Hugging Face's logo

gclose19
/
omnilingual-asr-ssl-hf

Safetensors
Model card Files
xet
Community
gclose19 commited on
Commit
315f55a
·
verified ·
1 Parent(s): 315f936

Upload folder using huggingface_hub

Browse files
Files changed (23) hide show
  1. .gitattributes +1 -0
  2. 1b/config.json +109 -0
  3. 1b/model.safetensors +3 -0
  4. 1b/preprocessor_config.json +8 -0
  5. 300m/config.json +109 -0
  6. 300m/model.safetensors +3 -0
  7. 300m/preprocessor_config.json +8 -0
  8. 3b/config.json +109 -0
  9. 3b/model-00001-of-00003.safetensors +3 -0
  10. 3b/model-00002-of-00003.safetensors +3 -0
  11. 3b/model-00003-of-00003.safetensors +3 -0
  12. 3b/model.safetensors.index.json +1013 -0
  13. 3b/preprocessor_config.json +8 -0
  14. 7b/config.json +109 -0
  15. 7b/model-00001-of-00006.safetensors +3 -0
  16. 7b/model-00002-of-00006.safetensors +3 -0
  17. 7b/model-00003-of-00006.safetensors +3 -0
  18. 7b/model-00004-of-00006.safetensors +3 -0
  19. 7b/model-00005-of-00006.safetensors +3 -0
  20. 7b/model-00006-of-00006.safetensors +3 -0
  21. 7b/model.safetensors.index.json +0 -0
  22. hf_wav2vec2_features.png +3 -0
  23. inference_hf.py +52 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ hf_wav2vec2_features.png filter=lfs diff=lfs merge=lfs -text
1b/config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.1,
3
+ "adapter_attn_dim": null,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForPreTraining"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": true,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "dtype": "float32",
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_dropout_prob": 0.1,
60
+ "hidden_size": 1280,
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 5120,
63
+ "layer_norm_eps": 1e-05,
64
+ "layerdrop": 0.0,
65
+ "mask_feature_length": 10,
66
+ "mask_feature_min_masks": 0,
67
+ "mask_feature_prob": 0.0,
68
+ "mask_time_length": 10,
69
+ "mask_time_min_masks": 2,
70
+ "mask_time_prob": 0.05,
71
+ "model_type": "wav2vec2",
72
+ "num_adapter_layers": 3,
73
+ "num_attention_heads": 16,
74
+ "num_codevector_groups": 2,
75
+ "num_codevectors_per_group": 320,
76
+ "num_conv_pos_embedding_groups": 16,
77
+ "num_conv_pos_embeddings": 128,
78
+ "num_feat_extract_layers": 7,
79
+ "num_hidden_layers": 48,
80
+ "num_negatives": 100,
81
+ "output_hidden_size": 1280,
82
+ "pad_token_id": 0,
83
+ "proj_codevector_dim": 1024,
84
+ "tdnn_dilation": [
85
+ 1,
86
+ 2,
87
+ 3,
88
+ 1,
89
+ 1
90
+ ],
91
+ "tdnn_dim": [
92
+ 512,
93
+ 512,
94
+ 512,
95
+ 512,
96
+ 1500
97
+ ],
98
+ "tdnn_kernel": [
99
+ 5,
100
+ 3,
101
+ 3,
102
+ 1,
103
+ 1
104
+ ],
105
+ "transformers_version": "4.57.1",
106
+ "use_weighted_layer_sum": false,
107
+ "vocab_size": 32,
108
+ "xvector_output_dim": 512
109
+ }
1b/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4878ccac61364da98e2e579ce75fd1f587e3f23be3ef7bd1f1fd7d6abb395e97
3
+ size 3860784616
1b/preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": false,
7
+ "sampling_rate": 16000
8
+ }
300m/config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.1,
3
+ "adapter_attn_dim": null,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForPreTraining"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": true,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "dtype": "float32",
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_dropout_prob": 0.1,
60
+ "hidden_size": 1024,
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 4096,
63
+ "layer_norm_eps": 1e-05,
64
+ "layerdrop": 0.0,
65
+ "mask_feature_length": 10,
66
+ "mask_feature_min_masks": 0,
67
+ "mask_feature_prob": 0.0,
68
+ "mask_time_length": 10,
69
+ "mask_time_min_masks": 2,
70
+ "mask_time_prob": 0.05,
71
+ "model_type": "wav2vec2",
72
+ "num_adapter_layers": 3,
73
+ "num_attention_heads": 16,
74
+ "num_codevector_groups": 2,
75
+ "num_codevectors_per_group": 320,
76
+ "num_conv_pos_embedding_groups": 16,
77
+ "num_conv_pos_embeddings": 128,
78
+ "num_feat_extract_layers": 7,
79
+ "num_hidden_layers": 24,
80
+ "num_negatives": 100,
81
+ "output_hidden_size": 1024,
82
+ "pad_token_id": 0,
83
+ "proj_codevector_dim": 768,
84
+ "tdnn_dilation": [
85
+ 1,
86
+ 2,
87
+ 3,
88
+ 1,
89
+ 1
90
+ ],
91
+ "tdnn_dim": [
92
+ 512,
93
+ 512,
94
+ 512,
95
+ 512,
96
+ 1500
97
+ ],
98
+ "tdnn_kernel": [
99
+ 5,
100
+ 3,
101
+ 3,
102
+ 1,
103
+ 1
104
+ ],
105
+ "transformers_version": "4.57.1",
106
+ "use_weighted_layer_sum": false,
107
+ "vocab_size": 32,
108
+ "xvector_output_dim": 512
109
+ }
300m/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0de75980f4eb40f337277a136b4952aa3a71a5b2e574587c5d99f4f60097d6c
3
+ size 1269615448
300m/preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": false,
7
+ "sampling_rate": 16000
8
+ }
3b/config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.1,
3
+ "adapter_attn_dim": null,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForPreTraining"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": true,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "dtype": "float32",
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_dropout_prob": 0.1,
60
+ "hidden_size": 2048,
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 8192,
63
+ "layer_norm_eps": 1e-05,
64
+ "layerdrop": 0.0,
65
+ "mask_feature_length": 10,
66
+ "mask_feature_min_masks": 0,
67
+ "mask_feature_prob": 0.0,
68
+ "mask_time_length": 10,
69
+ "mask_time_min_masks": 2,
70
+ "mask_time_prob": 0.05,
71
+ "model_type": "wav2vec2",
72
+ "num_adapter_layers": 3,
73
+ "num_attention_heads": 16,
74
+ "num_codevector_groups": 2,
75
+ "num_codevectors_per_group": 320,
76
+ "num_conv_pos_embedding_groups": 16,
77
+ "num_conv_pos_embeddings": 128,
78
+ "num_feat_extract_layers": 7,
79
+ "num_hidden_layers": 60,
80
+ "num_negatives": 100,
81
+ "output_hidden_size": 2048,
82
+ "pad_token_id": 0,
83
+ "proj_codevector_dim": 1024,
84
+ "tdnn_dilation": [
85
+ 1,
86
+ 2,
87
+ 3,
88
+ 1,
89
+ 1
90
+ ],
91
+ "tdnn_dim": [
92
+ 512,
93
+ 512,
94
+ 512,
95
+ 512,
96
+ 1500
97
+ ],
98
+ "tdnn_kernel": [
99
+ 5,
100
+ 3,
101
+ 3,
102
+ 1,
103
+ 1
104
+ ],
105
+ "transformers_version": "4.57.1",
106
+ "use_weighted_layer_sum": false,
107
+ "vocab_size": 32,
108
+ "xvector_output_dim": 512
109
+ }
3b/model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:088d9e81a8a0dd64dfd14d69628517ad2065f570df6ba316beb452143971b9d9
3
+ size 4989745384
3b/model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a591bfad83d4a57af087f57e03cd7ca275c13eaaff2dfd17d34a8f14c9d9781
3
+ size 4968743656
3b/model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c89afa8bb3851342f44a056d6e3bb7e958478e471cf848fa3affa8431375d2d1
3
+ size 2296759504
3b/model.safetensors.index.json ADDED
@@ -0,0 +1,1013 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_parameters": 3063780608,
4
+ "total_size": 12255122432
5
+ },
6
+ "weight_map": {
7
+ "project_hid.bias": "model-00003-of-00003.safetensors",
8
+ "project_hid.weight": "model-00003-of-00003.safetensors",
9
+ "project_q.bias": "model-00003-of-00003.safetensors",
10
+ "project_q.weight": "model-00003-of-00003.safetensors",
11
+ "quantizer.codevectors": "model-00003-of-00003.safetensors",
12
+ "quantizer.weight_proj.bias": "model-00003-of-00003.safetensors",
13
+ "quantizer.weight_proj.weight": "model-00003-of-00003.safetensors",
14
+ "wav2vec2.encoder.layer_norm.bias": "model-00001-of-00003.safetensors",
15
+ "wav2vec2.encoder.layer_norm.weight": "model-00001-of-00003.safetensors",
16
+ "wav2vec2.encoder.layers.0.attention.k_proj.bias": "model-00001-of-00003.safetensors",
17
+ "wav2vec2.encoder.layers.0.attention.k_proj.weight": "model-00001-of-00003.safetensors",
18
+ "wav2vec2.encoder.layers.0.attention.out_proj.bias": "model-00001-of-00003.safetensors",
19
+ "wav2vec2.encoder.layers.0.attention.out_proj.weight": "model-00001-of-00003.safetensors",
20
+ "wav2vec2.encoder.layers.0.attention.q_proj.bias": "model-00001-of-00003.safetensors",
21
+ "wav2vec2.encoder.layers.0.attention.q_proj.weight": "model-00001-of-00003.safetensors",
22
+ "wav2vec2.encoder.layers.0.attention.v_proj.bias": "model-00001-of-00003.safetensors",
23
+ "wav2vec2.encoder.layers.0.attention.v_proj.weight": "model-00001-of-00003.safetensors",
24
+ "wav2vec2.encoder.layers.0.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
25
+ "wav2vec2.encoder.layers.0.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
26
+ "wav2vec2.encoder.layers.0.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
27
+ "wav2vec2.encoder.layers.0.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
28
+ "wav2vec2.encoder.layers.0.final_layer_norm.bias": "model-00001-of-00003.safetensors",
29
+ "wav2vec2.encoder.layers.0.final_layer_norm.weight": "model-00001-of-00003.safetensors",
30
+ "wav2vec2.encoder.layers.0.layer_norm.bias": "model-00001-of-00003.safetensors",
31
+ "wav2vec2.encoder.layers.0.layer_norm.weight": "model-00001-of-00003.safetensors",
32
+ "wav2vec2.encoder.layers.1.attention.k_proj.bias": "model-00001-of-00003.safetensors",
33
+ "wav2vec2.encoder.layers.1.attention.k_proj.weight": "model-00001-of-00003.safetensors",
34
+ "wav2vec2.encoder.layers.1.attention.out_proj.bias": "model-00001-of-00003.safetensors",
35
+ "wav2vec2.encoder.layers.1.attention.out_proj.weight": "model-00001-of-00003.safetensors",
36
+ "wav2vec2.encoder.layers.1.attention.q_proj.bias": "model-00001-of-00003.safetensors",
37
+ "wav2vec2.encoder.layers.1.attention.q_proj.weight": "model-00001-of-00003.safetensors",
38
+ "wav2vec2.encoder.layers.1.attention.v_proj.bias": "model-00001-of-00003.safetensors",
39
+ "wav2vec2.encoder.layers.1.attention.v_proj.weight": "model-00001-of-00003.safetensors",
40
+ "wav2vec2.encoder.layers.1.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
41
+ "wav2vec2.encoder.layers.1.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
42
+ "wav2vec2.encoder.layers.1.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
43
+ "wav2vec2.encoder.layers.1.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
44
+ "wav2vec2.encoder.layers.1.final_layer_norm.bias": "model-00001-of-00003.safetensors",
45
+ "wav2vec2.encoder.layers.1.final_layer_norm.weight": "model-00001-of-00003.safetensors",
46
+ "wav2vec2.encoder.layers.1.layer_norm.bias": "model-00001-of-00003.safetensors",
47
+ "wav2vec2.encoder.layers.1.layer_norm.weight": "model-00001-of-00003.safetensors",
48
+ "wav2vec2.encoder.layers.10.attention.k_proj.bias": "model-00001-of-00003.safetensors",
49
+ "wav2vec2.encoder.layers.10.attention.k_proj.weight": "model-00001-of-00003.safetensors",
50
+ "wav2vec2.encoder.layers.10.attention.out_proj.bias": "model-00001-of-00003.safetensors",
51
+ "wav2vec2.encoder.layers.10.attention.out_proj.weight": "model-00001-of-00003.safetensors",
52
+ "wav2vec2.encoder.layers.10.attention.q_proj.bias": "model-00001-of-00003.safetensors",
53
+ "wav2vec2.encoder.layers.10.attention.q_proj.weight": "model-00001-of-00003.safetensors",
54
+ "wav2vec2.encoder.layers.10.attention.v_proj.bias": "model-00001-of-00003.safetensors",
55
+ "wav2vec2.encoder.layers.10.attention.v_proj.weight": "model-00001-of-00003.safetensors",
56
+ "wav2vec2.encoder.layers.10.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
57
+ "wav2vec2.encoder.layers.10.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
58
+ "wav2vec2.encoder.layers.10.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
59
+ "wav2vec2.encoder.layers.10.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
60
+ "wav2vec2.encoder.layers.10.final_layer_norm.bias": "model-00001-of-00003.safetensors",
61
+ "wav2vec2.encoder.layers.10.final_layer_norm.weight": "model-00001-of-00003.safetensors",
62
+ "wav2vec2.encoder.layers.10.layer_norm.bias": "model-00001-of-00003.safetensors",
63
+ "wav2vec2.encoder.layers.10.layer_norm.weight": "model-00001-of-00003.safetensors",
64
+ "wav2vec2.encoder.layers.11.attention.k_proj.bias": "model-00001-of-00003.safetensors",
65
+ "wav2vec2.encoder.layers.11.attention.k_proj.weight": "model-00001-of-00003.safetensors",
66
+ "wav2vec2.encoder.layers.11.attention.out_proj.bias": "model-00001-of-00003.safetensors",
67
+ "wav2vec2.encoder.layers.11.attention.out_proj.weight": "model-00001-of-00003.safetensors",
68
+ "wav2vec2.encoder.layers.11.attention.q_proj.bias": "model-00001-of-00003.safetensors",
69
+ "wav2vec2.encoder.layers.11.attention.q_proj.weight": "model-00001-of-00003.safetensors",
70
+ "wav2vec2.encoder.layers.11.attention.v_proj.bias": "model-00001-of-00003.safetensors",
71
+ "wav2vec2.encoder.layers.11.attention.v_proj.weight": "model-00001-of-00003.safetensors",
72
+ "wav2vec2.encoder.layers.11.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
73
+ "wav2vec2.encoder.layers.11.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
74
+ "wav2vec2.encoder.layers.11.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
75
+ "wav2vec2.encoder.layers.11.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
76
+ "wav2vec2.encoder.layers.11.final_layer_norm.bias": "model-00001-of-00003.safetensors",
77
+ "wav2vec2.encoder.layers.11.final_layer_norm.weight": "model-00001-of-00003.safetensors",
78
+ "wav2vec2.encoder.layers.11.layer_norm.bias": "model-00001-of-00003.safetensors",
79
+ "wav2vec2.encoder.layers.11.layer_norm.weight": "model-00001-of-00003.safetensors",
80
+ "wav2vec2.encoder.layers.12.attention.k_proj.bias": "model-00001-of-00003.safetensors",
81
+ "wav2vec2.encoder.layers.12.attention.k_proj.weight": "model-00001-of-00003.safetensors",
82
+ "wav2vec2.encoder.layers.12.attention.out_proj.bias": "model-00001-of-00003.safetensors",
83
+ "wav2vec2.encoder.layers.12.attention.out_proj.weight": "model-00001-of-00003.safetensors",
84
+ "wav2vec2.encoder.layers.12.attention.q_proj.bias": "model-00001-of-00003.safetensors",
85
+ "wav2vec2.encoder.layers.12.attention.q_proj.weight": "model-00001-of-00003.safetensors",
86
+ "wav2vec2.encoder.layers.12.attention.v_proj.bias": "model-00001-of-00003.safetensors",
87
+ "wav2vec2.encoder.layers.12.attention.v_proj.weight": "model-00001-of-00003.safetensors",
88
+ "wav2vec2.encoder.layers.12.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
89
+ "wav2vec2.encoder.layers.12.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
90
+ "wav2vec2.encoder.layers.12.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
91
+ "wav2vec2.encoder.layers.12.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
92
+ "wav2vec2.encoder.layers.12.final_layer_norm.bias": "model-00001-of-00003.safetensors",
93
+ "wav2vec2.encoder.layers.12.final_layer_norm.weight": "model-00001-of-00003.safetensors",
94
+ "wav2vec2.encoder.layers.12.layer_norm.bias": "model-00001-of-00003.safetensors",
95
+ "wav2vec2.encoder.layers.12.layer_norm.weight": "model-00001-of-00003.safetensors",
96
+ "wav2vec2.encoder.layers.13.attention.k_proj.bias": "model-00001-of-00003.safetensors",
97
+ "wav2vec2.encoder.layers.13.attention.k_proj.weight": "model-00001-of-00003.safetensors",
98
+ "wav2vec2.encoder.layers.13.attention.out_proj.bias": "model-00001-of-00003.safetensors",
99
+ "wav2vec2.encoder.layers.13.attention.out_proj.weight": "model-00001-of-00003.safetensors",
100
+ "wav2vec2.encoder.layers.13.attention.q_proj.bias": "model-00001-of-00003.safetensors",
101
+ "wav2vec2.encoder.layers.13.attention.q_proj.weight": "model-00001-of-00003.safetensors",
102
+ "wav2vec2.encoder.layers.13.attention.v_proj.bias": "model-00001-of-00003.safetensors",
103
+ "wav2vec2.encoder.layers.13.attention.v_proj.weight": "model-00001-of-00003.safetensors",
104
+ "wav2vec2.encoder.layers.13.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
105
+ "wav2vec2.encoder.layers.13.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
106
+ "wav2vec2.encoder.layers.13.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
107
+ "wav2vec2.encoder.layers.13.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
108
+ "wav2vec2.encoder.layers.13.final_layer_norm.bias": "model-00001-of-00003.safetensors",
109
+ "wav2vec2.encoder.layers.13.final_layer_norm.weight": "model-00001-of-00003.safetensors",
110
+ "wav2vec2.encoder.layers.13.layer_norm.bias": "model-00001-of-00003.safetensors",
111
+ "wav2vec2.encoder.layers.13.layer_norm.weight": "model-00001-of-00003.safetensors",
112
+ "wav2vec2.encoder.layers.14.attention.k_proj.bias": "model-00001-of-00003.safetensors",
113
+ "wav2vec2.encoder.layers.14.attention.k_proj.weight": "model-00001-of-00003.safetensors",
114
+ "wav2vec2.encoder.layers.14.attention.out_proj.bias": "model-00001-of-00003.safetensors",
115
+ "wav2vec2.encoder.layers.14.attention.out_proj.weight": "model-00001-of-00003.safetensors",
116
+ "wav2vec2.encoder.layers.14.attention.q_proj.bias": "model-00001-of-00003.safetensors",
117
+ "wav2vec2.encoder.layers.14.attention.q_proj.weight": "model-00001-of-00003.safetensors",
118
+ "wav2vec2.encoder.layers.14.attention.v_proj.bias": "model-00001-of-00003.safetensors",
119
+ "wav2vec2.encoder.layers.14.attention.v_proj.weight": "model-00001-of-00003.safetensors",
120
+ "wav2vec2.encoder.layers.14.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
121
+ "wav2vec2.encoder.layers.14.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
122
+ "wav2vec2.encoder.layers.14.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
123
+ "wav2vec2.encoder.layers.14.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
124
+ "wav2vec2.encoder.layers.14.final_layer_norm.bias": "model-00001-of-00003.safetensors",
125
+ "wav2vec2.encoder.layers.14.final_layer_norm.weight": "model-00001-of-00003.safetensors",
126
+ "wav2vec2.encoder.layers.14.layer_norm.bias": "model-00001-of-00003.safetensors",
127
+ "wav2vec2.encoder.layers.14.layer_norm.weight": "model-00001-of-00003.safetensors",
128
+ "wav2vec2.encoder.layers.15.attention.k_proj.bias": "model-00001-of-00003.safetensors",
129
+ "wav2vec2.encoder.layers.15.attention.k_proj.weight": "model-00001-of-00003.safetensors",
130
+ "wav2vec2.encoder.layers.15.attention.out_proj.bias": "model-00001-of-00003.safetensors",
131
+ "wav2vec2.encoder.layers.15.attention.out_proj.weight": "model-00001-of-00003.safetensors",
132
+ "wav2vec2.encoder.layers.15.attention.q_proj.bias": "model-00001-of-00003.safetensors",
133
+ "wav2vec2.encoder.layers.15.attention.q_proj.weight": "model-00001-of-00003.safetensors",
134
+ "wav2vec2.encoder.layers.15.attention.v_proj.bias": "model-00001-of-00003.safetensors",
135
+ "wav2vec2.encoder.layers.15.attention.v_proj.weight": "model-00001-of-00003.safetensors",
136
+ "wav2vec2.encoder.layers.15.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
137
+ "wav2vec2.encoder.layers.15.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
138
+ "wav2vec2.encoder.layers.15.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
139
+ "wav2vec2.encoder.layers.15.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
140
+ "wav2vec2.encoder.layers.15.final_layer_norm.bias": "model-00001-of-00003.safetensors",
141
+ "wav2vec2.encoder.layers.15.final_layer_norm.weight": "model-00001-of-00003.safetensors",
142
+ "wav2vec2.encoder.layers.15.layer_norm.bias": "model-00001-of-00003.safetensors",
143
+ "wav2vec2.encoder.layers.15.layer_norm.weight": "model-00001-of-00003.safetensors",
144
+ "wav2vec2.encoder.layers.16.attention.k_proj.bias": "model-00001-of-00003.safetensors",
145
+ "wav2vec2.encoder.layers.16.attention.k_proj.weight": "model-00001-of-00003.safetensors",
146
+ "wav2vec2.encoder.layers.16.attention.out_proj.bias": "model-00001-of-00003.safetensors",
147
+ "wav2vec2.encoder.layers.16.attention.out_proj.weight": "model-00001-of-00003.safetensors",
148
+ "wav2vec2.encoder.layers.16.attention.q_proj.bias": "model-00001-of-00003.safetensors",
149
+ "wav2vec2.encoder.layers.16.attention.q_proj.weight": "model-00001-of-00003.safetensors",
150
+ "wav2vec2.encoder.layers.16.attention.v_proj.bias": "model-00001-of-00003.safetensors",
151
+ "wav2vec2.encoder.layers.16.attention.v_proj.weight": "model-00001-of-00003.safetensors",
152
+ "wav2vec2.encoder.layers.16.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
153
+ "wav2vec2.encoder.layers.16.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
154
+ "wav2vec2.encoder.layers.16.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
155
+ "wav2vec2.encoder.layers.16.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
156
+ "wav2vec2.encoder.layers.16.final_layer_norm.bias": "model-00001-of-00003.safetensors",
157
+ "wav2vec2.encoder.layers.16.final_layer_norm.weight": "model-00001-of-00003.safetensors",
158
+ "wav2vec2.encoder.layers.16.layer_norm.bias": "model-00001-of-00003.safetensors",
159
+ "wav2vec2.encoder.layers.16.layer_norm.weight": "model-00001-of-00003.safetensors",
160
+ "wav2vec2.encoder.layers.17.attention.k_proj.bias": "model-00001-of-00003.safetensors",
161
+ "wav2vec2.encoder.layers.17.attention.k_proj.weight": "model-00001-of-00003.safetensors",
162
+ "wav2vec2.encoder.layers.17.attention.out_proj.bias": "model-00001-of-00003.safetensors",
163
+ "wav2vec2.encoder.layers.17.attention.out_proj.weight": "model-00001-of-00003.safetensors",
164
+ "wav2vec2.encoder.layers.17.attention.q_proj.bias": "model-00001-of-00003.safetensors",
165
+ "wav2vec2.encoder.layers.17.attention.q_proj.weight": "model-00001-of-00003.safetensors",
166
+ "wav2vec2.encoder.layers.17.attention.v_proj.bias": "model-00001-of-00003.safetensors",
167
+ "wav2vec2.encoder.layers.17.attention.v_proj.weight": "model-00001-of-00003.safetensors",
168
+ "wav2vec2.encoder.layers.17.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
169
+ "wav2vec2.encoder.layers.17.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
170
+ "wav2vec2.encoder.layers.17.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
171
+ "wav2vec2.encoder.layers.17.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
172
+ "wav2vec2.encoder.layers.17.final_layer_norm.bias": "model-00001-of-00003.safetensors",
173
+ "wav2vec2.encoder.layers.17.final_layer_norm.weight": "model-00001-of-00003.safetensors",
174
+ "wav2vec2.encoder.layers.17.layer_norm.bias": "model-00001-of-00003.safetensors",
175
+ "wav2vec2.encoder.layers.17.layer_norm.weight": "model-00001-of-00003.safetensors",
176
+ "wav2vec2.encoder.layers.18.attention.k_proj.bias": "model-00001-of-00003.safetensors",
177
+ "wav2vec2.encoder.layers.18.attention.k_proj.weight": "model-00001-of-00003.safetensors",
178
+ "wav2vec2.encoder.layers.18.attention.out_proj.bias": "model-00001-of-00003.safetensors",
179
+ "wav2vec2.encoder.layers.18.attention.out_proj.weight": "model-00001-of-00003.safetensors",
180
+ "wav2vec2.encoder.layers.18.attention.q_proj.bias": "model-00001-of-00003.safetensors",
181
+ "wav2vec2.encoder.layers.18.attention.q_proj.weight": "model-00001-of-00003.safetensors",
182
+ "wav2vec2.encoder.layers.18.attention.v_proj.bias": "model-00001-of-00003.safetensors",
183
+ "wav2vec2.encoder.layers.18.attention.v_proj.weight": "model-00001-of-00003.safetensors",
184
+ "wav2vec2.encoder.layers.18.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
185
+ "wav2vec2.encoder.layers.18.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
186
+ "wav2vec2.encoder.layers.18.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
187
+ "wav2vec2.encoder.layers.18.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
188
+ "wav2vec2.encoder.layers.18.final_layer_norm.bias": "model-00001-of-00003.safetensors",
189
+ "wav2vec2.encoder.layers.18.final_layer_norm.weight": "model-00001-of-00003.safetensors",
190
+ "wav2vec2.encoder.layers.18.layer_norm.bias": "model-00001-of-00003.safetensors",
191
+ "wav2vec2.encoder.layers.18.layer_norm.weight": "model-00001-of-00003.safetensors",
192
+ "wav2vec2.encoder.layers.19.attention.k_proj.bias": "model-00001-of-00003.safetensors",
193
+ "wav2vec2.encoder.layers.19.attention.k_proj.weight": "model-00001-of-00003.safetensors",
194
+ "wav2vec2.encoder.layers.19.attention.out_proj.bias": "model-00001-of-00003.safetensors",
195
+ "wav2vec2.encoder.layers.19.attention.out_proj.weight": "model-00001-of-00003.safetensors",
196
+ "wav2vec2.encoder.layers.19.attention.q_proj.bias": "model-00001-of-00003.safetensors",
197
+ "wav2vec2.encoder.layers.19.attention.q_proj.weight": "model-00001-of-00003.safetensors",
198
+ "wav2vec2.encoder.layers.19.attention.v_proj.bias": "model-00001-of-00003.safetensors",
199
+ "wav2vec2.encoder.layers.19.attention.v_proj.weight": "model-00001-of-00003.safetensors",
200
+ "wav2vec2.encoder.layers.19.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
201
+ "wav2vec2.encoder.layers.19.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
202
+ "wav2vec2.encoder.layers.19.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
203
+ "wav2vec2.encoder.layers.19.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
204
+ "wav2vec2.encoder.layers.19.final_layer_norm.bias": "model-00001-of-00003.safetensors",
205
+ "wav2vec2.encoder.layers.19.final_layer_norm.weight": "model-00001-of-00003.safetensors",
206
+ "wav2vec2.encoder.layers.19.layer_norm.bias": "model-00001-of-00003.safetensors",
207
+ "wav2vec2.encoder.layers.19.layer_norm.weight": "model-00001-of-00003.safetensors",
208
+ "wav2vec2.encoder.layers.2.attention.k_proj.bias": "model-00001-of-00003.safetensors",
209
+ "wav2vec2.encoder.layers.2.attention.k_proj.weight": "model-00001-of-00003.safetensors",
210
+ "wav2vec2.encoder.layers.2.attention.out_proj.bias": "model-00001-of-00003.safetensors",
211
+ "wav2vec2.encoder.layers.2.attention.out_proj.weight": "model-00001-of-00003.safetensors",
212
+ "wav2vec2.encoder.layers.2.attention.q_proj.bias": "model-00001-of-00003.safetensors",
213
+ "wav2vec2.encoder.layers.2.attention.q_proj.weight": "model-00001-of-00003.safetensors",
214
+ "wav2vec2.encoder.layers.2.attention.v_proj.bias": "model-00001-of-00003.safetensors",
215
+ "wav2vec2.encoder.layers.2.attention.v_proj.weight": "model-00001-of-00003.safetensors",
216
+ "wav2vec2.encoder.layers.2.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
217
+ "wav2vec2.encoder.layers.2.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
218
+ "wav2vec2.encoder.layers.2.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
219
+ "wav2vec2.encoder.layers.2.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
220
+ "wav2vec2.encoder.layers.2.final_layer_norm.bias": "model-00001-of-00003.safetensors",
221
+ "wav2vec2.encoder.layers.2.final_layer_norm.weight": "model-00001-of-00003.safetensors",
222
+ "wav2vec2.encoder.layers.2.layer_norm.bias": "model-00001-of-00003.safetensors",
223
+ "wav2vec2.encoder.layers.2.layer_norm.weight": "model-00001-of-00003.safetensors",
224
+ "wav2vec2.encoder.layers.20.attention.k_proj.bias": "model-00001-of-00003.safetensors",
225
+ "wav2vec2.encoder.layers.20.attention.k_proj.weight": "model-00001-of-00003.safetensors",
226
+ "wav2vec2.encoder.layers.20.attention.out_proj.bias": "model-00001-of-00003.safetensors",
227
+ "wav2vec2.encoder.layers.20.attention.out_proj.weight": "model-00001-of-00003.safetensors",
228
+ "wav2vec2.encoder.layers.20.attention.q_proj.bias": "model-00001-of-00003.safetensors",
229
+ "wav2vec2.encoder.layers.20.attention.q_proj.weight": "model-00001-of-00003.safetensors",
230
+ "wav2vec2.encoder.layers.20.attention.v_proj.bias": "model-00001-of-00003.safetensors",
231
+ "wav2vec2.encoder.layers.20.attention.v_proj.weight": "model-00001-of-00003.safetensors",
232
+ "wav2vec2.encoder.layers.20.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
233
+ "wav2vec2.encoder.layers.20.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
234
+ "wav2vec2.encoder.layers.20.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
235
+ "wav2vec2.encoder.layers.20.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
236
+ "wav2vec2.encoder.layers.20.final_layer_norm.bias": "model-00001-of-00003.safetensors",
237
+ "wav2vec2.encoder.layers.20.final_layer_norm.weight": "model-00001-of-00003.safetensors",
238
+ "wav2vec2.encoder.layers.20.layer_norm.bias": "model-00001-of-00003.safetensors",
239
+ "wav2vec2.encoder.layers.20.layer_norm.weight": "model-00001-of-00003.safetensors",
240
+ "wav2vec2.encoder.layers.21.attention.k_proj.bias": "model-00001-of-00003.safetensors",
241
+ "wav2vec2.encoder.layers.21.attention.k_proj.weight": "model-00001-of-00003.safetensors",
242
+ "wav2vec2.encoder.layers.21.attention.out_proj.bias": "model-00001-of-00003.safetensors",
243
+ "wav2vec2.encoder.layers.21.attention.out_proj.weight": "model-00001-of-00003.safetensors",
244
+ "wav2vec2.encoder.layers.21.attention.q_proj.bias": "model-00001-of-00003.safetensors",
245
+ "wav2vec2.encoder.layers.21.attention.q_proj.weight": "model-00001-of-00003.safetensors",
246
+ "wav2vec2.encoder.layers.21.attention.v_proj.bias": "model-00001-of-00003.safetensors",
247
+ "wav2vec2.encoder.layers.21.attention.v_proj.weight": "model-00001-of-00003.safetensors",
248
+ "wav2vec2.encoder.layers.21.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
249
+ "wav2vec2.encoder.layers.21.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
250
+ "wav2vec2.encoder.layers.21.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
251
+ "wav2vec2.encoder.layers.21.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
252
+ "wav2vec2.encoder.layers.21.final_layer_norm.bias": "model-00001-of-00003.safetensors",
253
+ "wav2vec2.encoder.layers.21.final_layer_norm.weight": "model-00001-of-00003.safetensors",
254
+ "wav2vec2.encoder.layers.21.layer_norm.bias": "model-00001-of-00003.safetensors",
255
+ "wav2vec2.encoder.layers.21.layer_norm.weight": "model-00001-of-00003.safetensors",
256
+ "wav2vec2.encoder.layers.22.attention.k_proj.bias": "model-00001-of-00003.safetensors",
257
+ "wav2vec2.encoder.layers.22.attention.k_proj.weight": "model-00001-of-00003.safetensors",
258
+ "wav2vec2.encoder.layers.22.attention.out_proj.bias": "model-00001-of-00003.safetensors",
259
+ "wav2vec2.encoder.layers.22.attention.out_proj.weight": "model-00001-of-00003.safetensors",
260
+ "wav2vec2.encoder.layers.22.attention.q_proj.bias": "model-00001-of-00003.safetensors",
261
+ "wav2vec2.encoder.layers.22.attention.q_proj.weight": "model-00001-of-00003.safetensors",
262
+ "wav2vec2.encoder.layers.22.attention.v_proj.bias": "model-00001-of-00003.safetensors",
263
+ "wav2vec2.encoder.layers.22.attention.v_proj.weight": "model-00001-of-00003.safetensors",
264
+ "wav2vec2.encoder.layers.22.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
265
+ "wav2vec2.encoder.layers.22.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
266
+ "wav2vec2.encoder.layers.22.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
267
+ "wav2vec2.encoder.layers.22.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
268
+ "wav2vec2.encoder.layers.22.final_layer_norm.bias": "model-00001-of-00003.safetensors",
269
+ "wav2vec2.encoder.layers.22.final_layer_norm.weight": "model-00001-of-00003.safetensors",
270
+ "wav2vec2.encoder.layers.22.layer_norm.bias": "model-00001-of-00003.safetensors",
271
+ "wav2vec2.encoder.layers.22.layer_norm.weight": "model-00001-of-00003.safetensors",
272
+ "wav2vec2.encoder.layers.23.attention.k_proj.bias": "model-00001-of-00003.safetensors",
273
+ "wav2vec2.encoder.layers.23.attention.k_proj.weight": "model-00001-of-00003.safetensors",
274
+ "wav2vec2.encoder.layers.23.attention.out_proj.bias": "model-00001-of-00003.safetensors",
275
+ "wav2vec2.encoder.layers.23.attention.out_proj.weight": "model-00001-of-00003.safetensors",
276
+ "wav2vec2.encoder.layers.23.attention.q_proj.bias": "model-00001-of-00003.safetensors",
277
+ "wav2vec2.encoder.layers.23.attention.q_proj.weight": "model-00001-of-00003.safetensors",
278
+ "wav2vec2.encoder.layers.23.attention.v_proj.bias": "model-00001-of-00003.safetensors",
279
+ "wav2vec2.encoder.layers.23.attention.v_proj.weight": "model-00001-of-00003.safetensors",
280
+ "wav2vec2.encoder.layers.23.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
281
+ "wav2vec2.encoder.layers.23.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
282
+ "wav2vec2.encoder.layers.23.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
283
+ "wav2vec2.encoder.layers.23.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
284
+ "wav2vec2.encoder.layers.23.final_layer_norm.bias": "model-00001-of-00003.safetensors",
285
+ "wav2vec2.encoder.layers.23.final_layer_norm.weight": "model-00001-of-00003.safetensors",
286
+ "wav2vec2.encoder.layers.23.layer_norm.bias": "model-00001-of-00003.safetensors",
287
+ "wav2vec2.encoder.layers.23.layer_norm.weight": "model-00001-of-00003.safetensors",
288
+ "wav2vec2.encoder.layers.24.attention.k_proj.bias": "model-00002-of-00003.safetensors",
289
+ "wav2vec2.encoder.layers.24.attention.k_proj.weight": "model-00002-of-00003.safetensors",
290
+ "wav2vec2.encoder.layers.24.attention.out_proj.bias": "model-00002-of-00003.safetensors",
291
+ "wav2vec2.encoder.layers.24.attention.out_proj.weight": "model-00002-of-00003.safetensors",
292
+ "wav2vec2.encoder.layers.24.attention.q_proj.bias": "model-00002-of-00003.safetensors",
293
+ "wav2vec2.encoder.layers.24.attention.q_proj.weight": "model-00002-of-00003.safetensors",
294
+ "wav2vec2.encoder.layers.24.attention.v_proj.bias": "model-00002-of-00003.safetensors",
295
+ "wav2vec2.encoder.layers.24.attention.v_proj.weight": "model-00002-of-00003.safetensors",
296
+ "wav2vec2.encoder.layers.24.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
297
+ "wav2vec2.encoder.layers.24.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
298
+ "wav2vec2.encoder.layers.24.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
299
+ "wav2vec2.encoder.layers.24.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
300
+ "wav2vec2.encoder.layers.24.final_layer_norm.bias": "model-00002-of-00003.safetensors",
301
+ "wav2vec2.encoder.layers.24.final_layer_norm.weight": "model-00002-of-00003.safetensors",
302
+ "wav2vec2.encoder.layers.24.layer_norm.bias": "model-00002-of-00003.safetensors",
303
+ "wav2vec2.encoder.layers.24.layer_norm.weight": "model-00002-of-00003.safetensors",
304
+ "wav2vec2.encoder.layers.25.attention.k_proj.bias": "model-00002-of-00003.safetensors",
305
+ "wav2vec2.encoder.layers.25.attention.k_proj.weight": "model-00002-of-00003.safetensors",
306
+ "wav2vec2.encoder.layers.25.attention.out_proj.bias": "model-00002-of-00003.safetensors",
307
+ "wav2vec2.encoder.layers.25.attention.out_proj.weight": "model-00002-of-00003.safetensors",
308
+ "wav2vec2.encoder.layers.25.attention.q_proj.bias": "model-00002-of-00003.safetensors",
309
+ "wav2vec2.encoder.layers.25.attention.q_proj.weight": "model-00002-of-00003.safetensors",
310
+ "wav2vec2.encoder.layers.25.attention.v_proj.bias": "model-00002-of-00003.safetensors",
311
+ "wav2vec2.encoder.layers.25.attention.v_proj.weight": "model-00002-of-00003.safetensors",
312
+ "wav2vec2.encoder.layers.25.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
313
+ "wav2vec2.encoder.layers.25.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
314
+ "wav2vec2.encoder.layers.25.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
315
+ "wav2vec2.encoder.layers.25.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
316
+ "wav2vec2.encoder.layers.25.final_layer_norm.bias": "model-00002-of-00003.safetensors",
317
+ "wav2vec2.encoder.layers.25.final_layer_norm.weight": "model-00002-of-00003.safetensors",
318
+ "wav2vec2.encoder.layers.25.layer_norm.bias": "model-00002-of-00003.safetensors",
319
+ "wav2vec2.encoder.layers.25.layer_norm.weight": "model-00002-of-00003.safetensors",
320
+ "wav2vec2.encoder.layers.26.attention.k_proj.bias": "model-00002-of-00003.safetensors",
321
+ "wav2vec2.encoder.layers.26.attention.k_proj.weight": "model-00002-of-00003.safetensors",
322
+ "wav2vec2.encoder.layers.26.attention.out_proj.bias": "model-00002-of-00003.safetensors",
323
+ "wav2vec2.encoder.layers.26.attention.out_proj.weight": "model-00002-of-00003.safetensors",
324
+ "wav2vec2.encoder.layers.26.attention.q_proj.bias": "model-00002-of-00003.safetensors",
325
+ "wav2vec2.encoder.layers.26.attention.q_proj.weight": "model-00002-of-00003.safetensors",
326
+ "wav2vec2.encoder.layers.26.attention.v_proj.bias": "model-00002-of-00003.safetensors",
327
+ "wav2vec2.encoder.layers.26.attention.v_proj.weight": "model-00002-of-00003.safetensors",
328
+ "wav2vec2.encoder.layers.26.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
329
+ "wav2vec2.encoder.layers.26.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
330
+ "wav2vec2.encoder.layers.26.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
331
+ "wav2vec2.encoder.layers.26.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
332
+ "wav2vec2.encoder.layers.26.final_layer_norm.bias": "model-00002-of-00003.safetensors",
333
+ "wav2vec2.encoder.layers.26.final_layer_norm.weight": "model-00002-of-00003.safetensors",
334
+ "wav2vec2.encoder.layers.26.layer_norm.bias": "model-00002-of-00003.safetensors",
335
+ "wav2vec2.encoder.layers.26.layer_norm.weight": "model-00002-of-00003.safetensors",
336
+ "wav2vec2.encoder.layers.27.attention.k_proj.bias": "model-00002-of-00003.safetensors",
337
+ "wav2vec2.encoder.layers.27.attention.k_proj.weight": "model-00002-of-00003.safetensors",
338
+ "wav2vec2.encoder.layers.27.attention.out_proj.bias": "model-00002-of-00003.safetensors",
339
+ "wav2vec2.encoder.layers.27.attention.out_proj.weight": "model-00002-of-00003.safetensors",
340
+ "wav2vec2.encoder.layers.27.attention.q_proj.bias": "model-00002-of-00003.safetensors",
341
+ "wav2vec2.encoder.layers.27.attention.q_proj.weight": "model-00002-of-00003.safetensors",
342
+ "wav2vec2.encoder.layers.27.attention.v_proj.bias": "model-00002-of-00003.safetensors",
343
+ "wav2vec2.encoder.layers.27.attention.v_proj.weight": "model-00002-of-00003.safetensors",
344
+ "wav2vec2.encoder.layers.27.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
345
+ "wav2vec2.encoder.layers.27.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
346
+ "wav2vec2.encoder.layers.27.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
347
+ "wav2vec2.encoder.layers.27.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
348
+ "wav2vec2.encoder.layers.27.final_layer_norm.bias": "model-00002-of-00003.safetensors",
349
+ "wav2vec2.encoder.layers.27.final_layer_norm.weight": "model-00002-of-00003.safetensors",
350
+ "wav2vec2.encoder.layers.27.layer_norm.bias": "model-00002-of-00003.safetensors",
351
+ "wav2vec2.encoder.layers.27.layer_norm.weight": "model-00002-of-00003.safetensors",
352
+ "wav2vec2.encoder.layers.28.attention.k_proj.bias": "model-00002-of-00003.safetensors",
353
+ "wav2vec2.encoder.layers.28.attention.k_proj.weight": "model-00002-of-00003.safetensors",
354
+ "wav2vec2.encoder.layers.28.attention.out_proj.bias": "model-00002-of-00003.safetensors",
355
+ "wav2vec2.encoder.layers.28.attention.out_proj.weight": "model-00002-of-00003.safetensors",
356
+ "wav2vec2.encoder.layers.28.attention.q_proj.bias": "model-00002-of-00003.safetensors",
357
+ "wav2vec2.encoder.layers.28.attention.q_proj.weight": "model-00002-of-00003.safetensors",
358
+ "wav2vec2.encoder.layers.28.attention.v_proj.bias": "model-00002-of-00003.safetensors",
359
+ "wav2vec2.encoder.layers.28.attention.v_proj.weight": "model-00002-of-00003.safetensors",
360
+ "wav2vec2.encoder.layers.28.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
361
+ "wav2vec2.encoder.layers.28.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
362
+ "wav2vec2.encoder.layers.28.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
363
+ "wav2vec2.encoder.layers.28.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
364
+ "wav2vec2.encoder.layers.28.final_layer_norm.bias": "model-00002-of-00003.safetensors",
365
+ "wav2vec2.encoder.layers.28.final_layer_norm.weight": "model-00002-of-00003.safetensors",
366
+ "wav2vec2.encoder.layers.28.layer_norm.bias": "model-00002-of-00003.safetensors",
367
+ "wav2vec2.encoder.layers.28.layer_norm.weight": "model-00002-of-00003.safetensors",
368
+ "wav2vec2.encoder.layers.29.attention.k_proj.bias": "model-00002-of-00003.safetensors",
369
+ "wav2vec2.encoder.layers.29.attention.k_proj.weight": "model-00002-of-00003.safetensors",
370
+ "wav2vec2.encoder.layers.29.attention.out_proj.bias": "model-00002-of-00003.safetensors",
371
+ "wav2vec2.encoder.layers.29.attention.out_proj.weight": "model-00002-of-00003.safetensors",
372
+ "wav2vec2.encoder.layers.29.attention.q_proj.bias": "model-00002-of-00003.safetensors",
373
+ "wav2vec2.encoder.layers.29.attention.q_proj.weight": "model-00002-of-00003.safetensors",
374
+ "wav2vec2.encoder.layers.29.attention.v_proj.bias": "model-00002-of-00003.safetensors",
375
+ "wav2vec2.encoder.layers.29.attention.v_proj.weight": "model-00002-of-00003.safetensors",
376
+ "wav2vec2.encoder.layers.29.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
377
+ "wav2vec2.encoder.layers.29.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
378
+ "wav2vec2.encoder.layers.29.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
379
+ "wav2vec2.encoder.layers.29.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
380
+ "wav2vec2.encoder.layers.29.final_layer_norm.bias": "model-00002-of-00003.safetensors",
381
+ "wav2vec2.encoder.layers.29.final_layer_norm.weight": "model-00002-of-00003.safetensors",
382
+ "wav2vec2.encoder.layers.29.layer_norm.bias": "model-00002-of-00003.safetensors",
383
+ "wav2vec2.encoder.layers.29.layer_norm.weight": "model-00002-of-00003.safetensors",
384
+ "wav2vec2.encoder.layers.3.attention.k_proj.bias": "model-00001-of-00003.safetensors",
385
+ "wav2vec2.encoder.layers.3.attention.k_proj.weight": "model-00001-of-00003.safetensors",
386
+ "wav2vec2.encoder.layers.3.attention.out_proj.bias": "model-00001-of-00003.safetensors",
387
+ "wav2vec2.encoder.layers.3.attention.out_proj.weight": "model-00001-of-00003.safetensors",
388
+ "wav2vec2.encoder.layers.3.attention.q_proj.bias": "model-00001-of-00003.safetensors",
389
+ "wav2vec2.encoder.layers.3.attention.q_proj.weight": "model-00001-of-00003.safetensors",
390
+ "wav2vec2.encoder.layers.3.attention.v_proj.bias": "model-00001-of-00003.safetensors",
391
+ "wav2vec2.encoder.layers.3.attention.v_proj.weight": "model-00001-of-00003.safetensors",
392
+ "wav2vec2.encoder.layers.3.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
393
+ "wav2vec2.encoder.layers.3.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
394
+ "wav2vec2.encoder.layers.3.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
395
+ "wav2vec2.encoder.layers.3.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
396
+ "wav2vec2.encoder.layers.3.final_layer_norm.bias": "model-00001-of-00003.safetensors",
397
+ "wav2vec2.encoder.layers.3.final_layer_norm.weight": "model-00001-of-00003.safetensors",
398
+ "wav2vec2.encoder.layers.3.layer_norm.bias": "model-00001-of-00003.safetensors",
399
+ "wav2vec2.encoder.layers.3.layer_norm.weight": "model-00001-of-00003.safetensors",
400
+ "wav2vec2.encoder.layers.30.attention.k_proj.bias": "model-00002-of-00003.safetensors",
401
+ "wav2vec2.encoder.layers.30.attention.k_proj.weight": "model-00002-of-00003.safetensors",
402
+ "wav2vec2.encoder.layers.30.attention.out_proj.bias": "model-00002-of-00003.safetensors",
403
+ "wav2vec2.encoder.layers.30.attention.out_proj.weight": "model-00002-of-00003.safetensors",
404
+ "wav2vec2.encoder.layers.30.attention.q_proj.bias": "model-00002-of-00003.safetensors",
405
+ "wav2vec2.encoder.layers.30.attention.q_proj.weight": "model-00002-of-00003.safetensors",
406
+ "wav2vec2.encoder.layers.30.attention.v_proj.bias": "model-00002-of-00003.safetensors",
407
+ "wav2vec2.encoder.layers.30.attention.v_proj.weight": "model-00002-of-00003.safetensors",
408
+ "wav2vec2.encoder.layers.30.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
409
+ "wav2vec2.encoder.layers.30.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
410
+ "wav2vec2.encoder.layers.30.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
411
+ "wav2vec2.encoder.layers.30.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
412
+ "wav2vec2.encoder.layers.30.final_layer_norm.bias": "model-00002-of-00003.safetensors",
413
+ "wav2vec2.encoder.layers.30.final_layer_norm.weight": "model-00002-of-00003.safetensors",
414
+ "wav2vec2.encoder.layers.30.layer_norm.bias": "model-00002-of-00003.safetensors",
415
+ "wav2vec2.encoder.layers.30.layer_norm.weight": "model-00002-of-00003.safetensors",
416
+ "wav2vec2.encoder.layers.31.attention.k_proj.bias": "model-00002-of-00003.safetensors",
417
+ "wav2vec2.encoder.layers.31.attention.k_proj.weight": "model-00002-of-00003.safetensors",
418
+ "wav2vec2.encoder.layers.31.attention.out_proj.bias": "model-00002-of-00003.safetensors",
419
+ "wav2vec2.encoder.layers.31.attention.out_proj.weight": "model-00002-of-00003.safetensors",
420
+ "wav2vec2.encoder.layers.31.attention.q_proj.bias": "model-00002-of-00003.safetensors",
421
+ "wav2vec2.encoder.layers.31.attention.q_proj.weight": "model-00002-of-00003.safetensors",
422
+ "wav2vec2.encoder.layers.31.attention.v_proj.bias": "model-00002-of-00003.safetensors",
423
+ "wav2vec2.encoder.layers.31.attention.v_proj.weight": "model-00002-of-00003.safetensors",
424
+ "wav2vec2.encoder.layers.31.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
425
+ "wav2vec2.encoder.layers.31.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
426
+ "wav2vec2.encoder.layers.31.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
427
+ "wav2vec2.encoder.layers.31.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
428
+ "wav2vec2.encoder.layers.31.final_layer_norm.bias": "model-00002-of-00003.safetensors",
429
+ "wav2vec2.encoder.layers.31.final_layer_norm.weight": "model-00002-of-00003.safetensors",
430
+ "wav2vec2.encoder.layers.31.layer_norm.bias": "model-00002-of-00003.safetensors",
431
+ "wav2vec2.encoder.layers.31.layer_norm.weight": "model-00002-of-00003.safetensors",
432
+ "wav2vec2.encoder.layers.32.attention.k_proj.bias": "model-00002-of-00003.safetensors",
433
+ "wav2vec2.encoder.layers.32.attention.k_proj.weight": "model-00002-of-00003.safetensors",
434
+ "wav2vec2.encoder.layers.32.attention.out_proj.bias": "model-00002-of-00003.safetensors",
435
+ "wav2vec2.encoder.layers.32.attention.out_proj.weight": "model-00002-of-00003.safetensors",
436
+ "wav2vec2.encoder.layers.32.attention.q_proj.bias": "model-00002-of-00003.safetensors",
437
+ "wav2vec2.encoder.layers.32.attention.q_proj.weight": "model-00002-of-00003.safetensors",
438
+ "wav2vec2.encoder.layers.32.attention.v_proj.bias": "model-00002-of-00003.safetensors",
439
+ "wav2vec2.encoder.layers.32.attention.v_proj.weight": "model-00002-of-00003.safetensors",
440
+ "wav2vec2.encoder.layers.32.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
441
+ "wav2vec2.encoder.layers.32.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
442
+ "wav2vec2.encoder.layers.32.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
443
+ "wav2vec2.encoder.layers.32.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
444
+ "wav2vec2.encoder.layers.32.final_layer_norm.bias": "model-00002-of-00003.safetensors",
445
+ "wav2vec2.encoder.layers.32.final_layer_norm.weight": "model-00002-of-00003.safetensors",
446
+ "wav2vec2.encoder.layers.32.layer_norm.bias": "model-00002-of-00003.safetensors",
447
+ "wav2vec2.encoder.layers.32.layer_norm.weight": "model-00002-of-00003.safetensors",
448
+ "wav2vec2.encoder.layers.33.attention.k_proj.bias": "model-00002-of-00003.safetensors",
449
+ "wav2vec2.encoder.layers.33.attention.k_proj.weight": "model-00002-of-00003.safetensors",
450
+ "wav2vec2.encoder.layers.33.attention.out_proj.bias": "model-00002-of-00003.safetensors",
451
+ "wav2vec2.encoder.layers.33.attention.out_proj.weight": "model-00002-of-00003.safetensors",
452
+ "wav2vec2.encoder.layers.33.attention.q_proj.bias": "model-00002-of-00003.safetensors",
453
+ "wav2vec2.encoder.layers.33.attention.q_proj.weight": "model-00002-of-00003.safetensors",
454
+ "wav2vec2.encoder.layers.33.attention.v_proj.bias": "model-00002-of-00003.safetensors",
455
+ "wav2vec2.encoder.layers.33.attention.v_proj.weight": "model-00002-of-00003.safetensors",
456
+ "wav2vec2.encoder.layers.33.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
457
+ "wav2vec2.encoder.layers.33.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
458
+ "wav2vec2.encoder.layers.33.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
459
+ "wav2vec2.encoder.layers.33.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
460
+ "wav2vec2.encoder.layers.33.final_layer_norm.bias": "model-00002-of-00003.safetensors",
461
+ "wav2vec2.encoder.layers.33.final_layer_norm.weight": "model-00002-of-00003.safetensors",
462
+ "wav2vec2.encoder.layers.33.layer_norm.bias": "model-00002-of-00003.safetensors",
463
+ "wav2vec2.encoder.layers.33.layer_norm.weight": "model-00002-of-00003.safetensors",
464
+ "wav2vec2.encoder.layers.34.attention.k_proj.bias": "model-00002-of-00003.safetensors",
465
+ "wav2vec2.encoder.layers.34.attention.k_proj.weight": "model-00002-of-00003.safetensors",
466
+ "wav2vec2.encoder.layers.34.attention.out_proj.bias": "model-00002-of-00003.safetensors",
467
+ "wav2vec2.encoder.layers.34.attention.out_proj.weight": "model-00002-of-00003.safetensors",
468
+ "wav2vec2.encoder.layers.34.attention.q_proj.bias": "model-00002-of-00003.safetensors",
469
+ "wav2vec2.encoder.layers.34.attention.q_proj.weight": "model-00002-of-00003.safetensors",
470
+ "wav2vec2.encoder.layers.34.attention.v_proj.bias": "model-00002-of-00003.safetensors",
471
+ "wav2vec2.encoder.layers.34.attention.v_proj.weight": "model-00002-of-00003.safetensors",
472
+ "wav2vec2.encoder.layers.34.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
473
+ "wav2vec2.encoder.layers.34.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
474
+ "wav2vec2.encoder.layers.34.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
475
+ "wav2vec2.encoder.layers.34.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
476
+ "wav2vec2.encoder.layers.34.final_layer_norm.bias": "model-00002-of-00003.safetensors",
477
+ "wav2vec2.encoder.layers.34.final_layer_norm.weight": "model-00002-of-00003.safetensors",
478
+ "wav2vec2.encoder.layers.34.layer_norm.bias": "model-00002-of-00003.safetensors",
479
+ "wav2vec2.encoder.layers.34.layer_norm.weight": "model-00002-of-00003.safetensors",
480
+ "wav2vec2.encoder.layers.35.attention.k_proj.bias": "model-00002-of-00003.safetensors",
481
+ "wav2vec2.encoder.layers.35.attention.k_proj.weight": "model-00002-of-00003.safetensors",
482
+ "wav2vec2.encoder.layers.35.attention.out_proj.bias": "model-00002-of-00003.safetensors",
483
+ "wav2vec2.encoder.layers.35.attention.out_proj.weight": "model-00002-of-00003.safetensors",
484
+ "wav2vec2.encoder.layers.35.attention.q_proj.bias": "model-00002-of-00003.safetensors",
485
+ "wav2vec2.encoder.layers.35.attention.q_proj.weight": "model-00002-of-00003.safetensors",
486
+ "wav2vec2.encoder.layers.35.attention.v_proj.bias": "model-00002-of-00003.safetensors",
487
+ "wav2vec2.encoder.layers.35.attention.v_proj.weight": "model-00002-of-00003.safetensors",
488
+ "wav2vec2.encoder.layers.35.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
489
+ "wav2vec2.encoder.layers.35.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
490
+ "wav2vec2.encoder.layers.35.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
491
+ "wav2vec2.encoder.layers.35.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
492
+ "wav2vec2.encoder.layers.35.final_layer_norm.bias": "model-00002-of-00003.safetensors",
493
+ "wav2vec2.encoder.layers.35.final_layer_norm.weight": "model-00002-of-00003.safetensors",
494
+ "wav2vec2.encoder.layers.35.layer_norm.bias": "model-00002-of-00003.safetensors",
495
+ "wav2vec2.encoder.layers.35.layer_norm.weight": "model-00002-of-00003.safetensors",
496
+ "wav2vec2.encoder.layers.36.attention.k_proj.bias": "model-00002-of-00003.safetensors",
497
+ "wav2vec2.encoder.layers.36.attention.k_proj.weight": "model-00002-of-00003.safetensors",
498
+ "wav2vec2.encoder.layers.36.attention.out_proj.bias": "model-00002-of-00003.safetensors",
499
+ "wav2vec2.encoder.layers.36.attention.out_proj.weight": "model-00002-of-00003.safetensors",
500
+ "wav2vec2.encoder.layers.36.attention.q_proj.bias": "model-00002-of-00003.safetensors",
501
+ "wav2vec2.encoder.layers.36.attention.q_proj.weight": "model-00002-of-00003.safetensors",
502
+ "wav2vec2.encoder.layers.36.attention.v_proj.bias": "model-00002-of-00003.safetensors",
503
+ "wav2vec2.encoder.layers.36.attention.v_proj.weight": "model-00002-of-00003.safetensors",
504
+ "wav2vec2.encoder.layers.36.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
505
+ "wav2vec2.encoder.layers.36.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
506
+ "wav2vec2.encoder.layers.36.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
507
+ "wav2vec2.encoder.layers.36.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
508
+ "wav2vec2.encoder.layers.36.final_layer_norm.bias": "model-00002-of-00003.safetensors",
509
+ "wav2vec2.encoder.layers.36.final_layer_norm.weight": "model-00002-of-00003.safetensors",
510
+ "wav2vec2.encoder.layers.36.layer_norm.bias": "model-00002-of-00003.safetensors",
511
+ "wav2vec2.encoder.layers.36.layer_norm.weight": "model-00002-of-00003.safetensors",
512
+ "wav2vec2.encoder.layers.37.attention.k_proj.bias": "model-00002-of-00003.safetensors",
513
+ "wav2vec2.encoder.layers.37.attention.k_proj.weight": "model-00002-of-00003.safetensors",
514
+ "wav2vec2.encoder.layers.37.attention.out_proj.bias": "model-00002-of-00003.safetensors",
515
+ "wav2vec2.encoder.layers.37.attention.out_proj.weight": "model-00002-of-00003.safetensors",
516
+ "wav2vec2.encoder.layers.37.attention.q_proj.bias": "model-00002-of-00003.safetensors",
517
+ "wav2vec2.encoder.layers.37.attention.q_proj.weight": "model-00002-of-00003.safetensors",
518
+ "wav2vec2.encoder.layers.37.attention.v_proj.bias": "model-00002-of-00003.safetensors",
519
+ "wav2vec2.encoder.layers.37.attention.v_proj.weight": "model-00002-of-00003.safetensors",
520
+ "wav2vec2.encoder.layers.37.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
521
+ "wav2vec2.encoder.layers.37.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
522
+ "wav2vec2.encoder.layers.37.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
523
+ "wav2vec2.encoder.layers.37.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
524
+ "wav2vec2.encoder.layers.37.final_layer_norm.bias": "model-00002-of-00003.safetensors",
525
+ "wav2vec2.encoder.layers.37.final_layer_norm.weight": "model-00002-of-00003.safetensors",
526
+ "wav2vec2.encoder.layers.37.layer_norm.bias": "model-00002-of-00003.safetensors",
527
+ "wav2vec2.encoder.layers.37.layer_norm.weight": "model-00002-of-00003.safetensors",
528
+ "wav2vec2.encoder.layers.38.attention.k_proj.bias": "model-00002-of-00003.safetensors",
529
+ "wav2vec2.encoder.layers.38.attention.k_proj.weight": "model-00002-of-00003.safetensors",
530
+ "wav2vec2.encoder.layers.38.attention.out_proj.bias": "model-00002-of-00003.safetensors",
531
+ "wav2vec2.encoder.layers.38.attention.out_proj.weight": "model-00002-of-00003.safetensors",
532
+ "wav2vec2.encoder.layers.38.attention.q_proj.bias": "model-00002-of-00003.safetensors",
533
+ "wav2vec2.encoder.layers.38.attention.q_proj.weight": "model-00002-of-00003.safetensors",
534
+ "wav2vec2.encoder.layers.38.attention.v_proj.bias": "model-00002-of-00003.safetensors",
535
+ "wav2vec2.encoder.layers.38.attention.v_proj.weight": "model-00002-of-00003.safetensors",
536
+ "wav2vec2.encoder.layers.38.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
537
+ "wav2vec2.encoder.layers.38.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
538
+ "wav2vec2.encoder.layers.38.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
539
+ "wav2vec2.encoder.layers.38.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
540
+ "wav2vec2.encoder.layers.38.final_layer_norm.bias": "model-00002-of-00003.safetensors",
541
+ "wav2vec2.encoder.layers.38.final_layer_norm.weight": "model-00002-of-00003.safetensors",
542
+ "wav2vec2.encoder.layers.38.layer_norm.bias": "model-00002-of-00003.safetensors",
543
+ "wav2vec2.encoder.layers.38.layer_norm.weight": "model-00002-of-00003.safetensors",
544
+ "wav2vec2.encoder.layers.39.attention.k_proj.bias": "model-00002-of-00003.safetensors",
545
+ "wav2vec2.encoder.layers.39.attention.k_proj.weight": "model-00002-of-00003.safetensors",
546
+ "wav2vec2.encoder.layers.39.attention.out_proj.bias": "model-00002-of-00003.safetensors",
547
+ "wav2vec2.encoder.layers.39.attention.out_proj.weight": "model-00002-of-00003.safetensors",
548
+ "wav2vec2.encoder.layers.39.attention.q_proj.bias": "model-00002-of-00003.safetensors",
549
+ "wav2vec2.encoder.layers.39.attention.q_proj.weight": "model-00002-of-00003.safetensors",
550
+ "wav2vec2.encoder.layers.39.attention.v_proj.bias": "model-00002-of-00003.safetensors",
551
+ "wav2vec2.encoder.layers.39.attention.v_proj.weight": "model-00002-of-00003.safetensors",
552
+ "wav2vec2.encoder.layers.39.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
553
+ "wav2vec2.encoder.layers.39.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
554
+ "wav2vec2.encoder.layers.39.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
555
+ "wav2vec2.encoder.layers.39.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
556
+ "wav2vec2.encoder.layers.39.final_layer_norm.bias": "model-00002-of-00003.safetensors",
557
+ "wav2vec2.encoder.layers.39.final_layer_norm.weight": "model-00002-of-00003.safetensors",
558
+ "wav2vec2.encoder.layers.39.layer_norm.bias": "model-00002-of-00003.safetensors",
559
+ "wav2vec2.encoder.layers.39.layer_norm.weight": "model-00002-of-00003.safetensors",
560
+ "wav2vec2.encoder.layers.4.attention.k_proj.bias": "model-00001-of-00003.safetensors",
561
+ "wav2vec2.encoder.layers.4.attention.k_proj.weight": "model-00001-of-00003.safetensors",
562
+ "wav2vec2.encoder.layers.4.attention.out_proj.bias": "model-00001-of-00003.safetensors",
563
+ "wav2vec2.encoder.layers.4.attention.out_proj.weight": "model-00001-of-00003.safetensors",
564
+ "wav2vec2.encoder.layers.4.attention.q_proj.bias": "model-00001-of-00003.safetensors",
565
+ "wav2vec2.encoder.layers.4.attention.q_proj.weight": "model-00001-of-00003.safetensors",
566
+ "wav2vec2.encoder.layers.4.attention.v_proj.bias": "model-00001-of-00003.safetensors",
567
+ "wav2vec2.encoder.layers.4.attention.v_proj.weight": "model-00001-of-00003.safetensors",
568
+ "wav2vec2.encoder.layers.4.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
569
+ "wav2vec2.encoder.layers.4.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
570
+ "wav2vec2.encoder.layers.4.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
571
+ "wav2vec2.encoder.layers.4.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
572
+ "wav2vec2.encoder.layers.4.final_layer_norm.bias": "model-00001-of-00003.safetensors",
573
+ "wav2vec2.encoder.layers.4.final_layer_norm.weight": "model-00001-of-00003.safetensors",
574
+ "wav2vec2.encoder.layers.4.layer_norm.bias": "model-00001-of-00003.safetensors",
575
+ "wav2vec2.encoder.layers.4.layer_norm.weight": "model-00001-of-00003.safetensors",
576
+ "wav2vec2.encoder.layers.40.attention.k_proj.bias": "model-00002-of-00003.safetensors",
577
+ "wav2vec2.encoder.layers.40.attention.k_proj.weight": "model-00002-of-00003.safetensors",
578
+ "wav2vec2.encoder.layers.40.attention.out_proj.bias": "model-00002-of-00003.safetensors",
579
+ "wav2vec2.encoder.layers.40.attention.out_proj.weight": "model-00002-of-00003.safetensors",
580
+ "wav2vec2.encoder.layers.40.attention.q_proj.bias": "model-00002-of-00003.safetensors",
581
+ "wav2vec2.encoder.layers.40.attention.q_proj.weight": "model-00002-of-00003.safetensors",
582
+ "wav2vec2.encoder.layers.40.attention.v_proj.bias": "model-00002-of-00003.safetensors",
583
+ "wav2vec2.encoder.layers.40.attention.v_proj.weight": "model-00002-of-00003.safetensors",
584
+ "wav2vec2.encoder.layers.40.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
585
+ "wav2vec2.encoder.layers.40.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
586
+ "wav2vec2.encoder.layers.40.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
587
+ "wav2vec2.encoder.layers.40.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
588
+ "wav2vec2.encoder.layers.40.final_layer_norm.bias": "model-00002-of-00003.safetensors",
589
+ "wav2vec2.encoder.layers.40.final_layer_norm.weight": "model-00002-of-00003.safetensors",
590
+ "wav2vec2.encoder.layers.40.layer_norm.bias": "model-00002-of-00003.safetensors",
591
+ "wav2vec2.encoder.layers.40.layer_norm.weight": "model-00002-of-00003.safetensors",
592
+ "wav2vec2.encoder.layers.41.attention.k_proj.bias": "model-00002-of-00003.safetensors",
593
+ "wav2vec2.encoder.layers.41.attention.k_proj.weight": "model-00002-of-00003.safetensors",
594
+ "wav2vec2.encoder.layers.41.attention.out_proj.bias": "model-00002-of-00003.safetensors",
595
+ "wav2vec2.encoder.layers.41.attention.out_proj.weight": "model-00002-of-00003.safetensors",
596
+ "wav2vec2.encoder.layers.41.attention.q_proj.bias": "model-00002-of-00003.safetensors",
597
+ "wav2vec2.encoder.layers.41.attention.q_proj.weight": "model-00002-of-00003.safetensors",
598
+ "wav2vec2.encoder.layers.41.attention.v_proj.bias": "model-00002-of-00003.safetensors",
599
+ "wav2vec2.encoder.layers.41.attention.v_proj.weight": "model-00002-of-00003.safetensors",
600
+ "wav2vec2.encoder.layers.41.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
601
+ "wav2vec2.encoder.layers.41.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
602
+ "wav2vec2.encoder.layers.41.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
603
+ "wav2vec2.encoder.layers.41.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
604
+ "wav2vec2.encoder.layers.41.final_layer_norm.bias": "model-00002-of-00003.safetensors",
605
+ "wav2vec2.encoder.layers.41.final_layer_norm.weight": "model-00002-of-00003.safetensors",
606
+ "wav2vec2.encoder.layers.41.layer_norm.bias": "model-00002-of-00003.safetensors",
607
+ "wav2vec2.encoder.layers.41.layer_norm.weight": "model-00002-of-00003.safetensors",
608
+ "wav2vec2.encoder.layers.42.attention.k_proj.bias": "model-00002-of-00003.safetensors",
609
+ "wav2vec2.encoder.layers.42.attention.k_proj.weight": "model-00002-of-00003.safetensors",
610
+ "wav2vec2.encoder.layers.42.attention.out_proj.bias": "model-00002-of-00003.safetensors",
611
+ "wav2vec2.encoder.layers.42.attention.out_proj.weight": "model-00002-of-00003.safetensors",
612
+ "wav2vec2.encoder.layers.42.attention.q_proj.bias": "model-00002-of-00003.safetensors",
613
+ "wav2vec2.encoder.layers.42.attention.q_proj.weight": "model-00002-of-00003.safetensors",
614
+ "wav2vec2.encoder.layers.42.attention.v_proj.bias": "model-00002-of-00003.safetensors",
615
+ "wav2vec2.encoder.layers.42.attention.v_proj.weight": "model-00002-of-00003.safetensors",
616
+ "wav2vec2.encoder.layers.42.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
617
+ "wav2vec2.encoder.layers.42.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
618
+ "wav2vec2.encoder.layers.42.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
619
+ "wav2vec2.encoder.layers.42.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
620
+ "wav2vec2.encoder.layers.42.final_layer_norm.bias": "model-00002-of-00003.safetensors",
621
+ "wav2vec2.encoder.layers.42.final_layer_norm.weight": "model-00002-of-00003.safetensors",
622
+ "wav2vec2.encoder.layers.42.layer_norm.bias": "model-00002-of-00003.safetensors",
623
+ "wav2vec2.encoder.layers.42.layer_norm.weight": "model-00002-of-00003.safetensors",
624
+ "wav2vec2.encoder.layers.43.attention.k_proj.bias": "model-00002-of-00003.safetensors",
625
+ "wav2vec2.encoder.layers.43.attention.k_proj.weight": "model-00002-of-00003.safetensors",
626
+ "wav2vec2.encoder.layers.43.attention.out_proj.bias": "model-00002-of-00003.safetensors",
627
+ "wav2vec2.encoder.layers.43.attention.out_proj.weight": "model-00002-of-00003.safetensors",
628
+ "wav2vec2.encoder.layers.43.attention.q_proj.bias": "model-00002-of-00003.safetensors",
629
+ "wav2vec2.encoder.layers.43.attention.q_proj.weight": "model-00002-of-00003.safetensors",
630
+ "wav2vec2.encoder.layers.43.attention.v_proj.bias": "model-00002-of-00003.safetensors",
631
+ "wav2vec2.encoder.layers.43.attention.v_proj.weight": "model-00002-of-00003.safetensors",
632
+ "wav2vec2.encoder.layers.43.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
633
+ "wav2vec2.encoder.layers.43.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
634
+ "wav2vec2.encoder.layers.43.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
635
+ "wav2vec2.encoder.layers.43.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
636
+ "wav2vec2.encoder.layers.43.final_layer_norm.bias": "model-00002-of-00003.safetensors",
637
+ "wav2vec2.encoder.layers.43.final_layer_norm.weight": "model-00002-of-00003.safetensors",
638
+ "wav2vec2.encoder.layers.43.layer_norm.bias": "model-00002-of-00003.safetensors",
639
+ "wav2vec2.encoder.layers.43.layer_norm.weight": "model-00002-of-00003.safetensors",
640
+ "wav2vec2.encoder.layers.44.attention.k_proj.bias": "model-00002-of-00003.safetensors",
641
+ "wav2vec2.encoder.layers.44.attention.k_proj.weight": "model-00002-of-00003.safetensors",
642
+ "wav2vec2.encoder.layers.44.attention.out_proj.bias": "model-00002-of-00003.safetensors",
643
+ "wav2vec2.encoder.layers.44.attention.out_proj.weight": "model-00002-of-00003.safetensors",
644
+ "wav2vec2.encoder.layers.44.attention.q_proj.bias": "model-00002-of-00003.safetensors",
645
+ "wav2vec2.encoder.layers.44.attention.q_proj.weight": "model-00002-of-00003.safetensors",
646
+ "wav2vec2.encoder.layers.44.attention.v_proj.bias": "model-00002-of-00003.safetensors",
647
+ "wav2vec2.encoder.layers.44.attention.v_proj.weight": "model-00002-of-00003.safetensors",
648
+ "wav2vec2.encoder.layers.44.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
649
+ "wav2vec2.encoder.layers.44.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
650
+ "wav2vec2.encoder.layers.44.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
651
+ "wav2vec2.encoder.layers.44.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
652
+ "wav2vec2.encoder.layers.44.final_layer_norm.bias": "model-00002-of-00003.safetensors",
653
+ "wav2vec2.encoder.layers.44.final_layer_norm.weight": "model-00002-of-00003.safetensors",
654
+ "wav2vec2.encoder.layers.44.layer_norm.bias": "model-00002-of-00003.safetensors",
655
+ "wav2vec2.encoder.layers.44.layer_norm.weight": "model-00002-of-00003.safetensors",
656
+ "wav2vec2.encoder.layers.45.attention.k_proj.bias": "model-00002-of-00003.safetensors",
657
+ "wav2vec2.encoder.layers.45.attention.k_proj.weight": "model-00002-of-00003.safetensors",
658
+ "wav2vec2.encoder.layers.45.attention.out_proj.bias": "model-00002-of-00003.safetensors",
659
+ "wav2vec2.encoder.layers.45.attention.out_proj.weight": "model-00002-of-00003.safetensors",
660
+ "wav2vec2.encoder.layers.45.attention.q_proj.bias": "model-00002-of-00003.safetensors",
661
+ "wav2vec2.encoder.layers.45.attention.q_proj.weight": "model-00002-of-00003.safetensors",
662
+ "wav2vec2.encoder.layers.45.attention.v_proj.bias": "model-00002-of-00003.safetensors",
663
+ "wav2vec2.encoder.layers.45.attention.v_proj.weight": "model-00002-of-00003.safetensors",
664
+ "wav2vec2.encoder.layers.45.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
665
+ "wav2vec2.encoder.layers.45.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
666
+ "wav2vec2.encoder.layers.45.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
667
+ "wav2vec2.encoder.layers.45.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
668
+ "wav2vec2.encoder.layers.45.final_layer_norm.bias": "model-00002-of-00003.safetensors",
669
+ "wav2vec2.encoder.layers.45.final_layer_norm.weight": "model-00002-of-00003.safetensors",
670
+ "wav2vec2.encoder.layers.45.layer_norm.bias": "model-00002-of-00003.safetensors",
671
+ "wav2vec2.encoder.layers.45.layer_norm.weight": "model-00002-of-00003.safetensors",
672
+ "wav2vec2.encoder.layers.46.attention.k_proj.bias": "model-00002-of-00003.safetensors",
673
+ "wav2vec2.encoder.layers.46.attention.k_proj.weight": "model-00002-of-00003.safetensors",
674
+ "wav2vec2.encoder.layers.46.attention.out_proj.bias": "model-00002-of-00003.safetensors",
675
+ "wav2vec2.encoder.layers.46.attention.out_proj.weight": "model-00002-of-00003.safetensors",
676
+ "wav2vec2.encoder.layers.46.attention.q_proj.bias": "model-00002-of-00003.safetensors",
677
+ "wav2vec2.encoder.layers.46.attention.q_proj.weight": "model-00002-of-00003.safetensors",
678
+ "wav2vec2.encoder.layers.46.attention.v_proj.bias": "model-00002-of-00003.safetensors",
679
+ "wav2vec2.encoder.layers.46.attention.v_proj.weight": "model-00002-of-00003.safetensors",
680
+ "wav2vec2.encoder.layers.46.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
681
+ "wav2vec2.encoder.layers.46.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
682
+ "wav2vec2.encoder.layers.46.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
683
+ "wav2vec2.encoder.layers.46.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
684
+ "wav2vec2.encoder.layers.46.final_layer_norm.bias": "model-00002-of-00003.safetensors",
685
+ "wav2vec2.encoder.layers.46.final_layer_norm.weight": "model-00002-of-00003.safetensors",
686
+ "wav2vec2.encoder.layers.46.layer_norm.bias": "model-00002-of-00003.safetensors",
687
+ "wav2vec2.encoder.layers.46.layer_norm.weight": "model-00002-of-00003.safetensors",
688
+ "wav2vec2.encoder.layers.47.attention.k_proj.bias": "model-00002-of-00003.safetensors",
689
+ "wav2vec2.encoder.layers.47.attention.k_proj.weight": "model-00002-of-00003.safetensors",
690
+ "wav2vec2.encoder.layers.47.attention.out_proj.bias": "model-00002-of-00003.safetensors",
691
+ "wav2vec2.encoder.layers.47.attention.out_proj.weight": "model-00002-of-00003.safetensors",
692
+ "wav2vec2.encoder.layers.47.attention.q_proj.bias": "model-00002-of-00003.safetensors",
693
+ "wav2vec2.encoder.layers.47.attention.q_proj.weight": "model-00002-of-00003.safetensors",
694
+ "wav2vec2.encoder.layers.47.attention.v_proj.bias": "model-00002-of-00003.safetensors",
695
+ "wav2vec2.encoder.layers.47.attention.v_proj.weight": "model-00002-of-00003.safetensors",
696
+ "wav2vec2.encoder.layers.47.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
697
+ "wav2vec2.encoder.layers.47.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
698
+ "wav2vec2.encoder.layers.47.feed_forward.output_dense.bias": "model-00002-of-00003.safetensors",
699
+ "wav2vec2.encoder.layers.47.feed_forward.output_dense.weight": "model-00002-of-00003.safetensors",
700
+ "wav2vec2.encoder.layers.47.final_layer_norm.bias": "model-00002-of-00003.safetensors",
701
+ "wav2vec2.encoder.layers.47.final_layer_norm.weight": "model-00002-of-00003.safetensors",
702
+ "wav2vec2.encoder.layers.47.layer_norm.bias": "model-00002-of-00003.safetensors",
703
+ "wav2vec2.encoder.layers.47.layer_norm.weight": "model-00002-of-00003.safetensors",
704
+ "wav2vec2.encoder.layers.48.attention.k_proj.bias": "model-00002-of-00003.safetensors",
705
+ "wav2vec2.encoder.layers.48.attention.k_proj.weight": "model-00002-of-00003.safetensors",
706
+ "wav2vec2.encoder.layers.48.attention.out_proj.bias": "model-00002-of-00003.safetensors",
707
+ "wav2vec2.encoder.layers.48.attention.out_proj.weight": "model-00002-of-00003.safetensors",
708
+ "wav2vec2.encoder.layers.48.attention.q_proj.bias": "model-00002-of-00003.safetensors",
709
+ "wav2vec2.encoder.layers.48.attention.q_proj.weight": "model-00002-of-00003.safetensors",
710
+ "wav2vec2.encoder.layers.48.attention.v_proj.bias": "model-00002-of-00003.safetensors",
711
+ "wav2vec2.encoder.layers.48.attention.v_proj.weight": "model-00002-of-00003.safetensors",
712
+ "wav2vec2.encoder.layers.48.feed_forward.intermediate_dense.bias": "model-00002-of-00003.safetensors",
713
+ "wav2vec2.encoder.layers.48.feed_forward.intermediate_dense.weight": "model-00002-of-00003.safetensors",
714
+ "wav2vec2.encoder.layers.48.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
715
+ "wav2vec2.encoder.layers.48.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
716
+ "wav2vec2.encoder.layers.48.final_layer_norm.bias": "model-00003-of-00003.safetensors",
717
+ "wav2vec2.encoder.layers.48.final_layer_norm.weight": "model-00003-of-00003.safetensors",
718
+ "wav2vec2.encoder.layers.48.layer_norm.bias": "model-00002-of-00003.safetensors",
719
+ "wav2vec2.encoder.layers.48.layer_norm.weight": "model-00002-of-00003.safetensors",
720
+ "wav2vec2.encoder.layers.49.attention.k_proj.bias": "model-00003-of-00003.safetensors",
721
+ "wav2vec2.encoder.layers.49.attention.k_proj.weight": "model-00003-of-00003.safetensors",
722
+ "wav2vec2.encoder.layers.49.attention.out_proj.bias": "model-00003-of-00003.safetensors",
723
+ "wav2vec2.encoder.layers.49.attention.out_proj.weight": "model-00003-of-00003.safetensors",
724
+ "wav2vec2.encoder.layers.49.attention.q_proj.bias": "model-00003-of-00003.safetensors",
725
+ "wav2vec2.encoder.layers.49.attention.q_proj.weight": "model-00003-of-00003.safetensors",
726
+ "wav2vec2.encoder.layers.49.attention.v_proj.bias": "model-00003-of-00003.safetensors",
727
+ "wav2vec2.encoder.layers.49.attention.v_proj.weight": "model-00003-of-00003.safetensors",
728
+ "wav2vec2.encoder.layers.49.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
729
+ "wav2vec2.encoder.layers.49.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
730
+ "wav2vec2.encoder.layers.49.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
731
+ "wav2vec2.encoder.layers.49.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
732
+ "wav2vec2.encoder.layers.49.final_layer_norm.bias": "model-00003-of-00003.safetensors",
733
+ "wav2vec2.encoder.layers.49.final_layer_norm.weight": "model-00003-of-00003.safetensors",
734
+ "wav2vec2.encoder.layers.49.layer_norm.bias": "model-00003-of-00003.safetensors",
735
+ "wav2vec2.encoder.layers.49.layer_norm.weight": "model-00003-of-00003.safetensors",
736
+ "wav2vec2.encoder.layers.5.attention.k_proj.bias": "model-00001-of-00003.safetensors",
737
+ "wav2vec2.encoder.layers.5.attention.k_proj.weight": "model-00001-of-00003.safetensors",
738
+ "wav2vec2.encoder.layers.5.attention.out_proj.bias": "model-00001-of-00003.safetensors",
739
+ "wav2vec2.encoder.layers.5.attention.out_proj.weight": "model-00001-of-00003.safetensors",
740
+ "wav2vec2.encoder.layers.5.attention.q_proj.bias": "model-00001-of-00003.safetensors",
741
+ "wav2vec2.encoder.layers.5.attention.q_proj.weight": "model-00001-of-00003.safetensors",
742
+ "wav2vec2.encoder.layers.5.attention.v_proj.bias": "model-00001-of-00003.safetensors",
743
+ "wav2vec2.encoder.layers.5.attention.v_proj.weight": "model-00001-of-00003.safetensors",
744
+ "wav2vec2.encoder.layers.5.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
745
+ "wav2vec2.encoder.layers.5.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
746
+ "wav2vec2.encoder.layers.5.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
747
+ "wav2vec2.encoder.layers.5.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
748
+ "wav2vec2.encoder.layers.5.final_layer_norm.bias": "model-00001-of-00003.safetensors",
749
+ "wav2vec2.encoder.layers.5.final_layer_norm.weight": "model-00001-of-00003.safetensors",
750
+ "wav2vec2.encoder.layers.5.layer_norm.bias": "model-00001-of-00003.safetensors",
751
+ "wav2vec2.encoder.layers.5.layer_norm.weight": "model-00001-of-00003.safetensors",
752
+ "wav2vec2.encoder.layers.50.attention.k_proj.bias": "model-00003-of-00003.safetensors",
753
+ "wav2vec2.encoder.layers.50.attention.k_proj.weight": "model-00003-of-00003.safetensors",
754
+ "wav2vec2.encoder.layers.50.attention.out_proj.bias": "model-00003-of-00003.safetensors",
755
+ "wav2vec2.encoder.layers.50.attention.out_proj.weight": "model-00003-of-00003.safetensors",
756
+ "wav2vec2.encoder.layers.50.attention.q_proj.bias": "model-00003-of-00003.safetensors",
757
+ "wav2vec2.encoder.layers.50.attention.q_proj.weight": "model-00003-of-00003.safetensors",
758
+ "wav2vec2.encoder.layers.50.attention.v_proj.bias": "model-00003-of-00003.safetensors",
759
+ "wav2vec2.encoder.layers.50.attention.v_proj.weight": "model-00003-of-00003.safetensors",
760
+ "wav2vec2.encoder.layers.50.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
761
+ "wav2vec2.encoder.layers.50.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
762
+ "wav2vec2.encoder.layers.50.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
763
+ "wav2vec2.encoder.layers.50.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
764
+ "wav2vec2.encoder.layers.50.final_layer_norm.bias": "model-00003-of-00003.safetensors",
765
+ "wav2vec2.encoder.layers.50.final_layer_norm.weight": "model-00003-of-00003.safetensors",
766
+ "wav2vec2.encoder.layers.50.layer_norm.bias": "model-00003-of-00003.safetensors",
767
+ "wav2vec2.encoder.layers.50.layer_norm.weight": "model-00003-of-00003.safetensors",
768
+ "wav2vec2.encoder.layers.51.attention.k_proj.bias": "model-00003-of-00003.safetensors",
769
+ "wav2vec2.encoder.layers.51.attention.k_proj.weight": "model-00003-of-00003.safetensors",
770
+ "wav2vec2.encoder.layers.51.attention.out_proj.bias": "model-00003-of-00003.safetensors",
771
+ "wav2vec2.encoder.layers.51.attention.out_proj.weight": "model-00003-of-00003.safetensors",
772
+ "wav2vec2.encoder.layers.51.attention.q_proj.bias": "model-00003-of-00003.safetensors",
773
+ "wav2vec2.encoder.layers.51.attention.q_proj.weight": "model-00003-of-00003.safetensors",
774
+ "wav2vec2.encoder.layers.51.attention.v_proj.bias": "model-00003-of-00003.safetensors",
775
+ "wav2vec2.encoder.layers.51.attention.v_proj.weight": "model-00003-of-00003.safetensors",
776
+ "wav2vec2.encoder.layers.51.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
777
+ "wav2vec2.encoder.layers.51.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
778
+ "wav2vec2.encoder.layers.51.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
779
+ "wav2vec2.encoder.layers.51.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
780
+ "wav2vec2.encoder.layers.51.final_layer_norm.bias": "model-00003-of-00003.safetensors",
781
+ "wav2vec2.encoder.layers.51.final_layer_norm.weight": "model-00003-of-00003.safetensors",
782
+ "wav2vec2.encoder.layers.51.layer_norm.bias": "model-00003-of-00003.safetensors",
783
+ "wav2vec2.encoder.layers.51.layer_norm.weight": "model-00003-of-00003.safetensors",
784
+ "wav2vec2.encoder.layers.52.attention.k_proj.bias": "model-00003-of-00003.safetensors",
785
+ "wav2vec2.encoder.layers.52.attention.k_proj.weight": "model-00003-of-00003.safetensors",
786
+ "wav2vec2.encoder.layers.52.attention.out_proj.bias": "model-00003-of-00003.safetensors",
787
+ "wav2vec2.encoder.layers.52.attention.out_proj.weight": "model-00003-of-00003.safetensors",
788
+ "wav2vec2.encoder.layers.52.attention.q_proj.bias": "model-00003-of-00003.safetensors",
789
+ "wav2vec2.encoder.layers.52.attention.q_proj.weight": "model-00003-of-00003.safetensors",
790
+ "wav2vec2.encoder.layers.52.attention.v_proj.bias": "model-00003-of-00003.safetensors",
791
+ "wav2vec2.encoder.layers.52.attention.v_proj.weight": "model-00003-of-00003.safetensors",
792
+ "wav2vec2.encoder.layers.52.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
793
+ "wav2vec2.encoder.layers.52.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
794
+ "wav2vec2.encoder.layers.52.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
795
+ "wav2vec2.encoder.layers.52.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
796
+ "wav2vec2.encoder.layers.52.final_layer_norm.bias": "model-00003-of-00003.safetensors",
797
+ "wav2vec2.encoder.layers.52.final_layer_norm.weight": "model-00003-of-00003.safetensors",
798
+ "wav2vec2.encoder.layers.52.layer_norm.bias": "model-00003-of-00003.safetensors",
799
+ "wav2vec2.encoder.layers.52.layer_norm.weight": "model-00003-of-00003.safetensors",
800
+ "wav2vec2.encoder.layers.53.attention.k_proj.bias": "model-00003-of-00003.safetensors",
801
+ "wav2vec2.encoder.layers.53.attention.k_proj.weight": "model-00003-of-00003.safetensors",
802
+ "wav2vec2.encoder.layers.53.attention.out_proj.bias": "model-00003-of-00003.safetensors",
803
+ "wav2vec2.encoder.layers.53.attention.out_proj.weight": "model-00003-of-00003.safetensors",
804
+ "wav2vec2.encoder.layers.53.attention.q_proj.bias": "model-00003-of-00003.safetensors",
805
+ "wav2vec2.encoder.layers.53.attention.q_proj.weight": "model-00003-of-00003.safetensors",
806
+ "wav2vec2.encoder.layers.53.attention.v_proj.bias": "model-00003-of-00003.safetensors",
807
+ "wav2vec2.encoder.layers.53.attention.v_proj.weight": "model-00003-of-00003.safetensors",
808
+ "wav2vec2.encoder.layers.53.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
809
+ "wav2vec2.encoder.layers.53.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
810
+ "wav2vec2.encoder.layers.53.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
811
+ "wav2vec2.encoder.layers.53.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
812
+ "wav2vec2.encoder.layers.53.final_layer_norm.bias": "model-00003-of-00003.safetensors",
813
+ "wav2vec2.encoder.layers.53.final_layer_norm.weight": "model-00003-of-00003.safetensors",
814
+ "wav2vec2.encoder.layers.53.layer_norm.bias": "model-00003-of-00003.safetensors",
815
+ "wav2vec2.encoder.layers.53.layer_norm.weight": "model-00003-of-00003.safetensors",
816
+ "wav2vec2.encoder.layers.54.attention.k_proj.bias": "model-00003-of-00003.safetensors",
817
+ "wav2vec2.encoder.layers.54.attention.k_proj.weight": "model-00003-of-00003.safetensors",
818
+ "wav2vec2.encoder.layers.54.attention.out_proj.bias": "model-00003-of-00003.safetensors",
819
+ "wav2vec2.encoder.layers.54.attention.out_proj.weight": "model-00003-of-00003.safetensors",
820
+ "wav2vec2.encoder.layers.54.attention.q_proj.bias": "model-00003-of-00003.safetensors",
821
+ "wav2vec2.encoder.layers.54.attention.q_proj.weight": "model-00003-of-00003.safetensors",
822
+ "wav2vec2.encoder.layers.54.attention.v_proj.bias": "model-00003-of-00003.safetensors",
823
+ "wav2vec2.encoder.layers.54.attention.v_proj.weight": "model-00003-of-00003.safetensors",
824
+ "wav2vec2.encoder.layers.54.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
825
+ "wav2vec2.encoder.layers.54.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
826
+ "wav2vec2.encoder.layers.54.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
827
+ "wav2vec2.encoder.layers.54.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
828
+ "wav2vec2.encoder.layers.54.final_layer_norm.bias": "model-00003-of-00003.safetensors",
829
+ "wav2vec2.encoder.layers.54.final_layer_norm.weight": "model-00003-of-00003.safetensors",
830
+ "wav2vec2.encoder.layers.54.layer_norm.bias": "model-00003-of-00003.safetensors",
831
+ "wav2vec2.encoder.layers.54.layer_norm.weight": "model-00003-of-00003.safetensors",
832
+ "wav2vec2.encoder.layers.55.attention.k_proj.bias": "model-00003-of-00003.safetensors",
833
+ "wav2vec2.encoder.layers.55.attention.k_proj.weight": "model-00003-of-00003.safetensors",
834
+ "wav2vec2.encoder.layers.55.attention.out_proj.bias": "model-00003-of-00003.safetensors",
835
+ "wav2vec2.encoder.layers.55.attention.out_proj.weight": "model-00003-of-00003.safetensors",
836
+ "wav2vec2.encoder.layers.55.attention.q_proj.bias": "model-00003-of-00003.safetensors",
837
+ "wav2vec2.encoder.layers.55.attention.q_proj.weight": "model-00003-of-00003.safetensors",
838
+ "wav2vec2.encoder.layers.55.attention.v_proj.bias": "model-00003-of-00003.safetensors",
839
+ "wav2vec2.encoder.layers.55.attention.v_proj.weight": "model-00003-of-00003.safetensors",
840
+ "wav2vec2.encoder.layers.55.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
841
+ "wav2vec2.encoder.layers.55.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
842
+ "wav2vec2.encoder.layers.55.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
843
+ "wav2vec2.encoder.layers.55.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
844
+ "wav2vec2.encoder.layers.55.final_layer_norm.bias": "model-00003-of-00003.safetensors",
845
+ "wav2vec2.encoder.layers.55.final_layer_norm.weight": "model-00003-of-00003.safetensors",
846
+ "wav2vec2.encoder.layers.55.layer_norm.bias": "model-00003-of-00003.safetensors",
847
+ "wav2vec2.encoder.layers.55.layer_norm.weight": "model-00003-of-00003.safetensors",
848
+ "wav2vec2.encoder.layers.56.attention.k_proj.bias": "model-00003-of-00003.safetensors",
849
+ "wav2vec2.encoder.layers.56.attention.k_proj.weight": "model-00003-of-00003.safetensors",
850
+ "wav2vec2.encoder.layers.56.attention.out_proj.bias": "model-00003-of-00003.safetensors",
851
+ "wav2vec2.encoder.layers.56.attention.out_proj.weight": "model-00003-of-00003.safetensors",
852
+ "wav2vec2.encoder.layers.56.attention.q_proj.bias": "model-00003-of-00003.safetensors",
853
+ "wav2vec2.encoder.layers.56.attention.q_proj.weight": "model-00003-of-00003.safetensors",
854
+ "wav2vec2.encoder.layers.56.attention.v_proj.bias": "model-00003-of-00003.safetensors",
855
+ "wav2vec2.encoder.layers.56.attention.v_proj.weight": "model-00003-of-00003.safetensors",
856
+ "wav2vec2.encoder.layers.56.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
857
+ "wav2vec2.encoder.layers.56.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
858
+ "wav2vec2.encoder.layers.56.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
859
+ "wav2vec2.encoder.layers.56.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
860
+ "wav2vec2.encoder.layers.56.final_layer_norm.bias": "model-00003-of-00003.safetensors",
861
+ "wav2vec2.encoder.layers.56.final_layer_norm.weight": "model-00003-of-00003.safetensors",
862
+ "wav2vec2.encoder.layers.56.layer_norm.bias": "model-00003-of-00003.safetensors",
863
+ "wav2vec2.encoder.layers.56.layer_norm.weight": "model-00003-of-00003.safetensors",
864
+ "wav2vec2.encoder.layers.57.attention.k_proj.bias": "model-00003-of-00003.safetensors",
865
+ "wav2vec2.encoder.layers.57.attention.k_proj.weight": "model-00003-of-00003.safetensors",
866
+ "wav2vec2.encoder.layers.57.attention.out_proj.bias": "model-00003-of-00003.safetensors",
867
+ "wav2vec2.encoder.layers.57.attention.out_proj.weight": "model-00003-of-00003.safetensors",
868
+ "wav2vec2.encoder.layers.57.attention.q_proj.bias": "model-00003-of-00003.safetensors",
869
+ "wav2vec2.encoder.layers.57.attention.q_proj.weight": "model-00003-of-00003.safetensors",
870
+ "wav2vec2.encoder.layers.57.attention.v_proj.bias": "model-00003-of-00003.safetensors",
871
+ "wav2vec2.encoder.layers.57.attention.v_proj.weight": "model-00003-of-00003.safetensors",
872
+ "wav2vec2.encoder.layers.57.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
873
+ "wav2vec2.encoder.layers.57.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
874
+ "wav2vec2.encoder.layers.57.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
875
+ "wav2vec2.encoder.layers.57.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
876
+ "wav2vec2.encoder.layers.57.final_layer_norm.bias": "model-00003-of-00003.safetensors",
877
+ "wav2vec2.encoder.layers.57.final_layer_norm.weight": "model-00003-of-00003.safetensors",
878
+ "wav2vec2.encoder.layers.57.layer_norm.bias": "model-00003-of-00003.safetensors",
879
+ "wav2vec2.encoder.layers.57.layer_norm.weight": "model-00003-of-00003.safetensors",
880
+ "wav2vec2.encoder.layers.58.attention.k_proj.bias": "model-00003-of-00003.safetensors",
881
+ "wav2vec2.encoder.layers.58.attention.k_proj.weight": "model-00003-of-00003.safetensors",
882
+ "wav2vec2.encoder.layers.58.attention.out_proj.bias": "model-00003-of-00003.safetensors",
883
+ "wav2vec2.encoder.layers.58.attention.out_proj.weight": "model-00003-of-00003.safetensors",
884
+ "wav2vec2.encoder.layers.58.attention.q_proj.bias": "model-00003-of-00003.safetensors",
885
+ "wav2vec2.encoder.layers.58.attention.q_proj.weight": "model-00003-of-00003.safetensors",
886
+ "wav2vec2.encoder.layers.58.attention.v_proj.bias": "model-00003-of-00003.safetensors",
887
+ "wav2vec2.encoder.layers.58.attention.v_proj.weight": "model-00003-of-00003.safetensors",
888
+ "wav2vec2.encoder.layers.58.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
889
+ "wav2vec2.encoder.layers.58.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
890
+ "wav2vec2.encoder.layers.58.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
891
+ "wav2vec2.encoder.layers.58.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
892
+ "wav2vec2.encoder.layers.58.final_layer_norm.bias": "model-00003-of-00003.safetensors",
893
+ "wav2vec2.encoder.layers.58.final_layer_norm.weight": "model-00003-of-00003.safetensors",
894
+ "wav2vec2.encoder.layers.58.layer_norm.bias": "model-00003-of-00003.safetensors",
895
+ "wav2vec2.encoder.layers.58.layer_norm.weight": "model-00003-of-00003.safetensors",
896
+ "wav2vec2.encoder.layers.59.attention.k_proj.bias": "model-00003-of-00003.safetensors",
897
+ "wav2vec2.encoder.layers.59.attention.k_proj.weight": "model-00003-of-00003.safetensors",
898
+ "wav2vec2.encoder.layers.59.attention.out_proj.bias": "model-00003-of-00003.safetensors",
899
+ "wav2vec2.encoder.layers.59.attention.out_proj.weight": "model-00003-of-00003.safetensors",
900
+ "wav2vec2.encoder.layers.59.attention.q_proj.bias": "model-00003-of-00003.safetensors",
901
+ "wav2vec2.encoder.layers.59.attention.q_proj.weight": "model-00003-of-00003.safetensors",
902
+ "wav2vec2.encoder.layers.59.attention.v_proj.bias": "model-00003-of-00003.safetensors",
903
+ "wav2vec2.encoder.layers.59.attention.v_proj.weight": "model-00003-of-00003.safetensors",
904
+ "wav2vec2.encoder.layers.59.feed_forward.intermediate_dense.bias": "model-00003-of-00003.safetensors",
905
+ "wav2vec2.encoder.layers.59.feed_forward.intermediate_dense.weight": "model-00003-of-00003.safetensors",
906
+ "wav2vec2.encoder.layers.59.feed_forward.output_dense.bias": "model-00003-of-00003.safetensors",
907
+ "wav2vec2.encoder.layers.59.feed_forward.output_dense.weight": "model-00003-of-00003.safetensors",
908
+ "wav2vec2.encoder.layers.59.final_layer_norm.bias": "model-00003-of-00003.safetensors",
909
+ "wav2vec2.encoder.layers.59.final_layer_norm.weight": "model-00003-of-00003.safetensors",
910
+ "wav2vec2.encoder.layers.59.layer_norm.bias": "model-00003-of-00003.safetensors",
911
+ "wav2vec2.encoder.layers.59.layer_norm.weight": "model-00003-of-00003.safetensors",
912
+ "wav2vec2.encoder.layers.6.attention.k_proj.bias": "model-00001-of-00003.safetensors",
913
+ "wav2vec2.encoder.layers.6.attention.k_proj.weight": "model-00001-of-00003.safetensors",
914
+ "wav2vec2.encoder.layers.6.attention.out_proj.bias": "model-00001-of-00003.safetensors",
915
+ "wav2vec2.encoder.layers.6.attention.out_proj.weight": "model-00001-of-00003.safetensors",
916
+ "wav2vec2.encoder.layers.6.attention.q_proj.bias": "model-00001-of-00003.safetensors",
917
+ "wav2vec2.encoder.layers.6.attention.q_proj.weight": "model-00001-of-00003.safetensors",
918
+ "wav2vec2.encoder.layers.6.attention.v_proj.bias": "model-00001-of-00003.safetensors",
919
+ "wav2vec2.encoder.layers.6.attention.v_proj.weight": "model-00001-of-00003.safetensors",
920
+ "wav2vec2.encoder.layers.6.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
921
+ "wav2vec2.encoder.layers.6.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
922
+ "wav2vec2.encoder.layers.6.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
923
+ "wav2vec2.encoder.layers.6.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
924
+ "wav2vec2.encoder.layers.6.final_layer_norm.bias": "model-00001-of-00003.safetensors",
925
+ "wav2vec2.encoder.layers.6.final_layer_norm.weight": "model-00001-of-00003.safetensors",
926
+ "wav2vec2.encoder.layers.6.layer_norm.bias": "model-00001-of-00003.safetensors",
927
+ "wav2vec2.encoder.layers.6.layer_norm.weight": "model-00001-of-00003.safetensors",
928
+ "wav2vec2.encoder.layers.7.attention.k_proj.bias": "model-00001-of-00003.safetensors",
929
+ "wav2vec2.encoder.layers.7.attention.k_proj.weight": "model-00001-of-00003.safetensors",
930
+ "wav2vec2.encoder.layers.7.attention.out_proj.bias": "model-00001-of-00003.safetensors",
931
+ "wav2vec2.encoder.layers.7.attention.out_proj.weight": "model-00001-of-00003.safetensors",
932
+ "wav2vec2.encoder.layers.7.attention.q_proj.bias": "model-00001-of-00003.safetensors",
933
+ "wav2vec2.encoder.layers.7.attention.q_proj.weight": "model-00001-of-00003.safetensors",
934
+ "wav2vec2.encoder.layers.7.attention.v_proj.bias": "model-00001-of-00003.safetensors",
935
+ "wav2vec2.encoder.layers.7.attention.v_proj.weight": "model-00001-of-00003.safetensors",
936
+ "wav2vec2.encoder.layers.7.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
937
+ "wav2vec2.encoder.layers.7.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
938
+ "wav2vec2.encoder.layers.7.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
939
+ "wav2vec2.encoder.layers.7.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
940
+ "wav2vec2.encoder.layers.7.final_layer_norm.bias": "model-00001-of-00003.safetensors",
941
+ "wav2vec2.encoder.layers.7.final_layer_norm.weight": "model-00001-of-00003.safetensors",
942
+ "wav2vec2.encoder.layers.7.layer_norm.bias": "model-00001-of-00003.safetensors",
943
+ "wav2vec2.encoder.layers.7.layer_norm.weight": "model-00001-of-00003.safetensors",
944
+ "wav2vec2.encoder.layers.8.attention.k_proj.bias": "model-00001-of-00003.safetensors",
945
+ "wav2vec2.encoder.layers.8.attention.k_proj.weight": "model-00001-of-00003.safetensors",
946
+ "wav2vec2.encoder.layers.8.attention.out_proj.bias": "model-00001-of-00003.safetensors",
947
+ "wav2vec2.encoder.layers.8.attention.out_proj.weight": "model-00001-of-00003.safetensors",
948
+ "wav2vec2.encoder.layers.8.attention.q_proj.bias": "model-00001-of-00003.safetensors",
949
+ "wav2vec2.encoder.layers.8.attention.q_proj.weight": "model-00001-of-00003.safetensors",
950
+ "wav2vec2.encoder.layers.8.attention.v_proj.bias": "model-00001-of-00003.safetensors",
951
+ "wav2vec2.encoder.layers.8.attention.v_proj.weight": "model-00001-of-00003.safetensors",
952
+ "wav2vec2.encoder.layers.8.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
953
+ "wav2vec2.encoder.layers.8.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
954
+ "wav2vec2.encoder.layers.8.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
955
+ "wav2vec2.encoder.layers.8.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
956
+ "wav2vec2.encoder.layers.8.final_layer_norm.bias": "model-00001-of-00003.safetensors",
957
+ "wav2vec2.encoder.layers.8.final_layer_norm.weight": "model-00001-of-00003.safetensors",
958
+ "wav2vec2.encoder.layers.8.layer_norm.bias": "model-00001-of-00003.safetensors",
959
+ "wav2vec2.encoder.layers.8.layer_norm.weight": "model-00001-of-00003.safetensors",
960
+ "wav2vec2.encoder.layers.9.attention.k_proj.bias": "model-00001-of-00003.safetensors",
961
+ "wav2vec2.encoder.layers.9.attention.k_proj.weight": "model-00001-of-00003.safetensors",
962
+ "wav2vec2.encoder.layers.9.attention.out_proj.bias": "model-00001-of-00003.safetensors",
963
+ "wav2vec2.encoder.layers.9.attention.out_proj.weight": "model-00001-of-00003.safetensors",
964
+ "wav2vec2.encoder.layers.9.attention.q_proj.bias": "model-00001-of-00003.safetensors",
965
+ "wav2vec2.encoder.layers.9.attention.q_proj.weight": "model-00001-of-00003.safetensors",
966
+ "wav2vec2.encoder.layers.9.attention.v_proj.bias": "model-00001-of-00003.safetensors",
967
+ "wav2vec2.encoder.layers.9.attention.v_proj.weight": "model-00001-of-00003.safetensors",
968
+ "wav2vec2.encoder.layers.9.feed_forward.intermediate_dense.bias": "model-00001-of-00003.safetensors",
969
+ "wav2vec2.encoder.layers.9.feed_forward.intermediate_dense.weight": "model-00001-of-00003.safetensors",
970
+ "wav2vec2.encoder.layers.9.feed_forward.output_dense.bias": "model-00001-of-00003.safetensors",
971
+ "wav2vec2.encoder.layers.9.feed_forward.output_dense.weight": "model-00001-of-00003.safetensors",
972
+ "wav2vec2.encoder.layers.9.final_layer_norm.bias": "model-00001-of-00003.safetensors",
973
+ "wav2vec2.encoder.layers.9.final_layer_norm.weight": "model-00001-of-00003.safetensors",
974
+ "wav2vec2.encoder.layers.9.layer_norm.bias": "model-00001-of-00003.safetensors",
975
+ "wav2vec2.encoder.layers.9.layer_norm.weight": "model-00001-of-00003.safetensors",
976
+ "wav2vec2.encoder.pos_conv_embed.conv.bias": "model-00001-of-00003.safetensors",
977
+ "wav2vec2.encoder.pos_conv_embed.conv.parametrizations.weight.original0": "model-00001-of-00003.safetensors",
978
+ "wav2vec2.encoder.pos_conv_embed.conv.parametrizations.weight.original1": "model-00001-of-00003.safetensors",
979
+ "wav2vec2.feature_extractor.conv_layers.0.conv.bias": "model-00001-of-00003.safetensors",
980
+ "wav2vec2.feature_extractor.conv_layers.0.conv.weight": "model-00001-of-00003.safetensors",
981
+ "wav2vec2.feature_extractor.conv_layers.0.layer_norm.bias": "model-00001-of-00003.safetensors",
982
+ "wav2vec2.feature_extractor.conv_layers.0.layer_norm.weight": "model-00001-of-00003.safetensors",
983
+ "wav2vec2.feature_extractor.conv_layers.1.conv.bias": "model-00001-of-00003.safetensors",
984
+ "wav2vec2.feature_extractor.conv_layers.1.conv.weight": "model-00001-of-00003.safetensors",
985
+ "wav2vec2.feature_extractor.conv_layers.1.layer_norm.bias": "model-00001-of-00003.safetensors",
986
+ "wav2vec2.feature_extractor.conv_layers.1.layer_norm.weight": "model-00001-of-00003.safetensors",
987
+ "wav2vec2.feature_extractor.conv_layers.2.conv.bias": "model-00001-of-00003.safetensors",
988
+ "wav2vec2.feature_extractor.conv_layers.2.conv.weight": "model-00001-of-00003.safetensors",
989
+ "wav2vec2.feature_extractor.conv_layers.2.layer_norm.bias": "model-00001-of-00003.safetensors",
990
+ "wav2vec2.feature_extractor.conv_layers.2.layer_norm.weight": "model-00001-of-00003.safetensors",
991
+ "wav2vec2.feature_extractor.conv_layers.3.conv.bias": "model-00001-of-00003.safetensors",
992
+ "wav2vec2.feature_extractor.conv_layers.3.conv.weight": "model-00001-of-00003.safetensors",
993
+ "wav2vec2.feature_extractor.conv_layers.3.layer_norm.bias": "model-00001-of-00003.safetensors",
994
+ "wav2vec2.feature_extractor.conv_layers.3.layer_norm.weight": "model-00001-of-00003.safetensors",
995
+ "wav2vec2.feature_extractor.conv_layers.4.conv.bias": "model-00001-of-00003.safetensors",
996
+ "wav2vec2.feature_extractor.conv_layers.4.conv.weight": "model-00001-of-00003.safetensors",
997
+ "wav2vec2.feature_extractor.conv_layers.4.layer_norm.bias": "model-00001-of-00003.safetensors",
998
+ "wav2vec2.feature_extractor.conv_layers.4.layer_norm.weight": "model-00001-of-00003.safetensors",
999
+ "wav2vec2.feature_extractor.conv_layers.5.conv.bias": "model-00001-of-00003.safetensors",
1000
+ "wav2vec2.feature_extractor.conv_layers.5.conv.weight": "model-00001-of-00003.safetensors",
1001
+ "wav2vec2.feature_extractor.conv_layers.5.layer_norm.bias": "model-00001-of-00003.safetensors",
1002
+ "wav2vec2.feature_extractor.conv_layers.5.layer_norm.weight": "model-00001-of-00003.safetensors",
1003
+ "wav2vec2.feature_extractor.conv_layers.6.conv.bias": "model-00001-of-00003.safetensors",
1004
+ "wav2vec2.feature_extractor.conv_layers.6.conv.weight": "model-00001-of-00003.safetensors",
1005
+ "wav2vec2.feature_extractor.conv_layers.6.layer_norm.bias": "model-00001-of-00003.safetensors",
1006
+ "wav2vec2.feature_extractor.conv_layers.6.layer_norm.weight": "model-00001-of-00003.safetensors",
1007
+ "wav2vec2.feature_projection.layer_norm.bias": "model-00001-of-00003.safetensors",
1008
+ "wav2vec2.feature_projection.layer_norm.weight": "model-00001-of-00003.safetensors",
1009
+ "wav2vec2.feature_projection.projection.bias": "model-00001-of-00003.safetensors",
1010
+ "wav2vec2.feature_projection.projection.weight": "model-00001-of-00003.safetensors",
1011
+ "wav2vec2.masked_spec_embed": "model-00001-of-00003.safetensors"
1012
+ }
1013
+ }
3b/preprocessor_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_size": 1,
4
+ "padding_side": "right",
5
+ "padding_value": 0.0,
6
+ "return_attention_mask": false,
7
+ "sampling_rate": 16000
8
+ }
7b/config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.1,
3
+ "adapter_attn_dim": null,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForPreTraining"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 768,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": true,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": true,
48
+ "dtype": "float32",
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_dropout": 0.0,
52
+ "feat_extract_norm": "layer",
53
+ "feat_proj_dropout": 0.1,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.1,
56
+ "gradient_checkpointing": false,
57
+ "hidden_act": "gelu",
58
+ "hidden_dropout": 0.1,
59
+ "hidden_dropout_prob": 0.1,
60
+ "hidden_size": 2048,
61
+ "initializer_range": 0.02,
62
+ "intermediate_size": 8192,
63
+ "layer_norm_eps": 1e-05,
64
+ "layerdrop": 0.0,
65
+ "mask_feature_length": 10,
66
+ "mask_feature_min_masks": 0,
67
+ "mask_feature_prob": 0.0,
68
+ "mask_time_length": 10,
69
+ "mask_time_min_masks": 2,
70
+ "mask_time_prob": 0.05,
71
+ "model_type": "wav2vec2",
72
+ "num_adapter_layers": 3,
73
+ "num_attention_heads": 16,
74
+ "num_codevector_groups": 2,
75
+ "num_codevectors_per_group": 320,
76
+ "num_conv_pos_embedding_groups": 16,
77
+ "num_conv_pos_embeddings": 128,
78
+ "num_feat_extract_layers": 7,
79
+ "num_hidden_layers": 128,
80
+ "num_negatives": 100,
81
+ "output_hidden_size": 2048,
82
+ "pad_token_id": 0,
83
+ "proj_codevector_dim": 1024,
84
+ "tdnn_dilation": [
85
+ 1,
86
+ 2,
87
+ 3,
88
+ 1,
89
+ 1
90
+ ],
91
+ "tdnn_dim": [
92
+ 512,
93
+ 512,
94
+ 512,
95
+ 512,
96
+ 1500
97
+ ],
98
+ "tdnn_kernel": [
99
+ 5,
100
+ 3,
101
+ 3,
102
+ 1,
103
+ 1
104
+ ],
105
+ "transformers_version": "4.57.1",
106
+ "use_weighted_layer_sum": false,
107
+ "vocab_size": 32,
108
+ "xvector_output_dim": 512
109
+ }
7b/model-00001-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c83f6a36a469d135ef6f57b1dcc2581edca95a74b1fec8eb1392d2643f96df26
3
+ size 4989745384
7b/model-00002-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d2f71044a1b7d631719d98c9780a650bbf501d322a41b6cac9c42cb90170200
3
+ size 4968743656
7b/model-00003-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1338cf2169760cb4c7c9c741be51775582dd33d0e2e09910ce1f684fe0745466
3
+ size 4968735696
7b/model-00004-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a78c4fa0810702a9449c97c089954101df5dbd1d57fa7a301f43c1b088766b5b
3
+ size 4985504024
7b/model-00005-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:808893966b15e3b7ea1b8301994cc364c348dc427022cb5d1079ddfb4dbfc9ce
3
+ size 4951958368
7b/model-00006-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f6667ccd0fda67e9c68e1e3004abbde7ac258300009f8817a8cfb731a3290ab
3
+ size 1088148928
7b/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
hf_wav2vec2_features.png ADDED

Git LFS Details

  • SHA256: c9f5e7107877fe450bf458044d6bf241e4aba5b72d6dc961c698ebe3edd9a2fa
  • Pointer size: 131 Bytes
  • Size of remote file: 354 kB
inference_hf.py ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from transformers import Wav2Vec2FeatureExtractor, Wav2Vec2Model
2
+ import sys
3
+ import torch
4
+ import torchinfo
5
+ hf_path = sys.argv[1]
6
+ audio_file = sys.argv[2]
7
+ extract = Wav2Vec2FeatureExtractor.from_pretrained(hf_path)
8
+ hf_model = Wav2Vec2Model.from_pretrained(hf_path)
9
+
10
+ torchinfo.summary(hf_model)
11
+
12
+ hf_model.eval()
13
+
14
+ import torchaudio
15
+
16
+ waveform, sample_rate = torchaudio.load(audio_file)
17
+ if waveform.shape[0] > 1:
18
+ waveform = torch.mean(waveform, dim=0, keepdim=True)
19
+
20
+ with torch.no_grad():
21
+ feat = extract(waveform.squeeze().numpy(), sampling_rate=sample_rate, return_tensors="pt", padding=True)
22
+ out = hf_model(feat.input_values)
23
+ last_hidden_states = out.last_hidden_state
24
+ cnn_out = out.extract_features
25
+ print("Last hidden states shape:", last_hidden_states.shape)
26
+ print("CNN features shape:", cnn_out[-1].shape)
27
+
28
+
29
+ #normalize features for visualization
30
+ import numpy as np
31
+ last_hidden_states = (last_hidden_states - last_hidden_states.min()) / (last_hidden_states.max() - last_hidden_states.min())
32
+ cnn_out = [(feat - feat.min()) / (feat.max() - feat.min()) for feat in cnn_out]
33
+
34
+
35
+ #apply log scaling
36
+ last_hidden_states = torch.log1p(last_hidden_states * 100)
37
+ cnn_out = [torch.log1p(feat * 100) for feat in cnn_out]
38
+
39
+ #visualize output feature maps
40
+ import matplotlib.pyplot as plt
41
+ plt.figure(figsize=(12, 6))
42
+ plt.subplot(2, 1, 1)
43
+ plt.title("Last Hidden States")
44
+ plt.imshow(last_hidden_states[0].cpu().numpy().T, aspect='auto', origin='lower')
45
+ plt.colorbar()
46
+ plt.subplot(2, 1, 2)
47
+ plt.title("CNN Features")
48
+ plt.imshow(cnn_out[-1].cpu().numpy().T, aspect='auto', origin='lower')
49
+ plt.colorbar()
50
+ plt.tight_layout()
51
+ plt.savefig("hf_wav2vec2_features.png")
52
+ plt.close()