jethrowang
commited on
Commit
•
e8879d0
1
Parent(s):
9d89b8b
End of training
Browse files
README.md
CHANGED
@@ -8,19 +8,19 @@ tags:
|
|
8 |
datasets:
|
9 |
- formospeech/hat_asr_aligned
|
10 |
model-index:
|
11 |
-
- name: Whisper Tiny Hakka
|
12 |
results: []
|
13 |
---
|
14 |
|
15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
16 |
should probably proofread and complete it, then remove this comment. -->
|
17 |
|
18 |
-
# Whisper Tiny Hakka
|
19 |
|
20 |
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the HAT ASR Aligned dataset.
|
21 |
It achieves the following results on the evaluation set:
|
22 |
-
- Loss: 0.
|
23 |
-
- Cer:
|
24 |
|
25 |
## Model description
|
26 |
|
@@ -39,7 +39,7 @@ More information needed
|
|
39 |
### Training hyperparameters
|
40 |
|
41 |
The following hyperparameters were used during training:
|
42 |
-
- learning_rate:
|
43 |
- train_batch_size: 64
|
44 |
- eval_batch_size: 32
|
45 |
- seed: 42
|
@@ -53,26 +53,26 @@ The following hyperparameters were used during training:
|
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
55 |
|:-------------:|:-------:|:----:|:---------------:|:-------:|
|
56 |
-
|
|
57 |
-
| 0.
|
58 |
-
| 0.
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
-
| 0.
|
62 |
-
| 0.
|
63 |
-
| 0.
|
64 |
-
| 0.
|
65 |
-
| 0.
|
66 |
-
| 0.
|
67 |
-
| 0.
|
68 |
-
| 0.
|
69 |
-
| 0.
|
70 |
-
| 0.
|
71 |
-
| 0.
|
72 |
-
| 0.
|
73 |
-
| 0.
|
74 |
-
| 0.
|
75 |
-
| 0.
|
76 |
|
77 |
|
78 |
### Framework versions
|
|
|
8 |
datasets:
|
9 |
- formospeech/hat_asr_aligned
|
10 |
model-index:
|
11 |
+
- name: Whisper Tiny Hakka Simulated Webcam
|
12 |
results: []
|
13 |
---
|
14 |
|
15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
16 |
should probably proofread and complete it, then remove this comment. -->
|
17 |
|
18 |
+
# Whisper Tiny Hakka Simulated Webcam
|
19 |
|
20 |
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the HAT ASR Aligned dataset.
|
21 |
It achieves the following results on the evaluation set:
|
22 |
+
- Loss: 0.1323
|
23 |
+
- Cer: 7.3156
|
24 |
|
25 |
## Model description
|
26 |
|
|
|
39 |
### Training hyperparameters
|
40 |
|
41 |
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 0.0001
|
43 |
- train_batch_size: 64
|
44 |
- eval_batch_size: 32
|
45 |
- seed: 42
|
|
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
55 |
|:-------------:|:-------:|:----:|:---------------:|:-------:|
|
56 |
+
| 0.2495 | 0.9980 | 488 | 0.3384 | 23.3014 |
|
57 |
+
| 0.1352 | 1.9959 | 976 | 0.2472 | 16.6320 |
|
58 |
+
| 0.0854 | 2.9939 | 1464 | 0.2133 | 16.6389 |
|
59 |
+
| 0.0447 | 3.9918 | 1952 | 0.1956 | 22.4831 |
|
60 |
+
| 0.0282 | 4.9898 | 2440 | 0.1921 | 12.4096 |
|
61 |
+
| 0.0167 | 5.9877 | 2928 | 0.1670 | 10.9682 |
|
62 |
+
| 0.0115 | 6.9857 | 3416 | 0.1833 | 10.5590 |
|
63 |
+
| 0.0078 | 7.9836 | 3904 | 0.1591 | 9.6204 |
|
64 |
+
| 0.0057 | 8.9816 | 4392 | 0.1568 | 10.5324 |
|
65 |
+
| 0.0037 | 9.9796 | 4880 | 0.1684 | 10.1371 |
|
66 |
+
| 0.0036 | 10.9775 | 5368 | 0.1626 | 10.8352 |
|
67 |
+
| 0.0015 | 11.9755 | 5856 | 0.1451 | 10.4226 |
|
68 |
+
| 0.0016 | 12.9734 | 6344 | 0.1562 | 10.2099 |
|
69 |
+
| 0.0007 | 13.9714 | 6832 | 0.1575 | 8.7731 |
|
70 |
+
| 0.0004 | 14.9693 | 7320 | 0.1395 | 9.8597 |
|
71 |
+
| 0.0006 | 15.9673 | 7808 | 0.1421 | 8.3316 |
|
72 |
+
| 0.0003 | 16.9652 | 8296 | 0.1345 | 7.5433 |
|
73 |
+
| 0.0001 | 17.9632 | 8784 | 0.1322 | 7.8692 |
|
74 |
+
| 0.0001 | 18.9611 | 9272 | 0.1326 | 7.3630 |
|
75 |
+
| 0.0001 | 19.9591 | 9760 | 0.1323 | 7.3156 |
|
76 |
|
77 |
|
78 |
### Framework versions
|