Rajaram Sivaramakrishnan
commited on
Commit
•
ca1429f
1
Parent(s):
ef7575a
tiny changes to README.md
Browse files
README.md
CHANGED
@@ -39,7 +39,7 @@ import torchaudio
|
|
39 |
from datasets import load_dataset
|
40 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
41 |
|
42 |
-
test_dataset = load_dataset("common_voice", "
|
43 |
|
44 |
processor = Wav2Vec2Processor.from_pretrained("Rajaram1996/wav2vec2-large-xlsr-tamil")
|
45 |
model = Wav2Vec2ForCTC.from_pretrained("Rajaram1996/wav2vec2-large-xlsr-tamil")
|
@@ -76,7 +76,7 @@ from datasets import load_dataset, load_metric
|
|
76 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
77 |
import re
|
78 |
|
79 |
-
test_dataset = load_dataset("common_voice", "
|
80 |
|
81 |
wer = load_metric("wer")
|
82 |
|
@@ -109,7 +109,7 @@ def evaluate(batch):
|
|
109 |
return batch
|
110 |
|
111 |
result = test_dataset.map(evaluate, batched=True, batch_size=8)
|
112 |
-
print(wer.compute(predictions=result["predicted"], references=result["target"]))
|
113 |
```
|
114 |
|
115 |
**Test Result**: 70.72 %
|
|
|
39 |
from datasets import load_dataset
|
40 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
41 |
|
42 |
+
test_dataset = load_dataset("common_voice", "ta", split="test[:2%]")
|
43 |
|
44 |
processor = Wav2Vec2Processor.from_pretrained("Rajaram1996/wav2vec2-large-xlsr-tamil")
|
45 |
model = Wav2Vec2ForCTC.from_pretrained("Rajaram1996/wav2vec2-large-xlsr-tamil")
|
|
|
76 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
77 |
import re
|
78 |
|
79 |
+
test_dataset = load_dataset("common_voice", "ta", split="test")
|
80 |
|
81 |
wer = load_metric("wer")
|
82 |
|
|
|
109 |
return batch
|
110 |
|
111 |
result = test_dataset.map(evaluate, batched=True, batch_size=8)
|
112 |
+
print("WER: {:2f}".format(100 * wer.compute(predictions=result["predicted"], references=result["target"])))
|
113 |
```
|
114 |
|
115 |
**Test Result**: 70.72 %
|