Hugging Face's logo

EYEDOL
/
ZIN_V2

TensorBoard
Safetensors
whisper
Model card Files
xet
 Metrics Community
EYEDOL commited on
Commit
b39cc18
·
verified ·
1 Parent(s): 94b5a98

Upload folder using huggingface_hub

Browse files
Files changed (20) hide show
  1. checkpoint-1200/config.json +47 -0
  2. checkpoint-1200/generation_config.json +9 -0
  3. checkpoint-1200/model.safetensors +3 -0
  4. checkpoint-1200/optimizer.pt +3 -0
  5. checkpoint-1200/preprocessor_config.json +15 -0
  6. checkpoint-1200/rng_state.pth +3 -0
  7. checkpoint-1200/scaler.pt +3 -0
  8. checkpoint-1200/scheduler.pt +3 -0
  9. checkpoint-1200/trainer_state.json +103 -0
  10. checkpoint-1200/training_args.bin +3 -0
  11. checkpoint-2328/config.json +47 -0
  12. checkpoint-2328/generation_config.json +9 -0
  13. checkpoint-2328/model.safetensors +3 -0
  14. checkpoint-2328/optimizer.pt +3 -0
  15. checkpoint-2328/preprocessor_config.json +15 -0
  16. checkpoint-2328/rng_state.pth +3 -0
  17. checkpoint-2328/scaler.pt +3 -0
  18. checkpoint-2328/scheduler.pt +3 -0
  19. checkpoint-2328/trainer_state.json +156 -0
  20. checkpoint-2328/training_args.bin +3 -0
checkpoint-1200/config.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "gelu",
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": null,
10
+ "bos_token_id": 50257,
11
+ "classifier_proj_size": 256,
12
+ "d_model": 768,
13
+ "decoder_attention_heads": 12,
14
+ "decoder_ffn_dim": 3072,
15
+ "decoder_layerdrop": 0.0,
16
+ "decoder_layers": 12,
17
+ "decoder_start_token_id": 50258,
18
+ "dropout": 0.0,
19
+ "encoder_attention_heads": 12,
20
+ "encoder_ffn_dim": 3072,
21
+ "encoder_layerdrop": 0.0,
22
+ "encoder_layers": 12,
23
+ "eos_token_id": 50257,
24
+ "forced_decoder_ids": null,
25
+ "init_std": 0.02,
26
+ "is_encoder_decoder": true,
27
+ "mask_feature_length": 10,
28
+ "mask_feature_min_masks": 0,
29
+ "mask_feature_prob": 0.0,
30
+ "mask_time_length": 10,
31
+ "mask_time_min_masks": 2,
32
+ "mask_time_prob": 0.05,
33
+ "max_length": null,
34
+ "max_source_positions": 1500,
35
+ "max_target_positions": 448,
36
+ "median_filter_width": 7,
37
+ "model_type": "whisper",
38
+ "num_hidden_layers": 12,
39
+ "num_mel_bins": 80,
40
+ "pad_token_id": 50257,
41
+ "scale_embedding": false,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.52.4",
44
+ "use_cache": false,
45
+ "use_weighted_layer_sum": false,
46
+ "vocab_size": 51865
47
+ }
checkpoint-1200/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 50257,
3
+ "decoder_start_token_id": 50258,
4
+ "eos_token_id": 50257,
5
+ "pad_token_id": 50257,
6
+ "suppress_tokens": [],
7
+ "transformers_version": "4.52.4",
8
+ "use_cache": false
9
+ }
checkpoint-1200/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ba71b9e62b5800b5dafe054243fe673a96f5447603aa2ae7804a33c4c200d4e
3
+ size 966995080
checkpoint-1200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a965500d1c3d667ba0264291449f4cd2d73d84f49b3c61322654a02962425dc
3
+ size 1925064044
checkpoint-1200/preprocessor_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "dither": 0.0,
4
+ "feature_extractor_type": "WhisperFeatureExtractor",
5
+ "feature_size": 80,
6
+ "hop_length": 160,
7
+ "n_fft": 400,
8
+ "n_samples": 480000,
9
+ "nb_max_frames": 3000,
10
+ "padding_side": "right",
11
+ "padding_value": 0.0,
12
+ "processor_class": "WhisperProcessor",
13
+ "return_attention_mask": false,
14
+ "sampling_rate": 16000
15
+ }
checkpoint-1200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15757e15ae2125dc6266877632f0041f17251b3a6536a8bcd8f959aff3c5d07e
3
+ size 14244
checkpoint-1200/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8b42d28fe33743d9cde4b7b0af642a2360adfab9e80319e645acdf2f069825c
3
+ size 988
checkpoint-1200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f050e8bf6b11221ac8dd211761dee040815b2dcc4472d2bc2957c172cc318be
3
+ size 1064
checkpoint-1200/trainer_state.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1200,
3
+ "best_metric": 8.762886597938143,
4
+ "best_model_checkpoint": "./ZINASR/checkpoint-1200",
5
+ "epoch": 4.123711340206185,
6
+ "eval_steps": 400,
7
+ "global_step": 1200,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.6872852233676976,
14
+ "grad_norm": 5.297506809234619,
15
+ "learning_rate": 9.950000000000001e-06,
16
+ "loss": 0.1217,
17
+ "step": 200
18
+ },
19
+ {
20
+ "epoch": 1.3745704467353952,
21
+ "grad_norm": 1.6356655359268188,
22
+ "learning_rate": 9.06484962406015e-06,
23
+ "loss": 0.0826,
24
+ "step": 400
25
+ },
26
+ {
27
+ "epoch": 1.3745704467353952,
28
+ "eval_loss": 0.1176551878452301,
29
+ "eval_runtime": 1451.5411,
30
+ "eval_samples_per_second": 1.602,
31
+ "eval_steps_per_second": 0.2,
32
+ "eval_wer": 8.891752577319588,
33
+ "step": 400
34
+ },
35
+ {
36
+ "epoch": 2.0618556701030926,
37
+ "grad_norm": 0.9965283274650574,
38
+ "learning_rate": 8.134398496240602e-06,
39
+ "loss": 0.0483,
40
+ "step": 600
41
+ },
42
+ {
43
+ "epoch": 2.7491408934707904,
44
+ "grad_norm": 1.6627748012542725,
45
+ "learning_rate": 7.194548872180452e-06,
46
+ "loss": 0.018,
47
+ "step": 800
48
+ },
49
+ {
50
+ "epoch": 2.7491408934707904,
51
+ "eval_loss": 0.121592216193676,
52
+ "eval_runtime": 1448.1115,
53
+ "eval_samples_per_second": 1.606,
54
+ "eval_steps_per_second": 0.201,
55
+ "eval_wer": 9.211531118747613,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 3.436426116838488,
60
+ "grad_norm": 1.0129088163375854,
61
+ "learning_rate": 6.254699248120302e-06,
62
+ "loss": 0.0118,
63
+ "step": 1000
64
+ },
65
+ {
66
+ "epoch": 4.123711340206185,
67
+ "grad_norm": 0.18754775822162628,
68
+ "learning_rate": 5.314849624060151e-06,
69
+ "loss": 0.0067,
70
+ "step": 1200
71
+ },
72
+ {
73
+ "epoch": 4.123711340206185,
74
+ "eval_loss": 0.12416346371173859,
75
+ "eval_runtime": 1449.1081,
76
+ "eval_samples_per_second": 1.604,
77
+ "eval_steps_per_second": 0.201,
78
+ "eval_wer": 8.762886597938143,
79
+ "step": 1200
80
+ }
81
+ ],
82
+ "logging_steps": 200,
83
+ "max_steps": 2328,
84
+ "num_input_tokens_seen": 0,
85
+ "num_train_epochs": 8,
86
+ "save_steps": 1200,
87
+ "stateful_callbacks": {
88
+ "TrainerControl": {
89
+ "args": {
90
+ "should_epoch_stop": false,
91
+ "should_evaluate": false,
92
+ "should_log": false,
93
+ "should_save": true,
94
+ "should_training_stop": false
95
+ },
96
+ "attributes": {}
97
+ }
98
+ },
99
+ "total_flos": 1.106667293147136e+19,
100
+ "train_batch_size": 16,
101
+ "trial_name": null,
102
+ "trial_params": null
103
+ }
checkpoint-1200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:964fce096945ae0681473bea8dd3f3ffe4945d5b3044785de05444e43384da72
3
+ size 5432
checkpoint-2328/config.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "gelu",
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": null,
10
+ "bos_token_id": 50257,
11
+ "classifier_proj_size": 256,
12
+ "d_model": 768,
13
+ "decoder_attention_heads": 12,
14
+ "decoder_ffn_dim": 3072,
15
+ "decoder_layerdrop": 0.0,
16
+ "decoder_layers": 12,
17
+ "decoder_start_token_id": 50258,
18
+ "dropout": 0.0,
19
+ "encoder_attention_heads": 12,
20
+ "encoder_ffn_dim": 3072,
21
+ "encoder_layerdrop": 0.0,
22
+ "encoder_layers": 12,
23
+ "eos_token_id": 50257,
24
+ "forced_decoder_ids": null,
25
+ "init_std": 0.02,
26
+ "is_encoder_decoder": true,
27
+ "mask_feature_length": 10,
28
+ "mask_feature_min_masks": 0,
29
+ "mask_feature_prob": 0.0,
30
+ "mask_time_length": 10,
31
+ "mask_time_min_masks": 2,
32
+ "mask_time_prob": 0.05,
33
+ "max_length": null,
34
+ "max_source_positions": 1500,
35
+ "max_target_positions": 448,
36
+ "median_filter_width": 7,
37
+ "model_type": "whisper",
38
+ "num_hidden_layers": 12,
39
+ "num_mel_bins": 80,
40
+ "pad_token_id": 50257,
41
+ "scale_embedding": false,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.52.4",
44
+ "use_cache": false,
45
+ "use_weighted_layer_sum": false,
46
+ "vocab_size": 51865
47
+ }
checkpoint-2328/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 50257,
3
+ "decoder_start_token_id": 50258,
4
+ "eos_token_id": 50257,
5
+ "pad_token_id": 50257,
6
+ "suppress_tokens": [],
7
+ "transformers_version": "4.52.4",
8
+ "use_cache": false
9
+ }
checkpoint-2328/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44e354ff3166b3fc983404465852bcc6bc291598d36d5daf95ec4ace75b7e9a2
3
+ size 966995080
checkpoint-2328/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68465cde1b8b6332b5dc57990368ca62744c99789ba59b8c8b74ca0db0d3dccd
3
+ size 1925064044
checkpoint-2328/preprocessor_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "dither": 0.0,
4
+ "feature_extractor_type": "WhisperFeatureExtractor",
5
+ "feature_size": 80,
6
+ "hop_length": 160,
7
+ "n_fft": 400,
8
+ "n_samples": 480000,
9
+ "nb_max_frames": 3000,
10
+ "padding_side": "right",
11
+ "padding_value": 0.0,
12
+ "processor_class": "WhisperProcessor",
13
+ "return_attention_mask": false,
14
+ "sampling_rate": 16000
15
+ }
checkpoint-2328/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1b81ce693acc82e0bf3603f75a354e989f4b0ed93b1f310aabaa44f8e562515
3
+ size 14244
checkpoint-2328/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb3ea351c94f50d7065367b8de5e944fabd6aabfe1ebd101a27322806e136668
3
+ size 988
checkpoint-2328/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:257595079f1ef0c1b8769557aecd7927711673d6af72fe830d559873e280359c
3
+ size 1064
checkpoint-2328/trainer_state.json ADDED
@@ -0,0 +1,156 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1200,
3
+ "best_metric": 8.762886597938143,
4
+ "best_model_checkpoint": "./ZINASR/checkpoint-1200",
5
+ "epoch": 8.0,
6
+ "eval_steps": 400,
7
+ "global_step": 2328,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.6872852233676976,
14
+ "grad_norm": 5.297506809234619,
15
+ "learning_rate": 9.950000000000001e-06,
16
+ "loss": 0.1217,
17
+ "step": 200
18
+ },
19
+ {
20
+ "epoch": 1.3745704467353952,
21
+ "grad_norm": 1.6356655359268188,
22
+ "learning_rate": 9.06484962406015e-06,
23
+ "loss": 0.0826,
24
+ "step": 400
25
+ },
26
+ {
27
+ "epoch": 1.3745704467353952,
28
+ "eval_loss": 0.1176551878452301,
29
+ "eval_runtime": 1451.5411,
30
+ "eval_samples_per_second": 1.602,
31
+ "eval_steps_per_second": 0.2,
32
+ "eval_wer": 8.891752577319588,
33
+ "step": 400
34
+ },
35
+ {
36
+ "epoch": 2.0618556701030926,
37
+ "grad_norm": 0.9965283274650574,
38
+ "learning_rate": 8.134398496240602e-06,
39
+ "loss": 0.0483,
40
+ "step": 600
41
+ },
42
+ {
43
+ "epoch": 2.7491408934707904,
44
+ "grad_norm": 1.6627748012542725,
45
+ "learning_rate": 7.194548872180452e-06,
46
+ "loss": 0.018,
47
+ "step": 800
48
+ },
49
+ {
50
+ "epoch": 2.7491408934707904,
51
+ "eval_loss": 0.121592216193676,
52
+ "eval_runtime": 1448.1115,
53
+ "eval_samples_per_second": 1.606,
54
+ "eval_steps_per_second": 0.201,
55
+ "eval_wer": 9.211531118747613,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 3.436426116838488,
60
+ "grad_norm": 1.0129088163375854,
61
+ "learning_rate": 6.254699248120302e-06,
62
+ "loss": 0.0118,
63
+ "step": 1000
64
+ },
65
+ {
66
+ "epoch": 4.123711340206185,
67
+ "grad_norm": 0.18754775822162628,
68
+ "learning_rate": 5.314849624060151e-06,
69
+ "loss": 0.0067,
70
+ "step": 1200
71
+ },
72
+ {
73
+ "epoch": 4.123711340206185,
74
+ "eval_loss": 0.12416346371173859,
75
+ "eval_runtime": 1449.1081,
76
+ "eval_samples_per_second": 1.604,
77
+ "eval_steps_per_second": 0.201,
78
+ "eval_wer": 8.762886597938143,
79
+ "step": 1200
80
+ },
81
+ {
82
+ "epoch": 4.810996563573883,
83
+ "grad_norm": 0.20880936086177826,
84
+ "learning_rate": 4.3750000000000005e-06,
85
+ "loss": 0.0034,
86
+ "step": 1400
87
+ },
88
+ {
89
+ "epoch": 5.498281786941581,
90
+ "grad_norm": 0.06731250137090683,
91
+ "learning_rate": 3.43515037593985e-06,
92
+ "loss": 0.0022,
93
+ "step": 1600
94
+ },
95
+ {
96
+ "epoch": 5.498281786941581,
97
+ "eval_loss": 0.12590037286281586,
98
+ "eval_runtime": 1450.6425,
99
+ "eval_samples_per_second": 1.603,
100
+ "eval_steps_per_second": 0.201,
101
+ "eval_wer": 8.858342878961436,
102
+ "step": 1600
103
+ },
104
+ {
105
+ "epoch": 6.185567010309279,
106
+ "grad_norm": 0.056367386132478714,
107
+ "learning_rate": 2.4953007518796995e-06,
108
+ "loss": 0.0014,
109
+ "step": 1800
110
+ },
111
+ {
112
+ "epoch": 6.872852233676976,
113
+ "grad_norm": 0.0637979805469513,
114
+ "learning_rate": 1.5554511278195489e-06,
115
+ "loss": 0.0012,
116
+ "step": 2000
117
+ },
118
+ {
119
+ "epoch": 6.872852233676976,
120
+ "eval_loss": 0.12869131565093994,
121
+ "eval_runtime": 1446.4517,
122
+ "eval_samples_per_second": 1.607,
123
+ "eval_steps_per_second": 0.201,
124
+ "eval_wer": 8.858342878961436,
125
+ "step": 2000
126
+ },
127
+ {
128
+ "epoch": 7.560137457044673,
129
+ "grad_norm": 0.0420345701277256,
130
+ "learning_rate": 6.156015037593985e-07,
131
+ "loss": 0.001,
132
+ "step": 2200
133
+ }
134
+ ],
135
+ "logging_steps": 200,
136
+ "max_steps": 2328,
137
+ "num_input_tokens_seen": 0,
138
+ "num_train_epochs": 8,
139
+ "save_steps": 1200,
140
+ "stateful_callbacks": {
141
+ "TrainerControl": {
142
+ "args": {
143
+ "should_epoch_stop": false,
144
+ "should_evaluate": false,
145
+ "should_log": false,
146
+ "should_save": true,
147
+ "should_training_stop": true
148
+ },
149
+ "attributes": {}
150
+ }
151
+ },
152
+ "total_flos": 2.146844510060544e+19,
153
+ "train_batch_size": 16,
154
+ "trial_name": null,
155
+ "trial_params": null
156
+ }
checkpoint-2328/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:964fce096945ae0681473bea8dd3f3ffe4945d5b3044785de05444e43384da72
3
+ size 5432