Commit
·
5800c55
1
Parent(s):
02eba6b
Update README.md
Browse files
README.md
CHANGED
@@ -26,12 +26,6 @@ model-index:
|
|
26 |
- name: Test CER using LM
|
27 |
type: cer
|
28 |
value: 2.4147
|
29 |
-
- name: Test WER without LM
|
30 |
-
type: wer
|
31 |
-
value: 20.1776
|
32 |
-
- name: Test CER without LM
|
33 |
-
type: cer
|
34 |
-
value: 4.7205
|
35 |
---
|
36 |
|
37 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -53,6 +47,14 @@ Training is conducted in Google Colab, the training notebook provided in the rep
|
|
53 |
Language Model Created from texts from processed sentence in train + validation split of dataset (common voice 8.0 for Interlingua)
|
54 |
Evaluation is conducted in Notebook, you can see within the repo "notebook_evaluation_wav2vec2_ia.ipynb"
|
55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
### Training hyperparameters
|
57 |
|
58 |
The following hyperparameters were used during training:
|
|
|
26 |
- name: Test CER using LM
|
27 |
type: cer
|
28 |
value: 2.4147
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
---
|
30 |
|
31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
47 |
Language Model Created from texts from processed sentence in train + validation split of dataset (common voice 8.0 for Interlingua)
|
48 |
Evaluation is conducted in Notebook, you can see within the repo "notebook_evaluation_wav2vec2_ia.ipynb"
|
49 |
|
50 |
+
Test WER without LM
|
51 |
+
wer = 20.1776 %
|
52 |
+
cer = 4.7205 %
|
53 |
+
|
54 |
+
Test WER using
|
55 |
+
wer = 8.6074 %
|
56 |
+
cer = 2.4147 %
|
57 |
+
|
58 |
### Training hyperparameters
|
59 |
|
60 |
The following hyperparameters were used during training:
|