | 0.1558 | 0.9927 | - |
|
426 |
+
| 0.5395 | 1250 | 0.2527 | 0.1140 | 0.9961 | - |
|
427 |
+
| 0.6474 | 1500 | 0.1987 | 0.0859 | 0.9964 | - |
|
428 |
+
| 0.7553 | 1750 | 0.1617 | 0.0729 | 0.9959 | - |
|
429 |
+
| 0.8632 | 2000 | 0.1419 | 0.0562 | 0.9966 | - |
|
430 |
+
| 0.9711 | 2250 | 0.1132 | 0.0495 | 0.9968 | - |
|
431 |
+
| 1.0790 | 2500 | 0.1043 | 0.0429 | 0.9971 | - |
|
432 |
+
| 1.1869 | 2750 | 0.0947 | 0.0368 | 0.9978 | - |
|
433 |
+
| 1.2948 | 3000 | 0.0736 | 0.0367 | 0.9976 | - |
|
434 |
+
| 1.4027 | 3250 | 0.0661 | 0.0296 | 0.9978 | - |
|
435 |
+
| 1.5106 | 3500 | 0.0613 | 0.0279 | 0.9985 | - |
|
436 |
+
| 1.6185 | 3750 | 0.0607 | 0.0264 | 0.9983 | - |
|
437 |
+
| 1.7264 | 4000 | 0.0521 | 0.0238 | 0.9985 | - |
|
438 |
+
| 1.8343 | 4250 | 0.0495 | 0.0216 | 0.9985 | - |
|
439 |
+
| 1.9422 | 4500 | 0.0425 | 0.0211 | 0.9983 | - |
|
440 |
+
| 2.0501 | 4750 | 0.0428 | 0.0200 | 0.9983 | - |
|
441 |
+
| 2.1580 | 5000 | 0.0435 | 0.0190 | 0.9985 | - |
|
442 |
+
| 2.2659 | 5250 | 0.0393 | 0.0188 | 0.9983 | - |
|
443 |
+
| 2.3738 | 5500 | 0.0356 | 0.0182 | 0.9983 | - |
|
444 |
+
| 2.4817 | 5750 | 0.0351 | 0.0180 | 0.9988 | - |
|
445 |
+
| 2.5896 | 6000 | 0.0394 | 0.0181 | 0.9985 | - |
|
446 |
+
| 2.5973 | 6018 | - | - | - | 0.9973 |
|
447 |
+
|
448 |
+
|
449 |
+
### Framework Versions
|
450 |
+
- Python: 3.11.0
|
451 |
+
- Sentence Transformers: 3.0.1
|
452 |
+
- Transformers: 4.39.3
|
453 |
+
- PyTorch: 2.2.0+cu121
|
454 |
+
- Accelerate: 0.28.0
|
455 |
+
- Datasets: 2.20.0
|
456 |
+
- Tokenizers: 0.15.2
|
457 |
+
|
458 |
+
## Citation
|
459 |
+
|
460 |
+
### BibTeX
|
461 |
+
|
462 |
+
#### Sentence Transformers
|
463 |
+
```bibtex
|
464 |
+
@inproceedings{reimers-2019-sentence-bert,
|
465 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
466 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
467 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
468 |
+
month = "11",
|
469 |
+
year = "2019",
|
470 |
+
publisher = "Association for Computational Linguistics",
|
471 |
+
url = "https://arxiv.org/abs/1908.10084",
|
472 |
+
}
|
473 |
+
```
|
474 |
+
|
475 |
+
#### MultipleNegativesRankingLoss
|
476 |
+
```bibtex
|
477 |
+
@misc{henderson2017efficient,
|
478 |
+
title={Efficient Natural Language Response Suggestion for Smart Reply},
|
479 |
+
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},
|
480 |
+
year={2017},
|
481 |
+
eprint={1705.00652},
|
482 |
+
archivePrefix={arXiv},
|
483 |
+
primaryClass={cs.CL}
|
484 |
+
}
|
485 |
+
```
|
486 |
+
|
487 |
+
<!--
|
488 |
+
## Glossary
|
489 |
+
|
490 |
+
*Clearly define terms in order to be accessible across audiences.*
|
491 |
+
-->
|
492 |
+
|
493 |
+
<!--
|
494 |
+
## Model Card Authors
|
495 |
+
|
496 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
497 |
+
-->
|
498 |
+
|
499 |
+
<!--
|
500 |
+
## Model Card Contact
|
501 |
+
|
502 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
503 |
+
-->
|
config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "nreimers/MiniLM-L6-H384-uncased",
|
3 |
+
"architectures": [
|
4 |
+
"BertModel"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 384,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 1536,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 6,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.39.3",
|
23 |
+
"type_vocab_size": 2,
|
24 |
+
"use_cache": true,
|
25 |
+
"vocab_size": 30522
|
26 |
+
}
|
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 sha256:33754635dde4936564ba073942e483097ce955b1a54cc3a2f8876f74ab472212
|
3 |
+
size 90864192
|
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 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": true,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"model_max_length": 512,
|
50 |
+
"never_split": null,
|
51 |
+
"pad_token": "[PAD]",
|
52 |
+
"sep_token": "[SEP]",
|
53 |
+
"strip_accents": null,
|
54 |
+
"tokenize_chinese_chars": true,
|
55 |
+
"tokenizer_class": "BertTokenizer",
|
56 |
+
"unk_token": "[UNK]"
|
57 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|