dmusingu commited on
Commit
f3ce2c7
1 Parent(s): 59ee130

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -2
README.md CHANGED
@@ -45,10 +45,10 @@ More information needed
45
 
46
  ## Training and evaluation data
47
 
48
- The model was trained on version 7 of the Luganda dataset of Mozilla common voices dataset. We used the train and validation dataset for training and the test dataset for validation.
49
 
50
  ## Training procedure
51
- The model was finetuned for 10 epochs using a learning rate of 5e-05. We used the AdamW optimizer.
52
 
53
  ### Training hyperparameters
54
 
@@ -82,3 +82,36 @@ The following hyperparameters were used during training:
82
  - Pytorch 2.2.1+cu121
83
  - Datasets 2.17.0
84
  - Tokenizers 0.15.2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
45
 
46
  ## Training and evaluation data
47
 
48
+ The model was trained on version 7 of the Luganda dataset of Mozilla common voices dataset. We used the train and validation dataset for training and the test dataset for validation.
49
 
50
  ## Training procedure
51
+ We trained the model on a 32 GB V100 GPU for 10 epochs using a learning rate of 5e-05. We used the AdamW optimizer.
52
 
53
  ### Training hyperparameters
54
 
 
82
  - Pytorch 2.2.1+cu121
83
  - Datasets 2.17.0
84
  - Tokenizers 0.15.2
85
+
86
+ ### Usage
87
+ ```python
88
+ import torch
89
+ import torchaudio
90
+ from datasets import load_dataset
91
+ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
92
+
93
+ test_dataset = load_dataset("common_voice", "lg", split="test[:10]")
94
+
95
+ processor = Wav2Vec2Processor.from_pretrained("dmusingu/w2v-bert-2.0-luganda-CV-train-validation-7.0")
96
+ model = Wav2Vec2ForCTC.from_pretrained("dmusingu/w2v-bert-2.0-luganda-CV-train-validation-7.0")
97
+
98
+ resampler = torchaudio.transforms.Resample(48_000, 16_000)
99
+
100
+ # Preprocessing the datasets.
101
+ # We need to read the audio files as arrays
102
+ def speech_file_to_array_fn(batch):
103
+ speech_array, sampling_rate = torchaudio.load(batch["path"])
104
+ batch["speech"] = resampler(speech_array).squeeze().numpy()
105
+ return batch
106
+
107
+ test_dataset = test_dataset.map(speech_file_to_array_fn)
108
+ inputs = processor(test_dataset[:2]["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
109
+
110
+ with torch.no_grad():
111
+ logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
112
+
113
+ predicted_ids = torch.argmax(logits, dim=-1)
114
+
115
+ print("Prediction:", processor.batch_decode(predicted_ids))
116
+ print("Reference:", test_dataset["sentence"][:2])
117
+ ```