philschmid HF staff commited on
Commit
83ff874
1 Parent(s): 1af1cad

End of training

Browse files
README.md ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - amazon_reviews_multi
7
+ metrics:
8
+ - accuracy
9
+ - f1
10
+ model-index:
11
+ - name: distilbert-base-multilingual-cased-sentiment-2
12
+ results:
13
+ - task:
14
+ name: Text Classification
15
+ type: text-classification
16
+ dataset:
17
+ name: amazon_reviews_multi
18
+ type: amazon_reviews_multi
19
+ args: all_languages
20
+ metrics:
21
+ - name: Accuracy
22
+ type: accuracy
23
+ value: 0.7475666666666667
24
+ - name: F1
25
+ type: f1
26
+ value: 0.7475666666666667
27
+ ---
28
+
29
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
30
+ should probably proofread and complete it, then remove this comment. -->
31
+
32
+ # distilbert-base-multilingual-cased-sentiment-2
33
+
34
+ This model is a fine-tuned version of [distilbert-base-multilingual-cased](https://huggingface.co/distilbert-base-multilingual-cased) on the amazon_reviews_multi dataset.
35
+ It achieves the following results on the evaluation set:
36
+ - Loss: 0.6067
37
+ - Accuracy: 0.7476
38
+ - F1: 0.7476
39
+
40
+ ## Model description
41
+
42
+ More information needed
43
+
44
+ ## Intended uses & limitations
45
+
46
+ More information needed
47
+
48
+ ## Training and evaluation data
49
+
50
+ More information needed
51
+
52
+ ## Training procedure
53
+
54
+ ### Training hyperparameters
55
+
56
+ The following hyperparameters were used during training:
57
+ - learning_rate: 0.00024
58
+ - train_batch_size: 16
59
+ - eval_batch_size: 16
60
+ - seed: 33
61
+ - distributed_type: sagemaker_data_parallel
62
+ - num_devices: 8
63
+ - total_train_batch_size: 128
64
+ - total_eval_batch_size: 128
65
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
66
+ - lr_scheduler_type: linear
67
+ - lr_scheduler_warmup_steps: 500
68
+ - num_epochs: 3
69
+ - mixed_precision_training: Native AMP
70
+
71
+ ### Training results
72
+
73
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
74
+ |:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|
75
+ | 0.6885 | 0.53 | 5000 | 0.6532 | 0.7217 | 0.7217 |
76
+ | 0.6411 | 1.07 | 10000 | 0.6348 | 0.7319 | 0.7319 |
77
+ | 0.6057 | 1.6 | 15000 | 0.6186 | 0.7387 | 0.7387 |
78
+ | 0.5844 | 2.13 | 20000 | 0.6236 | 0.7449 | 0.7449 |
79
+ | 0.549 | 2.67 | 25000 | 0.6067 | 0.7476 | 0.7476 |
80
+
81
+
82
+ ### Framework versions
83
+
84
+ - Transformers 4.12.3
85
+ - Pytorch 1.9.1
86
+ - Datasets 1.15.1
87
+ - Tokenizers 0.10.3
logs/events.out.tfevents.1643033115.algo-1.507.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:29004857ff313d49f39686f228e7d1cddb6c63604844d45dd3dc2a56e7b0a526
3
- size 5886
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3dced5eb16a50ab3e43cc7e5e434991d3d7c34c75fe51a0eba4bc8a6118a6a5
3
+ size 6246
logs/events.out.tfevents.1643036921.algo-1.507.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c4c0c0c2675347ff44e70a22b6384b9421c9850902cca8cff91bb0e72638920
3
+ size 416