Update README.md
Browse files
README.md
CHANGED
@@ -4,20 +4,21 @@ license: mit
|
|
4 |
datasets:
|
5 |
- AlienKevin/mixed_cantonese_and_english_speech
|
6 |
- mozilla-foundation/common_voice_17_0
|
|
|
7 |
metrics:
|
8 |
- cer
|
9 |
base_model:
|
10 |
- openai/whisper-large-v3-turbo
|
11 |
---
|
12 |
|
13 |
-
CER:
|
14 |
|
15 |
transformers-4.46.3<br/>
|
16 |
|
17 |
Train Args:<br/>
|
18 |
per_device_train_batch_size=16,<br/>
|
19 |
gradient_accumulation_steps=1,<br/>
|
20 |
-
learning_rate=
|
21 |
gradient_checkpointing=True,<br/>
|
22 |
per_device_eval_batch_size=16,<br/>
|
23 |
generation_max_length=225,<br/>
|
|
|
4 |
datasets:
|
5 |
- AlienKevin/mixed_cantonese_and_english_speech
|
6 |
- mozilla-foundation/common_voice_17_0
|
7 |
+
- mozilla-foundation/common_voice_11_0
|
8 |
metrics:
|
9 |
- cer
|
10 |
base_model:
|
11 |
- openai/whisper-large-v3-turbo
|
12 |
---
|
13 |
|
14 |
+
CER: 13.7% <br/>
|
15 |
|
16 |
transformers-4.46.3<br/>
|
17 |
|
18 |
Train Args:<br/>
|
19 |
per_device_train_batch_size=16,<br/>
|
20 |
gradient_accumulation_steps=1,<br/>
|
21 |
+
learning_rate=1e-5,<br/>
|
22 |
gradient_checkpointing=True,<br/>
|
23 |
per_device_eval_batch_size=16,<br/>
|
24 |
generation_max_length=225,<br/>
|