Upload folder using huggingface_hub
Browse files- config.json +111 -0
- onnx/model.onnx +3 -0
- onnx/model_quantized.onnx +3 -0
- preprocessor_config.json +9 -0
- quantize_config.json +32 -0
config.json
ADDED
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/unispeech-sat-base-plus-sd",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"apply_spec_augment": true,
|
5 |
+
"architectures": [
|
6 |
+
"UniSpeechSatForAudioFrameClassification"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.1,
|
9 |
+
"bos_token_id": 1,
|
10 |
+
"classifier_proj_size": 256,
|
11 |
+
"codevector_dim": 256,
|
12 |
+
"contrastive_logits_temperature": 0.1,
|
13 |
+
"conv_bias": false,
|
14 |
+
"conv_dim": [
|
15 |
+
512,
|
16 |
+
512,
|
17 |
+
512,
|
18 |
+
512,
|
19 |
+
512,
|
20 |
+
512,
|
21 |
+
512
|
22 |
+
],
|
23 |
+
"conv_kernel": [
|
24 |
+
10,
|
25 |
+
3,
|
26 |
+
3,
|
27 |
+
3,
|
28 |
+
3,
|
29 |
+
2,
|
30 |
+
2
|
31 |
+
],
|
32 |
+
"conv_stride": [
|
33 |
+
5,
|
34 |
+
2,
|
35 |
+
2,
|
36 |
+
2,
|
37 |
+
2,
|
38 |
+
2,
|
39 |
+
2
|
40 |
+
],
|
41 |
+
"ctc_loss_reduction": "mean",
|
42 |
+
"ctc_zero_infinity": false,
|
43 |
+
"diversity_loss_weight": 0.1,
|
44 |
+
"do_stable_layer_norm": false,
|
45 |
+
"eos_token_id": 2,
|
46 |
+
"feat_extract_activation": "gelu",
|
47 |
+
"feat_extract_dropout": 0.0,
|
48 |
+
"feat_extract_norm": "group",
|
49 |
+
"feat_proj_dropout": 0.1,
|
50 |
+
"feat_quantizer_dropout": 0.0,
|
51 |
+
"final_dropout": 0.0,
|
52 |
+
"hidden_act": "gelu",
|
53 |
+
"hidden_dropout": 0.1,
|
54 |
+
"hidden_size": 768,
|
55 |
+
"id2label": {
|
56 |
+
"0": "speaker_0",
|
57 |
+
"1": "speaker_1"
|
58 |
+
},
|
59 |
+
"initializer_range": 0.02,
|
60 |
+
"intermediate_size": 3072,
|
61 |
+
"label2id": {
|
62 |
+
"speaker_0": 0,
|
63 |
+
"speaker_1": 1
|
64 |
+
},
|
65 |
+
"layer_norm_eps": 1e-05,
|
66 |
+
"layerdrop": 0.1,
|
67 |
+
"mask_feature_length": 10,
|
68 |
+
"mask_feature_min_masks": 0,
|
69 |
+
"mask_feature_prob": 0.0,
|
70 |
+
"mask_time_length": 10,
|
71 |
+
"mask_time_min_masks": 2,
|
72 |
+
"mask_time_prob": 0.05,
|
73 |
+
"model_type": "unispeech-sat",
|
74 |
+
"num_attention_heads": 12,
|
75 |
+
"num_clusters": 504,
|
76 |
+
"num_codevector_groups": 2,
|
77 |
+
"num_codevectors_per_group": 320,
|
78 |
+
"num_conv_pos_embedding_groups": 16,
|
79 |
+
"num_conv_pos_embeddings": 128,
|
80 |
+
"num_feat_extract_layers": 7,
|
81 |
+
"num_hidden_layers": 12,
|
82 |
+
"num_negatives": 100,
|
83 |
+
"pad_token_id": 0,
|
84 |
+
"proj_codevector_dim": 256,
|
85 |
+
"tdnn_dilation": [
|
86 |
+
1,
|
87 |
+
2,
|
88 |
+
3,
|
89 |
+
1,
|
90 |
+
1
|
91 |
+
],
|
92 |
+
"tdnn_dim": [
|
93 |
+
512,
|
94 |
+
512,
|
95 |
+
512,
|
96 |
+
512,
|
97 |
+
1500
|
98 |
+
],
|
99 |
+
"tdnn_kernel": [
|
100 |
+
5,
|
101 |
+
3,
|
102 |
+
3,
|
103 |
+
1,
|
104 |
+
1
|
105 |
+
],
|
106 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
107 |
+
"transformers_version": "4.33.2",
|
108 |
+
"use_weighted_layer_sum": true,
|
109 |
+
"vocab_size": 32,
|
110 |
+
"xvector_output_dim": 512
|
111 |
+
}
|
onnx/model.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0da7fd6d576fd3148dca442265adc4985aec50a4dcb5994205087c4f1ddb507a
|
3 |
+
size 377820930
|
onnx/model_quantized.onnx
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fda8f086addbf9ae06b0052b03cb18f055d5eaf9714f9bf32fbf2d704f010f9c
|
3 |
+
size 95301468
|
preprocessor_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": false,
|
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 |
+
}
|
quantize_config.json
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"per_channel": false,
|
3 |
+
"reduce_range": false,
|
4 |
+
"per_model_config": {
|
5 |
+
"model": {
|
6 |
+
"op_types": [
|
7 |
+
"Div",
|
8 |
+
"ReduceSum",
|
9 |
+
"Shape",
|
10 |
+
"Transpose",
|
11 |
+
"Concat",
|
12 |
+
"Gather",
|
13 |
+
"Sub",
|
14 |
+
"MatMul",
|
15 |
+
"Pow",
|
16 |
+
"Slice",
|
17 |
+
"Unsqueeze",
|
18 |
+
"Conv",
|
19 |
+
"Erf",
|
20 |
+
"InstanceNormalization",
|
21 |
+
"Constant",
|
22 |
+
"ReduceMean",
|
23 |
+
"Reshape",
|
24 |
+
"Softmax",
|
25 |
+
"Sqrt",
|
26 |
+
"Mul",
|
27 |
+
"Add"
|
28 |
+
],
|
29 |
+
"weight_type": "QUInt8"
|
30 |
+
}
|
31 |
+
}
|
32 |
+
}
|