napsternxg commited on
Commit
8f21fde
1 Parent(s): 38d94d2

WNUT 17 training

Browse files
README.md CHANGED
@@ -1,3 +1,89 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - wnut_17
7
+ model-index:
8
+ - name: fine_tune_bert_output
9
+ results: []
10
  ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # fine_tune_bert_output
16
+
17
+ This model is a fine-tuned version of [vinai/bertweet-base](https://huggingface.co/vinai/bertweet-base) on the wnut_17 dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.3504
20
+ - Overall Precision: 0.6850
21
+ - Overall Recall: 0.6196
22
+ - Overall F1: 0.6507
23
+ - Overall Accuracy: 0.9502
24
+ - Corporation F1: 0.2626
25
+ - Creative-work F1: 0.4460
26
+ - Group F1: 0.3692
27
+ - Location F1: 0.7283
28
+ - Person F1: 0.7928
29
+ - Product F1: 0.3852
30
+
31
+ ## Model description
32
+
33
+ More information needed
34
+
35
+ ## Intended uses & limitations
36
+
37
+ More information needed
38
+
39
+ ## Training and evaluation data
40
+
41
+ More information needed
42
+
43
+ ## Training procedure
44
+
45
+ ### Training hyperparameters
46
+
47
+ The following hyperparameters were used during training:
48
+ - learning_rate: 2e-05
49
+ - train_batch_size: 16
50
+ - eval_batch_size: 16
51
+ - seed: 42
52
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
+ - lr_scheduler_type: linear
54
+ - num_epochs: 100
55
+
56
+ ### Training results
57
+
58
+ | Training Loss | Epoch | Step | Validation Loss | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy | Corporation F1 | Creative-work F1 | Group F1 | Location F1 | Person F1 | Product F1 |
59
+ |:-------------:|:-----:|:----:|:---------------:|:-----------------:|:--------------:|:----------:|:----------------:|:--------------:|:----------------:|:--------:|:-----------:|:---------:|:----------:|
60
+ | 0.2954 | 1.0 | 213 | 0.4357 | 0.0 | 0.0 | 0.0 | 0.8979 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
61
+ | 0.1654 | 2.0 | 426 | 0.3435 | 0.5890 | 0.3929 | 0.4714 | 0.9307 | 0.0 | 0.0 | 0.0 | 0.4444 | 0.6181 | 0.0 |
62
+ | 0.1213 | 3.0 | 639 | 0.2981 | 0.6425 | 0.4776 | 0.5479 | 0.9380 | 0.1961 | 0.0565 | 0.1720 | 0.5507 | 0.7285 | 0.0 |
63
+ | 0.051 | 4.0 | 852 | 0.2958 | 0.7020 | 0.4892 | 0.5766 | 0.9430 | 0.2540 | 0.3458 | 0.2157 | 0.63 | 0.73 | 0.2388 |
64
+ | 0.0503 | 5.0 | 1065 | 0.3154 | 0.6659 | 0.5033 | 0.5733 | 0.9429 | 0.2373 | 0.3365 | 0.3212 | 0.6 | 0.7223 | 0.2896 |
65
+ | 0.026 | 6.0 | 1278 | 0.2787 | 0.6768 | 0.5706 | 0.6192 | 0.9479 | 0.256 | 0.4163 | 0.3699 | 0.7107 | 0.7717 | 0.3052 |
66
+ | 0.0271 | 7.0 | 1491 | 0.2940 | 0.7122 | 0.5714 | 0.6341 | 0.9486 | 0.25 | 0.3982 | 0.3676 | 0.7033 | 0.7717 | 0.3460 |
67
+ | 0.0239 | 8.0 | 1704 | 0.2872 | 0.6210 | 0.5839 | 0.6019 | 0.9461 | 0.2722 | 0.4188 | 0.2902 | 0.7389 | 0.7593 | 0.3597 |
68
+ | 0.0155 | 9.0 | 1917 | 0.2910 | 0.6863 | 0.5905 | 0.6348 | 0.9494 | 0.2623 | 0.4959 | 0.3433 | 0.6878 | 0.7708 | 0.3614 |
69
+ | 0.0122 | 10.0 | 2130 | 0.3067 | 0.6966 | 0.5797 | 0.6328 | 0.9490 | 0.2558 | 0.4609 | 0.3309 | 0.6842 | 0.7645 | 0.3835 |
70
+ | 0.0161 | 11.0 | 2343 | 0.2782 | 0.6637 | 0.6096 | 0.6355 | 0.9502 | 0.3103 | 0.4710 | 0.4275 | 0.6811 | 0.7688 | 0.4110 |
71
+ | 0.0232 | 12.0 | 2556 | 0.3123 | 0.6832 | 0.5822 | 0.6287 | 0.9495 | 0.3235 | 0.4722 | 0.4 | 0.7374 | 0.7568 | 0.4321 |
72
+ | 0.012 | 13.0 | 2769 | 0.3161 | 0.6663 | 0.5573 | 0.6070 | 0.9475 | 0.2128 | 0.4474 | 0.3289 | 0.7243 | 0.7400 | 0.4130 |
73
+ | 0.0051 | 14.0 | 2982 | 0.3241 | 0.7131 | 0.5864 | 0.6436 | 0.9498 | 0.3125 | 0.4839 | 0.3934 | 0.6952 | 0.7700 | 0.3802 |
74
+ | 0.0096 | 15.0 | 3195 | 0.3140 | 0.6924 | 0.6321 | 0.6609 | 0.9520 | 0.3036 | 0.4542 | 0.3576 | 0.7312 | 0.7953 | 0.432 |
75
+ | 0.0045 | 16.0 | 3408 | 0.3356 | 0.6917 | 0.6038 | 0.6448 | 0.9498 | 0.2899 | 0.4858 | 0.3582 | 0.6952 | 0.7840 | 0.4275 |
76
+ | 0.0066 | 17.0 | 3621 | 0.3308 | 0.6738 | 0.6262 | 0.6492 | 0.9510 | 0.2957 | 0.4333 | 0.3673 | 0.6989 | 0.7954 | 0.3320 |
77
+ | 0.0068 | 18.0 | 3834 | 0.3527 | 0.7157 | 0.6063 | 0.6565 | 0.9505 | 0.2353 | 0.4211 | 0.4058 | 0.7182 | 0.7991 | 0.3825 |
78
+ | 0.0033 | 19.0 | 4047 | 0.3522 | 0.7298 | 0.6013 | 0.6594 | 0.9514 | 0.2093 | 0.4454 | 0.4390 | 0.7006 | 0.7912 | 0.4000 |
79
+ | 0.0067 | 20.0 | 4260 | 0.3721 | 0.6859 | 0.5839 | 0.6308 | 0.9476 | 0.2774 | 0.3442 | 0.3881 | 0.7254 | 0.7844 | 0.3140 |
80
+ | 0.0083 | 21.0 | 4473 | 0.3504 | 0.6850 | 0.6196 | 0.6507 | 0.9502 | 0.2626 | 0.4460 | 0.3692 | 0.7283 | 0.7928 | 0.3852 |
81
+
82
+
83
+ ### Framework versions
84
+
85
+ - Transformers 4.17.0
86
+ - Pytorch 1.11.0+cu113
87
+ - Datasets 2.0.0
88
+ - Tokenizers 0.11.6
89
+
added_tokens.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"<mask>": 64000}
all_results.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 21.0,
3
+ "total_flos": 1641754006788696.0,
4
+ "train_corporation_f1": 0.9816971713810316,
5
+ "train_creative-work_f1": 0.9383886255924171,
6
+ "train_group_f1": 0.9757343550447,
7
+ "train_location_f1": 0.9845758354755784,
8
+ "train_loss": 0.008754712529480457,
9
+ "train_overall_accuracy": 0.9985107796641235,
10
+ "train_overall_f1": 0.9778966131907308,
11
+ "train_overall_precision": 0.9723502304147466,
12
+ "train_overall_recall": 0.9835066332018645,
13
+ "train_person_f1": 0.9893795416433762,
14
+ "train_product_f1": 0.9455337690631808,
15
+ "train_runtime": 16.7057,
16
+ "train_samples_per_second": 203.164,
17
+ "train_steps_per_second": 12.75,
18
+ "validation_corporation_f1": 0.3103448275862069,
19
+ "validation_creative-work_f1": 0.47104247104247104,
20
+ "validation_group_f1": 0.42748091603053434,
21
+ "validation_location_f1": 0.6810810810810811,
22
+ "validation_loss": 0.2782056927680969,
23
+ "validation_overall_accuracy": 0.9502437284508382,
24
+ "validation_overall_f1": 0.6354978354978356,
25
+ "validation_overall_precision": 0.6636528028933092,
26
+ "validation_overall_recall": 0.6096345514950167,
27
+ "validation_person_f1": 0.7687548337200311,
28
+ "validation_product_f1": 0.4110429447852761,
29
+ "validation_runtime": 6.5808,
30
+ "validation_samples_per_second": 153.325,
31
+ "validation_steps_per_second": 9.725
32
+ }
bpe.codes ADDED
The diff for this file is too large to render. See raw diff
 
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/home/jupyter/bertweet-base",
3
+ "architectures": [
4
+ "RobertaForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "gradient_checkpointing": false,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 768,
14
+ "id2label": {
15
+ "0": "O",
16
+ "1": "B-corporation",
17
+ "2": "I-corporation",
18
+ "3": "B-creative-work",
19
+ "4": "I-creative-work",
20
+ "5": "B-group",
21
+ "6": "I-group",
22
+ "7": "B-location",
23
+ "8": "I-location",
24
+ "9": "B-person",
25
+ "10": "I-person",
26
+ "11": "B-product",
27
+ "12": "I-product"
28
+ },
29
+ "initializer_range": 0.02,
30
+ "intermediate_size": 3072,
31
+ "label2id": {
32
+ "LABEL_0": 0,
33
+ "LABEL_1": 1,
34
+ "LABEL_10": 10,
35
+ "LABEL_11": 11,
36
+ "LABEL_12": 12,
37
+ "LABEL_2": 2,
38
+ "LABEL_3": 3,
39
+ "LABEL_4": 4,
40
+ "LABEL_5": 5,
41
+ "LABEL_6": 6,
42
+ "LABEL_7": 7,
43
+ "LABEL_8": 8,
44
+ "LABEL_9": 9
45
+ },
46
+ "layer_norm_eps": 1e-05,
47
+ "max_position_embeddings": 130,
48
+ "model_type": "roberta",
49
+ "num_attention_heads": 12,
50
+ "num_hidden_layers": 12,
51
+ "pad_token_id": 1,
52
+ "position_embedding_type": "absolute",
53
+ "tokenizer_class": "BertweetTokenizer",
54
+ "torch_dtype": "float32",
55
+ "transformers_version": "4.17.0",
56
+ "type_vocab_size": 1,
57
+ "use_cache": true,
58
+ "vocab_size": 64001
59
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8c03ec350e0b96e236fd640760bc89a45c2e3627b25d849d74254945c595857
3
+ size 537360049
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"normalization": false, "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "special_tokens_map_file": null, "name_or_path": "/home/jupyter/bertweet-base", "tokenizer_class": "BertweetTokenizer"}
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:259a9221284a45586f6726c8755c0afe9082933f4b71b4bec61c9f3dbafca40c
3
+ size 3055
vocab.txt ADDED
The diff for this file is too large to render. See raw diff