yonas commited on
Commit
12fb3ac
1 Parent(s): 0d2a49b

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +103 -0
README.md ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: cc-by-4.0
5
+ library_name: nemo
6
+ datasets:
7
+ - mozilla-foundation/common_voice_11_0
8
+ thumbnail: null
9
+ tags:
10
+ - automatic-speech-recognition
11
+ - speech
12
+ - Kinyarwanda
13
+ - audio
14
+ - CTC
15
+ - Conformer
16
+ - Transformer
17
+ - NeMo
18
+ - pytorch
19
+ model-index:
20
+ - name: stt_rw_conformer_ctc_large
21
+ results: []
22
+
23
+ ---
24
+
25
+
26
+ ## Model Overview
27
+
28
+ <DESCRIBE IN ONE LINE THE MODEL AND ITS USE>
29
+
30
+ ## NVIDIA NeMo: Training
31
+
32
+ To train, fine-tune or play with the model you will need to install [NVIDIA NeMo](https://github.com/NVIDIA/NeMo). We recommend you install it after you've installed latest Pytorch version.
33
+ ```
34
+ pip install nemo_toolkit['all']
35
+ ```
36
+
37
+ ## How to Use this Model
38
+
39
+ The model is available for use in the NeMo toolkit [3], and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset.
40
+
41
+ ### Automatically instantiate the model
42
+
43
+ ```python
44
+ import nemo.collections.asr as nemo_asr
45
+ asr_model = nemo_asr.models.ASRModel.from_pretrained("yonas/stt_rw_conformer_ctc_large")
46
+ ```
47
+
48
+ ### Transcribing using Python
49
+ First, let's get a sample
50
+ ```
51
+ wget https://dldata-public.s3.us-east-2.amazonaws.com/2086-149220-0033.wav
52
+ ```
53
+ Then simply do:
54
+ ```
55
+ asr_model.transcribe(['2086-149220-0033.wav'])
56
+ ```
57
+
58
+ ### Transcribing many audio files
59
+
60
+ ```shell
61
+ python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py pretrained_name="yonas/stt_rw_conformer_ctc_large" audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
62
+ ```
63
+
64
+ ### Input
65
+
66
+ This model accepts 16000 KHz Mono-channel Audio (wav files) as input.
67
+
68
+ ### Output
69
+
70
+ This model provides transcribed speech as a string for a given audio sample.
71
+
72
+ ## Model Architecture
73
+
74
+ <ADD SOME INFORMATION ABOUT THE ARCHITECTURE>
75
+
76
+ ## Training
77
+
78
+ <ADD INFORMATION ABOUT HOW THE MODEL WAS TRAINED - HOW MANY EPOCHS, AMOUNT OF COMPUTE ETC>
79
+
80
+ ### Datasets
81
+
82
+ <LIST THE NAME AND SPLITS OF DATASETS USED TO TRAIN THIS MODEL (ALONG WITH LANGUAGE AND ANY ADDITIONAL INFORMATION)>
83
+
84
+ ## Performance
85
+
86
+ <LIST THE SCORES OF THE MODEL -
87
+ OR
88
+ USE THE Hugging Face Evaluate LiBRARY TO UPLOAD METRICS>
89
+
90
+ ## Limitations
91
+
92
+ <DECLARE ANY POTENTIAL LIMITATIONS OF THE MODEL>
93
+
94
+ Eg:
95
+ Since this model was trained on publically available speech datasets, the performance of this model might degrade for speech which includes technical terms, or vernacular that the model has not been trained on. The model might also perform worse for accented speech.
96
+
97
+
98
+ ## References
99
+
100
+ <ADD ANY REFERENCES HERE AS NEEDED>
101
+
102
+ [1] [NVIDIA NeMo Toolkit](https://github.com/NVIDIA/NeMo)
103
+