speech-test commited on
Commit
32f8dbd
1 Parent(s): ab2ef41

Upload model

Browse files
Files changed (4) hide show
  1. README.md +93 -0
  2. config.json +91 -0
  3. preprocessor_config.json +9 -0
  4. pytorch_model.bin +3 -0
README.md ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: en
3
+ datasets:
4
+ - superb
5
+ tags:
6
+ - speech
7
+ - audio
8
+ - hubert
9
+ - audio-classification
10
+ license: apache-2.0
11
+ ---
12
+
13
+ # Hubert-Large for Emotion Recognition
14
+
15
+ ## Model description
16
+
17
+ This is a ported version of
18
+ [S3PRL's Hubert for the SUPERB Emotion Recognition task](https://github.com/s3prl/s3prl/tree/master/s3prl/downstream/emotion).
19
+
20
+ The base model is [hubert-large-ll60k](https://huggingface.co/facebook/hubert-large-ll60k), which is pretrained on 16kHz
21
+ sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz.
22
+
23
+ For more information refer to [SUPERB: Speech processing Universal PERformance Benchmark](https://arxiv.org/abs/2105.01051)
24
+
25
+ ## Task and dataset description
26
+
27
+ Emotion Recognition (ER) predicts an emotion class for each utterance. The most widely used ER dataset
28
+ [IEMOCAP](https://sail.usc.edu/iemocap/) is adopted, and we follow the conventional evaluation protocol:
29
+ we drop the unbalanced emotion classes to leave the final four classes with a similar amount of data points and
30
+ cross-validate on five folds of the standard splits.
31
+
32
+ For the original model's training and evaluation instructions refer to the
33
+ [S3PRL downstream task README](https://github.com/s3prl/s3prl/tree/master/s3prl/downstream#er-emotion-recognition).
34
+
35
+
36
+ ## Usage examples
37
+
38
+ You can use the model via the Audio Classification pipeline:
39
+ ```python
40
+ from datasets import load_dataset
41
+ from transformers import pipeline
42
+
43
+ dataset = load_dataset("anton-l/superb_demo", "er", split="session1")
44
+
45
+ classifier = pipeline("audio-classification", model="superb/hubert-large-superb-er")
46
+ labels = classifier(dataset[0]["file"], top_k=5)
47
+ ```
48
+
49
+ Or use the model directly:
50
+ ```python
51
+ import torch
52
+ import librosa
53
+ from datasets import load_dataset
54
+ from transformers import HubertForSequenceClassification, Wav2Vec2FeatureExtractor
55
+
56
+ def map_to_array(example):
57
+ speech, _ = librosa.load(example["file"], sr=16000, mono=True)
58
+ example["speech"] = speech
59
+ return example
60
+
61
+ # load a demo dataset and read audio files
62
+ dataset = load_dataset("anton-l/superb_demo", "er", split="session1")
63
+ dataset = dataset.map(map_to_array)
64
+
65
+ model = HubertForSequenceClassification.from_pretrained("superb/hubert-large-superb-er")
66
+ feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained("superb/hubert-large-superb-er")
67
+
68
+ # compute attention masks and normalize the waveform if needed
69
+ inputs = feature_extractor(dataset[:4]["speech"], sampling_rate=16000, padding=True, return_tensors="pt")
70
+
71
+ logits = model(**inputs).logits
72
+ predicted_ids = torch.argmax(logits, dim=-1)
73
+ labels = [model.config.id2label[_id] for _id in predicted_ids.tolist()]
74
+ ```
75
+
76
+ ## Eval results
77
+
78
+ The evaluation metric is accuracy.
79
+
80
+ | | **s3prl** | **transformers** |
81
+ |--------|-----------|------------------|
82
+ |**session1**| `0.6762` | `N/A` |
83
+
84
+ ### BibTeX entry and citation info
85
+
86
+ ```bibtex
87
+ @article{yang2021superb,
88
+ title={SUPERB: Speech processing Universal PERformance Benchmark},
89
+ author={Yang, Shu-wen and Chi, Po-Han and Chuang, Yung-Sung and Lai, Cheng-I Jeff and Lakhotia, Kushal and Lin, Yist Y and Liu, Andy T and Shi, Jiatong and Chang, Xuankai and Lin, Guan-Ting and others},
90
+ journal={arXiv preprint arXiv:2105.01051},
91
+ year={2021}
92
+ }
93
+ ```
config.json ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/hubert-large-ll60k",
3
+ "activation_dropout": 0.0,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": true,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "do_stable_layer_norm": true,
42
+ "eos_token_id": 2,
43
+ "feat_extract_activation": "gelu",
44
+ "feat_extract_dropout": 0.0,
45
+ "feat_extract_norm": "layer",
46
+ "feat_proj_dropout": 0.1,
47
+ "final_dropout": 0.0,
48
+ "gradient_checkpointing": false,
49
+ "hidden_act": "gelu",
50
+ "hidden_dropout": 0.1,
51
+ "hidden_size": 1024,
52
+ "id2label": {
53
+ "0": "neu",
54
+ "1": "hap",
55
+ "2": "ang",
56
+ "3": "sad"
57
+ },
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 4096,
60
+ "label2id": {
61
+ "ang": 2,
62
+ "hap": 1,
63
+ "neu": 0,
64
+ "sad": 3
65
+ },
66
+ "layer_norm_eps": 1e-05,
67
+ "layerdrop": 0.1,
68
+ "mask_channel_length": 10,
69
+ "mask_channel_min_space": 1,
70
+ "mask_channel_other": 0.0,
71
+ "mask_channel_prob": 0.0,
72
+ "mask_channel_selection": "static",
73
+ "mask_feature_length": 10,
74
+ "mask_feature_prob": 0.0,
75
+ "mask_time_length": 10,
76
+ "mask_time_min_space": 1,
77
+ "mask_time_other": 0.0,
78
+ "mask_time_prob": 0.075,
79
+ "mask_time_selection": "static",
80
+ "model_type": "hubert",
81
+ "num_attention_heads": 16,
82
+ "num_conv_pos_embedding_groups": 16,
83
+ "num_conv_pos_embeddings": 128,
84
+ "num_feat_extract_layers": 7,
85
+ "num_hidden_layers": 24,
86
+ "pad_token_id": 0,
87
+ "torch_dtype": "float32",
88
+ "transformers_version": "4.11.0.dev0",
89
+ "use_weighted_layer_sum": true,
90
+ "vocab_size": 32
91
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9e5f9386285300a3e5e7d7f6852f578fec7bb0cc75e89661a3f98db7112237d
3
+ size 1262970923