Ahmet commited on
Commit
1f33cc3
·
1 Parent(s): 072520c

update model

Browse files
README.md CHANGED
@@ -21,6 +21,12 @@ This model was adapted from [ytu-ce-cosmos/turkish-tiny-bert-uncased](https://hu
21
  - [nli_tr](https://huggingface.co/datasets/nli_tr)
22
  - [emrecan/stsb-mt-turkish](https://huggingface.co/datasets/emrecan/stsb-mt-turkish)
23
 
 
 
 
 
 
 
24
  ## Usage (Sentence-Transformers)
25
 
26
  Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
@@ -85,10 +91,10 @@ print(sentence_embeddings)
85
  Achieved results on the [STS-b](https://huggingface.co/datasets/emrecan/stsb-mt-turkish) test split are given below:
86
 
87
  ```txt
88
- Cosine-Similarity : Pearson: 0.6587 Spearman: 0.6370
89
- Manhattan-Distance: Pearson: 0.6293 Spearman: 0.6151
90
- Euclidean-Distance: Pearson: 0.6335 Spearman: 0.6186
91
- Dot-Product-Similarity: Pearson: 0.5972 Spearman: 0.5756
92
  ```
93
 
94
 
@@ -110,7 +116,7 @@ Parameters of the fit()-Method:
110
  ```
111
  {
112
  "epochs": 10,
113
- "evaluation_steps": 22,
114
  "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
115
  "max_grad_norm": 1,
116
  "optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
@@ -128,7 +134,7 @@ Parameters of the fit()-Method:
128
  ## Full Model Architecture
129
  ```
130
  SentenceTransformer(
131
- (0): Transformer({'max_seq_length': 75, 'do_lower_case': False}) with Transformer model: BertModel
132
  (1): Pooling({'word_embedding_dimension': 128, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
133
  )
134
  ```
 
21
  - [nli_tr](https://huggingface.co/datasets/nli_tr)
22
  - [emrecan/stsb-mt-turkish](https://huggingface.co/datasets/emrecan/stsb-mt-turkish)
23
 
24
+ :warning: **All texts were manually lowercased,** [as stated](https://huggingface.co/ytu-ce-cosmos/turkish-tiny-bert-uncased#%E2%9A%A0-uncased-use-requires-manual-lowercase-conversion) by the model's authors:
25
+
26
+ ```python
27
+ text.replace("I", "ı").lower()
28
+ ```
29
+
30
  ## Usage (Sentence-Transformers)
31
 
32
  Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
 
91
  Achieved results on the [STS-b](https://huggingface.co/datasets/emrecan/stsb-mt-turkish) test split are given below:
92
 
93
  ```txt
94
+ Cosine-Similarity : Pearson: 0.7515 Spearman: 0.7467
95
+ Manhattan-Distance: Pearson: 0.7404 Spearman: 0.7299
96
+ Euclidean-Distance: Pearson: 0.7415 Spearman: 0.7305
97
+ Dot-Product-Similarity: Pearson: 0.6395 Spearman: 0.6140
98
  ```
99
 
100
 
 
116
  ```
117
  {
118
  "epochs": 10,
119
+ "evaluation_steps": 574,
120
  "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
121
  "max_grad_norm": 1,
122
  "optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
 
134
  ## Full Model Architecture
135
  ```
136
  SentenceTransformer(
137
+ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
138
  (1): Pooling({'word_embedding_dimension': 128, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
139
  )
140
  ```
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "output/training_nli_v2_ytu-ce-cosmos-turkish-tiny-bert-uncased-2024-02-14_22-23-39\\",
3
  "architectures": [
4
  "BertModel"
5
  ],
 
1
  {
2
+ "_name_or_path": "output/ytu_ce_cosmos-turkish_tiny_bert_uncased-b128-e10-nli/",
3
  "architectures": [
4
  "BertModel"
5
  ],
config_sentence_transformers.json CHANGED
@@ -2,6 +2,6 @@
2
  "__version__": {
3
  "sentence_transformers": "2.2.2",
4
  "transformers": "4.28.0",
5
- "pytorch": "2.0.1+cu118"
6
  }
7
  }
 
2
  "__version__": {
3
  "sentence_transformers": "2.2.2",
4
  "transformers": "4.28.0",
5
+ "pytorch": "2.1.0+cu121"
6
  }
7
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1d1d69f5f19045cb2acaf5ce6bc758eb9ef0bf19bfda219104c995f90a416595
3
- size 18317545
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0392c2975e017cc3bf5d50e0ecb956eb9b7d99a1ac7a0b1d3b002e5167adb5f
3
+ size 18317990
sentence_bert_config.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "max_seq_length": 75,
3
  "do_lower_case": false
4
  }
 
1
  {
2
+ "max_seq_length": 256,
3
  "do_lower_case": false
4
  }
tokenizer.json CHANGED
@@ -2,7 +2,7 @@
2
  "version": "1.0",
3
  "truncation": {
4
  "direction": "Right",
5
- "max_length": 75,
6
  "strategy": "LongestFirst",
7
  "stride": 0
8
  },
 
2
  "version": "1.0",
3
  "truncation": {
4
  "direction": "Right",
5
+ "max_length": 256,
6
  "strategy": "LongestFirst",
7
  "stride": 0
8
  },