Commit ·
d136706
1
Parent(s): d1314f7
Update README.md
Browse files
README.md
CHANGED
|
@@ -10,7 +10,6 @@ datasets:
|
|
| 10 |
- WSJ-1
|
| 11 |
- National Singapore Corpus Part 1
|
| 12 |
- National Singapore Corpus Part 6
|
| 13 |
-
- mozilla-foundation/common_voice_7_0
|
| 14 |
thumbnail: null
|
| 15 |
tags:
|
| 16 |
- automatic-speech-recognition
|
|
@@ -45,7 +44,7 @@ model-index:
|
|
| 45 |
metrics:
|
| 46 |
- name: Test WER
|
| 47 |
type: wer
|
| 48 |
-
value:
|
| 49 |
- task:
|
| 50 |
type: Automatic Speech Recognition
|
| 51 |
name: automatic-speech-recognition
|
|
@@ -59,49 +58,7 @@ model-index:
|
|
| 59 |
metrics:
|
| 60 |
- name: Test WER
|
| 61 |
type: wer
|
| 62 |
-
value:
|
| 63 |
-
- task:
|
| 64 |
-
type: Automatic Speech Recognition
|
| 65 |
-
name: automatic-speech-recognition
|
| 66 |
-
dataset:
|
| 67 |
-
name: Multilingual LibriSpeech
|
| 68 |
-
type: facebook/multilingual_librispeech
|
| 69 |
-
config: english
|
| 70 |
-
split: test
|
| 71 |
-
args:
|
| 72 |
-
language: en
|
| 73 |
-
metrics:
|
| 74 |
-
- name: Test WER
|
| 75 |
-
type: wer
|
| 76 |
-
value: 7.2
|
| 77 |
-
- task:
|
| 78 |
-
type: Automatic Speech Recognition
|
| 79 |
-
name: automatic-speech-recognition
|
| 80 |
-
dataset:
|
| 81 |
-
name: Mozilla Common Voice 7.0
|
| 82 |
-
type: mozilla-foundation/common_voice_7_0
|
| 83 |
-
config: en
|
| 84 |
-
split: test
|
| 85 |
-
args:
|
| 86 |
-
language: en
|
| 87 |
-
metrics:
|
| 88 |
-
- name: Test WER
|
| 89 |
-
type: wer
|
| 90 |
-
value: 8.0
|
| 91 |
-
- task:
|
| 92 |
-
type: Automatic Speech Recognition
|
| 93 |
-
name: automatic-speech-recognition
|
| 94 |
-
dataset:
|
| 95 |
-
name: Mozilla Common Voice 8.0
|
| 96 |
-
type: mozilla-foundation/common_voice_8_0
|
| 97 |
-
config: en
|
| 98 |
-
split: test
|
| 99 |
-
args:
|
| 100 |
-
language: en
|
| 101 |
-
metrics:
|
| 102 |
-
- name: Test WER
|
| 103 |
-
type: wer
|
| 104 |
-
value: 9.48
|
| 105 |
- task:
|
| 106 |
type: Automatic Speech Recognition
|
| 107 |
name: automatic-speech-recognition
|
|
@@ -113,7 +70,7 @@ model-index:
|
|
| 113 |
metrics:
|
| 114 |
- name: Test WER
|
| 115 |
type: wer
|
| 116 |
-
value: 2.
|
| 117 |
- task:
|
| 118 |
type: Automatic Speech Recognition
|
| 119 |
name: automatic-speech-recognition
|
|
@@ -125,7 +82,7 @@ model-index:
|
|
| 125 |
metrics:
|
| 126 |
- name: Test WER
|
| 127 |
type: wer
|
| 128 |
-
value:
|
| 129 |
- task:
|
| 130 |
type: Automatic Speech Recognition
|
| 131 |
name: automatic-speech-recognition
|
|
@@ -137,7 +94,7 @@ model-index:
|
|
| 137 |
metrics:
|
| 138 |
- name: Test WER
|
| 139 |
type: wer
|
| 140 |
-
value:
|
| 141 |
---
|
| 142 |
# NVIDIA Streaming Citrinet 1024 (en-US)
|
| 143 |
|
|
@@ -222,7 +179,6 @@ All the models in this collection are trained on a composite dataset (NeMo ASRSE
|
|
| 222 |
- Switchboard-1 Dataset
|
| 223 |
- WSJ-0 and WSJ-1
|
| 224 |
- National Speech Corpus (Part 1, Part 6)
|
| 225 |
-
- Mozilla Common Voice (v7.0)
|
| 226 |
|
| 227 |
Note: older versions of the model may have trained on smaller set of datasets.
|
| 228 |
|
|
|
|
| 10 |
- WSJ-1
|
| 11 |
- National Singapore Corpus Part 1
|
| 12 |
- National Singapore Corpus Part 6
|
|
|
|
| 13 |
thumbnail: null
|
| 14 |
tags:
|
| 15 |
- automatic-speech-recognition
|
|
|
|
| 44 |
metrics:
|
| 45 |
- name: Test WER
|
| 46 |
type: wer
|
| 47 |
+
value: 3.4
|
| 48 |
- task:
|
| 49 |
type: Automatic Speech Recognition
|
| 50 |
name: automatic-speech-recognition
|
|
|
|
| 58 |
metrics:
|
| 59 |
- name: Test WER
|
| 60 |
type: wer
|
| 61 |
+
value: 7.6
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
- task:
|
| 63 |
type: Automatic Speech Recognition
|
| 64 |
name: automatic-speech-recognition
|
|
|
|
| 70 |
metrics:
|
| 71 |
- name: Test WER
|
| 72 |
type: wer
|
| 73 |
+
value: 2.5
|
| 74 |
- task:
|
| 75 |
type: Automatic Speech Recognition
|
| 76 |
name: automatic-speech-recognition
|
|
|
|
| 82 |
metrics:
|
| 83 |
- name: Test WER
|
| 84 |
type: wer
|
| 85 |
+
value: 4.0
|
| 86 |
- task:
|
| 87 |
type: Automatic Speech Recognition
|
| 88 |
name: automatic-speech-recognition
|
|
|
|
| 94 |
metrics:
|
| 95 |
- name: Test WER
|
| 96 |
type: wer
|
| 97 |
+
value: 6.2
|
| 98 |
---
|
| 99 |
# NVIDIA Streaming Citrinet 1024 (en-US)
|
| 100 |
|
|
|
|
| 179 |
- Switchboard-1 Dataset
|
| 180 |
- WSJ-0 and WSJ-1
|
| 181 |
- National Speech Corpus (Part 1, Part 6)
|
|
|
|
| 182 |
|
| 183 |
Note: older versions of the model may have trained on smaller set of datasets.
|
| 184 |
|