hf-transformers-bot commited on
Commit
5b6a336
1 Parent(s): a66a66c

Upload tiny models for SpeechT5Model

Browse files
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<ctc_blank>": 80,
3
+ "<mask>": 79
4
+ }
config.json ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.1,
3
+ "apply_spec_augment": true,
4
+ "architectures": [
5
+ "SpeechT5Model"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "bos_token_id": 0,
9
+ "conv_bias": false,
10
+ "conv_dim": [
11
+ 512,
12
+ 512,
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512
18
+ ],
19
+ "conv_kernel": [
20
+ 10,
21
+ 3,
22
+ 3,
23
+ 3,
24
+ 3,
25
+ 2,
26
+ 2
27
+ ],
28
+ "conv_stride": [
29
+ 5,
30
+ 2,
31
+ 2,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2
36
+ ],
37
+ "decoder_attention_heads": 2,
38
+ "decoder_ffn_dim": 4,
39
+ "decoder_layerdrop": 0.1,
40
+ "decoder_layers": 4,
41
+ "decoder_start_token_id": 2,
42
+ "encoder_attention_heads": 2,
43
+ "encoder_ffn_dim": 4,
44
+ "encoder_layerdrop": 0.1,
45
+ "encoder_layers": 4,
46
+ "encoder_max_relative_position": 160,
47
+ "eos_token_id": 2,
48
+ "feat_extract_activation": "gelu",
49
+ "feat_extract_norm": "group",
50
+ "feat_proj_dropout": 0.0,
51
+ "hidden_act": "gelu",
52
+ "hidden_dropout": 0.1,
53
+ "hidden_size": 24,
54
+ "initializer_range": 0.02,
55
+ "is_encoder_decoder": true,
56
+ "layer_norm_eps": 1e-05,
57
+ "mask_feature_length": 10,
58
+ "mask_feature_min_masks": 0,
59
+ "mask_feature_prob": 0.0,
60
+ "mask_time_length": 10,
61
+ "mask_time_min_masks": 2,
62
+ "mask_time_prob": 0.05,
63
+ "max_speech_positions": 4000,
64
+ "max_text_positions": 450,
65
+ "model_type": "speecht5",
66
+ "num_conv_pos_embedding_groups": 16,
67
+ "num_conv_pos_embeddings": 128,
68
+ "num_feat_extract_layers": 7,
69
+ "num_mel_bins": 80,
70
+ "pad_token_id": 1,
71
+ "positional_dropout": 0.1,
72
+ "reduction_factor": 2,
73
+ "scale_embedding": false,
74
+ "speaker_embedding_dim": 512,
75
+ "speech_decoder_postnet_dropout": 0.5,
76
+ "speech_decoder_postnet_kernel": 5,
77
+ "speech_decoder_postnet_layers": 5,
78
+ "speech_decoder_postnet_units": 256,
79
+ "speech_decoder_prenet_dropout": 0.5,
80
+ "speech_decoder_prenet_layers": 2,
81
+ "speech_decoder_prenet_units": 256,
82
+ "torch_dtype": "float32",
83
+ "transformers_version": "4.28.0.dev0",
84
+ "use_cache": true,
85
+ "vocab_size": 81
86
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": false,
3
+ "feature_extractor_type": "SpeechT5FeatureExtractor",
4
+ "feature_size": 1,
5
+ "fmax": 7600,
6
+ "fmin": 80,
7
+ "frame_signal_scale": 1.0,
8
+ "hop_length": 16,
9
+ "mel_floor": 1e-10,
10
+ "n_fft": 1024,
11
+ "n_freqs": 513,
12
+ "num_mel_bins": 80,
13
+ "padding_side": "right",
14
+ "padding_value": 0.0,
15
+ "processor_class": "SpeechT5Processor",
16
+ "reduction_factor": 2,
17
+ "return_attention_mask": true,
18
+ "sample_size": 1024,
19
+ "sample_stride": 256,
20
+ "sampling_rate": 16000,
21
+ "win_function": "hann_window",
22
+ "win_length": 64
23
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16557ffaa62ad40659c9b96b40c7c883922f6826f4ae4208e15208c430b00a07
3
+ size 195705
special_tokens_map.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": {
5
+ "content": "<mask>",
6
+ "lstrip": true,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ },
11
+ "pad_token": "<pad>",
12
+ "unk_token": "<unk>"
13
+ }
spm_char.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
3
+ size 238473
tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "clean_up_tokenization_spaces": true,
4
+ "eos_token": "</s>",
5
+ "model_max_length": 450,
6
+ "pad_token": "<pad>",
7
+ "processor_class": "SpeechT5Processor",
8
+ "sp_model_kwargs": {},
9
+ "special_tokens_map_file": "/home/runner/.cache/huggingface/hub/models--microsoft--speecht5_asr/snapshots/53615c10408485422e09a12cda191a747f4bbe34/special_tokens_map.json",
10
+ "tokenizer_class": "SpeechT5Tokenizer",
11
+ "unk_token": "<unk>"
12
+ }