Commit
•
71890ea
1
Parent(s):
16a289b
End of training
Browse files
README.md
CHANGED
@@ -5,6 +5,11 @@ license: mit
|
|
5 |
base_model: xlm-roberta-large
|
6 |
tags:
|
7 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
|
8 |
model-index:
|
9 |
- name: xlm-roberta-large-ner-demo
|
10 |
results: []
|
@@ -17,16 +22,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [xlm-roberta-large](https://huggingface.co/xlm-roberta-large) on the None dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
-
|
21 |
-
-
|
22 |
-
-
|
23 |
-
-
|
24 |
-
-
|
25 |
-
- eval_runtime: 99.382
|
26 |
-
- eval_samples_per_second: 25.568
|
27 |
-
- eval_steps_per_second: 0.805
|
28 |
-
- epoch: 2.0
|
29 |
-
- step: 954
|
30 |
|
31 |
## Model description
|
32 |
|
@@ -51,7 +51,18 @@ The following hyperparameters were used during training:
|
|
51 |
- seed: 42
|
52 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
53 |
- lr_scheduler_type: linear
|
54 |
-
- num_epochs:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
55 |
|
56 |
### Framework versions
|
57 |
|
|
|
5 |
base_model: xlm-roberta-large
|
6 |
tags:
|
7 |
- generated_from_trainer
|
8 |
+
metrics:
|
9 |
+
- precision
|
10 |
+
- recall
|
11 |
+
- f1
|
12 |
+
- accuracy
|
13 |
model-index:
|
14 |
- name: xlm-roberta-large-ner-demo
|
15 |
results: []
|
|
|
22 |
|
23 |
This model is a fine-tuned version of [xlm-roberta-large](https://huggingface.co/xlm-roberta-large) on the None dataset.
|
24 |
It achieves the following results on the evaluation set:
|
25 |
+
- Loss: 0.0976
|
26 |
+
- Precision: 0.9340
|
27 |
+
- Recall: 0.9404
|
28 |
+
- F1: 0.9372
|
29 |
+
- Accuracy: 0.9816
|
|
|
|
|
|
|
|
|
|
|
30 |
|
31 |
## Model description
|
32 |
|
|
|
51 |
- seed: 42
|
52 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
53 |
- lr_scheduler_type: linear
|
54 |
+
- num_epochs: 5
|
55 |
+
|
56 |
+
### Training results
|
57 |
+
|
58 |
+
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
59 |
+
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
60 |
+
| 0.1657 | 1.0 | 477 | 0.0866 | 0.8655 | 0.8978 | 0.8814 | 0.9752 |
|
61 |
+
| 0.0716 | 2.0 | 954 | 0.0801 | 0.9135 | 0.9283 | 0.9208 | 0.9796 |
|
62 |
+
| 0.0448 | 3.0 | 1431 | 0.0814 | 0.9244 | 0.9374 | 0.9309 | 0.9805 |
|
63 |
+
| 0.0283 | 4.0 | 1908 | 0.0870 | 0.9256 | 0.9367 | 0.9311 | 0.9808 |
|
64 |
+
| 0.017 | 5.0 | 2385 | 0.0976 | 0.9340 | 0.9404 | 0.9372 | 0.9816 |
|
65 |
+
|
66 |
|
67 |
### Framework versions
|
68 |
|