Hugging Face's logo

ihanif
/
whisper-medium-urdu

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Urdu
whisper
hf-asr-leaderboard
urdu
whisper-event
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
2
ihanif commited on
Commit
6552750
•
1 Parent(s): f43f2d8

add 200 and 300 steps checkpoints

Browse files
Files changed (18) hide show
  1. checkpoint-200/config.json +41 -0
  2. checkpoint-200/optimizer.pt +3 -0
  3. checkpoint-200/preprocessor_config.json +0 -0
  4. checkpoint-200/pytorch_model.bin +3 -0
  5. checkpoint-200/rng_state.pth +3 -0
  6. checkpoint-200/scaler.pt +3 -0
  7. checkpoint-200/scheduler.pt +3 -0
  8. checkpoint-200/trainer_state.json +154 -0
  9. checkpoint-200/training_args.bin +3 -0
  10. checkpoint-300/config.json +41 -0
  11. checkpoint-300/optimizer.pt +3 -0
  12. checkpoint-300/preprocessor_config.json +0 -0
  13. checkpoint-300/pytorch_model.bin +3 -0
  14. checkpoint-300/rng_state.pth +3 -0
  15. checkpoint-300/scaler.pt +3 -0
  16. checkpoint-300/scheduler.pt +3 -0
  17. checkpoint-300/trainer_state.json +223 -0
  18. checkpoint-300/training_args.bin +3 -0
checkpoint-200/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-medium",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": [
10
+ 220,
11
+ 50257
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "d_model": 1024,
15
+ "decoder_attention_heads": 16,
16
+ "decoder_ffn_dim": 4096,
17
+ "decoder_layerdrop": 0.0,
18
+ "decoder_layers": 24,
19
+ "decoder_start_token_id": 50258,
20
+ "dropout": 0.0,
21
+ "encoder_attention_heads": 16,
22
+ "encoder_ffn_dim": 4096,
23
+ "encoder_layerdrop": 0.0,
24
+ "encoder_layers": 24,
25
+ "eos_token_id": 50257,
26
+ "forced_decoder_ids": null,
27
+ "init_std": 0.02,
28
+ "is_encoder_decoder": true,
29
+ "max_length": 448,
30
+ "max_source_positions": 1500,
31
+ "max_target_positions": 448,
32
+ "model_type": "whisper",
33
+ "num_hidden_layers": 24,
34
+ "num_mel_bins": 80,
35
+ "pad_token_id": 50257,
36
+ "scale_embedding": false,
37
+ "torch_dtype": "float32",
38
+ "transformers_version": "4.26.0.dev0",
39
+ "use_cache": false,
40
+ "vocab_size": 51865
41
+ }
checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46c849ed61144e475f7c9fd37c178440b5e7e44e5946eb4296ada59843f4f3af
3
+ size 6111427735
checkpoint-200/preprocessor_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-200/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5eeaa4b4b2e7dc750a86f0904797221f08de5028239fc72f25bd276a772000c1
3
+ size 3055754841
checkpoint-200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e758aefe798f132b0d37bdfedb1199698614acb3147589871ad045c84e89f15
3
+ size 14575
checkpoint-200/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83982c4fb519a661cd918a859b2d62fb9cc25996a4f00819eba4d30c02e371bc
3
+ size 557
checkpoint-200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dae91f004b31a174caf7110ee316eb98ae014269026aaeee4bea97d9ad557066
3
+ size 627
checkpoint-200/trainer_state.json ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 26.980130911344357,
3
+ "best_model_checkpoint": "./checkpoint-200",
4
+ "epoch": 0.966183574879227,
5
+ "global_step": 200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.05,
12
+ "learning_rate": 2.0000000000000003e-06,
13
+ "loss": 0.9955,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.1,
18
+ "learning_rate": 4.5e-06,
19
+ "loss": 0.7842,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.14,
24
+ "learning_rate": 7e-06,
25
+ "loss": 0.533,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.19,
30
+ "learning_rate": 9.5e-06,
31
+ "loss": 0.5079,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.24,
36
+ "learning_rate": 9.5e-06,
37
+ "loss": 0.4739,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.29,
42
+ "learning_rate": 8.875e-06,
43
+ "loss": 0.4768,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.34,
48
+ "learning_rate": 8.25e-06,
49
+ "loss": 0.4344,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.39,
54
+ "learning_rate": 7.625e-06,
55
+ "loss": 0.4655,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.43,
60
+ "learning_rate": 7e-06,
61
+ "loss": 0.4108,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.48,
66
+ "learning_rate": 6.375e-06,
67
+ "loss": 0.3908,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.48,
72
+ "eval_loss": 0.5148195028305054,
73
+ "eval_runtime": 2783.9171,
74
+ "eval_samples_per_second": 1.186,
75
+ "eval_steps_per_second": 0.074,
76
+ "eval_wer": 30.706069173864208,
77
+ "step": 100
78
+ },
79
+ {
80
+ "epoch": 0.53,
81
+ "learning_rate": 5.75e-06,
82
+ "loss": 0.3863,
83
+ "step": 110
84
+ },
85
+ {
86
+ "epoch": 0.58,
87
+ "learning_rate": 5.125e-06,
88
+ "loss": 0.3902,
89
+ "step": 120
90
+ },
91
+ {
92
+ "epoch": 0.63,
93
+ "learning_rate": 4.5e-06,
94
+ "loss": 0.3788,
95
+ "step": 130
96
+ },
97
+ {
98
+ "epoch": 0.68,
99
+ "learning_rate": 3.875e-06,
100
+ "loss": 0.403,
101
+ "step": 140
102
+ },
103
+ {
104
+ "epoch": 0.72,
105
+ "learning_rate": 3.2500000000000002e-06,
106
+ "loss": 0.3781,
107
+ "step": 150
108
+ },
109
+ {
110
+ "epoch": 0.77,
111
+ "learning_rate": 2.6250000000000003e-06,
112
+ "loss": 0.4094,
113
+ "step": 160
114
+ },
115
+ {
116
+ "epoch": 0.82,
117
+ "learning_rate": 2.0000000000000003e-06,
118
+ "loss": 0.3732,
119
+ "step": 170
120
+ },
121
+ {
122
+ "epoch": 0.87,
123
+ "learning_rate": 1.3750000000000002e-06,
124
+ "loss": 0.4084,
125
+ "step": 180
126
+ },
127
+ {
128
+ "epoch": 0.92,
129
+ "learning_rate": 7.5e-07,
130
+ "loss": 0.3786,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 0.97,
135
+ "learning_rate": 1.2500000000000002e-07,
136
+ "loss": 0.3519,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.97,
141
+ "eval_loss": 0.46850404143333435,
142
+ "eval_runtime": 2556.1826,
143
+ "eval_samples_per_second": 1.292,
144
+ "eval_steps_per_second": 0.081,
145
+ "eval_wer": 26.980130911344357,
146
+ "step": 200
147
+ }
148
+ ],
149
+ "max_steps": 200,
150
+ "num_train_epochs": 1,
151
+ "total_flos": 6.531871408128e+18,
152
+ "trial_name": null,
153
+ "trial_params": null
154
+ }
checkpoint-200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b747d3d49e8fc76cb72a7ae75a9c9a2591850e63223d0fdc354d5c4826934c9f
3
+ size 3579
checkpoint-300/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-medium",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "architectures": [
6
+ "WhisperForConditionalGeneration"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "begin_suppress_tokens": [
10
+ 220,
11
+ 50257
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "d_model": 1024,
15
+ "decoder_attention_heads": 16,
16
+ "decoder_ffn_dim": 4096,
17
+ "decoder_layerdrop": 0.0,
18
+ "decoder_layers": 24,
19
+ "decoder_start_token_id": 50258,
20
+ "dropout": 0.0,
21
+ "encoder_attention_heads": 16,
22
+ "encoder_ffn_dim": 4096,
23
+ "encoder_layerdrop": 0.0,
24
+ "encoder_layers": 24,
25
+ "eos_token_id": 50257,
26
+ "forced_decoder_ids": null,
27
+ "init_std": 0.02,
28
+ "is_encoder_decoder": true,
29
+ "max_length": 448,
30
+ "max_source_positions": 1500,
31
+ "max_target_positions": 448,
32
+ "model_type": "whisper",
33
+ "num_hidden_layers": 24,
34
+ "num_mel_bins": 80,
35
+ "pad_token_id": 50257,
36
+ "scale_embedding": false,
37
+ "torch_dtype": "float32",
38
+ "transformers_version": "4.26.0.dev0",
39
+ "use_cache": false,
40
+ "vocab_size": 51865
41
+ }
checkpoint-300/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6cdc3f516e99dd3353b903812cc3585902381baa98c18d089e1414e098a8e343
3
+ size 6111428695
checkpoint-300/preprocessor_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-300/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc43fe323b4439210889594b2ed3475e474ed2a8ccd3f5663693c5ac4386193a
3
+ size 3055754841
checkpoint-300/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03d624647473db68b357c5c0f54f8c17a9116e1f4a09ada7d5c8862daea4c764
3
+ size 14575
checkpoint-300/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d99023c3b8262c002557a160ea1cf0d08b630526ae2703f38fd0cc772e24beec
3
+ size 557
checkpoint-300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3031653416187e46a54b348ab76e808487ff77c056044182f596a9cb45e30549
3
+ size 627
checkpoint-300/trainer_state.json ADDED
@@ -0,0 +1,223 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 26.980130911344357,
3
+ "best_model_checkpoint": "./checkpoint-200",
4
+ "epoch": 1.4492753623188406,
5
+ "global_step": 300,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.05,
12
+ "learning_rate": 2.0000000000000003e-06,
13
+ "loss": 0.9955,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.1,
18
+ "learning_rate": 4.5e-06,
19
+ "loss": 0.7842,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.14,
24
+ "learning_rate": 7e-06,
25
+ "loss": 0.533,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.19,
30
+ "learning_rate": 9.5e-06,
31
+ "loss": 0.5079,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.24,
36
+ "learning_rate": 9.5e-06,
37
+ "loss": 0.4739,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.29,
42
+ "learning_rate": 8.875e-06,
43
+ "loss": 0.4768,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.34,
48
+ "learning_rate": 8.25e-06,
49
+ "loss": 0.4344,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.39,
54
+ "learning_rate": 7.625e-06,
55
+ "loss": 0.4655,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.43,
60
+ "learning_rate": 7e-06,
61
+ "loss": 0.4108,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.48,
66
+ "learning_rate": 6.375e-06,
67
+ "loss": 0.3908,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.48,
72
+ "eval_loss": 0.5148195028305054,
73
+ "eval_runtime": 2783.9171,
74
+ "eval_samples_per_second": 1.186,
75
+ "eval_steps_per_second": 0.074,
76
+ "eval_wer": 30.706069173864208,
77
+ "step": 100
78
+ },
79
+ {
80
+ "epoch": 0.53,
81
+ "learning_rate": 5.75e-06,
82
+ "loss": 0.3863,
83
+ "step": 110
84
+ },
85
+ {
86
+ "epoch": 0.58,
87
+ "learning_rate": 5.125e-06,
88
+ "loss": 0.3902,
89
+ "step": 120
90
+ },
91
+ {
92
+ "epoch": 0.63,
93
+ "learning_rate": 4.5e-06,
94
+ "loss": 0.3788,
95
+ "step": 130
96
+ },
97
+ {
98
+ "epoch": 0.68,
99
+ "learning_rate": 3.875e-06,
100
+ "loss": 0.403,
101
+ "step": 140
102
+ },
103
+ {
104
+ "epoch": 0.72,
105
+ "learning_rate": 3.2500000000000002e-06,
106
+ "loss": 0.3781,
107
+ "step": 150
108
+ },
109
+ {
110
+ "epoch": 0.77,
111
+ "learning_rate": 2.6250000000000003e-06,
112
+ "loss": 0.4094,
113
+ "step": 160
114
+ },
115
+ {
116
+ "epoch": 0.82,
117
+ "learning_rate": 2.0000000000000003e-06,
118
+ "loss": 0.3732,
119
+ "step": 170
120
+ },
121
+ {
122
+ "epoch": 0.87,
123
+ "learning_rate": 1.3750000000000002e-06,
124
+ "loss": 0.4084,
125
+ "step": 180
126
+ },
127
+ {
128
+ "epoch": 0.92,
129
+ "learning_rate": 7.5e-07,
130
+ "loss": 0.3786,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 0.97,
135
+ "learning_rate": 1.2500000000000002e-07,
136
+ "loss": 0.3519,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.97,
141
+ "eval_loss": 0.46850404143333435,
142
+ "eval_runtime": 2556.1826,
143
+ "eval_samples_per_second": 1.292,
144
+ "eval_steps_per_second": 0.081,
145
+ "eval_wer": 26.980130911344357,
146
+ "step": 200
147
+ },
148
+ {
149
+ "epoch": 1.01,
150
+ "learning_rate": 3.538461538461539e-06,
151
+ "loss": 0.3721,
152
+ "step": 210
153
+ },
154
+ {
155
+ "epoch": 1.06,
156
+ "learning_rate": 3.153846153846154e-06,
157
+ "loss": 0.2392,
158
+ "step": 220
159
+ },
160
+ {
161
+ "epoch": 1.11,
162
+ "learning_rate": 2.7692307692307697e-06,
163
+ "loss": 0.2854,
164
+ "step": 230
165
+ },
166
+ {
167
+ "epoch": 1.16,
168
+ "learning_rate": 2.384615384615385e-06,
169
+ "loss": 0.2517,
170
+ "step": 240
171
+ },
172
+ {
173
+ "epoch": 1.21,
174
+ "learning_rate": 2.0000000000000003e-06,
175
+ "loss": 0.2506,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 1.26,
180
+ "learning_rate": 1.6153846153846157e-06,
181
+ "loss": 0.2462,
182
+ "step": 260
183
+ },
184
+ {
185
+ "epoch": 1.3,
186
+ "learning_rate": 1.230769230769231e-06,
187
+ "loss": 0.2179,
188
+ "step": 270
189
+ },
190
+ {
191
+ "epoch": 1.35,
192
+ "learning_rate": 8.461538461538463e-07,
193
+ "loss": 0.2579,
194
+ "step": 280
195
+ },
196
+ {
197
+ "epoch": 1.4,
198
+ "learning_rate": 4.615384615384616e-07,
199
+ "loss": 0.2446,
200
+ "step": 290
201
+ },
202
+ {
203
+ "epoch": 1.45,
204
+ "learning_rate": 7.692307692307694e-08,
205
+ "loss": 0.2426,
206
+ "step": 300
207
+ },
208
+ {
209
+ "epoch": 1.45,
210
+ "eval_loss": 0.4635893404483795,
211
+ "eval_runtime": 2857.1857,
212
+ "eval_samples_per_second": 1.156,
213
+ "eval_steps_per_second": 0.072,
214
+ "eval_wer": 28.502265773267748,
215
+ "step": 300
216
+ }
217
+ ],
218
+ "max_steps": 300,
219
+ "num_train_epochs": 2,
220
+ "total_flos": 9.77943622385664e+18,
221
+ "trial_name": null,
222
+ "trial_params": null
223
+ }
checkpoint-300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6406a19600ef55b665d6e036bbd71317a735105e08ff51d4b60484c5c5f0047a
3
+ size 3579