Hugging Face's logo

tachiwin
/
language_classification_checkpoints

Safetensors
whisper
Model card Files
xet
Community
ljcamargo commited on
Commit
3291d35
·
verified ·
1 Parent(s): df4cd5c

Training in progress, step 300, checkpoint

Browse files
Files changed (10) hide show
  1. last-checkpoint/config.json +49 -0
  2. last-checkpoint/generation_config.json +151 -0
  3. last-checkpoint/model.safetensors +3 -0
  4. last-checkpoint/optimizer.pt +3 -0
  5. last-checkpoint/preprocessor_config.json +15 -0
  6. last-checkpoint/rng_state.pth +3 -0
  7. last-checkpoint/scaler.pt +3 -0
  8. last-checkpoint/scheduler.pt +3 -0
  9. last-checkpoint/trainer_state.json +256 -0
  10. last-checkpoint/training_args.bin +3 -0
last-checkpoint/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": [
10
+ 220,
11
+ 50256
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "classifier_proj_size": 256,
15
+ "d_model": 1280,
16
+ "decoder_attention_heads": 20,
17
+ "decoder_ffn_dim": 5120,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 4,
20
+ "decoder_start_token_id": 50258,
21
+ "dropout": 0.0,
22
+ "dtype": "float32",
23
+ "encoder_attention_heads": 20,
24
+ "encoder_ffn_dim": 5120,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 32,
27
+ "eos_token_id": 50257,
28
+ "init_std": 0.02,
29
+ "is_encoder_decoder": true,
30
+ "mask_feature_length": 10,
31
+ "mask_feature_min_masks": 0,
32
+ "mask_feature_prob": 0.0,
33
+ "mask_time_length": 10,
34
+ "mask_time_min_masks": 2,
35
+ "mask_time_prob": 0.05,
36
+ "max_source_positions": 1500,
37
+ "max_target_positions": 448,
38
+ "median_filter_width": 7,
39
+ "model_type": "whisper",
40
+ "num_hidden_layers": 32,
41
+ "num_mel_bins": 128,
42
+ "pad_token_id": 50257,
43
+ "scale_embedding": false,
44
+ "transformers_version": "4.56.2",
45
+ "unsloth_version": "2025.12.8",
46
+ "use_cache": true,
47
+ "use_weighted_layer_sum": false,
48
+ "vocab_size": 52303
49
+ }
last-checkpoint/generation_config.json ADDED
@@ -0,0 +1,151 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 2,
5
+ 4
6
+ ],
7
+ [
8
+ 2,
9
+ 11
10
+ ],
11
+ [
12
+ 3,
13
+ 3
14
+ ],
15
+ [
16
+ 3,
17
+ 6
18
+ ],
19
+ [
20
+ 3,
21
+ 11
22
+ ],
23
+ [
24
+ 3,
25
+ 14
26
+ ]
27
+ ],
28
+ "begin_suppress_tokens": [
29
+ 220,
30
+ 50257
31
+ ],
32
+ "bos_token_id": 50257,
33
+ "decoder_start_token_id": 50258,
34
+ "eos_token_id": 50257,
35
+ "forced_decoder_ids": null,
36
+ "is_multilingual": true,
37
+ "lang_to_id": {
38
+ "<|af|>": 50327,
39
+ "<|am|>": 50334,
40
+ "<|ar|>": 50272,
41
+ "<|as|>": 50350,
42
+ "<|az|>": 50304,
43
+ "<|ba|>": 50355,
44
+ "<|be|>": 50330,
45
+ "<|bg|>": 50292,
46
+ "<|bn|>": 50302,
47
+ "<|bo|>": 50347,
48
+ "<|br|>": 50309,
49
+ "<|bs|>": 50315,
50
+ "<|ca|>": 50270,
51
+ "<|cs|>": 50283,
52
+ "<|cy|>": 50297,
53
+ "<|da|>": 50285,
54
+ "<|de|>": 50261,
55
+ "<|el|>": 50281,
56
+ "<|en|>": 50259,
57
+ "<|es|>": 50262,
58
+ "<|et|>": 50307,
59
+ "<|eu|>": 50310,
60
+ "<|fa|>": 50300,
61
+ "<|fi|>": 50277,
62
+ "<|fo|>": 50338,
63
+ "<|fr|>": 50265,
64
+ "<|gl|>": 50319,
65
+ "<|gu|>": 50333,
66
+ "<|haw|>": 50352,
67
+ "<|ha|>": 50354,
68
+ "<|he|>": 50279,
69
+ "<|hi|>": 50276,
70
+ "<|hr|>": 50291,
71
+ "<|ht|>": 50339,
72
+ "<|hu|>": 50286,
73
+ "<|hy|>": 50312,
74
+ "<|id|>": 50275,
75
+ "<|is|>": 50311,
76
+ "<|it|>": 50274,
77
+ "<|ja|>": 50266,
78
+ "<|jw|>": 50356,
79
+ "<|ka|>": 50329,
80
+ "<|kk|>": 50316,
81
+ "<|km|>": 50323,
82
+ "<|kn|>": 50306,
83
+ "<|ko|>": 50264,
84
+ "<|la|>": 50294,
85
+ "<|lb|>": 50345,
86
+ "<|ln|>": 50353,
87
+ "<|lo|>": 50336,
88
+ "<|lt|>": 50293,
89
+ "<|lv|>": 50301,
90
+ "<|mg|>": 50349,
91
+ "<|mi|>": 50295,
92
+ "<|mk|>": 50308,
93
+ "<|ml|>": 50296,
94
+ "<|mn|>": 50314,
95
+ "<|mr|>": 50320,
96
+ "<|ms|>": 50282,
97
+ "<|mt|>": 50343,
98
+ "<|my|>": 50346,
99
+ "<|ne|>": 50313,
100
+ "<|nl|>": 50271,
101
+ "<|nn|>": 50342,
102
+ "<|no|>": 50288,
103
+ "<|oc|>": 50328,
104
+ "<|pa|>": 50321,
105
+ "<|pl|>": 50269,
106
+ "<|ps|>": 50340,
107
+ "<|pt|>": 50267,
108
+ "<|ro|>": 50284,
109
+ "<|ru|>": 50263,
110
+ "<|sa|>": 50344,
111
+ "<|sd|>": 50332,
112
+ "<|si|>": 50322,
113
+ "<|sk|>": 50298,
114
+ "<|sl|>": 50305,
115
+ "<|sn|>": 50324,
116
+ "<|so|>": 50326,
117
+ "<|sq|>": 50317,
118
+ "<|sr|>": 50303,
119
+ "<|su|>": 50357,
120
+ "<|sv|>": 50273,
121
+ "<|sw|>": 50318,
122
+ "<|ta|>": 50287,
123
+ "<|te|>": 50299,
124
+ "<|tg|>": 50331,
125
+ "<|th|>": 50289,
126
+ "<|tk|>": 50341,
127
+ "<|tl|>": 50348,
128
+ "<|tr|>": 50268,
129
+ "<|tt|>": 50351,
130
+ "<|uk|>": 50280,
131
+ "<|ur|>": 50290,
132
+ "<|uz|>": 50337,
133
+ "<|vi|>": 50278,
134
+ "<|yi|>": 50335,
135
+ "<|yo|>": 50325,
136
+ "<|yue|>": 50358,
137
+ "<|zh|>": 50260
138
+ },
139
+ "max_initial_timestamp_index": 50,
140
+ "max_length": 448,
141
+ "no_timestamps_token_id": 50364,
142
+ "pad_token_id": 50257,
143
+ "prev_sot_token_id": 50362,
144
+ "return_timestamps": false,
145
+ "suppress_tokens": [],
146
+ "task_to_id": {
147
+ "transcribe": 50360,
148
+ "translate": 50359
149
+ },
150
+ "transformers_version": "4.56.2"
151
+ }
last-checkpoint/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85363bc3f8b3be52c66fdbb216c5e681b65a1619d094204c96d8f21f92b55366
3
+ size 3237818848
last-checkpoint/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d688d8141ac8f14abfd6d9138d0d80e88cec420d7c794332c18006fdab3debd
3
+ size 2061550039
last-checkpoint/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": 128,
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
+ }
last-checkpoint/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1beae2217e3edc7a0dbdeb472e91f228cbe255351a62c3bb1d01db55e8b2bd92
3
+ size 14645
last-checkpoint/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc1a0da602f8abf4bf342932694d528cc1f0baa4d5027de58ad34f4d9855d085
3
+ size 1383
last-checkpoint/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e457036843badbfd4728f955bda9fcb0eece56c68befb6d6871ad9773fdafcb3
3
+ size 1465
last-checkpoint/trainer_state.json ADDED
@@ -0,0 +1,256 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.13333333333333333,
6
+ "eval_steps": 300,
7
+ "global_step": 300,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0044444444444444444,
14
+ "grad_norm": 240.13470458984375,
15
+ "learning_rate": 2.7272727272727273e-05,
16
+ "loss": 11.9773,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.008888888888888889,
21
+ "grad_norm": 34.34966278076172,
22
+ "learning_rate": 7.272727272727273e-05,
23
+ "loss": 9.6995,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.013333333333333334,
28
+ "grad_norm": 22.52939224243164,
29
+ "learning_rate": 0.0001181818181818182,
30
+ "loss": 6.6389,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.017777777777777778,
35
+ "grad_norm": 19.9622745513916,
36
+ "learning_rate": 0.00016363636363636366,
37
+ "loss": 3.7452,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.022222222222222223,
42
+ "grad_norm": 16.391326904296875,
43
+ "learning_rate": 0.0001999995943808681,
44
+ "loss": 3.7348,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.02666666666666667,
49
+ "grad_norm": 13.753044128417969,
50
+ "learning_rate": 0.00019998539805675495,
51
+ "loss": 3.3315,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.03111111111111111,
56
+ "grad_norm": 12.171226501464844,
57
+ "learning_rate": 0.00019995092406646422,
58
+ "loss": 3.2473,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.035555555555555556,
63
+ "grad_norm": 14.022675514221191,
64
+ "learning_rate": 0.00019989617940153752,
65
+ "loss": 3.1642,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.04,
70
+ "grad_norm": 10.995219230651855,
71
+ "learning_rate": 0.0001998211751645364,
72
+ "loss": 2.9791,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.044444444444444446,
77
+ "grad_norm": 13.217508316040039,
78
+ "learning_rate": 0.00019972592656679084,
79
+ "loss": 3.0747,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.04888888888888889,
84
+ "grad_norm": 13.215702056884766,
85
+ "learning_rate": 0.0001996104529253142,
86
+ "loss": 2.8731,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.05333333333333334,
91
+ "grad_norm": 12.901390075683594,
92
+ "learning_rate": 0.00019947477765888558,
93
+ "loss": 2.6857,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.057777777777777775,
98
+ "grad_norm": 10.362764358520508,
99
+ "learning_rate": 0.00019931892828330047,
100
+ "loss": 2.6396,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.06222222222222222,
105
+ "grad_norm": 17.187591552734375,
106
+ "learning_rate": 0.00019914293640579023,
107
+ "loss": 2.4217,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.06666666666666667,
112
+ "grad_norm": 11.482383728027344,
113
+ "learning_rate": 0.00019894683771861206,
114
+ "loss": 2.2934,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.07111111111111111,
119
+ "grad_norm": 9.790170669555664,
120
+ "learning_rate": 0.00019873067199181033,
121
+ "loss": 2.372,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.07555555555555556,
126
+ "grad_norm": 12.26194953918457,
127
+ "learning_rate": 0.00019849448306515094,
128
+ "loss": 2.2495,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.08,
133
+ "grad_norm": 10.313677787780762,
134
+ "learning_rate": 0.00019823831883923046,
135
+ "loss": 2.2395,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.08444444444444445,
140
+ "grad_norm": 12.778143882751465,
141
+ "learning_rate": 0.00019796223126576138,
142
+ "loss": 2.2095,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.08888888888888889,
147
+ "grad_norm": 10.047595024108887,
148
+ "learning_rate": 0.00019766627633703622,
149
+ "loss": 2.1296,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.09333333333333334,
154
+ "grad_norm": 13.92696475982666,
155
+ "learning_rate": 0.00019735051407457176,
156
+ "loss": 2.0027,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.09777777777777778,
161
+ "grad_norm": 13.361489295959473,
162
+ "learning_rate": 0.00019701500851693635,
163
+ "loss": 1.8503,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.10222222222222223,
168
+ "grad_norm": 11.596494674682617,
169
+ "learning_rate": 0.00019665982770676255,
170
+ "loss": 2.0362,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.10666666666666667,
175
+ "grad_norm": 15.200052261352539,
176
+ "learning_rate": 0.00019628504367694754,
177
+ "loss": 1.9158,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.1111111111111111,
182
+ "grad_norm": 14.349342346191406,
183
+ "learning_rate": 0.00019589073243604454,
184
+ "loss": 1.8403,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.11555555555555555,
189
+ "grad_norm": 14.814216613769531,
190
+ "learning_rate": 0.00019547697395284766,
191
+ "loss": 1.9161,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.12,
196
+ "grad_norm": 8.870625495910645,
197
+ "learning_rate": 0.0001950438521401738,
198
+ "loss": 1.7764,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.12444444444444444,
203
+ "grad_norm": 10.462891578674316,
204
+ "learning_rate": 0.0001945914548378446,
205
+ "loss": 1.9073,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.1288888888888889,
210
+ "grad_norm": 10.39869499206543,
211
+ "learning_rate": 0.00019411987379487196,
212
+ "loss": 1.7828,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.13333333333333333,
217
+ "grad_norm": 10.365347862243652,
218
+ "learning_rate": 0.0001936292046508508,
219
+ "loss": 1.7332,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.13333333333333333,
224
+ "eval_code_acc": 0.0,
225
+ "eval_exact_match": 0.0,
226
+ "eval_family_acc": 0.0,
227
+ "eval_loss": 1.4483628273010254,
228
+ "eval_runtime": 275.4346,
229
+ "eval_samples_per_second": 1.815,
230
+ "eval_steps_per_second": 1.815,
231
+ "eval_super_acc": 0.0,
232
+ "step": 300
233
+ }
234
+ ],
235
+ "logging_steps": 10,
236
+ "max_steps": 2250,
237
+ "num_input_tokens_seen": 0,
238
+ "num_train_epochs": 1,
239
+ "save_steps": 300,
240
+ "stateful_callbacks": {
241
+ "TrainerControl": {
242
+ "args": {
243
+ "should_epoch_stop": false,
244
+ "should_evaluate": false,
245
+ "should_log": false,
246
+ "should_save": true,
247
+ "should_training_stop": false
248
+ },
249
+ "attributes": {}
250
+ }
251
+ },
252
+ "total_flos": 8.183765532672e+18,
253
+ "train_batch_size": 4,
254
+ "trial_name": null,
255
+ "trial_params": null
256
+ }
last-checkpoint/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1391ef8e316083e07c8d8583f470718cb59500095cf3cf7ff701fcd881f0dd7b
3
+ size 5969