Add new SentenceTransformer model.
Browse files
README.md
CHANGED
@@ -1,201 +1,577 @@
|
|
1 |
---
|
2 |
base_model: colorfulscoop/sbert-base-ja
|
3 |
-
|
4 |
-
|
5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
---
|
7 |
|
8 |
-
#
|
9 |
-
|
10 |
-
<!-- Provide a quick summary of what the model is/does. -->
|
11 |
-
|
12 |
|
|
|
13 |
|
14 |
## Model Details
|
15 |
|
16 |
### Model Description
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
-
|
19 |
-
|
20 |
-
Generates similarity embeddings
|
21 |
-
|
22 |
-
- **Developed by:** [More Information Needed]
|
23 |
-
- **Funded by [optional]:** [More Information Needed]
|
24 |
-
- **Shared by [optional]:** [More Information Needed]
|
25 |
-
- **Model type:** [More Information Needed]
|
26 |
-
- **Language(s) (NLP):** ja
|
27 |
-
- **License:** cc-by-sa-4.0
|
28 |
-
- **Finetuned from model [optional]:** colorfulscoop/sbert-base-ja
|
29 |
-
|
30 |
-
### Model Sources [optional]
|
31 |
-
|
32 |
-
<!-- Provide the basic links for the model. -->
|
33 |
-
|
34 |
-
- **Repository:** [More Information Needed]
|
35 |
-
- **Paper [optional]:** [More Information Needed]
|
36 |
-
- **Demo [optional]:** [More Information Needed]
|
37 |
-
|
38 |
-
## Uses
|
39 |
-
|
40 |
-
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
41 |
|
42 |
-
|
|
|
|
|
43 |
|
44 |
-
|
45 |
|
46 |
-
|
|
|
|
|
|
|
|
|
|
|
47 |
|
48 |
-
|
49 |
|
50 |
-
|
51 |
|
52 |
-
|
53 |
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
[More Information Needed]
|
59 |
|
60 |
-
|
|
|
|
|
61 |
|
62 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
|
64 |
-
|
65 |
-
|
66 |
-
|
|
|
|
|
67 |
|
68 |
-
<!--
|
|
|
69 |
|
70 |
-
|
71 |
|
72 |
-
|
73 |
-
|
74 |
-
Use the code below to get started with the model.
|
75 |
-
|
76 |
-
[More Information Needed]
|
77 |
-
|
78 |
-
## Training Details
|
79 |
|
80 |
-
|
|
|
81 |
|
82 |
-
|
83 |
|
84 |
-
|
85 |
|
86 |
-
|
|
|
87 |
|
88 |
-
<!--
|
89 |
-
|
90 |
-
#### Preprocessing [optional]
|
91 |
-
|
92 |
-
[More Information Needed]
|
93 |
-
|
94 |
-
|
95 |
-
#### Training Hyperparameters
|
96 |
-
|
97 |
-
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
98 |
-
|
99 |
-
#### Speeds, Sizes, Times [optional]
|
100 |
-
|
101 |
-
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
102 |
|
103 |
-
|
|
|
104 |
|
105 |
## Evaluation
|
106 |
|
107 |
-
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
-
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
119 |
-
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
-
|
126 |
-
|
127 |
-
|
128 |
-
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
|
139 |
-
|
140 |
-
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
-
[More Information Needed]
|
160 |
-
|
161 |
-
### Compute Infrastructure
|
162 |
-
|
163 |
-
[More Information Needed]
|
164 |
-
|
165 |
-
#### Hardware
|
166 |
-
|
167 |
-
[More Information Needed]
|
168 |
-
|
169 |
-
#### Software
|
170 |
-
|
171 |
-
[More Information Needed]
|
172 |
-
|
173 |
-
## Citation [optional]
|
174 |
-
|
175 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
176 |
-
|
177 |
-
**BibTeX:**
|
178 |
-
|
179 |
-
[More Information Needed]
|
180 |
-
|
181 |
-
**APA:**
|
182 |
-
|
183 |
-
[More Information Needed]
|
184 |
-
|
185 |
-
## Glossary [optional]
|
186 |
-
|
187 |
-
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
188 |
-
|
189 |
-
[More Information Needed]
|
190 |
-
|
191 |
-
## More Information [optional]
|
192 |
-
|
193 |
-
[More Information Needed]
|
194 |
|
195 |
-
|
|
|
196 |
|
197 |
-
|
198 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
199 |
## Model Card Contact
|
200 |
|
201 |
-
|
|
|
|
1 |
---
|
2 |
base_model: colorfulscoop/sbert-base-ja
|
3 |
+
library_name: sentence-transformers
|
4 |
+
metrics:
|
5 |
+
- cosine_accuracy
|
6 |
+
- cosine_accuracy_threshold
|
7 |
+
- cosine_f1
|
8 |
+
- cosine_f1_threshold
|
9 |
+
- cosine_precision
|
10 |
+
- cosine_recall
|
11 |
+
- cosine_ap
|
12 |
+
- dot_accuracy
|
13 |
+
- dot_accuracy_threshold
|
14 |
+
- dot_f1
|
15 |
+
- dot_f1_threshold
|
16 |
+
- dot_precision
|
17 |
+
- dot_recall
|
18 |
+
- dot_ap
|
19 |
+
- manhattan_accuracy
|
20 |
+
- manhattan_accuracy_threshold
|
21 |
+
- manhattan_f1
|
22 |
+
- manhattan_f1_threshold
|
23 |
+
- manhattan_precision
|
24 |
+
- manhattan_recall
|
25 |
+
- manhattan_ap
|
26 |
+
- euclidean_accuracy
|
27 |
+
- euclidean_accuracy_threshold
|
28 |
+
- euclidean_f1
|
29 |
+
- euclidean_f1_threshold
|
30 |
+
- euclidean_precision
|
31 |
+
- euclidean_recall
|
32 |
+
- euclidean_ap
|
33 |
+
- max_accuracy
|
34 |
+
- max_accuracy_threshold
|
35 |
+
- max_f1
|
36 |
+
- max_f1_threshold
|
37 |
+
- max_precision
|
38 |
+
- max_recall
|
39 |
+
- max_ap
|
40 |
+
pipeline_tag: sentence-similarity
|
41 |
+
tags:
|
42 |
+
- sentence-transformers
|
43 |
+
- sentence-similarity
|
44 |
+
- feature-extraction
|
45 |
+
- generated_from_trainer
|
46 |
+
- dataset_size:5330
|
47 |
+
- loss:CoSENTLoss
|
48 |
+
widget:
|
49 |
+
- source_sentence: 家族 は 日陰 の 公園 で 午後 を 楽しんで い ます 。
|
50 |
+
sentences:
|
51 |
+
- 家族 は 公園 で ピクニック を して い ます 。
|
52 |
+
- 人 が 会衆 に 説教 して い ます 。
|
53 |
+
- 青い シャツ を 着た 子供 は フリスビー を 投げ ます 。
|
54 |
+
- source_sentence: 雪 が 溶けて 、 小さな 男の子 が 裏庭 で 犬 と 遊んで い ます 。
|
55 |
+
sentences:
|
56 |
+
- 男性 は スポーツ を 見て い ます 。
|
57 |
+
- 女性 は ビーチ サンダル を 運んで い ます 。
|
58 |
+
- 吹雪 の 雲 の 後ろ から 太陽 が 出た 後 、 小さな 男の子 と 彼 の 犬 が 外 で 遊ぶ 。
|
59 |
+
- source_sentence: 日焼け した パンツ を 着た 男性 と 、 膝 ブレース を 着た 女性 。
|
60 |
+
sentences:
|
61 |
+
- 二 人 が 踊って い ます 。
|
62 |
+
- バック グラウンド に 空 の フィールド が あり ます 。
|
63 |
+
- 3 人 の 若い 女の子 が そこ に 向かって いる 噴水 が あり ます 。
|
64 |
+
- source_sentence: 3 人 の 子供 が 映画 を 見て い ます 。
|
65 |
+
sentences:
|
66 |
+
- で歩いて いる 女性 を 見て 食品 ベンダー
|
67 |
+
- スケート ボード に 乗った 男 は 、 歓声 を 上げる 観客 に 手 の 込んだ ジャンプ を し ます 。
|
68 |
+
- 素晴らしい 映画 を 見て いる 人間 。
|
69 |
+
- source_sentence: 2 人 の 消防 士 が 消防 車 に 向かって 歩いて い ます 。
|
70 |
+
sentences:
|
71 |
+
- 2 人 の 消防 士 が トラック に 行く
|
72 |
+
- 男 は 彼 の ne の 盗難 を 非難 し ます 。
|
73 |
+
- 岩 だらけ の 海岸 を 歩いて いる と 、 波 を 見て いる 男性 が い ます 。
|
74 |
+
model-index:
|
75 |
+
- name: SentenceTransformer based on colorfulscoop/sbert-base-ja
|
76 |
+
results:
|
77 |
+
- task:
|
78 |
+
type: binary-classification
|
79 |
+
name: Binary Classification
|
80 |
+
dataset:
|
81 |
+
name: custom arc semantics data jp
|
82 |
+
type: custom-arc-semantics-data-jp
|
83 |
+
metrics:
|
84 |
+
- type: cosine_accuracy
|
85 |
+
value: 0.7196998123827392
|
86 |
+
name: Cosine Accuracy
|
87 |
+
- type: cosine_accuracy_threshold
|
88 |
+
value: 0.955563485622406
|
89 |
+
name: Cosine Accuracy Threshold
|
90 |
+
- type: cosine_f1
|
91 |
+
value: 0.8188111721174504
|
92 |
+
name: Cosine F1
|
93 |
+
- type: cosine_f1_threshold
|
94 |
+
value: 0.9209792017936707
|
95 |
+
name: Cosine F1 Threshold
|
96 |
+
- type: cosine_precision
|
97 |
+
value: 0.7104391052195526
|
98 |
+
name: Cosine Precision
|
99 |
+
- type: cosine_recall
|
100 |
+
value: 0.9661971830985916
|
101 |
+
name: Cosine Recall
|
102 |
+
- type: cosine_ap
|
103 |
+
value: 0.7547086304231496
|
104 |
+
name: Cosine Ap
|
105 |
+
- type: dot_accuracy
|
106 |
+
value: 0.7204502814258912
|
107 |
+
name: Dot Accuracy
|
108 |
+
- type: dot_accuracy_threshold
|
109 |
+
value: 611.2391357421875
|
110 |
+
name: Dot Accuracy Threshold
|
111 |
+
- type: dot_f1
|
112 |
+
value: 0.818549346016647
|
113 |
+
name: Dot F1
|
114 |
+
- type: dot_f1_threshold
|
115 |
+
value: 588.0654296875
|
116 |
+
name: Dot F1 Threshold
|
117 |
+
- type: dot_precision
|
118 |
+
value: 0.7082304526748971
|
119 |
+
name: Dot Precision
|
120 |
+
- type: dot_recall
|
121 |
+
value: 0.9695774647887324
|
122 |
+
name: Dot Recall
|
123 |
+
- type: dot_ap
|
124 |
+
value: 0.7795651338156695
|
125 |
+
name: Dot Ap
|
126 |
+
- type: manhattan_accuracy
|
127 |
+
value: 0.7193245778611632
|
128 |
+
name: Manhattan Accuracy
|
129 |
+
- type: manhattan_accuracy_threshold
|
130 |
+
value: 177.08966064453125
|
131 |
+
name: Manhattan Accuracy Threshold
|
132 |
+
- type: manhattan_f1
|
133 |
+
value: 0.8182469548602818
|
134 |
+
name: Manhattan F1
|
135 |
+
- type: manhattan_f1_threshold
|
136 |
+
value: 223.81216430664062
|
137 |
+
name: Manhattan F1 Threshold
|
138 |
+
- type: manhattan_precision
|
139 |
+
value: 0.7101990049751243
|
140 |
+
name: Manhattan Precision
|
141 |
+
- type: manhattan_recall
|
142 |
+
value: 0.9650704225352112
|
143 |
+
name: Manhattan Recall
|
144 |
+
- type: manhattan_ap
|
145 |
+
value: 0.7546997956004545
|
146 |
+
name: Manhattan Ap
|
147 |
+
- type: euclidean_accuracy
|
148 |
+
value: 0.7196998123827392
|
149 |
+
name: Euclidean Accuracy
|
150 |
+
- type: euclidean_accuracy_threshold
|
151 |
+
value: 7.550699710845947
|
152 |
+
name: Euclidean Accuracy Threshold
|
153 |
+
- type: euclidean_f1
|
154 |
+
value: 0.8188111721174504
|
155 |
+
name: Euclidean F1
|
156 |
+
- type: euclidean_f1_threshold
|
157 |
+
value: 10.068297386169434
|
158 |
+
name: Euclidean F1 Threshold
|
159 |
+
- type: euclidean_precision
|
160 |
+
value: 0.7104391052195526
|
161 |
+
name: Euclidean Precision
|
162 |
+
- type: euclidean_recall
|
163 |
+
value: 0.9661971830985916
|
164 |
+
name: Euclidean Recall
|
165 |
+
- type: euclidean_ap
|
166 |
+
value: 0.7545919416671392
|
167 |
+
name: Euclidean Ap
|
168 |
+
- type: max_accuracy
|
169 |
+
value: 0.7204502814258912
|
170 |
+
name: Max Accuracy
|
171 |
+
- type: max_accuracy_threshold
|
172 |
+
value: 611.2391357421875
|
173 |
+
name: Max Accuracy Threshold
|
174 |
+
- type: max_f1
|
175 |
+
value: 0.8188111721174504
|
176 |
+
name: Max F1
|
177 |
+
- type: max_f1_threshold
|
178 |
+
value: 588.0654296875
|
179 |
+
name: Max F1 Threshold
|
180 |
+
- type: max_precision
|
181 |
+
value: 0.7104391052195526
|
182 |
+
name: Max Precision
|
183 |
+
- type: max_recall
|
184 |
+
value: 0.9695774647887324
|
185 |
+
name: Max Recall
|
186 |
+
- type: max_ap
|
187 |
+
value: 0.7795651338156695
|
188 |
+
name: Max Ap
|
189 |
---
|
190 |
|
191 |
+
# SentenceTransformer based on colorfulscoop/sbert-base-ja
|
|
|
|
|
|
|
192 |
|
193 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [colorfulscoop/sbert-base-ja](https://huggingface.co/colorfulscoop/sbert-base-ja) on the csv dataset. 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.
|
194 |
|
195 |
## Model Details
|
196 |
|
197 |
### Model Description
|
198 |
+
- **Model Type:** Sentence Transformer
|
199 |
+
- **Base model:** [colorfulscoop/sbert-base-ja](https://huggingface.co/colorfulscoop/sbert-base-ja) <!-- at revision ecb8a98cd5176719ff7ab0d770a27420118732cf -->
|
200 |
+
- **Maximum Sequence Length:** 512 tokens
|
201 |
+
- **Output Dimensionality:** 768 tokens
|
202 |
+
- **Similarity Function:** Cosine Similarity
|
203 |
+
- **Training Dataset:**
|
204 |
+
- csv
|
205 |
+
<!-- - **Language:** Unknown -->
|
206 |
+
<!-- - **License:** Unknown -->
|
207 |
|
208 |
+
### Model Sources
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
209 |
|
210 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
211 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
212 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
213 |
|
214 |
+
### Full Model Architecture
|
215 |
|
216 |
+
```
|
217 |
+
SentenceTransformer(
|
218 |
+
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
|
219 |
+
(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})
|
220 |
+
)
|
221 |
+
```
|
222 |
|
223 |
+
## Usage
|
224 |
|
225 |
+
### Direct Usage (Sentence Transformers)
|
226 |
|
227 |
+
First install the Sentence Transformers library:
|
228 |
|
229 |
+
```bash
|
230 |
+
pip install -U sentence-transformers
|
231 |
+
```
|
|
|
|
|
232 |
|
233 |
+
Then you can load this model and run inference.
|
234 |
+
```python
|
235 |
+
from sentence_transformers import SentenceTransformer
|
236 |
|
237 |
+
# Download from the 🤗 Hub
|
238 |
+
model = SentenceTransformer("sentence_transformers_model_id")
|
239 |
+
# Run inference
|
240 |
+
sentences = [
|
241 |
+
'2 人 の 消防 士 が 消防 車 に 向かって 歩いて い ます 。',
|
242 |
+
'2 人 の 消防 士 が トラック に 行く',
|
243 |
+
'岩 だらけ の 海岸 を 歩いて いる と 、 波 を 見て いる 男性 が い ます 。',
|
244 |
+
]
|
245 |
+
embeddings = model.encode(sentences)
|
246 |
+
print(embeddings.shape)
|
247 |
+
# [3, 768]
|
248 |
|
249 |
+
# Get the similarity scores for the embeddings
|
250 |
+
similarities = model.similarity(embeddings, embeddings)
|
251 |
+
print(similarities.shape)
|
252 |
+
# [3, 3]
|
253 |
+
```
|
254 |
|
255 |
+
<!--
|
256 |
+
### Direct Usage (Transformers)
|
257 |
|
258 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
259 |
|
260 |
+
</details>
|
261 |
+
-->
|
|
|
|
|
|
|
|
|
|
|
262 |
|
263 |
+
<!--
|
264 |
+
### Downstream Usage (Sentence Transformers)
|
265 |
|
266 |
+
You can finetune this model on your own dataset.
|
267 |
|
268 |
+
<details><summary>Click to expand</summary>
|
269 |
|
270 |
+
</details>
|
271 |
+
-->
|
272 |
|
273 |
+
<!--
|
274 |
+
### Out-of-Scope Use
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
275 |
|
276 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
277 |
+
-->
|
278 |
|
279 |
## Evaluation
|
280 |
|
281 |
+
### Metrics
|
282 |
+
|
283 |
+
#### Binary Classification
|
284 |
+
* Dataset: `custom-arc-semantics-data-jp`
|
285 |
+
* Evaluated with [<code>BinaryClassificationEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.BinaryClassificationEvaluator)
|
286 |
+
|
287 |
+
| Metric | Value |
|
288 |
+
|:-----------------------------|:-----------|
|
289 |
+
| cosine_accuracy | 0.7197 |
|
290 |
+
| cosine_accuracy_threshold | 0.9556 |
|
291 |
+
| cosine_f1 | 0.8188 |
|
292 |
+
| cosine_f1_threshold | 0.921 |
|
293 |
+
| cosine_precision | 0.7104 |
|
294 |
+
| cosine_recall | 0.9662 |
|
295 |
+
| cosine_ap | 0.7547 |
|
296 |
+
| dot_accuracy | 0.7205 |
|
297 |
+
| dot_accuracy_threshold | 611.2391 |
|
298 |
+
| dot_f1 | 0.8185 |
|
299 |
+
| dot_f1_threshold | 588.0654 |
|
300 |
+
| dot_precision | 0.7082 |
|
301 |
+
| dot_recall | 0.9696 |
|
302 |
+
| dot_ap | 0.7796 |
|
303 |
+
| manhattan_accuracy | 0.7193 |
|
304 |
+
| manhattan_accuracy_threshold | 177.0897 |
|
305 |
+
| manhattan_f1 | 0.8182 |
|
306 |
+
| manhattan_f1_threshold | 223.8122 |
|
307 |
+
| manhattan_precision | 0.7102 |
|
308 |
+
| manhattan_recall | 0.9651 |
|
309 |
+
| manhattan_ap | 0.7547 |
|
310 |
+
| euclidean_accuracy | 0.7197 |
|
311 |
+
| euclidean_accuracy_threshold | 7.5507 |
|
312 |
+
| euclidean_f1 | 0.8188 |
|
313 |
+
| euclidean_f1_threshold | 10.0683 |
|
314 |
+
| euclidean_precision | 0.7104 |
|
315 |
+
| euclidean_recall | 0.9662 |
|
316 |
+
| euclidean_ap | 0.7546 |
|
317 |
+
| max_accuracy | 0.7205 |
|
318 |
+
| max_accuracy_threshold | 611.2391 |
|
319 |
+
| max_f1 | 0.8188 |
|
320 |
+
| max_f1_threshold | 588.0654 |
|
321 |
+
| max_precision | 0.7104 |
|
322 |
+
| max_recall | 0.9696 |
|
323 |
+
| **max_ap** | **0.7796** |
|
324 |
+
|
325 |
+
<!--
|
326 |
+
## Bias, Risks and Limitations
|
327 |
+
|
328 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
329 |
+
-->
|
330 |
+
|
331 |
+
<!--
|
332 |
+
### Recommendations
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
333 |
|
334 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
335 |
+
-->
|
336 |
|
337 |
+
## Training Details
|
338 |
|
339 |
+
### Training Dataset
|
340 |
+
|
341 |
+
#### csv
|
342 |
+
|
343 |
+
* Dataset: csv
|
344 |
+
* Size: 5,330 training samples
|
345 |
+
* Columns: <code>text1</code>, <code>text2</code>, and <code>label</code>
|
346 |
+
* Approximate statistics based on the first 1000 samples:
|
347 |
+
| | text1 | text2 | label |
|
348 |
+
|:--------|:------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:------------------------------------------------|
|
349 |
+
| type | string | string | int |
|
350 |
+
| details | <ul><li>min: 12 tokens</li><li>mean: 35.93 tokens</li><li>max: 102 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 22.72 tokens</li><li>max: 66 tokens</li></ul> | <ul><li>0: ~32.50%</li><li>1: ~67.50%</li></ul> |
|
351 |
+
* Samples:
|
352 |
+
| text1 | text2 | label |
|
353 |
+
|:------------------------------------------------------------------------------|:---------------------------------------------------------------|:---------------|
|
354 |
+
| <code>明るい 茶色 の 犬 は 、 フェンス で 囲ま れた エリア で 赤 茶色 の 犬 と ボール を プレー して い ます 。</code> | <code>犬 は フェンス で 囲ま れた エリア で ボール を プレー し ます</code> | <code>0</code> |
|
355 |
+
| <code>青い 服 を 着た 人 が バイク に 乗って い ます 。</code> | <code>その 人 は 、 地元 の 大学 で 授業 を 受ける ため に バイク に 乗って い ます 。</code> | <code>1</code> |
|
356 |
+
| <code>真っ白な 女性 が ピンク と 紫 の 花 の バナー の 前 に 買い物 袋 を 持って 立って い ます 。</code> | <code>女性 は フットボール の 試合 で すべて 赤 を 着て い ます 。</code> | <code>1</code> |
|
357 |
+
* Loss: [<code>CoSENTLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
|
358 |
+
```json
|
359 |
+
{
|
360 |
+
"scale": 20.0,
|
361 |
+
"similarity_fct": "pairwise_cos_sim"
|
362 |
+
}
|
363 |
+
```
|
364 |
+
|
365 |
+
### Evaluation Dataset
|
366 |
+
|
367 |
+
#### csv
|
368 |
+
|
369 |
+
* Dataset: csv
|
370 |
+
* Size: 5,330 evaluation samples
|
371 |
+
* Columns: <code>text1</code>, <code>text2</code>, and <code>label</code>
|
372 |
+
* Approximate statistics based on the first 1000 samples:
|
373 |
+
| | text1 | text2 | label |
|
374 |
+
|:--------|:------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:------------------------------------------------|
|
375 |
+
| type | string | string | int |
|
376 |
+
| details | <ul><li>min: 11 tokens</li><li>mean: 35.77 tokens</li><li>max: 109 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 22.73 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>0: ~32.60%</li><li>1: ~67.40%</li></ul> |
|
377 |
+
* Samples:
|
378 |
+
| text1 | text2 | label |
|
379 |
+
|:----------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------|:---------------|
|
380 |
+
| <code>眼鏡 を かけた 男 は 、 「 FREE WORD 」 と いう 標識 を 持って い ます 。</code> | <code>人 は 眼鏡 を 外して い ます 。</code> | <code>1</code> |
|
381 |
+
| <code>麦わら 帽子 を かぶった 2 人 の 男 、 青い シャツ 、 日焼け した 帽子 の 男 、 赤い 格子 縞 の シャツ 、 両方 と も 前 に バスケット を 持って 、 未 舗装 の 道路 の 脇 に 座って い ます 。</code> | <code>帽子 を かぶった 2 人 の 男 が 道路 の 脇 に 座って いる</code> | <code>0</code> |
|
382 |
+
| <code>傘 を 持って 歩く 女性 。</code> | <code>女性 が 歩いて い ます 。</code> | <code>0</code> |
|
383 |
+
* Loss: [<code>CoSENTLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
|
384 |
+
```json
|
385 |
+
{
|
386 |
+
"scale": 20.0,
|
387 |
+
"similarity_fct": "pairwise_cos_sim"
|
388 |
+
}
|
389 |
+
```
|
390 |
+
|
391 |
+
### Training Hyperparameters
|
392 |
+
#### Non-Default Hyperparameters
|
393 |
+
|
394 |
+
- `eval_strategy`: epoch
|
395 |
+
- `learning_rate`: 2e-05
|
396 |
+
- `num_train_epochs`: 1
|
397 |
+
- `warmup_ratio`: 0.4
|
398 |
+
- `fp16`: True
|
399 |
+
- `batch_sampler`: no_duplicates
|
400 |
+
|
401 |
+
#### All Hyperparameters
|
402 |
+
<details><summary>Click to expand</summary>
|
403 |
+
|
404 |
+
- `overwrite_output_dir`: False
|
405 |
+
- `do_predict`: False
|
406 |
+
- `eval_strategy`: epoch
|
407 |
+
- `prediction_loss_only`: True
|
408 |
+
- `per_device_train_batch_size`: 8
|
409 |
+
- `per_device_eval_batch_size`: 8
|
410 |
+
- `per_gpu_train_batch_size`: None
|
411 |
+
- `per_gpu_eval_batch_size`: None
|
412 |
+
- `gradient_accumulation_steps`: 1
|
413 |
+
- `eval_accumulation_steps`: None
|
414 |
+
- `torch_empty_cache_steps`: None
|
415 |
+
- `learning_rate`: 2e-05
|
416 |
+
- `weight_decay`: 0.0
|
417 |
+
- `adam_beta1`: 0.9
|
418 |
+
- `adam_beta2`: 0.999
|
419 |
+
- `adam_epsilon`: 1e-08
|
420 |
+
- `max_grad_norm`: 1.0
|
421 |
+
- `num_train_epochs`: 1
|
422 |
+
- `max_steps`: -1
|
423 |
+
- `lr_scheduler_type`: linear
|
424 |
+
- `lr_scheduler_kwargs`: {}
|
425 |
+
- `warmup_ratio`: 0.4
|
426 |
+
- `warmup_steps`: 0
|
427 |
+
- `log_level`: passive
|
428 |
+
- `log_level_replica`: warning
|
429 |
+
- `log_on_each_node`: True
|
430 |
+
- `logging_nan_inf_filter`: True
|
431 |
+
- `save_safetensors`: True
|
432 |
+
- `save_on_each_node`: False
|
433 |
+
- `save_only_model`: False
|
434 |
+
- `restore_callback_states_from_checkpoint`: False
|
435 |
+
- `no_cuda`: False
|
436 |
+
- `use_cpu`: False
|
437 |
+
- `use_mps_device`: False
|
438 |
+
- `seed`: 42
|
439 |
+
- `data_seed`: None
|
440 |
+
- `jit_mode_eval`: False
|
441 |
+
- `use_ipex`: False
|
442 |
+
- `bf16`: False
|
443 |
+
- `fp16`: True
|
444 |
+
- `fp16_opt_level`: O1
|
445 |
+
- `half_precision_backend`: auto
|
446 |
+
- `bf16_full_eval`: False
|
447 |
+
- `fp16_full_eval`: False
|
448 |
+
- `tf32`: None
|
449 |
+
- `local_rank`: 0
|
450 |
+
- `ddp_backend`: None
|
451 |
+
- `tpu_num_cores`: None
|
452 |
+
- `tpu_metrics_debug`: False
|
453 |
+
- `debug`: []
|
454 |
+
- `dataloader_drop_last`: False
|
455 |
+
- `dataloader_num_workers`: 0
|
456 |
+
- `dataloader_prefetch_factor`: None
|
457 |
+
- `past_index`: -1
|
458 |
+
- `disable_tqdm`: False
|
459 |
+
- `remove_unused_columns`: True
|
460 |
+
- `label_names`: None
|
461 |
+
- `load_best_model_at_end`: False
|
462 |
+
- `ignore_data_skip`: False
|
463 |
+
- `fsdp`: []
|
464 |
+
- `fsdp_min_num_params`: 0
|
465 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
466 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
467 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
|
468 |
+
- `deepspeed`: None
|
469 |
+
- `label_smoothing_factor`: 0.0
|
470 |
+
- `optim`: adamw_torch
|
471 |
+
- `optim_args`: None
|
472 |
+
- `adafactor`: False
|
473 |
+
- `group_by_length`: False
|
474 |
+
- `length_column_name`: length
|
475 |
+
- `ddp_find_unused_parameters`: None
|
476 |
+
- `ddp_bucket_cap_mb`: None
|
477 |
+
- `ddp_broadcast_buffers`: False
|
478 |
+
- `dataloader_pin_memory`: True
|
479 |
+
- `dataloader_persistent_workers`: False
|
480 |
+
- `skip_memory_metrics`: True
|
481 |
+
- `use_legacy_prediction_loop`: False
|
482 |
+
- `push_to_hub`: False
|
483 |
+
- `resume_from_checkpoint`: None
|
484 |
+
- `hub_model_id`: None
|
485 |
+
- `hub_strategy`: every_save
|
486 |
+
- `hub_private_repo`: False
|
487 |
+
- `hub_always_push`: False
|
488 |
+
- `gradient_checkpointing`: False
|
489 |
+
- `gradient_checkpointing_kwargs`: None
|
490 |
+
- `include_inputs_for_metrics`: False
|
491 |
+
- `eval_do_concat_batches`: True
|
492 |
+
- `fp16_backend`: auto
|
493 |
+
- `push_to_hub_model_id`: None
|
494 |
+
- `push_to_hub_organization`: None
|
495 |
+
- `mp_parameters`:
|
496 |
+
- `auto_find_batch_size`: False
|
497 |
+
- `full_determinism`: False
|
498 |
+
- `torchdynamo`: None
|
499 |
+
- `ray_scope`: last
|
500 |
+
- `ddp_timeout`: 1800
|
501 |
+
- `torch_compile`: False
|
502 |
+
- `torch_compile_backend`: None
|
503 |
+
- `torch_compile_mode`: None
|
504 |
+
- `dispatch_batches`: None
|
505 |
+
- `split_batches`: None
|
506 |
+
- `include_tokens_per_second`: False
|
507 |
+
- `include_num_input_tokens_seen`: False
|
508 |
+
- `neftune_noise_alpha`: None
|
509 |
+
- `optim_target_modules`: None
|
510 |
+
- `batch_eval_metrics`: False
|
511 |
+
- `eval_on_start`: False
|
512 |
+
- `eval_use_gather_object`: False
|
513 |
+
- `batch_sampler`: no_duplicates
|
514 |
+
- `multi_dataset_batch_sampler`: proportional
|
515 |
+
|
516 |
+
</details>
|
517 |
+
|
518 |
+
### Training Logs
|
519 |
+
| Epoch | Step | Training Loss | loss | custom-arc-semantics-data-jp_max_ap |
|
520 |
+
|:-----:|:----:|:-------------:|:------:|:-----------------------------------:|
|
521 |
+
| 1.0 | 334 | 3.415 | 2.3727 | 0.7796 |
|
522 |
+
|
523 |
+
|
524 |
+
### Framework Versions
|
525 |
+
- Python: 3.10.14
|
526 |
+
- Sentence Transformers: 3.1.0
|
527 |
+
- Transformers: 4.44.2
|
528 |
+
- PyTorch: 2.4.1+cu121
|
529 |
+
- Accelerate: 0.34.2
|
530 |
+
- Datasets: 2.20.0
|
531 |
+
- Tokenizers: 0.19.1
|
532 |
+
|
533 |
+
## Citation
|
534 |
+
|
535 |
+
### BibTeX
|
536 |
+
|
537 |
+
#### Sentence Transformers
|
538 |
+
```bibtex
|
539 |
+
@inproceedings{reimers-2019-sentence-bert,
|
540 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
541 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
542 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
543 |
+
month = "11",
|
544 |
+
year = "2019",
|
545 |
+
publisher = "Association for Computational Linguistics",
|
546 |
+
url = "https://arxiv.org/abs/1908.10084",
|
547 |
+
}
|
548 |
+
```
|
549 |
+
|
550 |
+
#### CoSENTLoss
|
551 |
+
```bibtex
|
552 |
+
@online{kexuefm-8847,
|
553 |
+
title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
|
554 |
+
author={Su Jianlin},
|
555 |
+
year={2022},
|
556 |
+
month={Jan},
|
557 |
+
url={https://kexue.fm/archives/8847},
|
558 |
+
}
|
559 |
+
```
|
560 |
+
|
561 |
+
<!--
|
562 |
+
## Glossary
|
563 |
+
|
564 |
+
*Clearly define terms in order to be accessible across audiences.*
|
565 |
+
-->
|
566 |
+
|
567 |
+
<!--
|
568 |
+
## Model Card Authors
|
569 |
+
|
570 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
571 |
+
-->
|
572 |
+
|
573 |
+
<!--
|
574 |
## Model Card Contact
|
575 |
|
576 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
577 |
+
-->
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 442491744
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4eb1a02b8262293fcf8abdf762d4bf31c669f3a846a8dc2d230dcfe6159a4bfa
|
3 |
size 442491744
|
runs/Sep17_02-56-11_default/events.out.tfevents.1726541772.default.7712.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:73a3b1e96b15e121a71ec7fcd387bbc7d80a8ec351b5afcc88ac597306ab9761
|
3 |
+
size 8199
|