haseong8012
commited on
Commit
•
10edb00
1
Parent(s):
60d82bd
End of training
Browse files- README.md +127 -0
- pytorch_model.bin +1 -1
README.md
ADDED
@@ -0,0 +1,127 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
datasets:
|
7 |
+
- zeroth_korean
|
8 |
+
metrics:
|
9 |
+
- wer
|
10 |
+
model-index:
|
11 |
+
- name: wav2vec2-large-xlsr-53-fine-tune_korean_byAILAB
|
12 |
+
results:
|
13 |
+
- task:
|
14 |
+
name: Automatic Speech Recognition
|
15 |
+
type: automatic-speech-recognition
|
16 |
+
dataset:
|
17 |
+
name: zeroth_korean
|
18 |
+
type: zeroth_korean
|
19 |
+
config: clean
|
20 |
+
split: test
|
21 |
+
args: clean
|
22 |
+
metrics:
|
23 |
+
- name: Wer
|
24 |
+
type: wer
|
25 |
+
value: 0.8577021532901672
|
26 |
+
---
|
27 |
+
|
28 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
+
should probably proofread and complete it, then remove this comment. -->
|
30 |
+
|
31 |
+
# wav2vec2-large-xlsr-53-fine-tune_korean_byAILAB
|
32 |
+
|
33 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the zeroth_korean dataset.
|
34 |
+
It achieves the following results on the evaluation set:
|
35 |
+
- Loss: 1.4406
|
36 |
+
- Wer: 0.8577
|
37 |
+
|
38 |
+
## Model description
|
39 |
+
|
40 |
+
More information needed
|
41 |
+
|
42 |
+
## Intended uses & limitations
|
43 |
+
|
44 |
+
More information needed
|
45 |
+
|
46 |
+
## Training and evaluation data
|
47 |
+
|
48 |
+
More information needed
|
49 |
+
|
50 |
+
## Training procedure
|
51 |
+
|
52 |
+
### Training hyperparameters
|
53 |
+
|
54 |
+
The following hyperparameters were used during training:
|
55 |
+
- learning_rate: 0.0003
|
56 |
+
- train_batch_size: 16
|
57 |
+
- eval_batch_size: 8
|
58 |
+
- seed: 42
|
59 |
+
- gradient_accumulation_steps: 2
|
60 |
+
- total_train_batch_size: 32
|
61 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
62 |
+
- lr_scheduler_type: linear
|
63 |
+
- lr_scheduler_warmup_steps: 500
|
64 |
+
- num_epochs: 50
|
65 |
+
|
66 |
+
### Training results
|
67 |
+
|
68 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
69 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
70 |
+
| No log | 0.99 | 38 | 54.3133 | 1.0 |
|
71 |
+
| No log | 2.0 | 77 | 33.5397 | 1.0 |
|
72 |
+
| No log | 2.99 | 115 | 19.6459 | 1.0 |
|
73 |
+
| No log | 4.0 | 154 | 11.1346 | 1.0 |
|
74 |
+
| No log | 4.99 | 192 | 5.8854 | 1.0 |
|
75 |
+
| No log | 6.0 | 231 | 4.8784 | 1.0 |
|
76 |
+
| No log | 6.99 | 269 | 4.8369 | 1.0 |
|
77 |
+
| No log | 8.0 | 308 | 4.8535 | 1.0 |
|
78 |
+
| No log | 8.99 | 346 | 4.8388 | 1.0 |
|
79 |
+
| No log | 10.0 | 385 | 4.8360 | 1.0 |
|
80 |
+
| 15.1801 | 10.99 | 423 | 4.7653 | 1.0 |
|
81 |
+
| 15.1801 | 12.0 | 462 | 4.7385 | 1.0 |
|
82 |
+
| 15.1801 | 12.99 | 500 | 4.6927 | 0.9989 |
|
83 |
+
| 15.1801 | 14.0 | 539 | 4.6673 | 0.9991 |
|
84 |
+
| 15.1801 | 14.99 | 577 | 4.6948 | 0.9991 |
|
85 |
+
| 15.1801 | 16.0 | 616 | 4.6713 | 0.9991 |
|
86 |
+
| 15.1801 | 16.99 | 654 | 4.6603 | 1.0 |
|
87 |
+
| 15.1801 | 18.0 | 693 | 4.6428 | 0.9995 |
|
88 |
+
| 15.1801 | 18.99 | 731 | 4.6520 | 0.9994 |
|
89 |
+
| 15.1801 | 20.0 | 770 | 4.6554 | 0.9967 |
|
90 |
+
| 4.3888 | 20.99 | 808 | 4.6054 | 0.9998 |
|
91 |
+
| 4.3888 | 22.0 | 847 | 4.5723 | 0.9976 |
|
92 |
+
| 4.3888 | 22.99 | 885 | 4.4586 | 0.9967 |
|
93 |
+
| 4.3888 | 24.0 | 924 | 4.2547 | 0.9934 |
|
94 |
+
| 4.3888 | 24.99 | 962 | 3.6554 | 0.9931 |
|
95 |
+
| 4.3888 | 26.0 | 1001 | 2.8387 | 1.0084 |
|
96 |
+
| 4.3888 | 26.99 | 1039 | 2.4191 | 1.0551 |
|
97 |
+
| 4.3888 | 28.0 | 1078 | 2.0997 | 1.0197 |
|
98 |
+
| 4.3888 | 28.99 | 1116 | 2.0103 | 1.0176 |
|
99 |
+
| 4.3888 | 30.0 | 1155 | 1.8189 | 0.9461 |
|
100 |
+
| 4.3888 | 30.99 | 1193 | 1.7623 | 0.9726 |
|
101 |
+
| 2.7217 | 32.0 | 1232 | 1.7383 | 0.9976 |
|
102 |
+
| 2.7217 | 32.99 | 1270 | 1.6522 | 0.9584 |
|
103 |
+
| 2.7217 | 34.0 | 1309 | 1.5558 | 0.9193 |
|
104 |
+
| 2.7217 | 34.99 | 1347 | 1.5811 | 0.9440 |
|
105 |
+
| 2.7217 | 36.0 | 1386 | 1.5208 | 0.9158 |
|
106 |
+
| 2.7217 | 36.99 | 1424 | 1.5088 | 0.9038 |
|
107 |
+
| 2.7217 | 38.0 | 1463 | 1.5039 | 0.9086 |
|
108 |
+
| 2.7217 | 38.99 | 1501 | 1.4853 | 0.8987 |
|
109 |
+
| 2.7217 | 40.0 | 1540 | 1.4799 | 0.8847 |
|
110 |
+
| 2.7217 | 40.99 | 1578 | 1.4259 | 0.8694 |
|
111 |
+
| 0.7635 | 42.0 | 1617 | 1.4878 | 0.8883 |
|
112 |
+
| 0.7635 | 42.99 | 1655 | 1.4394 | 0.8693 |
|
113 |
+
| 0.7635 | 44.0 | 1694 | 1.4623 | 0.8743 |
|
114 |
+
| 0.7635 | 44.99 | 1732 | 1.4495 | 0.8710 |
|
115 |
+
| 0.7635 | 46.0 | 1771 | 1.4463 | 0.8655 |
|
116 |
+
| 0.7635 | 46.99 | 1809 | 1.4553 | 0.8704 |
|
117 |
+
| 0.7635 | 48.0 | 1848 | 1.4500 | 0.8646 |
|
118 |
+
| 0.7635 | 48.99 | 1886 | 1.4387 | 0.8566 |
|
119 |
+
| 0.7635 | 49.35 | 1900 | 1.4406 | 0.8577 |
|
120 |
+
|
121 |
+
|
122 |
+
### Framework versions
|
123 |
+
|
124 |
+
- Transformers 4.33.2
|
125 |
+
- Pytorch 1.12.1
|
126 |
+
- Datasets 2.14.5
|
127 |
+
- Tokenizers 0.13.3
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1266847409
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:da8b7d81865ea8161cc3919bd41c35a26012f01531fb88c6aa3bbb66002b82ba
|
3 |
size 1266847409
|