Hugging Face's logo

jziebura
/
youth-slang-sentiment-classifier

Text Classification
Transformers
Safetensors
Polish
bert
Generated from Trainer
Model card Files
xet
Community
jziebura commited on
Commit
d78d4ef
·
verified ·
1 Parent(s): 169aa5c

End of training

Browse files
Files changed (3) hide show
  1. README.md +36 -39
  2. model.safetensors +1 -1
  3. tokenizer.json +1 -6
README.md CHANGED
@@ -17,30 +17,27 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  # finetuning
19
 
20
- This model is a fine-tuned version of [allegro/herbert-base-cased](https://huggingface.co/allegro/herbert-base-cased) on a [jziebura/polish_youth_slang_classification](https://huggingface.co/datasets/jziebura/polish_youth_slang_classification) dataset.
21
-
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.7785
24
- - Accuracy: 0.7072
25
- - F1: 0.7071
26
- - F1 Macro: 0.6961
27
 
28
  ## Model description
29
 
30
- The model is part of the experiments conducted during the creation of my master's thesis titled: *"A language model analyzing Polish youth slang"*.
31
- It was fine-tuned to classify the sentiment of the Polish youth slang into three categories: negative, neutral or ambiguous, and positive.
 
32
 
 
33
 
34
  ## Training and evaluation data
35
 
36
- All data comes from the [jziebura/polish_youth_slang_classification](https://huggingface.co/datasets/jziebura/polish_youth_slang_classification) dataset
37
 
38
  ## Training procedure
39
 
40
- The hyperparameters were selected from those recommended in the [BERT introduction paper](https://arxiv.org/abs/1810.04805) and then optimized using the Optuna backend.
41
-
42
- The HPO and fine-tuning were both conducted on the Google Colab platform on their free-tier T4 GPU instances.
43
-
44
  ### Training hyperparameters
45
 
46
  The following hyperparameters were used during training:
@@ -56,32 +53,32 @@ The following hyperparameters were used during training:
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | F1 Macro |
58
  |:-------------:|:------:|:----:|:---------------:|:--------:|:------:|:--------:|
59
- | No log | 0 | 0 | 1.0997 | 0.3432 | 0.3520 | 0.3095 |
60
- | 1.0454 | 0.1176 | 32 | 0.9823 | 0.5037 | 0.3374 | 0.2233 |
61
- | 0.9738 | 0.2353 | 64 | 0.8940 | 0.5849 | 0.5037 | 0.4512 |
62
- | 0.888 | 0.3529 | 96 | 0.8146 | 0.6199 | 0.6078 | 0.5914 |
63
- | 0.8244 | 0.4706 | 128 | 0.8177 | 0.6144 | 0.6177 | 0.6067 |
64
- | 0.8112 | 0.5882 | 160 | 0.7941 | 0.6236 | 0.6282 | 0.6216 |
65
- | 0.7136 | 0.7059 | 192 | 0.7872 | 0.6531 | 0.6246 | 0.6045 |
66
- | 0.7729 | 0.8235 | 224 | 0.7487 | 0.6568 | 0.6566 | 0.6451 |
67
- | 0.7293 | 0.9412 | 256 | 0.7410 | 0.6624 | 0.6505 | 0.6333 |
68
- | 0.7 | 1.0588 | 288 | 0.7267 | 0.6771 | 0.6637 | 0.6466 |
69
- | 0.5851 | 1.1765 | 320 | 0.7670 | 0.6679 | 0.6682 | 0.6571 |
70
- | 0.6227 | 1.2941 | 352 | 0.7830 | 0.6605 | 0.6273 | 0.6049 |
71
- | 0.5306 | 1.4118 | 384 | 0.8415 | 0.6679 | 0.6699 | 0.6641 |
72
- | 0.543 | 1.5294 | 416 | 0.7516 | 0.6697 | 0.6701 | 0.6617 |
73
- | 0.5577 | 1.6471 | 448 | 0.7589 | 0.6716 | 0.6704 | 0.6612 |
74
- | 0.596 | 1.7647 | 480 | 0.7646 | 0.6753 | 0.6752 | 0.6685 |
75
- | 0.5602 | 1.8824 | 512 | 0.7413 | 0.6771 | 0.6775 | 0.6735 |
76
- | 0.5304 | 2.0 | 544 | 0.7547 | 0.6697 | 0.6666 | 0.6584 |
77
- | 0.3825 | 2.1176 | 576 | 0.8180 | 0.6679 | 0.6673 | 0.6607 |
78
- | 0.3699 | 2.2353 | 608 | 0.8897 | 0.6679 | 0.6703 | 0.6682 |
79
- | 0.3564 | 2.3529 | 640 | 0.8247 | 0.6697 | 0.6670 | 0.6594 |
80
- | 0.3608 | 2.4706 | 672 | 0.8349 | 0.6734 | 0.6715 | 0.6666 |
81
- | 0.3246 | 2.5882 | 704 | 0.8496 | 0.6771 | 0.6746 | 0.6672 |
82
- | 0.3398 | 2.7059 | 736 | 0.8569 | 0.6790 | 0.6783 | 0.6732 |
83
- | 0.3966 | 2.8235 | 768 | 0.8525 | 0.6771 | 0.6739 | 0.6649 |
84
- | 0.3501 | 2.9412 | 800 | 0.8471 | 0.6771 | 0.6760 | 0.6698 |
85
 
86
 
87
  ### Framework versions
 
17
 
18
  # finetuning
19
 
20
+ This model is a fine-tuned version of [allegro/herbert-base-cased](https://huggingface.co/allegro/herbert-base-cased) on an unknown dataset.
 
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 0.7289
23
+ - Accuracy: 0.7127
24
+ - F1: 0.7110
25
+ - F1 Macro: 0.6977
26
 
27
  ## Model description
28
 
29
+ More information needed
30
+
31
+ ## Intended uses & limitations
32
 
33
+ More information needed
34
 
35
  ## Training and evaluation data
36
 
37
+ More information needed
38
 
39
  ## Training procedure
40
 
 
 
 
 
41
  ### Training hyperparameters
42
 
43
  The following hyperparameters were used during training:
 
53
 
54
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | F1 Macro |
55
  |:-------------:|:------:|:----:|:---------------:|:--------:|:------:|:--------:|
56
+ | No log | 0 | 0 | 1.1025 | 0.3118 | 0.2848 | 0.2324 |
57
+ | 1.0546 | 0.1176 | 32 | 1.0001 | 0.5037 | 0.3374 | 0.2233 |
58
+ | 0.9406 | 0.2353 | 64 | 0.8969 | 0.5849 | 0.5654 | 0.5371 |
59
+ | 0.8885 | 0.3529 | 96 | 0.8430 | 0.6015 | 0.6074 | 0.6090 |
60
+ | 0.8452 | 0.4706 | 128 | 0.8230 | 0.6218 | 0.6173 | 0.5990 |
61
+ | 0.8208 | 0.5882 | 160 | 0.8393 | 0.6107 | 0.6125 | 0.5982 |
62
+ | 0.7182 | 0.7059 | 192 | 0.7848 | 0.6605 | 0.6504 | 0.6324 |
63
+ | 0.7644 | 0.8235 | 224 | 0.7708 | 0.6587 | 0.6516 | 0.6347 |
64
+ | 0.7211 | 0.9412 | 256 | 0.7734 | 0.6642 | 0.6440 | 0.6155 |
65
+ | 0.7182 | 1.0588 | 288 | 0.7423 | 0.6863 | 0.6761 | 0.6534 |
66
+ | 0.578 | 1.1765 | 320 | 0.7521 | 0.6661 | 0.6637 | 0.6503 |
67
+ | 0.6434 | 1.2941 | 352 | 0.7673 | 0.6771 | 0.6570 | 0.6373 |
68
+ | 0.5519 | 1.4118 | 384 | 0.8297 | 0.6513 | 0.6560 | 0.6548 |
69
+ | 0.5714 | 1.5294 | 416 | 0.7851 | 0.6531 | 0.6556 | 0.6472 |
70
+ | 0.583 | 1.6471 | 448 | 0.7941 | 0.6587 | 0.6585 | 0.6472 |
71
+ | 0.6426 | 1.7647 | 480 | 0.7596 | 0.6605 | 0.6623 | 0.6575 |
72
+ | 0.5681 | 1.8824 | 512 | 0.7831 | 0.6679 | 0.6672 | 0.6567 |
73
+ | 0.5424 | 2.0 | 544 | 0.7885 | 0.6439 | 0.6470 | 0.6472 |
74
+ | 0.4013 | 2.1176 | 576 | 0.8117 | 0.6771 | 0.6780 | 0.6696 |
75
+ | 0.369 | 2.2353 | 608 | 0.8527 | 0.6845 | 0.6856 | 0.6792 |
76
+ | 0.3405 | 2.3529 | 640 | 0.8640 | 0.6697 | 0.6652 | 0.6553 |
77
+ | 0.3633 | 2.4706 | 672 | 0.8678 | 0.6753 | 0.6682 | 0.6545 |
78
+ | 0.3827 | 2.5882 | 704 | 0.8551 | 0.6679 | 0.6649 | 0.6577 |
79
+ | 0.3826 | 2.7059 | 736 | 0.8680 | 0.6790 | 0.6770 | 0.6708 |
80
+ | 0.4146 | 2.8235 | 768 | 0.8515 | 0.6808 | 0.6801 | 0.6744 |
81
+ | 0.3592 | 2.9412 | 800 | 0.8463 | 0.6771 | 0.6762 | 0.6711 |
82
 
83
 
84
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:29530d3b4253a1caae0c237f7e1143a37f52d4f09756258657d31f6a163ca1df
3
  size 497804292
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6ccb91bbf3b49558989cd45acda6cd02d7769eba867df592e36552d0a4a0ea0
3
  size 497804292
tokenizer.json CHANGED
@@ -1,11 +1,6 @@
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
  {
 
1
  {
2
  "version": "1.0",
3
+ "truncation": null,
 
 
 
 
 
4
  "padding": null,
5
  "added_tokens": [
6
  {