Update README.md
Browse files
README.md
CHANGED
@@ -6,13 +6,13 @@ tags:
|
|
6 |
- automatic-speech-recognition
|
7 |
- mozilla-foundation/common_voice_8_0
|
8 |
- generated_from_trainer
|
9 |
-
-
|
10 |
- robust-speech-event
|
11 |
- model_for_talk
|
12 |
datasets:
|
13 |
- mozilla-foundation/common_voice_8_0
|
14 |
model-index:
|
15 |
-
- name: wav2vec2-large-xls-r-1b-
|
16 |
results:
|
17 |
- task:
|
18 |
name: Automatic Speech Recognition
|
@@ -20,51 +20,51 @@ model-index:
|
|
20 |
dataset:
|
21 |
name: Common Voice 8
|
22 |
type: mozilla-foundation/common_voice_8_0
|
23 |
-
args:
|
24 |
metrics:
|
25 |
- name: Test WER
|
26 |
type: wer
|
27 |
-
value:
|
28 |
- name: Test CER
|
29 |
type: cer
|
30 |
-
value:
|
31 |
- task:
|
32 |
name: Automatic Speech Recognition
|
33 |
type: automatic-speech-recognition
|
34 |
dataset:
|
35 |
name: Robust Speech Event - Dev Data
|
36 |
type: speech-recognition-community-v2/dev_data
|
37 |
-
args:
|
38 |
metrics:
|
39 |
- name: Test WER
|
40 |
type: wer
|
41 |
-
value:
|
42 |
- name: Test CER
|
43 |
type: cer
|
44 |
-
value:
|
45 |
---
|
46 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
47 |
should probably proofread and complete it, then remove this comment. -->
|
48 |
|
49 |
-
# wav2vec2-large-xls-r-1b-
|
50 |
|
51 |
-
This model is a fine-tuned version of [wav2vec2-large-xls-r-1b-
|
52 |
It achieves the following results on the test set:
|
53 |
-
- Loss: 0.
|
54 |
-
- Wer: 0.
|
55 |
|
56 |
-
Note that the above test results come from the original model without LM (language model) which can be found at https://huggingface.co/RuudVelo/wav2vec2-large-xls-r-1b-
|
57 |
|
58 |
## Model description
|
59 |
-
Model RuudVelo/wav2vec2-large-xls-r-1b-
|
60 |
## Intended uses & limitations
|
61 |
More information needed
|
62 |
## Training and evaluation data
|
63 |
-
Common Voice 8
|
64 |
## Training procedure
|
65 |
|
66 |
### Training hyperparameters
|
67 |
-
Parameters can be found in the run.sh file at https://huggingface.co/RuudVelo/wav2vec2-large-xls-r-1b-
|
68 |
|
69 |
### Framework versions
|
70 |
- Transformers 4.16.0.dev0
|
|
|
6 |
- automatic-speech-recognition
|
7 |
- mozilla-foundation/common_voice_8_0
|
8 |
- generated_from_trainer
|
9 |
+
- mt
|
10 |
- robust-speech-event
|
11 |
- model_for_talk
|
12 |
datasets:
|
13 |
- mozilla-foundation/common_voice_8_0
|
14 |
model-index:
|
15 |
+
- name: wav2vec2-large-xls-r-1b-cv8-mt-lm
|
16 |
results:
|
17 |
- task:
|
18 |
name: Automatic Speech Recognition
|
|
|
20 |
dataset:
|
21 |
name: Common Voice 8
|
22 |
type: mozilla-foundation/common_voice_8_0
|
23 |
+
args: mt
|
24 |
metrics:
|
25 |
- name: Test WER
|
26 |
type: wer
|
27 |
+
value: 15.88
|
28 |
- name: Test CER
|
29 |
type: cer
|
30 |
+
value: 3.65
|
31 |
- task:
|
32 |
name: Automatic Speech Recognition
|
33 |
type: automatic-speech-recognition
|
34 |
dataset:
|
35 |
name: Robust Speech Event - Dev Data
|
36 |
type: speech-recognition-community-v2/dev_data
|
37 |
+
args: mt
|
38 |
metrics:
|
39 |
- name: Test WER
|
40 |
type: wer
|
41 |
+
value:
|
42 |
- name: Test CER
|
43 |
type: cer
|
44 |
+
value:
|
45 |
---
|
46 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
47 |
should probably proofread and complete it, then remove this comment. -->
|
48 |
|
49 |
+
# wav2vec2-large-xls-r-1b-cv8-mt-lm
|
50 |
|
51 |
+
This model is a fine-tuned version of [wav2vec2-large-xls-r-1b-cv8-mt-lm](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the common_voice 8 dataset.
|
52 |
It achieves the following results on the test set:
|
53 |
+
- Loss: 0.2210
|
54 |
+
- Wer: 0.1974
|
55 |
|
56 |
+
Note that the above test results come from the original model without LM (language model) which can be found at https://huggingface.co/RuudVelo/wav2vec2-large-xls-r-1b-cv8-mt. The results with the LM model can be found on the right side of this model card.
|
57 |
|
58 |
## Model description
|
59 |
+
Model RuudVelo/wav2vec2-large-xls-r-1b-cv8-mt which has been improved with a KenLM 3-gram.
|
60 |
## Intended uses & limitations
|
61 |
More information needed
|
62 |
## Training and evaluation data
|
63 |
+
Common Voice 8 mt dataset has been used for the model
|
64 |
## Training procedure
|
65 |
|
66 |
### Training hyperparameters
|
67 |
+
Parameters can be found in the run.sh file at https://huggingface.co/RuudVelo/wav2vec2-large-xls-r-1b-cv8-mt-lm
|
68 |
|
69 |
### Framework versions
|
70 |
- Transformers 4.16.0.dev0
|