Hugging Face's logo

lnxdx
/
B1_1000_1e-4_hp-mehrdad

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
lnxdx commited on
Commit
9bea35d
•
1 Parent(s): 099b2c5

End of training

Browse files
Files changed (11) hide show
  1. README.md +72 -0
  2. all_results.json +9 -0
  3. config.json +116 -0
  4. model.safetensors +3 -0
  5. preprocessor_config.json +10 -0
  6. special_tokens_map.json +30 -0
  7. tokenizer_config.json +48 -0
  8. train_results.json +9 -0
  9. trainer_state.json +178 -0
  10. training_args.bin +3 -0
  11. vocab.json +42 -0
README.md ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: lnxdx/21_2500_1e-4_hp-mehrdad
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - wer
7
+ model-index:
8
+ - name: 23_3000_1e-4_hp-mehrdad
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # 23_3000_1e-4_hp-mehrdad
16
+
17
+ This model is a fine-tuned version of [lnxdx/21_2500_1e-4_hp-mehrdad](https://huggingface.co/lnxdx/21_2500_1e-4_hp-mehrdad) on the None dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.9137
20
+ - Wer: 0.2932
21
+
22
+ ## Model description
23
+
24
+ More information needed
25
+
26
+ ## Intended uses & limitations
27
+
28
+ More information needed
29
+
30
+ ## Training and evaluation data
31
+
32
+ More information needed
33
+
34
+ ## Training procedure
35
+
36
+ ### Training hyperparameters
37
+
38
+ The following hyperparameters were used during training:
39
+ - learning_rate: 0.0001
40
+ - train_batch_size: 8
41
+ - eval_batch_size: 8
42
+ - seed: 42
43
+ - gradient_accumulation_steps: 2
44
+ - total_train_batch_size: 16
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 500
48
+ - training_steps: 1000
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
55
+ | 0.7626 | 0.62 | 100 | 0.6772 | 0.3216 |
56
+ | 0.7702 | 1.25 | 200 | 0.6705 | 0.3178 |
57
+ | 0.7605 | 1.88 | 300 | 0.6425 | 0.3175 |
58
+ | 0.6617 | 2.5 | 400 | 0.7113 | 0.3029 |
59
+ | 0.661 | 3.12 | 500 | 0.7173 | 0.3020 |
60
+ | 0.6037 | 3.75 | 600 | 0.6833 | 0.3478 |
61
+ | 0.5343 | 4.38 | 700 | 0.8272 | 0.2959 |
62
+ | 0.5096 | 5.0 | 800 | 0.8571 | 0.2973 |
63
+ | 0.4471 | 5.62 | 900 | 0.9498 | 0.2909 |
64
+ | 0.4393 | 6.25 | 1000 | 0.9137 | 0.2932 |
65
+
66
+
67
+ ### Framework versions
68
+
69
+ - Transformers 4.35.2
70
+ - Pytorch 2.1.0+cu118
71
+ - Datasets 2.15.0
72
+ - Tokenizers 0.15.0
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.25,
3
+ "total_flos": 4.798362622332561e+18,
4
+ "train_loss": 0.6149942665100098,
5
+ "train_runtime": 5145.8365,
6
+ "train_samples": 2554,
7
+ "train_samples_per_second": 3.109,
8
+ "train_steps_per_second": 0.194
9
+ }
config.json ADDED
@@ -0,0 +1,116 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "lnxdx/21_2500_1e-4_hp-mehrdad",
3
+ "activation_dropout": 0.09216,
4
+ "adapter_attn_dim": null,
5
+ "adapter_kernel_size": 3,
6
+ "adapter_stride": 2,
7
+ "add_adapter": false,
8
+ "apply_spec_augment": true,
9
+ "architectures": [
10
+ "Wav2Vec2ForCTC"
11
+ ],
12
+ "attention_dropout": 0.05316,
13
+ "bos_token_id": 1,
14
+ "classifier_proj_size": 256,
15
+ "codevector_dim": 256,
16
+ "contrastive_logits_temperature": 0.1,
17
+ "conv_bias": true,
18
+ "conv_dim": [
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512,
25
+ 512
26
+ ],
27
+ "conv_kernel": [
28
+ 10,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 3,
33
+ 2,
34
+ 2
35
+ ],
36
+ "conv_stride": [
37
+ 5,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2,
43
+ 2
44
+ ],
45
+ "ctc_loss_reduction": "mean",
46
+ "ctc_zero_infinity": true,
47
+ "diversity_loss_weight": 0.1,
48
+ "do_stable_layer_norm": true,
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": 0.01249,
54
+ "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.0,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.01941,
58
+ "hidden_size": 1024,
59
+ "initializer_range": 0.02,
60
+ "intermediate_size": 4096,
61
+ "layer_norm_eps": 1e-05,
62
+ "layerdrop": 0.01377,
63
+ "mask_channel_length": 10,
64
+ "mask_channel_min_space": 1,
65
+ "mask_channel_other": 0.0,
66
+ "mask_channel_prob": 0.0,
67
+ "mask_channel_selection": "static",
68
+ "mask_feature_length": 10,
69
+ "mask_feature_min_masks": 0,
70
+ "mask_feature_prob": 0.0,
71
+ "mask_time_length": 10,
72
+ "mask_time_min_masks": 2,
73
+ "mask_time_min_space": 1,
74
+ "mask_time_other": 0.0,
75
+ "mask_time_prob": 0.04529,
76
+ "mask_time_selection": "static",
77
+ "model_type": "wav2vec2",
78
+ "num_adapter_layers": 3,
79
+ "num_attention_heads": 16,
80
+ "num_codevector_groups": 2,
81
+ "num_codevectors_per_group": 320,
82
+ "num_conv_pos_embedding_groups": 16,
83
+ "num_conv_pos_embeddings": 128,
84
+ "num_feat_extract_layers": 7,
85
+ "num_hidden_layers": 24,
86
+ "num_negatives": 100,
87
+ "output_hidden_size": 1024,
88
+ "pad_token_id": 0,
89
+ "proj_codevector_dim": 256,
90
+ "tdnn_dilation": [
91
+ 1,
92
+ 2,
93
+ 3,
94
+ 1,
95
+ 1
96
+ ],
97
+ "tdnn_dim": [
98
+ 512,
99
+ 512,
100
+ 512,
101
+ 512,
102
+ 1500
103
+ ],
104
+ "tdnn_kernel": [
105
+ 5,
106
+ 3,
107
+ 3,
108
+ 1,
109
+ 1
110
+ ],
111
+ "torch_dtype": "float32",
112
+ "transformers_version": "4.35.2",
113
+ "use_weighted_layer_sum": false,
114
+ "vocab_size": 40,
115
+ "xvector_output_dim": 512
116
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84dd0138ed195b1aec1a93db366343b6943a2cd19fe2e7c5a9ed3d3118fd7a53
3
+ size 1261971480
preprocessor_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "processor_class": "Wav2Vec2Processor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000
10
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": true,
5
+ "normalized": false,
6
+ "rstrip": true,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": true,
12
+ "normalized": false,
13
+ "rstrip": true,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<pad>",
18
+ "lstrip": true,
19
+ "normalized": false,
20
+ "rstrip": true,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": true,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<pad>",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "1": {
12
+ "content": "<s>",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": true,
22
+ "normalized": false,
23
+ "rstrip": true,
24
+ "single_word": false,
25
+ "special": false
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": true,
30
+ "normalized": false,
31
+ "rstrip": true,
32
+ "single_word": false,
33
+ "special": false
34
+ }
35
+ },
36
+ "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": true,
38
+ "do_lower_case": false,
39
+ "eos_token": "</s>",
40
+ "model_max_length": 1000000000000000019884624838656,
41
+ "pad_token": "<pad>",
42
+ "processor_class": "Wav2Vec2Processor",
43
+ "replace_word_delimiter_char": " ",
44
+ "target_lang": null,
45
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
46
+ "unk_token": "<unk>",
47
+ "word_delimiter_token": "|"
48
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.25,
3
+ "total_flos": 4.798362622332561e+18,
4
+ "train_loss": 0.6149942665100098,
5
+ "train_runtime": 5145.8365,
6
+ "train_samples": 2554,
7
+ "train_samples_per_second": 3.109,
8
+ "train_steps_per_second": 0.194
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,178 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.2908878504672897,
3
+ "best_model_checkpoint": "/content/drive/MyDrive/wav2vec2-large-xlsr-persian-asr-shemo_sm74/checkpoint-900",
4
+ "epoch": 6.25,
5
+ "eval_steps": 100,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.62,
13
+ "learning_rate": 1.9800000000000004e-05,
14
+ "loss": 0.7626,
15
+ "step": 100
16
+ },
17
+ {
18
+ "epoch": 0.62,
19
+ "eval_loss": 0.6772463321685791,
20
+ "eval_runtime": 36.0706,
21
+ "eval_samples_per_second": 7.873,
22
+ "eval_steps_per_second": 0.998,
23
+ "eval_wer": 0.321553738317757,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 1.25,
28
+ "learning_rate": 3.9800000000000005e-05,
29
+ "loss": 0.7702,
30
+ "step": 200
31
+ },
32
+ {
33
+ "epoch": 1.25,
34
+ "eval_loss": 0.6705306768417358,
35
+ "eval_runtime": 35.9659,
36
+ "eval_samples_per_second": 7.896,
37
+ "eval_steps_per_second": 1.001,
38
+ "eval_wer": 0.3177570093457944,
39
+ "step": 200
40
+ },
41
+ {
42
+ "epoch": 1.88,
43
+ "learning_rate": 5.9800000000000003e-05,
44
+ "loss": 0.7605,
45
+ "step": 300
46
+ },
47
+ {
48
+ "epoch": 1.88,
49
+ "eval_loss": 0.6425135731697083,
50
+ "eval_runtime": 36.0164,
51
+ "eval_samples_per_second": 7.885,
52
+ "eval_steps_per_second": 1.0,
53
+ "eval_wer": 0.317464953271028,
54
+ "step": 300
55
+ },
56
+ {
57
+ "epoch": 2.5,
58
+ "learning_rate": 7.98e-05,
59
+ "loss": 0.6617,
60
+ "step": 400
61
+ },
62
+ {
63
+ "epoch": 2.5,
64
+ "eval_loss": 0.7112990617752075,
65
+ "eval_runtime": 35.9864,
66
+ "eval_samples_per_second": 7.892,
67
+ "eval_steps_per_second": 1.0,
68
+ "eval_wer": 0.3028621495327103,
69
+ "step": 400
70
+ },
71
+ {
72
+ "epoch": 3.12,
73
+ "learning_rate": 9.98e-05,
74
+ "loss": 0.661,
75
+ "step": 500
76
+ },
77
+ {
78
+ "epoch": 3.12,
79
+ "eval_loss": 0.7172772288322449,
80
+ "eval_runtime": 35.6659,
81
+ "eval_samples_per_second": 7.963,
82
+ "eval_steps_per_second": 1.009,
83
+ "eval_wer": 0.3019859813084112,
84
+ "step": 500
85
+ },
86
+ {
87
+ "epoch": 3.75,
88
+ "learning_rate": 8.020000000000001e-05,
89
+ "loss": 0.6037,
90
+ "step": 600
91
+ },
92
+ {
93
+ "epoch": 3.75,
94
+ "eval_loss": 0.683303713798523,
95
+ "eval_runtime": 36.0387,
96
+ "eval_samples_per_second": 7.88,
97
+ "eval_steps_per_second": 0.999,
98
+ "eval_wer": 0.347838785046729,
99
+ "step": 600
100
+ },
101
+ {
102
+ "epoch": 4.38,
103
+ "learning_rate": 6.02e-05,
104
+ "loss": 0.5343,
105
+ "step": 700
106
+ },
107
+ {
108
+ "epoch": 4.38,
109
+ "eval_loss": 0.8271604180335999,
110
+ "eval_runtime": 35.168,
111
+ "eval_samples_per_second": 8.076,
112
+ "eval_steps_per_second": 1.024,
113
+ "eval_wer": 0.29585280373831774,
114
+ "step": 700
115
+ },
116
+ {
117
+ "epoch": 5.0,
118
+ "learning_rate": 4.02e-05,
119
+ "loss": 0.5096,
120
+ "step": 800
121
+ },
122
+ {
123
+ "epoch": 5.0,
124
+ "eval_loss": 0.8571420311927795,
125
+ "eval_runtime": 35.4406,
126
+ "eval_samples_per_second": 8.013,
127
+ "eval_steps_per_second": 1.016,
128
+ "eval_wer": 0.2973130841121495,
129
+ "step": 800
130
+ },
131
+ {
132
+ "epoch": 5.62,
133
+ "learning_rate": 2.0200000000000003e-05,
134
+ "loss": 0.4471,
135
+ "step": 900
136
+ },
137
+ {
138
+ "epoch": 5.62,
139
+ "eval_loss": 0.9498320817947388,
140
+ "eval_runtime": 37.0202,
141
+ "eval_samples_per_second": 7.671,
142
+ "eval_steps_per_second": 0.972,
143
+ "eval_wer": 0.2908878504672897,
144
+ "step": 900
145
+ },
146
+ {
147
+ "epoch": 6.25,
148
+ "learning_rate": 2.0000000000000002e-07,
149
+ "loss": 0.4393,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 6.25,
154
+ "eval_loss": 0.9137212038040161,
155
+ "eval_runtime": 36.1327,
156
+ "eval_samples_per_second": 7.86,
157
+ "eval_steps_per_second": 0.996,
158
+ "eval_wer": 0.2932242990654206,
159
+ "step": 1000
160
+ },
161
+ {
162
+ "epoch": 6.25,
163
+ "step": 1000,
164
+ "total_flos": 4.798362622332561e+18,
165
+ "train_loss": 0.6149942665100098,
166
+ "train_runtime": 5145.8365,
167
+ "train_samples_per_second": 3.109,
168
+ "train_steps_per_second": 0.194
169
+ }
170
+ ],
171
+ "logging_steps": 100,
172
+ "max_steps": 1000,
173
+ "num_train_epochs": 7,
174
+ "save_steps": 100,
175
+ "total_flos": 4.798362622332561e+18,
176
+ "trial_name": null,
177
+ "trial_params": null
178
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7493d12d17c4f847c0ddfe51a8f4999b94bf97dec3410643ca7062b54e89f47e
3
+ size 4664
vocab.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</s>": 2,
3
+ "<pad>": 0,
4
+ "<s>": 1,
5
+ "<unk>": 3,
6
+ "|": 4,
7
+ "آ": 5,
8
+ "ئ": 6,
9
+ "ا": 7,
10
+ "ب": 8,
11
+ "ت": 9,
12
+ "ث": 10,
13
+ "ج": 11,
14
+ "ح": 12,
15
+ "خ": 13,
16
+ "د": 14,
17
+ "ذ": 15,
18
+ "ر": 16,
19
+ "ز": 17,
20
+ "س": 18,
21
+ "ش": 19,
22
+ "ص": 20,
23
+ "ض": 21,
24
+ "ط": 22,
25
+ "ظ": 23,
26
+ "ع": 24,
27
+ "غ": 25,
28
+ "ف": 26,
29
+ "ق": 27,
30
+ "ل": 28,
31
+ "م": 29,
32
+ "ن": 30,
33
+ "ه": 31,
34
+ "و": 32,
35
+ "پ": 33,
36
+ "چ": 34,
37
+ "ژ": 35,
38
+ "ک": 36,
39
+ "گ": 37,
40
+ "ی": 38,
41
+ "‌": 39
42
+ }