Hugging Face's logo

thejosango
/
nuha-binary

Text Classification
Transformers
PyTorch
Arabic
bert
hate-speech
gender-based-violence
arabic
binary-classification
pilot
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
yazansh commited on
Commit
6aaa15d
·
verified ·
1 Parent(s): ff731bd

binary-39

Browse files
Files changed (4) hide show
  1. README.md +13 -23
  2. config.toml +6 -6
  3. pytorch_model.bin +1 -1
  4. training_args.bin +1 -1
README.md CHANGED
@@ -23,13 +23,13 @@ model-index:
23
  metrics:
24
  - name: F1
25
  type: f1
26
- value: 0.6928497604059006
27
  - name: Precision
28
  type: precision
29
- value: 0.7049713193116635
30
  - name: Recall
31
  type: recall
32
- value: 0.6811380011084427
33
  ---
34
 
35
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -39,10 +39,10 @@ should probably proofread and complete it, then remove this comment. -->
39
 
40
  This model is a fine-tuned version of [thejosango/nuha-mlm](https://huggingface.co/thejosango/nuha-mlm) on the nuha-dataset dataset.
41
  It achieves the following results on the evaluation set:
42
- - Loss: 0.4966
43
- - F1: 0.6928
44
- - Precision: 0.7050
45
- - Recall: 0.6811
46
  - Support: None
47
 
48
  ## Model description
@@ -63,31 +63,21 @@ More information needed
63
 
64
  The following hyperparameters were used during training:
65
  - learning_rate: 1e-05
66
- - train_batch_size: 32
67
- - eval_batch_size: 32
68
  - seed: 42
69
- - gradient_accumulation_steps: 2
70
- - total_train_batch_size: 64
71
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
72
  - lr_scheduler_type: constant
73
- - num_epochs: 5
74
  - label_smoothing_factor: 0.1
75
 
76
  ### Training results
77
 
78
  | Training Loss | Epoch | Step | Validation Loss | F1 | Precision | Recall | Support |
79
  |:-------------:|:-----:|:----:|:---------------:|:------:|:---------:|:------:|:-------:|
80
- | 0.5537 | 0.25 | 500 | 0.5298 | 0.6135 | 0.6830 | 0.5568 | None |
81
- | 0.5294 | 0.5 | 1000 | 0.5174 | 0.6260 | 0.7106 | 0.5594 | None |
82
- | 0.5236 | 0.75 | 1500 | 0.5129 | 0.6751 | 0.6752 | 0.6750 | None |
83
- | 0.5211 | 1.0 | 2000 | 0.5038 | 0.6656 | 0.7084 | 0.6277 | None |
84
- | 0.4972 | 1.26 | 2500 | 0.5034 | 0.6486 | 0.7288 | 0.5843 | None |
85
- | 0.4926 | 1.51 | 3000 | 0.5058 | 0.6944 | 0.6755 | 0.7144 | None |
86
- | 0.4914 | 1.76 | 3500 | 0.5006 | 0.6936 | 0.6901 | 0.6970 | None |
87
- | 0.4899 | 2.01 | 4000 | 0.5019 | 0.6881 | 0.7036 | 0.6732 | None |
88
- | 0.4715 | 2.26 | 4500 | 0.4987 | 0.6949 | 0.6977 | 0.6920 | None |
89
- | 0.474 | 2.51 | 5000 | 0.4990 | 0.6998 | 0.6922 | 0.7076 | None |
90
- | 0.4729 | 2.76 | 5500 | 0.4966 | 0.6928 | 0.7050 | 0.6811 | None |
91
 
92
 
93
  ### Framework versions
 
23
  metrics:
24
  - name: F1
25
  type: f1
26
+ value: 0.592617830777967
27
  - name: Precision
28
  type: precision
29
+ value: 0.7691627358490566
30
  - name: Recall
31
  type: recall
32
+ value: 0.481987807130981
33
  ---
34
 
35
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
39
 
40
  This model is a fine-tuned version of [thejosango/nuha-mlm](https://huggingface.co/thejosango/nuha-mlm) on the nuha-dataset dataset.
41
  It achieves the following results on the evaluation set:
42
+ - Loss: 0.4196
43
+ - F1: 0.5926
44
+ - Precision: 0.7692
45
+ - Recall: 0.4820
46
  - Support: None
47
 
48
  ## Model description
 
63
 
64
  The following hyperparameters were used during training:
65
  - learning_rate: 1e-05
66
+ - train_batch_size: 64
67
+ - eval_batch_size: 64
68
  - seed: 42
 
 
69
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
70
  - lr_scheduler_type: constant
71
+ - num_epochs: 1
72
  - label_smoothing_factor: 0.1
73
 
74
  ### Training results
75
 
76
  | Training Loss | Epoch | Step | Validation Loss | F1 | Precision | Recall | Support |
77
  |:-------------:|:-----:|:----:|:---------------:|:------:|:---------:|:------:|:-------:|
78
+ | 0.4508 | 0.25 | 500 | 0.4341 | 0.4407 | 0.7864 | 0.3061 | None |
79
+ | 0.4317 | 0.5 | 1000 | 0.4229 | 0.4936 | 0.8037 | 0.3562 | None |
80
+ | 0.4289 | 0.75 | 1500 | 0.4196 | 0.5926 | 0.7692 | 0.4820 | None |
 
 
 
 
 
 
 
 
81
 
82
 
83
  ### Framework versions
config.toml CHANGED
@@ -1,5 +1,5 @@
1
  [experiment]
2
- name = "binary-38"
3
  type = "binary"
4
 
5
 
@@ -16,15 +16,15 @@ revision = "ce20f497544665775129f9ff5b3cd2a3e350dce8"
16
 
17
 
18
  [training]
19
- num_train_epochs = 5
20
  warmup_steps = 0
21
  lr_scheduler_type = "constant"
22
  learning_rate = 1e-5
23
- per_device_train_batch_size = 32
24
- per_device_eval_batch_size = 32
25
- gradient_accumulation_steps = 2
26
  weight_decay = 0.00
27
  label_smoothing_factor = 0.1
28
- weighted_loss = false
29
  early_stopping_patience = 5
30
  early_stopping_threshold = 0.005
 
1
  [experiment]
2
+ name = "binary-39"
3
  type = "binary"
4
 
5
 
 
16
 
17
 
18
  [training]
19
+ num_train_epochs = 1
20
  warmup_steps = 0
21
  lr_scheduler_type = "constant"
22
  learning_rate = 1e-5
23
+ per_device_train_batch_size = 64
24
+ per_device_eval_batch_size = 64
25
+ gradient_accumulation_steps = 1
26
  weight_decay = 0.00
27
  label_smoothing_factor = 0.1
28
+ weighted_loss = true
29
  early_stopping_patience = 5
30
  early_stopping_threshold = 0.005
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c3eceddd77e5a00819ab30f4f9f90180c97198b12fefd284d24ecf49c2b80ce5
3
  size 540847921
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99972b963c59dacfc2e2c98109726ef6bbffac594bb8e422cdb7bc65758e593c
3
  size 540847921
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:08daaa9929d5ac6d9774c9e26f13cc06d212873de4e3172ec8b81ea97d9d5d6b
3
  size 4091
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b50e314607e1472325db96b6790c30fded1af537fb57a1e3c4cacc16a62b2d08
3
  size 4091