Hugging Face's logo

thejosango
/
nuha-ajp-binary

Text Classification
Transformers
PyTorch
Safetensors
Arabic
bert
hate-speech
gender-based-violence
arabic
binary-classification
pilot
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
4
yazansh commited on
Commit
99b45f1
·
1 Parent(s): e1d4ffd

binary-20

Browse files
Files changed (5) hide show
  1. README.md +18 -19
  2. config.toml +6 -6
  3. pytorch_model.bin +1 -1
  4. tokenizer.json +6 -1
  5. training_args.bin +1 -1
README.md CHANGED
@@ -23,13 +23,13 @@ model-index:
23
  metrics:
24
  - name: F1
25
  type: f1
26
- value: 0.6647587898609976
27
  - name: Precision
28
  type: precision
29
- value: 0.5703262013328657
30
  - name: Recall
31
  type: recall
32
- value: 0.7966682998530132
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: 0.8563
43
- - F1: 0.6648
44
- - Precision: 0.5703
45
- - Recall: 0.7967
46
  - Support: None
47
 
48
  ## Model description
@@ -62,27 +62,26 @@ More information needed
62
  ### Training hyperparameters
63
 
64
  The following hyperparameters were used during training:
65
- - learning_rate: 2e-05
66
- - train_batch_size: 32
67
  - eval_batch_size: 32
68
  - seed: 42
69
  - gradient_accumulation_steps: 4
70
- - total_train_batch_size: 128
71
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
72
- - lr_scheduler_type: constant
73
  - lr_scheduler_warmup_steps: 1000.0
74
- - num_epochs: 10
75
  - label_smoothing_factor: 0.1
76
 
77
  ### Training results
78
 
79
- | Training Loss | Epoch | Step | F1 | Validation Loss | Precision | Recall | Support |
80
- |:-------------:|:-----:|:----:|:------:|:---------------:|:---------:|:------:|:-------:|
81
- | 2.4774 | 1.96 | 500 | 0.6241 | 0.8233 | 0.6118 | 0.6369 | None |
82
- | 1.4697 | 3.92 | 1000 | 0.6530 | 0.8029 | 0.6085 | 0.7046 | None |
83
- | 1.35 | 5.88 | 1500 | 0.8659 | 0.6592 | 0.5726 | 0.7766 | None |
84
- | 1.2811 | 7.84 | 2000 | 0.8426 | 0.6305 | 0.6559 | 0.6071 | None |
85
- | 1.1341 | 9.8 | 2500 | 0.8563 | 0.6648 | 0.5703 | 0.7967 | None |
86
 
87
 
88
  ### Framework versions
 
23
  metrics:
24
  - name: F1
25
  type: f1
26
+ value: 0.6302113631956563
27
  - name: Precision
28
  type: precision
29
+ value: 0.4972460220318237
30
  - name: Recall
31
  type: recall
32
+ value: 0.8602435150873478
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: 1.0884
43
+ - F1: 0.6302
44
+ - Precision: 0.4972
45
+ - Recall: 0.8602
46
  - Support: None
47
 
48
  ## Model description
 
62
  ### Training hyperparameters
63
 
64
  The following hyperparameters were used during training:
65
+ - learning_rate: 1e-05
66
+ - train_batch_size: 16
67
  - eval_batch_size: 32
68
  - seed: 42
69
  - gradient_accumulation_steps: 4
70
+ - total_train_batch_size: 64
71
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
72
+ - lr_scheduler_type: constant_with_warmup
73
  - lr_scheduler_warmup_steps: 1000.0
74
+ - num_epochs: 5
75
  - label_smoothing_factor: 0.1
76
 
77
  ### Training results
78
 
79
+ | Training Loss | Epoch | Step | Validation Loss | F1 | Precision | Recall | Support |
80
+ |:-------------:|:-----:|:----:|:---------------:|:------:|:---------:|:------:|:-------:|
81
+ | 4.7382 | 1.06 | 500 | 1.6112 | 0.5093 | 0.5664 | 0.4627 | None |
82
+ | 2.8127 | 2.12 | 1000 | 1.4358 | 0.6255 | 0.4994 | 0.8370 | None |
83
+ | 2.0837 | 3.18 | 1500 | 1.0886 | 0.6362 | 0.5187 | 0.8227 | None |
84
+ | 1.6086 | 4.24 | 2000 | 1.0884 | 0.6302 | 0.4972 | 0.8602 | None |
 
85
 
86
 
87
  ### Framework versions
config.toml CHANGED
@@ -1,5 +1,5 @@
1
  [experiment]
2
- name = "binary-19"
3
  type = "binary"
4
 
5
 
@@ -23,14 +23,14 @@ classifier_dropout = 0.2
23
 
24
 
25
  [training]
26
- num_train_epochs = 10
27
  warmup_steps = 1e3
28
- lr_scheduler_type = "constant"
29
- learning_rate = 2e-5
30
- per_device_train_batch_size = 32
31
  per_device_eval_batch_size = 32
32
  gradient_accumulation_steps = 4
33
- weight_decay = 0.01
34
  label_smoothing_factor = 0.1
35
  weighted_loss = false
36
  resample_data = true
 
1
  [experiment]
2
+ name = "binary-20"
3
  type = "binary"
4
 
5
 
 
23
 
24
 
25
  [training]
26
+ num_train_epochs = 5
27
  warmup_steps = 1e3
28
+ lr_scheduler_type = "constant_with_warmup"
29
+ learning_rate = 1e-5
30
+ per_device_train_batch_size = 16
31
  per_device_eval_batch_size = 32
32
  gradient_accumulation_steps = 4
33
+ weight_decay = 0.05
34
  label_smoothing_factor = 0.1
35
  weighted_loss = false
36
  resample_data = true
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2a9521a0283f141feb882e687a4691c70cd2828ccc3f01aba026eae1d19ae68d
3
  size 540847921
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59b0e5bbf20a6b167f71da35e9552c419eaea22968c91db1542e7179ca0ce3a0
3
  size 540847921
tokenizer.json CHANGED
@@ -1,6 +1,11 @@
1
  {
2
  "version": "1.0",
3
- "truncation": null,
 
 
 
 
 
4
  "padding": null,
5
  "added_tokens": [
6
  {
 
1
  {
2
  "version": "1.0",
3
+ "truncation": {
4
+ "direction": "Right",
5
+ "max_length": 512,
6
+ "strategy": "LongestFirst",
7
+ "stride": 0
8
+ },
9
  "padding": null,
10
  "added_tokens": [
11
  {
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bc2182eb44154c5f51c0d5b6960fb585b6faee8e43ed6f4a420742f141dac216
3
  size 4091
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a36be302c6e8f2422e4c79fa435fc84ec8b153a6ccfe8f5bb5f0297e17d46015
3
  size 4091