atasoglu commited on
Commit
8a28ca8
1 Parent(s): 5c92d2f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -5
README.md CHANGED
@@ -6,14 +6,21 @@ tags:
6
  - feature-extraction
7
  - sentence-similarity
8
  - transformers
9
-
 
 
 
 
 
10
  ---
11
 
12
  # atasoglu/mbert-base-cased-nli-stsb-tr
13
 
14
  This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
15
 
16
- <!--- Describe your model here -->
 
 
17
 
18
  ## Usage (Sentence-Transformers)
19
 
@@ -76,10 +83,14 @@ print(sentence_embeddings)
76
 
77
  ## Evaluation Results
78
 
79
- <!--- Describe how your model was evaluated -->
80
-
81
- For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=atasoglu/mbert-base-cased-nli-stsb-tr)
82
 
 
 
 
 
 
 
83
 
84
  ## Training
85
  The model was trained with the parameters:
 
6
  - feature-extraction
7
  - sentence-similarity
8
  - transformers
9
+ license: apache-2.0
10
+ datasets:
11
+ - nli_tr
12
+ - emrecan/stsb-mt-turkish
13
+ language:
14
+ - tr
15
  ---
16
 
17
  # atasoglu/mbert-base-cased-nli-stsb-tr
18
 
19
  This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
20
 
21
+ This model was adapted from [google-bert/bert-base-multilingual-cased](https://huggingface.co/google-bert/bert-base-multilingual-cased) and fine-tuned on these datasets:
22
+ - [nli_tr](https://huggingface.co/datasets/nli_tr)
23
+ - [emrecan/stsb-mt-turkish](https://huggingface.co/datasets/emrecan/stsb-mt-turkish)
24
 
25
  ## Usage (Sentence-Transformers)
26
 
 
83
 
84
  ## Evaluation Results
85
 
86
+ Achieved results on the [STS-b](https://huggingface.co/datasets/emrecan/stsb-mt-turkish) test split are given below:
 
 
87
 
88
+ ```txt
89
+ Cosine-Similarity : Pearson: 0.8152 Spearman: 0.8130
90
+ Manhattan-Distance: Pearson: 0.8049 Spearman: 0.8128
91
+ Euclidean-Distance: Pearson: 0.8049 Spearman: 0.8126
92
+ Dot-Product-Similarity: Pearson: 0.7878 Spearman: 0.7822
93
+ ```
94
 
95
  ## Training
96
  The model was trained with the parameters: