Update info

README.md

@@ -6,7 +6,6 @@ tags:
 - speech
 - audio
 - hubert
-- s3prl
 license: apache-2.0
 ---
 
@@ -16,13 +15,12 @@ license: apache-2.0
 
 This is a ported version of [S3PRL's Hubert for the SUPERB Intent Classification task](https://github.com/s3prl/s3prl/tree/master/s3prl/downstream/fluent_commands).
 
-The base model is [hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960). 
-It is pretrained on 16kHz sampled speech audio. 
-When using the model make sure that your speech input is also sampled at 16Khz. 
+The base model is [hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960), which is pretrained on 16kHz 
+sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz. 
 
 For more information refer to [SUPERB: Speech processing Universal PERformance Benchmark](https://arxiv.org/abs/2105.01051)
 
-## Task description
+## Task and dataset description
 
 Intent Classification (IC) classifies utterances into predefined classes to determine the intent of
 speakers. SUPERB uses the 
@@ -38,20 +36,26 @@ For the original model's training and evaluation instructions refer to the
 You can use the model directly like so:
 ```python
 import torch
-import numpy as np
+import librosa
 from datasets import load_dataset
 from transformers import HubertForSequenceClassification, Wav2Vec2FeatureExtractor
 
-# TODO: replace with the official superb dataset
-superb_ks = load_dataset("anton-l/superb_dummy", "ic", split="test")
+def map_to_array(example):
+    speech, _ = librosa.load(example["file"], sr=16000, mono=True)
+    example["speech"] = speech
+    return example
+
+# load a demo dataset and read audio files
+dataset = load_dataset("anton-l/superb_demo", "ic", split="test")
+dataset = dataset.map(map_to_array)
+
 model = HubertForSequenceClassification.from_pretrained("superb/hubert-base-superb-ic")
 feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained("superb/hubert-base-superb-ic")
 
-audio = np.array(superb_ks[0]["speech"])
 # compute attention masks and normalize the waveform if needed
-inputs = feature_extractor(audio, sampling_rate=16_000, return_tensors="pt")
+inputs = feature_extractor(dataset[:4]["speech"], sampling_rate=16000, padding=True, return_tensors="pt")
 
-logits = model(**inputs).logits[0]
+logits = model(**inputs).logits
 
 action_ids = torch.argmax(logits[:, :6], dim=-1).tolist()
 action_labels = [model.config.id2label[_id] for _id in action_ids]
@@ -67,9 +71,9 @@ location_labels = [model.config.id2label[_id + 20] for _id in location_ids]
 
 The evaluation metric is accuracy.
 
-|      | `s3prl` | `transformers` |
-|------|---------|----------------|
-|`test`| TBA     | TBA            |
+|        | **s3prl** | **transformers** |
+|--------|-----------|------------------|
+|**test**| `0.9834`  | `N/A`         |
 
 ### BibTeX entry and citation info