Hugging Face's logo

Gizachew
/
hubert-agum960-amharic

Transformers
TensorBoard
Safetensors
hubert
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Gizachew commited on
Commit
8a2672c
•
1 Parent(s): c9bef9a

End of training

Browse files
Files changed (10) hide show
  1. README.md +69 -0
  2. all_results.json +9 -0
  3. config.json +93 -0
  4. model.safetensors +3 -0
  5. preprocessor_config.json +9 -0
  6. runs/Jun15_12-00-10_4255ba7a45a2/events.out.tfevents.1718452832.4255ba7a45a2.35.0 +3 -0
  7. runs/Jun15_12-02-40_4255ba7a45a2/events.out.tfevents.1718452968.4255ba7a45a2.35.1 +3 -0
  8. train_results.json +9 -0
  9. trainer_state.json +447 -0
  10. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/hubert-large-ls960-ft
4
+ tags:
5
+ - generated_from_trainer
6
+ metrics:
7
+ - accuracy
8
+ model-index:
9
+ - name: hubert-agum960-amharic
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # hubert-agum960-amharic
17
+
18
+ This model is a fine-tuned version of [facebook/hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft) on the None dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 0.4469
21
+ - Accuracy: 0.9260
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 1e-05
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 4
43
+ - seed: 42
44
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
+ - lr_scheduler_type: linear
46
+ - num_epochs: 5
47
+ - mixed_precision_training: Native AMP
48
+
49
+ ### Training results
50
+
51
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
52
+ |:-------------:|:------:|:----:|:---------------:|:--------:|
53
+ | 1.0337 | 0.1664 | 500 | 0.9555 | 0.6861 |
54
+ | 0.5792 | 0.3328 | 1000 | 0.6325 | 0.8535 |
55
+ | 0.3656 | 0.4992 | 1500 | 0.5913 | 0.8789 |
56
+ | 0.33 | 0.6656 | 2000 | 0.4296 | 0.9118 |
57
+ | 0.2212 | 0.8319 | 2500 | 0.4698 | 0.9155 |
58
+ | 0.364 | 0.9983 | 3000 | 0.4107 | 0.9133 |
59
+ | 0.2371 | 1.1647 | 3500 | 0.5081 | 0.9118 |
60
+ | 0.3086 | 1.3311 | 4000 | 0.4347 | 0.9275 |
61
+ | 0.1531 | 1.4975 | 4500 | 0.4469 | 0.9260 |
62
+
63
+
64
+ ### Framework versions
65
+
66
+ - Transformers 4.41.2
67
+ - Pytorch 2.1.2
68
+ - Datasets 2.19.2
69
+ - Tokenizers 0.19.1
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.497504159733777,
3
+ "total_flos": 2.060256870141264e+18,
4
+ "train_loss": 0.4611078107621935,
5
+ "train_runtime": 3912.8249,
6
+ "train_samples": 12018,
7
+ "train_samples_per_second": 15.357,
8
+ "train_steps_per_second": 3.84
9
+ }
config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/hubert-large-ls960-ft",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": true,
5
+ "architectures": [
6
+ "HubertForSpeechClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
10
+ "classifier_proj_size": 256,
11
+ "conv_bias": true,
12
+ "conv_dim": [
13
+ 512,
14
+ 512,
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512
20
+ ],
21
+ "conv_kernel": [
22
+ 10,
23
+ 3,
24
+ 3,
25
+ 3,
26
+ 3,
27
+ 2,
28
+ 2
29
+ ],
30
+ "conv_stride": [
31
+ 5,
32
+ 2,
33
+ 2,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2
38
+ ],
39
+ "ctc_loss_reduction": "sum",
40
+ "ctc_zero_infinity": false,
41
+ "diversity_loss_weight": 0.1,
42
+ "do_stable_layer_norm": true,
43
+ "eos_token_id": 2,
44
+ "feat_extract_activation": "gelu",
45
+ "feat_extract_dropout": 0.0,
46
+ "feat_extract_norm": "layer",
47
+ "feat_proj_dropout": 0.1,
48
+ "feat_proj_layer_norm": true,
49
+ "final_dropout": 0.1,
50
+ "finetuning_task": "hubert_clf",
51
+ "gradient_checkpointing": false,
52
+ "hidden_act": "gelu",
53
+ "hidden_dropout": 0.1,
54
+ "hidden_dropout_prob": 0.1,
55
+ "hidden_size": 1024,
56
+ "id2label": {
57
+ "0": "01Neutral",
58
+ "1": "02Fearful",
59
+ "2": "03Happy",
60
+ "3": "04Sad",
61
+ "4": "05Angry"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 4096,
65
+ "label2id": {
66
+ "01Neutral": 0,
67
+ "02Fearful": 1,
68
+ "03Happy": 2,
69
+ "04Sad": 3,
70
+ "05Angry": 4
71
+ },
72
+ "layer_norm_eps": 1e-05,
73
+ "layerdrop": 0.1,
74
+ "mask_feature_length": 10,
75
+ "mask_feature_min_masks": 0,
76
+ "mask_feature_prob": 0.0,
77
+ "mask_time_length": 10,
78
+ "mask_time_min_masks": 2,
79
+ "mask_time_prob": 0.05,
80
+ "model_type": "hubert",
81
+ "num_attention_heads": 16,
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
+ "pad_token_id": 0,
87
+ "pooling_mode": "mean",
88
+ "problem_type": "single_label_classification",
89
+ "torch_dtype": "float32",
90
+ "transformers_version": "4.41.2",
91
+ "use_weighted_layer_sum": false,
92
+ "vocab_size": 32
93
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e4e2f75ffdee20cd0081352b3a7a2b9bfda1c2a1f2eec4dccb4cc5b3327c44a
3
+ size 1266025756
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
runs/Jun15_12-00-10_4255ba7a45a2/events.out.tfevents.1718452832.4255ba7a45a2.35.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30c75af62866c501d31b85188040b9615f1a2290240953a459bf47dfb0ae6fd8
3
+ size 6062
runs/Jun15_12-02-40_4255ba7a45a2/events.out.tfevents.1718452968.4255ba7a45a2.35.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:442a73a6de596e35ac5b29a3c3925d37f14150a0eb2237f679bebedf534cd97c
3
+ size 18862
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.497504159733777,
3
+ "total_flos": 2.060256870141264e+18,
4
+ "train_loss": 0.4611078107621935,
5
+ "train_runtime": 3912.8249,
6
+ "train_samples": 12018,
7
+ "train_samples_per_second": 15.357,
8
+ "train_steps_per_second": 3.84
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,447 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.4106574058532715,
3
+ "best_model_checkpoint": "/kaggle/working/hubert-agum960-amharic/checkpoint-3000",
4
+ "epoch": 1.497504159733777,
5
+ "eval_steps": 500,
6
+ "global_step": 4500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.033277870216306155,
13
+ "grad_norm": 3.5415408611297607,
14
+ "learning_rate": 9.933444259567388e-06,
15
+ "loss": 1.6,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.06655574043261231,
20
+ "grad_norm": 3.9917523860931396,
21
+ "learning_rate": 9.866888519134776e-06,
22
+ "loss": 1.5237,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.09983361064891846,
27
+ "grad_norm": NaN,
28
+ "learning_rate": 9.80099833610649e-06,
29
+ "loss": 1.3107,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.13311148086522462,
34
+ "grad_norm": 19.04720687866211,
35
+ "learning_rate": 9.735108153078204e-06,
36
+ "loss": 1.1818,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.16638935108153077,
41
+ "grad_norm": 4.309573650360107,
42
+ "learning_rate": 9.668552412645591e-06,
43
+ "loss": 1.0337,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.16638935108153077,
48
+ "eval_accuracy": 0.6860986351966858,
49
+ "eval_loss": 0.9554809927940369,
50
+ "eval_runtime": 125.3513,
51
+ "eval_samples_per_second": 10.674,
52
+ "eval_steps_per_second": 2.672,
53
+ "step": 500
54
+ },
55
+ {
56
+ "epoch": 0.19966722129783693,
57
+ "grad_norm": 5.9266791343688965,
58
+ "learning_rate": 9.601996672212978e-06,
59
+ "loss": 0.8586,
60
+ "step": 600
61
+ },
62
+ {
63
+ "epoch": 0.23294509151414308,
64
+ "grad_norm": 25.942941665649414,
65
+ "learning_rate": 9.535440931780367e-06,
66
+ "loss": 0.7985,
67
+ "step": 700
68
+ },
69
+ {
70
+ "epoch": 0.26622296173044924,
71
+ "grad_norm": 23.755361557006836,
72
+ "learning_rate": 9.469550748752082e-06,
73
+ "loss": 0.6843,
74
+ "step": 800
75
+ },
76
+ {
77
+ "epoch": 0.2995008319467554,
78
+ "grad_norm": 5.648043155670166,
79
+ "learning_rate": 9.402995008319469e-06,
80
+ "loss": 0.5446,
81
+ "step": 900
82
+ },
83
+ {
84
+ "epoch": 0.33277870216306155,
85
+ "grad_norm": 1.2758762836456299,
86
+ "learning_rate": 9.337104825291182e-06,
87
+ "loss": 0.5792,
88
+ "step": 1000
89
+ },
90
+ {
91
+ "epoch": 0.33277870216306155,
92
+ "eval_accuracy": 0.853512704372406,
93
+ "eval_loss": 0.6325268745422363,
94
+ "eval_runtime": 121.8954,
95
+ "eval_samples_per_second": 10.977,
96
+ "eval_steps_per_second": 2.748,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.36605657237936773,
101
+ "grad_norm": 1.7034721374511719,
102
+ "learning_rate": 9.27054908485857e-06,
103
+ "loss": 0.4254,
104
+ "step": 1100
105
+ },
106
+ {
107
+ "epoch": 0.39933444259567386,
108
+ "grad_norm": 0.4905766546726227,
109
+ "learning_rate": 9.203993344425958e-06,
110
+ "loss": 0.5024,
111
+ "step": 1200
112
+ },
113
+ {
114
+ "epoch": 0.43261231281198004,
115
+ "grad_norm": 1.4882396459579468,
116
+ "learning_rate": 9.137437603993346e-06,
117
+ "loss": 0.4714,
118
+ "step": 1300
119
+ },
120
+ {
121
+ "epoch": 0.46589018302828616,
122
+ "grad_norm": 0.21027016639709473,
123
+ "learning_rate": 9.070881863560733e-06,
124
+ "loss": 0.4184,
125
+ "step": 1400
126
+ },
127
+ {
128
+ "epoch": 0.49916805324459235,
129
+ "grad_norm": 0.40890273451805115,
130
+ "learning_rate": 9.004326123128122e-06,
131
+ "loss": 0.3656,
132
+ "step": 1500
133
+ },
134
+ {
135
+ "epoch": 0.49916805324459235,
136
+ "eval_accuracy": 0.878923773765564,
137
+ "eval_loss": 0.5912803411483765,
138
+ "eval_runtime": 121.7902,
139
+ "eval_samples_per_second": 10.986,
140
+ "eval_steps_per_second": 2.751,
141
+ "step": 1500
142
+ },
143
+ {
144
+ "epoch": 0.5324459234608985,
145
+ "grad_norm": 27.935806274414062,
146
+ "learning_rate": 8.937770382695509e-06,
147
+ "loss": 0.4205,
148
+ "step": 1600
149
+ },
150
+ {
151
+ "epoch": 0.5657237936772047,
152
+ "grad_norm": 13.320003509521484,
153
+ "learning_rate": 8.871880199667222e-06,
154
+ "loss": 0.3586,
155
+ "step": 1700
156
+ },
157
+ {
158
+ "epoch": 0.5990016638935108,
159
+ "grad_norm": 1.6185659170150757,
160
+ "learning_rate": 8.80532445923461e-06,
161
+ "loss": 0.3782,
162
+ "step": 1800
163
+ },
164
+ {
165
+ "epoch": 0.632279534109817,
166
+ "grad_norm": 1.0840280055999756,
167
+ "learning_rate": 8.738768718801997e-06,
168
+ "loss": 0.3585,
169
+ "step": 1900
170
+ },
171
+ {
172
+ "epoch": 0.6655574043261231,
173
+ "grad_norm": 0.10571889579296112,
174
+ "learning_rate": 8.672212978369386e-06,
175
+ "loss": 0.33,
176
+ "step": 2000
177
+ },
178
+ {
179
+ "epoch": 0.6655574043261231,
180
+ "eval_accuracy": 0.9118086695671082,
181
+ "eval_loss": 0.4296289384365082,
182
+ "eval_runtime": 121.9384,
183
+ "eval_samples_per_second": 10.973,
184
+ "eval_steps_per_second": 2.747,
185
+ "step": 2000
186
+ },
187
+ {
188
+ "epoch": 0.6988352745424293,
189
+ "grad_norm": 12.672430038452148,
190
+ "learning_rate": 8.605657237936773e-06,
191
+ "loss": 0.4664,
192
+ "step": 2100
193
+ },
194
+ {
195
+ "epoch": 0.7321131447587355,
196
+ "grad_norm": 0.23700322210788727,
197
+ "learning_rate": 8.539101497504162e-06,
198
+ "loss": 0.364,
199
+ "step": 2200
200
+ },
201
+ {
202
+ "epoch": 0.7653910149750416,
203
+ "grad_norm": 0.07070857286453247,
204
+ "learning_rate": 8.472545757071549e-06,
205
+ "loss": 0.3883,
206
+ "step": 2300
207
+ },
208
+ {
209
+ "epoch": 0.7986688851913477,
210
+ "grad_norm": 0.022693803533911705,
211
+ "learning_rate": 8.405990016638937e-06,
212
+ "loss": 0.365,
213
+ "step": 2400
214
+ },
215
+ {
216
+ "epoch": 0.831946755407654,
217
+ "grad_norm": 46.31725311279297,
218
+ "learning_rate": 8.339434276206322e-06,
219
+ "loss": 0.2212,
220
+ "step": 2500
221
+ },
222
+ {
223
+ "epoch": 0.831946755407654,
224
+ "eval_accuracy": 0.9155455827713013,
225
+ "eval_loss": 0.4698057472705841,
226
+ "eval_runtime": 122.1469,
227
+ "eval_samples_per_second": 10.954,
228
+ "eval_steps_per_second": 2.743,
229
+ "step": 2500
230
+ },
231
+ {
232
+ "epoch": 0.8652246256239601,
233
+ "grad_norm": 0.03843013569712639,
234
+ "learning_rate": 8.272878535773711e-06,
235
+ "loss": 0.2667,
236
+ "step": 2600
237
+ },
238
+ {
239
+ "epoch": 0.8985024958402662,
240
+ "grad_norm": 0.8355897068977356,
241
+ "learning_rate": 8.206322795341098e-06,
242
+ "loss": 0.2647,
243
+ "step": 2700
244
+ },
245
+ {
246
+ "epoch": 0.9317803660565723,
247
+ "grad_norm": 71.40643310546875,
248
+ "learning_rate": 8.139767054908487e-06,
249
+ "loss": 0.2689,
250
+ "step": 2800
251
+ },
252
+ {
253
+ "epoch": 0.9650582362728786,
254
+ "grad_norm": 0.15802563726902008,
255
+ "learning_rate": 8.073211314475874e-06,
256
+ "loss": 0.2999,
257
+ "step": 2900
258
+ },
259
+ {
260
+ "epoch": 0.9983361064891847,
261
+ "grad_norm": 0.03220776841044426,
262
+ "learning_rate": 8.006655574043262e-06,
263
+ "loss": 0.364,
264
+ "step": 3000
265
+ },
266
+ {
267
+ "epoch": 0.9983361064891847,
268
+ "eval_accuracy": 0.9133034348487854,
269
+ "eval_loss": 0.4106574058532715,
270
+ "eval_runtime": 121.8932,
271
+ "eval_samples_per_second": 10.977,
272
+ "eval_steps_per_second": 2.748,
273
+ "step": 3000
274
+ },
275
+ {
276
+ "epoch": 1.0316139767054908,
277
+ "grad_norm": 14.846766471862793,
278
+ "learning_rate": 7.940099833610649e-06,
279
+ "loss": 0.2743,
280
+ "step": 3100
281
+ },
282
+ {
283
+ "epoch": 1.064891846921797,
284
+ "grad_norm": 5.4475579261779785,
285
+ "learning_rate": 7.873544093178038e-06,
286
+ "loss": 0.2405,
287
+ "step": 3200
288
+ },
289
+ {
290
+ "epoch": 1.098169717138103,
291
+ "grad_norm": 0.051481593400239944,
292
+ "learning_rate": 7.806988352745425e-06,
293
+ "loss": 0.2283,
294
+ "step": 3300
295
+ },
296
+ {
297
+ "epoch": 1.1314475873544092,
298
+ "grad_norm": 0.10480177402496338,
299
+ "learning_rate": 7.740432612312813e-06,
300
+ "loss": 0.2049,
301
+ "step": 3400
302
+ },
303
+ {
304
+ "epoch": 1.1647254575707155,
305
+ "grad_norm": 0.2395573854446411,
306
+ "learning_rate": 7.6738768718802e-06,
307
+ "loss": 0.2371,
308
+ "step": 3500
309
+ },
310
+ {
311
+ "epoch": 1.1647254575707155,
312
+ "eval_accuracy": 0.9118086695671082,
313
+ "eval_loss": 0.5081421732902527,
314
+ "eval_runtime": 122.1747,
315
+ "eval_samples_per_second": 10.952,
316
+ "eval_steps_per_second": 2.742,
317
+ "step": 3500
318
+ },
319
+ {
320
+ "epoch": 1.1980033277870217,
321
+ "grad_norm": 0.08652403950691223,
322
+ "learning_rate": 7.607321131447588e-06,
323
+ "loss": 0.1947,
324
+ "step": 3600
325
+ },
326
+ {
327
+ "epoch": 1.2312811980033278,
328
+ "grad_norm": 0.01758364401757717,
329
+ "learning_rate": 7.540765391014976e-06,
330
+ "loss": 0.2981,
331
+ "step": 3700
332
+ },
333
+ {
334
+ "epoch": 1.264559068219634,
335
+ "grad_norm": 0.023226574063301086,
336
+ "learning_rate": 7.4742096505823635e-06,
337
+ "loss": 0.2315,
338
+ "step": 3800
339
+ },
340
+ {
341
+ "epoch": 1.29783693843594,
342
+ "grad_norm": 2.320756673812866,
343
+ "learning_rate": 7.407653910149751e-06,
344
+ "loss": 0.1545,
345
+ "step": 3900
346
+ },
347
+ {
348
+ "epoch": 1.3311148086522462,
349
+ "grad_norm": 0.02829299308359623,
350
+ "learning_rate": 7.341098169717139e-06,
351
+ "loss": 0.3086,
352
+ "step": 4000
353
+ },
354
+ {
355
+ "epoch": 1.3311148086522462,
356
+ "eval_accuracy": 0.927503764629364,
357
+ "eval_loss": 0.43469786643981934,
358
+ "eval_runtime": 122.0442,
359
+ "eval_samples_per_second": 10.963,
360
+ "eval_steps_per_second": 2.745,
361
+ "step": 4000
362
+ },
363
+ {
364
+ "epoch": 1.3643926788685525,
365
+ "grad_norm": 0.563818097114563,
366
+ "learning_rate": 7.274542429284527e-06,
367
+ "loss": 0.203,
368
+ "step": 4100
369
+ },
370
+ {
371
+ "epoch": 1.3976705490848587,
372
+ "grad_norm": 0.032793425023555756,
373
+ "learning_rate": 7.207986688851915e-06,
374
+ "loss": 0.2317,
375
+ "step": 4200
376
+ },
377
+ {
378
+ "epoch": 1.4309484193011648,
379
+ "grad_norm": 13.284598350524902,
380
+ "learning_rate": 7.1414309484193024e-06,
381
+ "loss": 0.1721,
382
+ "step": 4300
383
+ },
384
+ {
385
+ "epoch": 1.464226289517471,
386
+ "grad_norm": 7.5517897605896,
387
+ "learning_rate": 7.0748752079866885e-06,
388
+ "loss": 0.2042,
389
+ "step": 4400
390
+ },
391
+ {
392
+ "epoch": 1.497504159733777,
393
+ "grad_norm": 0.04595191031694412,
394
+ "learning_rate": 7.008319467554076e-06,
395
+ "loss": 0.1531,
396
+ "step": 4500
397
+ },
398
+ {
399
+ "epoch": 1.497504159733777,
400
+ "eval_accuracy": 0.926008939743042,
401
+ "eval_loss": 0.4468592703342438,
402
+ "eval_runtime": 122.6586,
403
+ "eval_samples_per_second": 10.908,
404
+ "eval_steps_per_second": 2.731,
405
+ "step": 4500
406
+ },
407
+ {
408
+ "epoch": 1.497504159733777,
409
+ "step": 4500,
410
+ "total_flos": 2.060256870141264e+18,
411
+ "train_loss": 0.4611078107621935,
412
+ "train_runtime": 3912.8249,
413
+ "train_samples_per_second": 15.357,
414
+ "train_steps_per_second": 3.84
415
+ }
416
+ ],
417
+ "logging_steps": 100,
418
+ "max_steps": 15025,
419
+ "num_input_tokens_seen": 0,
420
+ "num_train_epochs": 5,
421
+ "save_steps": 500,
422
+ "stateful_callbacks": {
423
+ "EarlyStoppingCallback": {
424
+ "args": {
425
+ "early_stopping_patience": 3,
426
+ "early_stopping_threshold": 0.0
427
+ },
428
+ "attributes": {
429
+ "early_stopping_patience_counter": 0
430
+ }
431
+ },
432
+ "TrainerControl": {
433
+ "args": {
434
+ "should_epoch_stop": false,
435
+ "should_evaluate": false,
436
+ "should_log": false,
437
+ "should_save": true,
438
+ "should_training_stop": true
439
+ },
440
+ "attributes": {}
441
+ }
442
+ },
443
+ "total_flos": 2.060256870141264e+18,
444
+ "train_batch_size": 4,
445
+ "trial_name": null,
446
+ "trial_params": null
447
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2e39203d964af082866179c6a0ff14c4c5db0bbec38e0af23ba7e07899d988c
3
+ size 5176