Add new SentenceTransformer model.
Browse files- README.md +41 -109
- config.json +1 -1
- model.safetensors +1 -1
README.md
CHANGED
@@ -6,42 +6,43 @@ tags:
|
|
6 |
- sentence-similarity
|
7 |
- feature-extraction
|
8 |
- generated_from_trainer
|
9 |
-
- dataset_size:
|
10 |
-
- loss:
|
11 |
base_model: sentence-transformers/all-MiniLM-L6-v2
|
12 |
datasets: []
|
13 |
widget:
|
14 |
-
- source_sentence:
|
15 |
-
with NOAA and SARSAT.
|
16 |
sentences:
|
17 |
-
-
|
18 |
-
-
|
19 |
-
|
20 |
-
-
|
|
|
|
|
21 |
sentences:
|
22 |
-
-
|
23 |
-
|
24 |
-
-
|
25 |
-
|
|
|
|
|
|
|
26 |
sentences:
|
27 |
-
-
|
28 |
-
|
29 |
-
-
|
30 |
-
|
31 |
-
|
32 |
-
Rules 12, 16, and 40.
|
33 |
sentences:
|
34 |
-
-
|
35 |
-
|
36 |
-
-
|
37 |
-
|
38 |
-
- source_sentence: The publication was named after Sir James Joynton Smith.
|
39 |
sentences:
|
40 |
-
-
|
41 |
-
|
42 |
-
|
43 |
-
-
|
44 |
-
Smith.
|
45 |
pipeline_tag: sentence-similarity
|
46 |
---
|
47 |
|
@@ -95,9 +96,9 @@ from sentence_transformers import SentenceTransformer
|
|
95 |
model = SentenceTransformer("LeoChiuu/all-MiniLM-L6-v2-negations")
|
96 |
# Run inference
|
97 |
sentences = [
|
98 |
-
'The
|
99 |
-
'
|
100 |
-
|
101 |
]
|
102 |
embeddings = model.encode(sentences)
|
103 |
print(embeddings.shape)
|
@@ -152,25 +153,23 @@ You can finetune this model on your own dataset.
|
|
152 |
#### Unnamed Dataset
|
153 |
|
154 |
|
155 |
-
* Size:
|
156 |
* Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
|
157 |
* Approximate statistics based on the first 1000 samples:
|
158 |
| | sentence_0 | sentence_1 | label |
|
159 |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:------------------------------------------------|
|
160 |
| type | string | string | int |
|
161 |
-
| details | <ul><li>min:
|
162 |
* Samples:
|
163 |
-
| sentence_0
|
164 |
-
|
165 |
-
| <code>
|
166 |
-
| <code>
|
167 |
-
| <code>
|
168 |
-
* Loss: [<code>
|
169 |
```json
|
170 |
{
|
171 |
-
"
|
172 |
-
"margin": 0.5,
|
173 |
-
"size_average": true
|
174 |
}
|
175 |
```
|
176 |
|
@@ -293,59 +292,6 @@ You can finetune this model on your own dataset.
|
|
293 |
|
294 |
</details>
|
295 |
|
296 |
-
### Training Logs
|
297 |
-
| Epoch | Step | Training Loss |
|
298 |
-
|:------:|:-----:|:-------------:|
|
299 |
-
| 0.2068 | 500 | 0.0353 |
|
300 |
-
| 0.4136 | 1000 | 0.0307 |
|
301 |
-
| 0.6203 | 1500 | 0.0234 |
|
302 |
-
| 0.8271 | 2000 | 0.0187 |
|
303 |
-
| 1.0339 | 2500 | 0.0152 |
|
304 |
-
| 1.2407 | 3000 | 0.0134 |
|
305 |
-
| 1.4475 | 3500 | 0.0123 |
|
306 |
-
| 1.6543 | 4000 | 0.0111 |
|
307 |
-
| 1.8610 | 4500 | 0.0107 |
|
308 |
-
| 2.0678 | 5000 | 0.0097 |
|
309 |
-
| 2.2746 | 5500 | 0.0096 |
|
310 |
-
| 2.4814 | 6000 | 0.0091 |
|
311 |
-
| 2.6882 | 6500 | 0.0087 |
|
312 |
-
| 2.8950 | 7000 | 0.0086 |
|
313 |
-
| 3.1017 | 7500 | 0.0075 |
|
314 |
-
| 3.3085 | 8000 | 0.008 |
|
315 |
-
| 3.5153 | 8500 | 0.0074 |
|
316 |
-
| 3.7221 | 9000 | 0.007 |
|
317 |
-
| 3.9289 | 9500 | 0.007 |
|
318 |
-
| 4.1356 | 10000 | 0.0063 |
|
319 |
-
| 4.3424 | 10500 | 0.0068 |
|
320 |
-
| 4.5492 | 11000 | 0.0061 |
|
321 |
-
| 4.7560 | 11500 | 0.0059 |
|
322 |
-
| 4.9628 | 12000 | 0.0056 |
|
323 |
-
| 5.1696 | 12500 | 0.0052 |
|
324 |
-
| 5.3763 | 13000 | 0.0055 |
|
325 |
-
| 5.5831 | 13500 | 0.0051 |
|
326 |
-
| 5.7899 | 14000 | 0.005 |
|
327 |
-
| 5.9967 | 14500 | 0.0047 |
|
328 |
-
| 6.2035 | 15000 | 0.0046 |
|
329 |
-
| 6.4103 | 15500 | 0.0047 |
|
330 |
-
| 6.6170 | 16000 | 0.0044 |
|
331 |
-
| 6.8238 | 16500 | 0.0044 |
|
332 |
-
| 7.0306 | 17000 | 0.0041 |
|
333 |
-
| 7.2374 | 17500 | 0.004 |
|
334 |
-
| 7.4442 | 18000 | 0.0044 |
|
335 |
-
| 7.6510 | 18500 | 0.0039 |
|
336 |
-
| 7.8577 | 19000 | 0.0038 |
|
337 |
-
| 8.0645 | 19500 | 0.0038 |
|
338 |
-
| 8.2713 | 20000 | 0.0037 |
|
339 |
-
| 8.4781 | 20500 | 0.0039 |
|
340 |
-
| 8.6849 | 21000 | 0.0037 |
|
341 |
-
| 8.8916 | 21500 | 0.0036 |
|
342 |
-
| 9.0984 | 22000 | 0.0034 |
|
343 |
-
| 9.3052 | 22500 | 0.0036 |
|
344 |
-
| 9.5120 | 23000 | 0.0035 |
|
345 |
-
| 9.7188 | 23500 | 0.0034 |
|
346 |
-
| 9.9256 | 24000 | 0.0035 |
|
347 |
-
|
348 |
-
|
349 |
### Framework Versions
|
350 |
- Python: 3.11.9
|
351 |
- Sentence Transformers: 3.0.1
|
@@ -372,20 +318,6 @@ You can finetune this model on your own dataset.
|
|
372 |
}
|
373 |
```
|
374 |
|
375 |
-
#### ContrastiveLoss
|
376 |
-
```bibtex
|
377 |
-
@inproceedings{hadsell2006dimensionality,
|
378 |
-
author={Hadsell, R. and Chopra, S. and LeCun, Y.},
|
379 |
-
booktitle={2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)},
|
380 |
-
title={Dimensionality Reduction by Learning an Invariant Mapping},
|
381 |
-
year={2006},
|
382 |
-
volume={2},
|
383 |
-
number={},
|
384 |
-
pages={1735-1742},
|
385 |
-
doi={10.1109/CVPR.2006.100}
|
386 |
-
}
|
387 |
-
```
|
388 |
-
|
389 |
<!--
|
390 |
## Glossary
|
391 |
|
|
|
6 |
- sentence-similarity
|
7 |
- feature-extraction
|
8 |
- generated_from_trainer
|
9 |
+
- dataset_size:75
|
10 |
+
- loss:CosineSimilarityLoss
|
11 |
base_model: sentence-transformers/all-MiniLM-L6-v2
|
12 |
datasets: []
|
13 |
widget:
|
14 |
+
- source_sentence: This store featured in the SavaCentre TV adverts in 1983.
|
|
|
15 |
sentences:
|
16 |
+
- I love the Scream movies and all horror movies and this one ranks way up there.
|
17 |
+
- Development of synchronous toothed-belts was halted by the Gilmer company prior
|
18 |
+
to 1940.
|
19 |
+
- This store was not featured in the SavaCentre TV promotions in 1983.
|
20 |
+
- source_sentence: In 2014, Nextgen earns KLAS Top Performance Honors for Ambulatory
|
21 |
+
RCM Services.
|
22 |
sentences:
|
23 |
+
- These strategies employ reporter transposon s and in vitro expression technology
|
24 |
+
(IVET).
|
25 |
+
- In 2014, Nextgen fails to achieve KLAS Top Performance Honors for Ambulatory RCM
|
26 |
+
Services.
|
27 |
+
- The film's sole bright spot was Jonah Hill (who will look almost unrecognizable
|
28 |
+
to fans of the recent Superbad due to the amount of weight he lost in the interim).
|
29 |
+
- source_sentence: E105 has never been implicated in atopic asthma.
|
30 |
sentences:
|
31 |
+
- E105 has been implicated in non-atopic asthma.
|
32 |
+
- The species is named in honor of the divorce of Sara Anderson and Malcolm Slaney.
|
33 |
+
- Each annex to a filed document is not required to have page numbering.
|
34 |
+
- source_sentence: Additionally, a church at San Lazaro in Orange Walk District escaped
|
35 |
+
all damage.
|
|
|
36 |
sentences:
|
37 |
+
- Kuwait has a reputation for being the central music influence of the GCC countries.
|
38 |
+
- Early settlers may have introduced it 4,000 years ago.
|
39 |
+
- Additionally, a church at San Lazaro in Orange Walk District suffered severe damage.
|
40 |
+
- source_sentence: The content in Australia is lower than in other reports.
|
|
|
41 |
sentences:
|
42 |
+
- Other reports also show a content lower than 0.1% in Australia.
|
43 |
+
- Commercial DNP is unable to be utilized as an antiseptic or as a non-selective
|
44 |
+
bioaccumulating pesticide.
|
45 |
+
- Installation of Halon systems is mandated by the European Union.
|
|
|
46 |
pipeline_tag: sentence-similarity
|
47 |
---
|
48 |
|
|
|
96 |
model = SentenceTransformer("LeoChiuu/all-MiniLM-L6-v2-negations")
|
97 |
# Run inference
|
98 |
sentences = [
|
99 |
+
'The content in Australia is lower than in other reports.',
|
100 |
+
'Other reports also show a content lower than 0.1% in Australia.',
|
101 |
+
'Installation of Halon systems is mandated by the European Union.',
|
102 |
]
|
103 |
embeddings = model.encode(sentences)
|
104 |
print(embeddings.shape)
|
|
|
153 |
#### Unnamed Dataset
|
154 |
|
155 |
|
156 |
+
* Size: 75 training samples
|
157 |
* Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
|
158 |
* Approximate statistics based on the first 1000 samples:
|
159 |
| | sentence_0 | sentence_1 | label |
|
160 |
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:------------------------------------------------|
|
161 |
| type | string | string | int |
|
162 |
+
| details | <ul><li>min: 9 tokens</li><li>mean: 16.36 tokens</li><li>max: 39 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 16.55 tokens</li><li>max: 43 tokens</li></ul> | <ul><li>0: ~61.33%</li><li>1: ~38.67%</li></ul> |
|
163 |
* Samples:
|
164 |
+
| sentence_0 | sentence_1 | label |
|
165 |
+
|:---------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------|:---------------|
|
166 |
+
| <code>It wasn't an inexpensive piece, but I would still have expected better quality.</code> | <code>It was an inexpensive piece, but I would still have expected better quality.</code> | <code>0</code> |
|
167 |
+
| <code>My name is noncrucial.</code> | <code>My name is important.</code> | <code>0</code> |
|
168 |
+
| <code>Hawthorne mostly wrote against his own religious belief.</code> | <code>Hawthorne wrote against his beliefs.</code> | <code>1</code> |
|
169 |
+
* Loss: [<code>CosineSimilarityLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosinesimilarityloss) with these parameters:
|
170 |
```json
|
171 |
{
|
172 |
+
"loss_fct": "torch.nn.modules.loss.MSELoss"
|
|
|
|
|
173 |
}
|
174 |
```
|
175 |
|
|
|
292 |
|
293 |
</details>
|
294 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
295 |
### Framework Versions
|
296 |
- Python: 3.11.9
|
297 |
- Sentence Transformers: 3.0.1
|
|
|
318 |
}
|
319 |
```
|
320 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
321 |
<!--
|
322 |
## Glossary
|
323 |
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"BertModel"
|
5 |
],
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "sentence-transformers/all-MiniLM-L6-v2",
|
3 |
"architectures": [
|
4 |
"BertModel"
|
5 |
],
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 90864192
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2b3f93fc93c0fbdf4be9f9217841543915515a6610212538520a608457a9d4a7
|
3 |
size 90864192
|