Yohhei commited on
Commit
e04039e
·
verified ·
1 Parent(s): f059177

Add new SentenceTransformer model.

Browse files
Files changed (2) hide show
  1. README.md +36 -40
  2. model.safetensors +1 -1
README.md CHANGED
@@ -65,49 +65,49 @@ model-index:
65
  type: intfloat/multilingual-e5-small
66
  metrics:
67
  - type: cosine_accuracy@1
68
- value: 0.488156820038116
69
  name: Cosine Accuracy@1
70
  - type: cosine_accuracy@3
71
- value: 0.6664851619929213
72
  name: Cosine Accuracy@3
73
  - type: cosine_accuracy@5
74
- value: 0.7465287231146202
75
  name: Cosine Accuracy@5
76
  - type: cosine_accuracy@10
77
- value: 0.8205826300027226
78
  name: Cosine Accuracy@10
79
  - type: cosine_precision@1
80
- value: 0.488156820038116
81
  name: Cosine Precision@1
82
  - type: cosine_precision@3
83
- value: 0.22216172066430712
84
  name: Cosine Precision@3
85
  - type: cosine_precision@5
86
- value: 0.149305744622924
87
  name: Cosine Precision@5
88
  - type: cosine_precision@10
89
- value: 0.08205826300027225
90
  name: Cosine Precision@10
91
  - type: cosine_recall@1
92
- value: 0.488156820038116
93
  name: Cosine Recall@1
94
  - type: cosine_recall@3
95
- value: 0.6664851619929213
96
  name: Cosine Recall@3
97
  - type: cosine_recall@5
98
- value: 0.7465287231146202
99
  name: Cosine Recall@5
100
  - type: cosine_recall@10
101
- value: 0.8205826300027226
102
  name: Cosine Recall@10
103
  - type: cosine_ndcg@10
104
- value: 0.6494972918003398
105
  name: Cosine Ndcg@10
106
  - type: cosine_mrr@10
107
- value: 0.5951477750552078
108
  name: Cosine Mrr@10
109
  - type: cosine_map@100
110
- value: 0.6015029417718801
111
  name: Cosine Map@100
112
  ---
113
 
@@ -207,23 +207,23 @@ You can finetune this model on your own dataset.
207
  * Dataset: `intfloat/multilingual-e5-small`
208
  * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
209
 
210
- | Metric | Value |
211
- |:--------------------|:-----------|
212
- | cosine_accuracy@1 | 0.4882 |
213
- | cosine_accuracy@3 | 0.6665 |
214
- | cosine_accuracy@5 | 0.7465 |
215
- | cosine_accuracy@10 | 0.8206 |
216
- | cosine_precision@1 | 0.4882 |
217
- | cosine_precision@3 | 0.2222 |
218
- | cosine_precision@5 | 0.1493 |
219
- | cosine_precision@10 | 0.0821 |
220
- | cosine_recall@1 | 0.4882 |
221
- | cosine_recall@3 | 0.6665 |
222
- | cosine_recall@5 | 0.7465 |
223
- | cosine_recall@10 | 0.8206 |
224
- | cosine_ndcg@10 | 0.6495 |
225
- | cosine_mrr@10 | 0.5951 |
226
- | **cosine_map@100** | **0.6015** |
227
 
228
  <!--
229
  ## Bias, Risks and Limitations
@@ -398,14 +398,10 @@ You can finetune this model on your own dataset.
398
  | Epoch | Step | Training Loss | intfloat/multilingual-e5-small_cosine_map@100 |
399
  |:------:|:----:|:-------------:|:---------------------------------------------:|
400
  | 0 | 0 | - | 0.4852 |
401
- | 0.2174 | 100 | 0.6285 | 0.5761 |
402
- | 0.4348 | 200 | 0.1429 | 0.5824 |
403
- | 0.6522 | 300 | 0.1281 | 0.5903 |
404
- | 0.8696 | 400 | 0.1197 | 0.5943 |
405
- | 0.2174 | 100 | 0.0523 | 0.5826 |
406
- | 0.4348 | 200 | 0.0407 | 0.5886 |
407
- | 0.6522 | 300 | 0.0555 | 0.5978 |
408
- | 0.8696 | 400 | 0.0779 | 0.6015 |
409
 
410
 
411
  ### Framework Versions
 
65
  type: intfloat/multilingual-e5-small
66
  metrics:
67
  - type: cosine_accuracy@1
68
+ value: 0.4778110536346311
69
  name: Cosine Accuracy@1
70
  - type: cosine_accuracy@3
71
+ value: 0.6634903348761231
72
  name: Cosine Accuracy@3
73
  - type: cosine_accuracy@5
74
+ value: 0.7359106997005173
75
  name: Cosine Accuracy@5
76
  - type: cosine_accuracy@10
77
+ value: 0.8222161720664307
78
  name: Cosine Accuracy@10
79
  - type: cosine_precision@1
80
+ value: 0.4778110536346311
81
  name: Cosine Precision@1
82
  - type: cosine_precision@3
83
+ value: 0.22116344495870766
84
  name: Cosine Precision@3
85
  - type: cosine_precision@5
86
+ value: 0.14718213994010343
87
  name: Cosine Precision@5
88
  - type: cosine_precision@10
89
+ value: 0.08222161720664307
90
  name: Cosine Precision@10
91
  - type: cosine_recall@1
92
+ value: 0.4778110536346311
93
  name: Cosine Recall@1
94
  - type: cosine_recall@3
95
+ value: 0.6634903348761231
96
  name: Cosine Recall@3
97
  - type: cosine_recall@5
98
+ value: 0.7359106997005173
99
  name: Cosine Recall@5
100
  - type: cosine_recall@10
101
+ value: 0.8222161720664307
102
  name: Cosine Recall@10
103
  - type: cosine_ndcg@10
104
+ value: 0.6444880621773137
105
  name: Cosine Ndcg@10
106
  - type: cosine_mrr@10
107
+ value: 0.5882212109818965
108
  name: Cosine Mrr@10
109
  - type: cosine_map@100
110
+ value: 0.5940274223886709
111
  name: Cosine Map@100
112
  ---
113
 
 
207
  * Dataset: `intfloat/multilingual-e5-small`
208
  * Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
209
 
210
+ | Metric | Value |
211
+ |:--------------------|:----------|
212
+ | cosine_accuracy@1 | 0.4778 |
213
+ | cosine_accuracy@3 | 0.6635 |
214
+ | cosine_accuracy@5 | 0.7359 |
215
+ | cosine_accuracy@10 | 0.8222 |
216
+ | cosine_precision@1 | 0.4778 |
217
+ | cosine_precision@3 | 0.2212 |
218
+ | cosine_precision@5 | 0.1472 |
219
+ | cosine_precision@10 | 0.0822 |
220
+ | cosine_recall@1 | 0.4778 |
221
+ | cosine_recall@3 | 0.6635 |
222
+ | cosine_recall@5 | 0.7359 |
223
+ | cosine_recall@10 | 0.8222 |
224
+ | cosine_ndcg@10 | 0.6445 |
225
+ | cosine_mrr@10 | 0.5882 |
226
+ | **cosine_map@100** | **0.594** |
227
 
228
  <!--
229
  ## Bias, Risks and Limitations
 
398
  | Epoch | Step | Training Loss | intfloat/multilingual-e5-small_cosine_map@100 |
399
  |:------:|:----:|:-------------:|:---------------------------------------------:|
400
  | 0 | 0 | - | 0.4852 |
401
+ | 0.2174 | 100 | 0.6285 | 0.5759 |
402
+ | 0.4348 | 200 | 0.1429 | 0.5827 |
403
+ | 0.6522 | 300 | 0.1281 | 0.5901 |
404
+ | 0.8696 | 400 | 0.1197 | 0.5940 |
 
 
 
 
405
 
406
 
407
  ### Framework Versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8c9fc025259374e57fa6eec2c9f5ec8f8aacd235436cc87de40d0064badc124b
3
  size 470637416
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f78b7dd90133f2d3c7488daed22b009fd32814bd7ba1a8cc19abb950128ac8e
3
  size 470637416