End of training
Browse files
README.md
CHANGED
@@ -5,36 +5,9 @@ tags:
|
|
5 |
- generated_from_trainer
|
6 |
datasets:
|
7 |
- ner
|
8 |
-
metrics:
|
9 |
-
- precision
|
10 |
-
- recall
|
11 |
-
- f1
|
12 |
-
- accuracy
|
13 |
model-index:
|
14 |
- name: Bert-NER
|
15 |
-
results:
|
16 |
-
- task:
|
17 |
-
name: Token Classification
|
18 |
-
type: token-classification
|
19 |
-
dataset:
|
20 |
-
name: ner
|
21 |
-
type: ner
|
22 |
-
config: indian_names
|
23 |
-
split: train
|
24 |
-
args: indian_names
|
25 |
-
metrics:
|
26 |
-
- name: Precision
|
27 |
-
type: precision
|
28 |
-
value: 0.9952566491614433
|
29 |
-
- name: Recall
|
30 |
-
type: recall
|
31 |
-
value: 0.9982668388499966
|
32 |
-
- name: F1
|
33 |
-
type: f1
|
34 |
-
value: 0.9967594713357425
|
35 |
-
- name: Accuracy
|
36 |
-
type: accuracy
|
37 |
-
value: 0.9982867445455388
|
38 |
---
|
39 |
|
40 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -43,12 +16,6 @@ should probably proofread and complete it, then remove this comment. -->
|
|
43 |
# Bert-NER
|
44 |
|
45 |
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the ner dataset.
|
46 |
-
It achieves the following results on the evaluation set:
|
47 |
-
- Loss: 0.0041
|
48 |
-
- Precision: 0.9953
|
49 |
-
- Recall: 0.9983
|
50 |
-
- F1: 0.9968
|
51 |
-
- Accuracy: 0.9983
|
52 |
|
53 |
## Model description
|
54 |
|
@@ -67,28 +34,19 @@ More information needed
|
|
67 |
### Training hyperparameters
|
68 |
|
69 |
The following hyperparameters were used during training:
|
70 |
-
- learning_rate:
|
71 |
- train_batch_size: 32
|
72 |
- eval_batch_size: 32
|
73 |
- seed: 42
|
74 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
75 |
- lr_scheduler_type: linear
|
76 |
-
- num_epochs:
|
77 |
|
78 |
### Training results
|
79 |
|
80 |
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
81 |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
82 |
-
| No log | 1.0 | 486 | 0.
|
83 |
-
| 0.04 | 2.0 | 972 | 0.0328 | 0.9846 | 0.9700 | 0.9772 | 0.9886 |
|
84 |
-
| 0.0362 | 3.0 | 1458 | 0.0241 | 0.9957 | 0.9719 | 0.9837 | 0.9918 |
|
85 |
-
| 0.0302 | 4.0 | 1944 | 0.0163 | 0.9970 | 0.9787 | 0.9877 | 0.9937 |
|
86 |
-
| 0.0249 | 5.0 | 2430 | 0.0134 | 0.9914 | 0.9888 | 0.9901 | 0.9949 |
|
87 |
-
| 0.0188 | 6.0 | 2916 | 0.0092 | 0.9929 | 0.9934 | 0.9932 | 0.9965 |
|
88 |
-
| 0.0145 | 7.0 | 3402 | 0.0070 | 0.9934 | 0.9952 | 0.9943 | 0.9971 |
|
89 |
-
| 0.0115 | 8.0 | 3888 | 0.0060 | 0.9941 | 0.9966 | 0.9954 | 0.9976 |
|
90 |
-
| 0.009 | 9.0 | 4374 | 0.0048 | 0.9955 | 0.9973 | 0.9964 | 0.9981 |
|
91 |
-
| 0.0072 | 10.0 | 4860 | 0.0041 | 0.9953 | 0.9983 | 0.9968 | 0.9983 |
|
92 |
|
93 |
|
94 |
### Framework versions
|
|
|
5 |
- generated_from_trainer
|
6 |
datasets:
|
7 |
- ner
|
|
|
|
|
|
|
|
|
|
|
8 |
model-index:
|
9 |
- name: Bert-NER
|
10 |
+
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
---
|
12 |
|
13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
16 |
# Bert-NER
|
17 |
|
18 |
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the ner dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
## Model description
|
21 |
|
|
|
34 |
### Training hyperparameters
|
35 |
|
36 |
The following hyperparameters were used during training:
|
37 |
+
- learning_rate: 1e-05
|
38 |
- train_batch_size: 32
|
39 |
- eval_batch_size: 32
|
40 |
- seed: 42
|
41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
42 |
- lr_scheduler_type: linear
|
43 |
+
- num_epochs: 1
|
44 |
|
45 |
### Training results
|
46 |
|
47 |
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
48 |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
49 |
+
| No log | 1.0 | 486 | 0.0032 | 0.9972 | 0.9978 | 0.9975 | 0.9987 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
|
51 |
|
52 |
### Framework versions
|