Update README.md

README.md

@@ -28,4 +28,38 @@ model-index:
 ---
 # Wav2vec 2.0 large VoxRex Swedish
 
-![WER during training](chart_1.svg "WER")
+Additionally pretrained and finetuned version of Facebooks [VoxPopuli-sv large](https://huggingface.co/facebook/wav2vec2-large-voxrex) model using Swedish radio broadcasts, NST and Common Voice data. Evalutation without a language model gives the following: WER for NST + Common Voice test set (2% of total sentences) is **3.40%**. WER for Common Voice test set is **10.72%** directly and **8.71%** with a 4-gram language model.
+
+When using this model, make sure that your speech input is sampled at 16kHz.
+
+## Training
+This model has additionally pretrained on 3500h of a mix of Swedish local radio broadcasts, audio books and other audio sources. It has been fine-tuned for 120000 updates on NST + CommonVoice<!-- and then for an additional 20000 updates on CommonVoice only. The additional fine-tuning on CommonVoice hurts performance on the NST+CommonVoice test set somewhat and, unsurprisingly, improves it on the CommonVoice test set. It seems to perform generally better though [citation needed]-->.
+
+![WER during training](chart_1.svg "WER")
+
+## Usage
+The model can be used directly (without a language model) as follows:
+```python
+import torch
+import torchaudio
+from datasets import load_dataset
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+test_dataset = load_dataset("common_voice", "sv-SE", split="test[:2%]").
+processor = Wav2Vec2Processor.from_pretrained("KBLab/wav2vec2-large-voxpopuli-sv-swedish")
+model = Wav2Vec2ForCTC.from_pretrained("KBLab/wav2vec2-large-voxpopuli-sv-swedish")
+resampler = torchaudio.transforms.Resample(48_000, 16_000)
+# Preprocessing the datasets.
+# We need to read the aduio files as arrays
+def speech_file_to_array_fn(batch):
+    speech_array, sampling_rate = torchaudio.load(batch["path"])
+    batch["speech"] = resampler(speech_array).squeeze().numpy()
+    return batch
+test_dataset = test_dataset.map(speech_file_to_array_fn)
+inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)
+with torch.no_grad():
+    logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
+predicted_ids = torch.argmax(logits, dim=-1)
+print("Prediction:", processor.batch_decode(predicted_ids))
+print("Reference:", test_dataset["sentence"][:2])
+```
+