Hugging Face's logo

BramVanroy
/
bert-base-multilingual-cased-hebban-reviews5

Text Classification
Transformers
PyTorch
TensorBoard
Dutch
bert
sentiment-analysis
dutch
text
Eval Results
Inference Endpoints
Model card Files Metrics Community
1
bert-base-multilingual-cased-hebban-reviews5
File size: 2,527 Bytes 
dce51a9
51c06b8
 
 
 
dce51a9
51c06b8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
dce51a9
51c06b8
0acc7d0
51c06b8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
873abe6
51c06b8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
 
---
datasets:
- BramVanroy/hebban-reviews
language:
- nl
license: mit
metrics:
- accuracy
- f1
- precision
- qwk
- recall
model-index:
- name: bert-base-multilingual-cased-hebban-reviews5
  results:
  - dataset:
      config: filtered_rating
      name: BramVanroy/hebban-reviews - filtered_rating - 2.0.0
      revision: 2.0.0
      split: test
      type: BramVanroy/hebban-reviews
    metrics:
    - name: Test accuracy
      type: accuracy
      value: 0.5898668639053254
    - name: Test f1
      type: f1
      value: 0.5899204480029937
    - name: Test precision
      type: precision
      value: 0.5971431895675179
    - name: Test qwk
      type: qwk
      value: 0.7050840079198698
    - name: Test recall
      type: recall
      value: 0.5898668639053254
    task:
      name: sentiment analysis
      type: text-classification
tags:
- sentiment-analysis
- dutch
- text
widget:
- text: Wauw, wat een leuk boek! Ik heb me er er goed mee vermaakt.
- text: Nee, deze vond ik niet goed. De auteur doet zijn best om je als lezer mee
    te trekken in het verhaal maar mij overtuigt het alleszins niet.
- text: Ik vind het niet slecht maar de schrijfstijl trekt me ook niet echt aan. Het
    wordt een beetje saai vanaf het vijfde hoofdstuk
---

# bert-base-multilingual-cased-hebban-reviews5

# Dataset
- dataset_name: BramVanroy/hebban-reviews
- dataset_config: filtered_rating
- dataset_revision: 2.0.0
- labelcolumn: review_rating0
- textcolumn: review_text_without_quotes

# Training
- optim: adamw_hf
- learning_rate: 5e-05
- per_device_train_batch_size: 64
- per_device_eval_batch_size: 64
- gradient_accumulation_steps: 1
- max_steps: 5001
- save_steps: 500
- metric_for_best_model: qwk

# Best checkedpoint based on validation
- best_metric: 0.697825193570947
- best_model_checkpoint: trained/hebban-reviews5/bert-base-multilingual-cased/checkpoint-4500

# Test results of best checkpoint
- accuracy: 0.5898668639053254
- f1: 0.5899204480029937
- precision: 0.5971431895675179
- qwk: 0.7050840079198698
- recall: 0.5898668639053254

## Confusion matrix

![cfm](fig/test_confusion_matrix.png)

## Normalized confusion matrix

![norm cfm](fig/test_confusion_matrix_norm.png)

# Environment
- cuda_capabilities: 8.0; 8.0
- cuda_device_count: 2
- cuda_devices: NVIDIA A100-SXM4-80GB; NVIDIA A100-SXM4-80GB
- finetuner_commit: 8159b4c1d5e66b36f68dd263299927ffb8670ebd
- platform: Linux-4.18.0-305.49.1.el8_4.x86_64-x86_64-with-glibc2.28
- python_version: 3.9.5
- toch_version: 1.10.0
- transformers_version: 4.21.0