update model card README.md
Browse files
README.md
CHANGED
@@ -1,41 +1,38 @@
|
|
1 |
---
|
2 |
-
language:
|
3 |
-
- be
|
4 |
license: apache-2.0
|
5 |
tags:
|
6 |
-
- whisper-event
|
7 |
- generated_from_trainer
|
8 |
datasets:
|
9 |
-
-
|
10 |
metrics:
|
11 |
- wer
|
12 |
model-index:
|
13 |
-
- name:
|
14 |
results:
|
15 |
- task:
|
16 |
name: Automatic Speech Recognition
|
17 |
type: automatic-speech-recognition
|
18 |
dataset:
|
19 |
-
name:
|
20 |
-
type:
|
21 |
config: be
|
22 |
split: validation
|
23 |
args: be
|
24 |
metrics:
|
25 |
- name: Wer
|
26 |
type: wer
|
27 |
-
value:
|
28 |
---
|
29 |
|
30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
31 |
should probably proofread and complete it, then remove this comment. -->
|
32 |
|
33 |
-
#
|
34 |
|
35 |
-
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the
|
36 |
It achieves the following results on the evaluation set:
|
37 |
-
- Loss: 0.
|
38 |
-
- Wer:
|
39 |
|
40 |
## Model description
|
41 |
|
@@ -61,23 +58,33 @@ The following hyperparameters were used during training:
|
|
61 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
62 |
- lr_scheduler_type: linear
|
63 |
- lr_scheduler_warmup_steps: 10
|
64 |
-
- training_steps:
|
65 |
- mixed_precision_training: Native AMP
|
66 |
|
67 |
### Training results
|
68 |
|
69 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
70 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
71 |
-
| 2.
|
72 |
-
| 1.
|
73 |
-
| 0.
|
74 |
-
| 0.
|
75 |
-
| 0.
|
76 |
-
| 0.
|
77 |
-
| 0.
|
78 |
-
| 0.
|
79 |
-
| 0.
|
80 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
81 |
|
82 |
|
83 |
### Framework versions
|
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
tags:
|
|
|
4 |
- generated_from_trainer
|
5 |
datasets:
|
6 |
+
- common_voice_11_0
|
7 |
metrics:
|
8 |
- wer
|
9 |
model-index:
|
10 |
+
- name: whisper-tiny-be-test
|
11 |
results:
|
12 |
- task:
|
13 |
name: Automatic Speech Recognition
|
14 |
type: automatic-speech-recognition
|
15 |
dataset:
|
16 |
+
name: common_voice_11_0
|
17 |
+
type: common_voice_11_0
|
18 |
config: be
|
19 |
split: validation
|
20 |
args: be
|
21 |
metrics:
|
22 |
- name: Wer
|
23 |
type: wer
|
24 |
+
value: 51.46520146520146
|
25 |
---
|
26 |
|
27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
28 |
should probably proofread and complete it, then remove this comment. -->
|
29 |
|
30 |
+
# whisper-tiny-be-test
|
31 |
|
32 |
+
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the common_voice_11_0 dataset.
|
33 |
It achieves the following results on the evaluation set:
|
34 |
+
- Loss: 0.4624
|
35 |
+
- Wer: 51.4652
|
36 |
|
37 |
## Model description
|
38 |
|
|
|
58 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
59 |
- lr_scheduler_type: linear
|
60 |
- lr_scheduler_warmup_steps: 10
|
61 |
+
- training_steps: 200
|
62 |
- mixed_precision_training: Native AMP
|
63 |
|
64 |
### Training results
|
65 |
|
66 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
67 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
68 |
+
| 2.5366 | 0.05 | 10 | 1.5402 | 94.5055 |
|
69 |
+
| 1.3721 | 0.1 | 20 | 1.0021 | 75.8242 |
|
70 |
+
| 0.9921 | 0.15 | 30 | 0.8322 | 75.0916 |
|
71 |
+
| 0.9844 | 0.2 | 40 | 0.8080 | 72.8938 |
|
72 |
+
| 0.7071 | 0.25 | 50 | 0.7862 | 77.2894 |
|
73 |
+
| 0.7998 | 0.3 | 60 | 0.7052 | 68.8645 |
|
74 |
+
| 0.6935 | 0.35 | 70 | 0.6781 | 64.2857 |
|
75 |
+
| 0.81 | 0.4 | 80 | 0.6341 | 63.5531 |
|
76 |
+
| 0.6133 | 0.45 | 90 | 0.6083 | 62.6374 |
|
77 |
+
| 0.6675 | 0.5 | 100 | 0.5851 | 62.8205 |
|
78 |
+
| 0.5577 | 0.55 | 110 | 0.5651 | 59.3407 |
|
79 |
+
| 0.6473 | 0.6 | 120 | 0.5638 | 58.0586 |
|
80 |
+
| 0.6018 | 0.65 | 130 | 0.5434 | 53.8462 |
|
81 |
+
| 0.5918 | 0.7 | 140 | 0.5385 | 54.9451 |
|
82 |
+
| 0.5654 | 0.75 | 150 | 0.5200 | 58.0586 |
|
83 |
+
| 0.587 | 0.8 | 160 | 0.4974 | 57.1429 |
|
84 |
+
| 0.6157 | 0.85 | 170 | 0.4834 | 53.2967 |
|
85 |
+
| 0.6803 | 0.9 | 180 | 0.4852 | 55.8608 |
|
86 |
+
| 0.4813 | 0.95 | 190 | 0.4686 | 51.2821 |
|
87 |
+
| 0.4952 | 1.0 | 200 | 0.4624 | 51.4652 |
|
88 |
|
89 |
|
90 |
### Framework versions
|
train.log
CHANGED
@@ -207,3 +207,5 @@
|
|
207 |
{'loss': 0.4813, 'learning_rate': 6.842105263157896e-06, 'epoch': 0.95}
|
208 |
{'eval_loss': 0.4685819447040558, 'eval_wer': 51.28205128205128, 'eval_runtime': 17.9367, 'eval_samples_per_second': 3.568, 'eval_steps_per_second': 0.112, 'epoch': 0.95}
|
209 |
{'loss': 0.4952, 'learning_rate': 1.5789473684210528e-06, 'epoch': 1.0}
|
|
|
|
|
|
207 |
{'loss': 0.4813, 'learning_rate': 6.842105263157896e-06, 'epoch': 0.95}
|
208 |
{'eval_loss': 0.4685819447040558, 'eval_wer': 51.28205128205128, 'eval_runtime': 17.9367, 'eval_samples_per_second': 3.568, 'eval_steps_per_second': 0.112, 'epoch': 0.95}
|
209 |
{'loss': 0.4952, 'learning_rate': 1.5789473684210528e-06, 'epoch': 1.0}
|
210 |
+
{'eval_loss': 0.4624484181404114, 'eval_wer': 51.46520146520146, 'eval_runtime': 19.165, 'eval_samples_per_second': 3.339, 'eval_steps_per_second': 0.104, 'epoch': 1.0}
|
211 |
+
{'train_runtime': 2053.4009, 'train_samples_per_second': 3.117, 'train_steps_per_second': 0.097, 'train_loss': 0.8012711083889008, 'epoch': 1.0}
|