Ertugrul commited on
Commit
490b3bf
1 Parent(s): ee1303e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +72 -0
README.md CHANGED
@@ -1,3 +1,75 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ language:
4
+ - tr
5
+ metrics:
6
+ - accuracy
7
+ - recall
8
+ - f1
9
+ library_name: transformers
10
+ pipeline_tag: text-classification
11
+ model-index:
12
+ - name: deprem_v12
13
+ results:
14
+ - task:
15
+ type: text-classification
16
+ dataset:
17
+ type: deprem_private_dataset_v1_2
18
+ name: deprem_private_dataset_v1_2
19
+ metrics:
20
+ - type: recall
21
+ value: 0.82
22
+ verified: false
23
+ - type: f1
24
+ value: 0.76
25
+ verified: false
26
+ widget:
27
+ - text: >-
28
+ acil acil acil antakyadan istanbula gitmek için antakya expoya ulaşmaya çalışan 19 kişilik bir aile için şehir içi ulaşım desteği istiyoruz. dışardalar üşüyorlar.iletebileceğiniz numaraları bekliyorum
29
+ example_title: Örnek
30
  ---
31
+
32
+
33
+
34
+ ## Eval Results
35
+ ```
36
+ precision recall f1-score support
37
+
38
+ Alakasiz 0.87 0.91 0.89 734
39
+ Barinma 0.79 0.89 0.84 207
40
+ Elektronik 0.69 0.83 0.75 130
41
+ Giysi 0.71 0.81 0.76 94
42
+ Kurtarma 0.82 0.85 0.83 362
43
+ Lojistik 0.57 0.67 0.62 112
44
+ Saglik 0.68 0.85 0.75 108
45
+ Su 0.56 0.76 0.64 78
46
+ Yagma 0.60 0.77 0.68 31
47
+ Yemek 0.71 0.89 0.79 117
48
+
49
+ micro avg 0.77 0.86 0.81 1973
50
+ macro avg 0.70 0.82 0.76 1973
51
+ weighted avg 0.78 0.86 0.82 1973
52
+ samples avg 0.83 0.88 0.84 1973
53
+
54
+
55
+ ```
56
+
57
+ ## Threshold:
58
+ - **Best Threshold:** 0.40
59
+
60
+ ```python
61
+ {'per_device_train_batch_size': 32,
62
+ 'per_device_eval_batch_size': 32,
63
+ 'learning_rate': 5.8679699888213376e-05,
64
+ 'weight_decay': 0.03530961718117487,
65
+ 'num_train_epochs': 4,
66
+ 'lr_scheduler_type': 'cosine',
67
+ 'warmup_steps': 40,
68
+ 'seed': 42,
69
+ 'fp16': True,
70
+ 'load_best_model_at_end': True,
71
+ 'metric_for_best_model': 'macro f1',
72
+ 'greater_is_better': True,
73
+ ```
74
+ ## Class Loss Weights
75
+ - Same as Anıl's approach.