houssamboukhalfa commited on
Commit
0829892
·
verified ·
1 Parent(s): 4dd20f6

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. telecom_marabert_large_full_pipeline/checkpoint-306/config.json +53 -0
  2. telecom_marabert_large_full_pipeline/checkpoint-306/model.safetensors +3 -0
  3. telecom_marabert_large_full_pipeline/checkpoint-306/optimizer.pt +3 -0
  4. telecom_marabert_large_full_pipeline/checkpoint-306/rng_state.pth +3 -0
  5. telecom_marabert_large_full_pipeline/checkpoint-306/scheduler.pt +3 -0
  6. telecom_marabert_large_full_pipeline/checkpoint-306/special_tokens_map.json +37 -0
  7. telecom_marabert_large_full_pipeline/checkpoint-306/tokenizer.json +0 -0
  8. telecom_marabert_large_full_pipeline/checkpoint-306/tokenizer_config.json +58 -0
  9. telecom_marabert_large_full_pipeline/checkpoint-306/trainer_state.json +271 -0
  10. telecom_marabert_large_full_pipeline/checkpoint-306/training_args.bin +3 -0
  11. telecom_marabert_large_full_pipeline/checkpoint-306/vocab.txt +0 -0
  12. telecom_marabert_large_full_pipeline/checkpoint-510/config.json +53 -0
  13. telecom_marabert_large_full_pipeline/checkpoint-510/model.safetensors +3 -0
  14. telecom_marabert_large_full_pipeline/checkpoint-510/optimizer.pt +3 -0
  15. telecom_marabert_large_full_pipeline/checkpoint-510/rng_state.pth +3 -0
  16. telecom_marabert_large_full_pipeline/checkpoint-510/scheduler.pt +3 -0
  17. telecom_marabert_large_full_pipeline/checkpoint-510/special_tokens_map.json +37 -0
  18. telecom_marabert_large_full_pipeline/checkpoint-510/tokenizer.json +0 -0
  19. telecom_marabert_large_full_pipeline/checkpoint-510/tokenizer_config.json +58 -0
  20. telecom_marabert_large_full_pipeline/checkpoint-510/trainer_state.json +423 -0
  21. telecom_marabert_large_full_pipeline/checkpoint-510/training_args.bin +3 -0
  22. telecom_marabert_large_full_pipeline/checkpoint-510/vocab.txt +0 -0
  23. telecom_marabert_large_full_pipeline/checkpoint-561/config.json +53 -0
  24. telecom_marabert_large_full_pipeline/checkpoint-561/model.safetensors +3 -0
  25. telecom_marabert_large_full_pipeline/checkpoint-561/optimizer.pt +3 -0
  26. telecom_marabert_large_full_pipeline/checkpoint-561/rng_state.pth +3 -0
  27. telecom_marabert_large_full_pipeline/checkpoint-561/scheduler.pt +3 -0
  28. telecom_marabert_large_full_pipeline/checkpoint-561/special_tokens_map.json +37 -0
  29. telecom_marabert_large_full_pipeline/checkpoint-561/tokenizer.json +0 -0
  30. telecom_marabert_large_full_pipeline/checkpoint-561/tokenizer_config.json +58 -0
  31. telecom_marabert_large_full_pipeline/checkpoint-561/trainer_state.json +461 -0
  32. telecom_marabert_large_full_pipeline/checkpoint-561/training_args.bin +3 -0
  33. telecom_marabert_large_full_pipeline/checkpoint-561/vocab.txt +0 -0
  34. telecom_marabert_large_full_pipeline/config.json +55 -0
  35. telecom_marabert_large_full_pipeline/model.safetensors +3 -0
  36. telecom_marabert_large_full_pipeline/special_tokens_map.json +37 -0
  37. telecom_marabert_large_full_pipeline/tokenizer.json +0 -0
  38. telecom_marabert_large_full_pipeline/tokenizer_config.json +58 -0
  39. telecom_marabert_large_full_pipeline/training_args.bin +3 -0
  40. telecom_marabert_large_full_pipeline/vocab.txt +0 -0
  41. telecom_marbertv2_cpt_ft/checkpoint-3500/config.json +54 -0
  42. telecom_marbertv2_cpt_ft/checkpoint-3500/model.safetensors +3 -0
  43. telecom_marbertv2_cpt_ft/checkpoint-3500/optimizer.pt +3 -0
  44. telecom_marbertv2_cpt_ft/checkpoint-3500/rng_state.pth +3 -0
  45. telecom_marbertv2_cpt_ft/checkpoint-3500/scheduler.pt +3 -0
  46. telecom_marbertv2_cpt_ft/checkpoint-3500/special_tokens_map.json +37 -0
  47. telecom_marbertv2_cpt_ft/checkpoint-3500/tokenizer.json +0 -0
  48. telecom_marbertv2_cpt_ft/checkpoint-3500/tokenizer_config.json +58 -0
  49. telecom_marbertv2_cpt_ft/checkpoint-3500/trainer_state.json +524 -0
  50. telecom_marbertv2_cpt_ft/checkpoint-3500/training_args.bin +3 -0
telecom_marabert_large_full_pipeline/checkpoint-306/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": 1,
15
+ "1": 2,
16
+ "2": 3,
17
+ "3": 4,
18
+ "4": 5,
19
+ "5": 6,
20
+ "6": 7,
21
+ "7": 8,
22
+ "8": 9
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "label2id": {
27
+ "1": 0,
28
+ "2": 1,
29
+ "3": 2,
30
+ "4": 3,
31
+ "5": 4,
32
+ "6": 5,
33
+ "7": 6,
34
+ "8": 7,
35
+ "9": 8
36
+ },
37
+ "layer_norm_eps": 1e-12,
38
+ "max_position_embeddings": 512,
39
+ "model_type": "bert",
40
+ "num_attention_heads": 12,
41
+ "num_hidden_layers": 12,
42
+ "pad_token_id": 0,
43
+ "pooler_fc_size": 768,
44
+ "pooler_num_attention_heads": 12,
45
+ "pooler_num_fc_layers": 3,
46
+ "pooler_size_per_head": 128,
47
+ "pooler_type": "first_token_transform",
48
+ "position_embedding_type": "absolute",
49
+ "transformers_version": "4.57.3",
50
+ "type_vocab_size": 2,
51
+ "use_cache": true,
52
+ "vocab_size": 100000
53
+ }
telecom_marabert_large_full_pipeline/checkpoint-306/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3bf93c2e7faae9340d928e41440193d5bb83b0ea4d4d8db23f0665daa351c9a
3
+ size 651416604
telecom_marabert_large_full_pipeline/checkpoint-306/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c88585a880caceeb42539d447fe0e0d8692e38284d60d47b5fec899949f109df
3
+ size 1302957451
telecom_marabert_large_full_pipeline/checkpoint-306/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e92e2dcc5d295c21cc9aa0879f8086bf662538a036a133a73c5d00d8d81bcb8
3
+ size 14645
telecom_marabert_large_full_pipeline/checkpoint-306/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa09c32451d4afc38682702ea79f3548b6c08816feab91a46088a926b1d8f2b0
3
+ size 1465
telecom_marabert_large_full_pipeline/checkpoint-306/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
telecom_marabert_large_full_pipeline/checkpoint-306/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/checkpoint-306/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "4": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
telecom_marabert_large_full_pipeline/checkpoint-306/trainer_state.json ADDED
@@ -0,0 +1,271 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 306,
3
+ "best_metric": 0.9036424022837064,
4
+ "best_model_checkpoint": "./telecom_marabert_large_full_pipeline/checkpoint-306",
5
+ "epoch": 6.0,
6
+ "eval_steps": 500,
7
+ "global_step": 306,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.49504950495049505,
14
+ "grad_norm": 1.8976224660873413,
15
+ "learning_rate": 1.5686274509803923e-06,
16
+ "loss": 2.1865,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.9900990099009901,
21
+ "grad_norm": 1.939752459526062,
22
+ "learning_rate": 3.2026143790849674e-06,
23
+ "loss": 2.1373,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 1.0,
28
+ "eval_accuracy": 0.4222222222222222,
29
+ "eval_f1_class_1": 0.8888888888888888,
30
+ "eval_f1_class_2": 0.2222222222222222,
31
+ "eval_f1_class_3": 0.0,
32
+ "eval_f1_class_4": 0.0,
33
+ "eval_f1_class_5": 0.0,
34
+ "eval_f1_class_6": 0.0,
35
+ "eval_f1_class_7": 0.9565217391304348,
36
+ "eval_f1_class_8": 0.43137254901960786,
37
+ "eval_f1_class_9": 0.38016528925619836,
38
+ "eval_f1_macro": 0.3199078542797058,
39
+ "eval_f1_weighted": 0.3570860377261989,
40
+ "eval_loss": 2.050368070602417,
41
+ "eval_precision_macro": 0.32134458721760306,
42
+ "eval_precision_weighted": 0.3565318720080625,
43
+ "eval_recall_macro": 0.375,
44
+ "eval_recall_weighted": 0.4222222222222222,
45
+ "eval_runtime": 0.4171,
46
+ "eval_samples_per_second": 431.574,
47
+ "eval_steps_per_second": 14.386,
48
+ "step": 51
49
+ },
50
+ {
51
+ "epoch": 1.4752475247524752,
52
+ "grad_norm": 2.7972664833068848,
53
+ "learning_rate": 4.836601307189543e-06,
54
+ "loss": 2.0164,
55
+ "step": 75
56
+ },
57
+ {
58
+ "epoch": 1.9702970297029703,
59
+ "grad_norm": 2.76436185836792,
60
+ "learning_rate": 6.470588235294119e-06,
61
+ "loss": 1.8001,
62
+ "step": 100
63
+ },
64
+ {
65
+ "epoch": 2.0,
66
+ "eval_accuracy": 0.8777777777777778,
67
+ "eval_f1_class_1": 1.0,
68
+ "eval_f1_class_2": 0.7567567567567568,
69
+ "eval_f1_class_3": 0.972972972972973,
70
+ "eval_f1_class_4": 0.8837209302325582,
71
+ "eval_f1_class_5": 0.8888888888888888,
72
+ "eval_f1_class_6": 0.6428571428571429,
73
+ "eval_f1_class_7": 0.96,
74
+ "eval_f1_class_8": 0.7755102040816326,
75
+ "eval_f1_class_9": 0.9387755102040817,
76
+ "eval_f1_macro": 0.8688313784437816,
77
+ "eval_f1_weighted": 0.8723569380169602,
78
+ "eval_loss": 1.5676754713058472,
79
+ "eval_precision_macro": 0.8769592612559364,
80
+ "eval_precision_weighted": 0.876579307875983,
81
+ "eval_recall_macro": 0.8712804195002337,
82
+ "eval_recall_weighted": 0.8777777777777778,
83
+ "eval_runtime": 0.3929,
84
+ "eval_samples_per_second": 458.187,
85
+ "eval_steps_per_second": 15.273,
86
+ "step": 102
87
+ },
88
+ {
89
+ "epoch": 2.4554455445544554,
90
+ "grad_norm": 3.865165948867798,
91
+ "learning_rate": 8.104575163398693e-06,
92
+ "loss": 1.5567,
93
+ "step": 125
94
+ },
95
+ {
96
+ "epoch": 2.9504950495049505,
97
+ "grad_norm": 3.0548179149627686,
98
+ "learning_rate": 9.738562091503268e-06,
99
+ "loss": 1.3614,
100
+ "step": 150
101
+ },
102
+ {
103
+ "epoch": 3.0,
104
+ "eval_accuracy": 0.9055555555555556,
105
+ "eval_f1_class_1": 1.0,
106
+ "eval_f1_class_2": 0.8292682926829268,
107
+ "eval_f1_class_3": 0.972972972972973,
108
+ "eval_f1_class_4": 0.926829268292683,
109
+ "eval_f1_class_5": 0.8888888888888888,
110
+ "eval_f1_class_6": 0.6896551724137931,
111
+ "eval_f1_class_7": 0.96,
112
+ "eval_f1_class_8": 0.851063829787234,
113
+ "eval_f1_class_9": 0.9583333333333334,
114
+ "eval_f1_macro": 0.8974457509302035,
115
+ "eval_f1_weighted": 0.9021962765230499,
116
+ "eval_loss": 1.1128010749816895,
117
+ "eval_precision_macro": 0.9019337828033481,
118
+ "eval_precision_weighted": 0.9043258303693086,
119
+ "eval_recall_macro": 0.8991126635089484,
120
+ "eval_recall_weighted": 0.9055555555555556,
121
+ "eval_runtime": 0.3913,
122
+ "eval_samples_per_second": 460.034,
123
+ "eval_steps_per_second": 15.334,
124
+ "step": 153
125
+ },
126
+ {
127
+ "epoch": 3.4356435643564356,
128
+ "grad_norm": 5.272692680358887,
129
+ "learning_rate": 9.847494553376908e-06,
130
+ "loss": 1.1806,
131
+ "step": 175
132
+ },
133
+ {
134
+ "epoch": 3.9306930693069306,
135
+ "grad_norm": 7.215332984924316,
136
+ "learning_rate": 9.665940450254176e-06,
137
+ "loss": 1.0304,
138
+ "step": 200
139
+ },
140
+ {
141
+ "epoch": 4.0,
142
+ "eval_accuracy": 0.9,
143
+ "eval_f1_class_1": 1.0,
144
+ "eval_f1_class_2": 0.7804878048780488,
145
+ "eval_f1_class_3": 1.0,
146
+ "eval_f1_class_4": 0.926829268292683,
147
+ "eval_f1_class_5": 0.88,
148
+ "eval_f1_class_6": 0.64,
149
+ "eval_f1_class_7": 0.9795918367346939,
150
+ "eval_f1_class_8": 0.8301886792452831,
151
+ "eval_f1_class_9": 0.9583333333333334,
152
+ "eval_f1_macro": 0.8883812136093379,
153
+ "eval_f1_weighted": 0.8941263361866708,
154
+ "eval_loss": 0.8936125636100769,
155
+ "eval_precision_macro": 0.9135504469987228,
156
+ "eval_precision_weighted": 0.9092049808429119,
157
+ "eval_recall_macro": 0.8870454164571812,
158
+ "eval_recall_weighted": 0.9,
159
+ "eval_runtime": 0.3826,
160
+ "eval_samples_per_second": 470.424,
161
+ "eval_steps_per_second": 15.681,
162
+ "step": 204
163
+ },
164
+ {
165
+ "epoch": 4.415841584158416,
166
+ "grad_norm": 4.240860939025879,
167
+ "learning_rate": 9.484386347131446e-06,
168
+ "loss": 0.926,
169
+ "step": 225
170
+ },
171
+ {
172
+ "epoch": 4.910891089108911,
173
+ "grad_norm": 7.547500133514404,
174
+ "learning_rate": 9.302832244008714e-06,
175
+ "loss": 0.8455,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 5.0,
180
+ "eval_accuracy": 0.8944444444444445,
181
+ "eval_f1_class_1": 0.975609756097561,
182
+ "eval_f1_class_2": 0.7619047619047619,
183
+ "eval_f1_class_3": 0.9444444444444444,
184
+ "eval_f1_class_4": 0.9,
185
+ "eval_f1_class_5": 0.8888888888888888,
186
+ "eval_f1_class_6": 0.7142857142857143,
187
+ "eval_f1_class_7": 0.96,
188
+ "eval_f1_class_8": 0.875,
189
+ "eval_f1_class_9": 0.9583333333333334,
190
+ "eval_f1_macro": 0.8864963221060782,
191
+ "eval_f1_weighted": 0.8915267453865015,
192
+ "eval_loss": 0.7933213710784912,
193
+ "eval_precision_macro": 0.895996040996041,
194
+ "eval_precision_weighted": 0.8972462352462351,
195
+ "eval_recall_macro": 0.8867832288110926,
196
+ "eval_recall_weighted": 0.8944444444444445,
197
+ "eval_runtime": 0.377,
198
+ "eval_samples_per_second": 477.469,
199
+ "eval_steps_per_second": 15.916,
200
+ "step": 255
201
+ },
202
+ {
203
+ "epoch": 5.396039603960396,
204
+ "grad_norm": 6.939261436462402,
205
+ "learning_rate": 9.121278140885986e-06,
206
+ "loss": 0.7949,
207
+ "step": 275
208
+ },
209
+ {
210
+ "epoch": 5.891089108910891,
211
+ "grad_norm": 4.395618915557861,
212
+ "learning_rate": 8.939724037763254e-06,
213
+ "loss": 0.7405,
214
+ "step": 300
215
+ },
216
+ {
217
+ "epoch": 6.0,
218
+ "eval_accuracy": 0.9055555555555556,
219
+ "eval_f1_class_1": 1.0,
220
+ "eval_f1_class_2": 0.7727272727272727,
221
+ "eval_f1_class_3": 1.0,
222
+ "eval_f1_class_4": 0.95,
223
+ "eval_f1_class_5": 0.9230769230769231,
224
+ "eval_f1_class_6": 0.6666666666666666,
225
+ "eval_f1_class_7": 1.0,
226
+ "eval_f1_class_8": 0.8260869565217391,
227
+ "eval_f1_class_9": 0.9583333333333334,
228
+ "eval_f1_macro": 0.8996545724806596,
229
+ "eval_f1_weighted": 0.9036424022837064,
230
+ "eval_loss": 0.7440095543861389,
231
+ "eval_precision_macro": 0.9038085988085989,
232
+ "eval_precision_weighted": 0.9062825692825692,
233
+ "eval_recall_macro": 0.900330987095693,
234
+ "eval_recall_weighted": 0.9055555555555556,
235
+ "eval_runtime": 0.3904,
236
+ "eval_samples_per_second": 461.066,
237
+ "eval_steps_per_second": 15.369,
238
+ "step": 306
239
+ }
240
+ ],
241
+ "logging_steps": 25,
242
+ "max_steps": 1530,
243
+ "num_input_tokens_seen": 0,
244
+ "num_train_epochs": 30,
245
+ "save_steps": 500,
246
+ "stateful_callbacks": {
247
+ "EarlyStoppingCallback": {
248
+ "args": {
249
+ "early_stopping_patience": 5,
250
+ "early_stopping_threshold": 0.001
251
+ },
252
+ "attributes": {
253
+ "early_stopping_patience_counter": 0
254
+ }
255
+ },
256
+ "TrainerControl": {
257
+ "args": {
258
+ "should_epoch_stop": false,
259
+ "should_evaluate": false,
260
+ "should_log": false,
261
+ "should_save": true,
262
+ "should_training_stop": false
263
+ },
264
+ "attributes": {}
265
+ }
266
+ },
267
+ "total_flos": 1271695651826688.0,
268
+ "train_batch_size": 16,
269
+ "trial_name": null,
270
+ "trial_params": null
271
+ }
telecom_marabert_large_full_pipeline/checkpoint-306/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f988eeda8a1fc14cbf1fe7dca7e292179f39245156399c748eeca11dad745279
3
+ size 5905
telecom_marabert_large_full_pipeline/checkpoint-306/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/checkpoint-510/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": 1,
15
+ "1": 2,
16
+ "2": 3,
17
+ "3": 4,
18
+ "4": 5,
19
+ "5": 6,
20
+ "6": 7,
21
+ "7": 8,
22
+ "8": 9
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "label2id": {
27
+ "1": 0,
28
+ "2": 1,
29
+ "3": 2,
30
+ "4": 3,
31
+ "5": 4,
32
+ "6": 5,
33
+ "7": 6,
34
+ "8": 7,
35
+ "9": 8
36
+ },
37
+ "layer_norm_eps": 1e-12,
38
+ "max_position_embeddings": 512,
39
+ "model_type": "bert",
40
+ "num_attention_heads": 12,
41
+ "num_hidden_layers": 12,
42
+ "pad_token_id": 0,
43
+ "pooler_fc_size": 768,
44
+ "pooler_num_attention_heads": 12,
45
+ "pooler_num_fc_layers": 3,
46
+ "pooler_size_per_head": 128,
47
+ "pooler_type": "first_token_transform",
48
+ "position_embedding_type": "absolute",
49
+ "transformers_version": "4.57.3",
50
+ "type_vocab_size": 2,
51
+ "use_cache": true,
52
+ "vocab_size": 100000
53
+ }
telecom_marabert_large_full_pipeline/checkpoint-510/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d7b7d1afd5fac889130f20f6431f75351392d805881f3e90bfad94e2edf6f30
3
+ size 651416604
telecom_marabert_large_full_pipeline/checkpoint-510/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5b361e1b5b6f8bf028956269943121feee160c5cc6613c7b650102f43eec5a1
3
+ size 1302957451
telecom_marabert_large_full_pipeline/checkpoint-510/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63318b574675192537283918cbdb2e0438735fa8b9850588e9d1cecc374bb37c
3
+ size 14645
telecom_marabert_large_full_pipeline/checkpoint-510/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09bab529778f79d8de23a01f2c591c9d1a9d2d6426f4d98745f818c1b69f1ca0
3
+ size 1465
telecom_marabert_large_full_pipeline/checkpoint-510/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
telecom_marabert_large_full_pipeline/checkpoint-510/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/checkpoint-510/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "4": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
telecom_marabert_large_full_pipeline/checkpoint-510/trainer_state.json ADDED
@@ -0,0 +1,423 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 306,
3
+ "best_metric": 0.9036424022837064,
4
+ "best_model_checkpoint": "./telecom_marabert_large_full_pipeline/checkpoint-306",
5
+ "epoch": 10.0,
6
+ "eval_steps": 500,
7
+ "global_step": 510,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.49504950495049505,
14
+ "grad_norm": 1.8976224660873413,
15
+ "learning_rate": 1.5686274509803923e-06,
16
+ "loss": 2.1865,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.9900990099009901,
21
+ "grad_norm": 1.939752459526062,
22
+ "learning_rate": 3.2026143790849674e-06,
23
+ "loss": 2.1373,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 1.0,
28
+ "eval_accuracy": 0.4222222222222222,
29
+ "eval_f1_class_1": 0.8888888888888888,
30
+ "eval_f1_class_2": 0.2222222222222222,
31
+ "eval_f1_class_3": 0.0,
32
+ "eval_f1_class_4": 0.0,
33
+ "eval_f1_class_5": 0.0,
34
+ "eval_f1_class_6": 0.0,
35
+ "eval_f1_class_7": 0.9565217391304348,
36
+ "eval_f1_class_8": 0.43137254901960786,
37
+ "eval_f1_class_9": 0.38016528925619836,
38
+ "eval_f1_macro": 0.3199078542797058,
39
+ "eval_f1_weighted": 0.3570860377261989,
40
+ "eval_loss": 2.050368070602417,
41
+ "eval_precision_macro": 0.32134458721760306,
42
+ "eval_precision_weighted": 0.3565318720080625,
43
+ "eval_recall_macro": 0.375,
44
+ "eval_recall_weighted": 0.4222222222222222,
45
+ "eval_runtime": 0.4171,
46
+ "eval_samples_per_second": 431.574,
47
+ "eval_steps_per_second": 14.386,
48
+ "step": 51
49
+ },
50
+ {
51
+ "epoch": 1.4752475247524752,
52
+ "grad_norm": 2.7972664833068848,
53
+ "learning_rate": 4.836601307189543e-06,
54
+ "loss": 2.0164,
55
+ "step": 75
56
+ },
57
+ {
58
+ "epoch": 1.9702970297029703,
59
+ "grad_norm": 2.76436185836792,
60
+ "learning_rate": 6.470588235294119e-06,
61
+ "loss": 1.8001,
62
+ "step": 100
63
+ },
64
+ {
65
+ "epoch": 2.0,
66
+ "eval_accuracy": 0.8777777777777778,
67
+ "eval_f1_class_1": 1.0,
68
+ "eval_f1_class_2": 0.7567567567567568,
69
+ "eval_f1_class_3": 0.972972972972973,
70
+ "eval_f1_class_4": 0.8837209302325582,
71
+ "eval_f1_class_5": 0.8888888888888888,
72
+ "eval_f1_class_6": 0.6428571428571429,
73
+ "eval_f1_class_7": 0.96,
74
+ "eval_f1_class_8": 0.7755102040816326,
75
+ "eval_f1_class_9": 0.9387755102040817,
76
+ "eval_f1_macro": 0.8688313784437816,
77
+ "eval_f1_weighted": 0.8723569380169602,
78
+ "eval_loss": 1.5676754713058472,
79
+ "eval_precision_macro": 0.8769592612559364,
80
+ "eval_precision_weighted": 0.876579307875983,
81
+ "eval_recall_macro": 0.8712804195002337,
82
+ "eval_recall_weighted": 0.8777777777777778,
83
+ "eval_runtime": 0.3929,
84
+ "eval_samples_per_second": 458.187,
85
+ "eval_steps_per_second": 15.273,
86
+ "step": 102
87
+ },
88
+ {
89
+ "epoch": 2.4554455445544554,
90
+ "grad_norm": 3.865165948867798,
91
+ "learning_rate": 8.104575163398693e-06,
92
+ "loss": 1.5567,
93
+ "step": 125
94
+ },
95
+ {
96
+ "epoch": 2.9504950495049505,
97
+ "grad_norm": 3.0548179149627686,
98
+ "learning_rate": 9.738562091503268e-06,
99
+ "loss": 1.3614,
100
+ "step": 150
101
+ },
102
+ {
103
+ "epoch": 3.0,
104
+ "eval_accuracy": 0.9055555555555556,
105
+ "eval_f1_class_1": 1.0,
106
+ "eval_f1_class_2": 0.8292682926829268,
107
+ "eval_f1_class_3": 0.972972972972973,
108
+ "eval_f1_class_4": 0.926829268292683,
109
+ "eval_f1_class_5": 0.8888888888888888,
110
+ "eval_f1_class_6": 0.6896551724137931,
111
+ "eval_f1_class_7": 0.96,
112
+ "eval_f1_class_8": 0.851063829787234,
113
+ "eval_f1_class_9": 0.9583333333333334,
114
+ "eval_f1_macro": 0.8974457509302035,
115
+ "eval_f1_weighted": 0.9021962765230499,
116
+ "eval_loss": 1.1128010749816895,
117
+ "eval_precision_macro": 0.9019337828033481,
118
+ "eval_precision_weighted": 0.9043258303693086,
119
+ "eval_recall_macro": 0.8991126635089484,
120
+ "eval_recall_weighted": 0.9055555555555556,
121
+ "eval_runtime": 0.3913,
122
+ "eval_samples_per_second": 460.034,
123
+ "eval_steps_per_second": 15.334,
124
+ "step": 153
125
+ },
126
+ {
127
+ "epoch": 3.4356435643564356,
128
+ "grad_norm": 5.272692680358887,
129
+ "learning_rate": 9.847494553376908e-06,
130
+ "loss": 1.1806,
131
+ "step": 175
132
+ },
133
+ {
134
+ "epoch": 3.9306930693069306,
135
+ "grad_norm": 7.215332984924316,
136
+ "learning_rate": 9.665940450254176e-06,
137
+ "loss": 1.0304,
138
+ "step": 200
139
+ },
140
+ {
141
+ "epoch": 4.0,
142
+ "eval_accuracy": 0.9,
143
+ "eval_f1_class_1": 1.0,
144
+ "eval_f1_class_2": 0.7804878048780488,
145
+ "eval_f1_class_3": 1.0,
146
+ "eval_f1_class_4": 0.926829268292683,
147
+ "eval_f1_class_5": 0.88,
148
+ "eval_f1_class_6": 0.64,
149
+ "eval_f1_class_7": 0.9795918367346939,
150
+ "eval_f1_class_8": 0.8301886792452831,
151
+ "eval_f1_class_9": 0.9583333333333334,
152
+ "eval_f1_macro": 0.8883812136093379,
153
+ "eval_f1_weighted": 0.8941263361866708,
154
+ "eval_loss": 0.8936125636100769,
155
+ "eval_precision_macro": 0.9135504469987228,
156
+ "eval_precision_weighted": 0.9092049808429119,
157
+ "eval_recall_macro": 0.8870454164571812,
158
+ "eval_recall_weighted": 0.9,
159
+ "eval_runtime": 0.3826,
160
+ "eval_samples_per_second": 470.424,
161
+ "eval_steps_per_second": 15.681,
162
+ "step": 204
163
+ },
164
+ {
165
+ "epoch": 4.415841584158416,
166
+ "grad_norm": 4.240860939025879,
167
+ "learning_rate": 9.484386347131446e-06,
168
+ "loss": 0.926,
169
+ "step": 225
170
+ },
171
+ {
172
+ "epoch": 4.910891089108911,
173
+ "grad_norm": 7.547500133514404,
174
+ "learning_rate": 9.302832244008714e-06,
175
+ "loss": 0.8455,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 5.0,
180
+ "eval_accuracy": 0.8944444444444445,
181
+ "eval_f1_class_1": 0.975609756097561,
182
+ "eval_f1_class_2": 0.7619047619047619,
183
+ "eval_f1_class_3": 0.9444444444444444,
184
+ "eval_f1_class_4": 0.9,
185
+ "eval_f1_class_5": 0.8888888888888888,
186
+ "eval_f1_class_6": 0.7142857142857143,
187
+ "eval_f1_class_7": 0.96,
188
+ "eval_f1_class_8": 0.875,
189
+ "eval_f1_class_9": 0.9583333333333334,
190
+ "eval_f1_macro": 0.8864963221060782,
191
+ "eval_f1_weighted": 0.8915267453865015,
192
+ "eval_loss": 0.7933213710784912,
193
+ "eval_precision_macro": 0.895996040996041,
194
+ "eval_precision_weighted": 0.8972462352462351,
195
+ "eval_recall_macro": 0.8867832288110926,
196
+ "eval_recall_weighted": 0.8944444444444445,
197
+ "eval_runtime": 0.377,
198
+ "eval_samples_per_second": 477.469,
199
+ "eval_steps_per_second": 15.916,
200
+ "step": 255
201
+ },
202
+ {
203
+ "epoch": 5.396039603960396,
204
+ "grad_norm": 6.939261436462402,
205
+ "learning_rate": 9.121278140885986e-06,
206
+ "loss": 0.7949,
207
+ "step": 275
208
+ },
209
+ {
210
+ "epoch": 5.891089108910891,
211
+ "grad_norm": 4.395618915557861,
212
+ "learning_rate": 8.939724037763254e-06,
213
+ "loss": 0.7405,
214
+ "step": 300
215
+ },
216
+ {
217
+ "epoch": 6.0,
218
+ "eval_accuracy": 0.9055555555555556,
219
+ "eval_f1_class_1": 1.0,
220
+ "eval_f1_class_2": 0.7727272727272727,
221
+ "eval_f1_class_3": 1.0,
222
+ "eval_f1_class_4": 0.95,
223
+ "eval_f1_class_5": 0.9230769230769231,
224
+ "eval_f1_class_6": 0.6666666666666666,
225
+ "eval_f1_class_7": 1.0,
226
+ "eval_f1_class_8": 0.8260869565217391,
227
+ "eval_f1_class_9": 0.9583333333333334,
228
+ "eval_f1_macro": 0.8996545724806596,
229
+ "eval_f1_weighted": 0.9036424022837064,
230
+ "eval_loss": 0.7440095543861389,
231
+ "eval_precision_macro": 0.9038085988085989,
232
+ "eval_precision_weighted": 0.9062825692825692,
233
+ "eval_recall_macro": 0.900330987095693,
234
+ "eval_recall_weighted": 0.9055555555555556,
235
+ "eval_runtime": 0.3904,
236
+ "eval_samples_per_second": 461.066,
237
+ "eval_steps_per_second": 15.369,
238
+ "step": 306
239
+ },
240
+ {
241
+ "epoch": 6.376237623762377,
242
+ "grad_norm": 5.959896087646484,
243
+ "learning_rate": 8.758169934640524e-06,
244
+ "loss": 0.7125,
245
+ "step": 325
246
+ },
247
+ {
248
+ "epoch": 6.871287128712871,
249
+ "grad_norm": 5.87659215927124,
250
+ "learning_rate": 8.576615831517793e-06,
251
+ "loss": 0.673,
252
+ "step": 350
253
+ },
254
+ {
255
+ "epoch": 7.0,
256
+ "eval_accuracy": 0.9055555555555556,
257
+ "eval_f1_class_1": 0.975609756097561,
258
+ "eval_f1_class_2": 0.7368421052631579,
259
+ "eval_f1_class_3": 0.972972972972973,
260
+ "eval_f1_class_4": 0.8888888888888888,
261
+ "eval_f1_class_5": 0.9230769230769231,
262
+ "eval_f1_class_6": 0.7407407407407407,
263
+ "eval_f1_class_7": 1.0,
264
+ "eval_f1_class_8": 0.88,
265
+ "eval_f1_class_9": 0.9583333333333334,
266
+ "eval_f1_macro": 0.8973849689303974,
267
+ "eval_f1_weighted": 0.9014864486152104,
268
+ "eval_loss": 0.7316537499427795,
269
+ "eval_precision_macro": 0.9132654999321667,
270
+ "eval_precision_weighted": 0.9115048161714827,
271
+ "eval_recall_macro": 0.8972608116570964,
272
+ "eval_recall_weighted": 0.9055555555555556,
273
+ "eval_runtime": 0.3857,
274
+ "eval_samples_per_second": 466.734,
275
+ "eval_steps_per_second": 15.558,
276
+ "step": 357
277
+ },
278
+ {
279
+ "epoch": 7.356435643564357,
280
+ "grad_norm": 5.4788947105407715,
281
+ "learning_rate": 8.395061728395062e-06,
282
+ "loss": 0.6209,
283
+ "step": 375
284
+ },
285
+ {
286
+ "epoch": 7.851485148514851,
287
+ "grad_norm": 5.614921569824219,
288
+ "learning_rate": 8.213507625272332e-06,
289
+ "loss": 0.6277,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 8.0,
294
+ "eval_accuracy": 0.8944444444444445,
295
+ "eval_f1_class_1": 0.975609756097561,
296
+ "eval_f1_class_2": 0.7142857142857143,
297
+ "eval_f1_class_3": 0.9444444444444444,
298
+ "eval_f1_class_4": 0.8837209302325582,
299
+ "eval_f1_class_5": 0.8888888888888888,
300
+ "eval_f1_class_6": 0.6923076923076923,
301
+ "eval_f1_class_7": 1.0,
302
+ "eval_f1_class_8": 0.8979591836734694,
303
+ "eval_f1_class_9": 0.9583333333333334,
304
+ "eval_f1_macro": 0.8839499936959624,
305
+ "eval_f1_weighted": 0.8907458103132077,
306
+ "eval_loss": 0.7260087132453918,
307
+ "eval_precision_macro": 0.9019365497626368,
308
+ "eval_precision_weighted": 0.9034921053181922,
309
+ "eval_recall_macro": 0.8848769107283039,
310
+ "eval_recall_weighted": 0.8944444444444445,
311
+ "eval_runtime": 0.3921,
312
+ "eval_samples_per_second": 459.098,
313
+ "eval_steps_per_second": 15.303,
314
+ "step": 408
315
+ },
316
+ {
317
+ "epoch": 8.336633663366337,
318
+ "grad_norm": 0.9043681025505066,
319
+ "learning_rate": 8.031953522149601e-06,
320
+ "loss": 0.5791,
321
+ "step": 425
322
+ },
323
+ {
324
+ "epoch": 8.831683168316832,
325
+ "grad_norm": 8.910099983215332,
326
+ "learning_rate": 7.850399419026871e-06,
327
+ "loss": 0.5777,
328
+ "step": 450
329
+ },
330
+ {
331
+ "epoch": 9.0,
332
+ "eval_accuracy": 0.9,
333
+ "eval_f1_class_1": 1.0,
334
+ "eval_f1_class_2": 0.7317073170731707,
335
+ "eval_f1_class_3": 0.972972972972973,
336
+ "eval_f1_class_4": 0.8888888888888888,
337
+ "eval_f1_class_5": 0.96,
338
+ "eval_f1_class_6": 0.7142857142857143,
339
+ "eval_f1_class_7": 1.0,
340
+ "eval_f1_class_8": 0.8571428571428571,
341
+ "eval_f1_class_9": 0.9361702127659575,
342
+ "eval_f1_macro": 0.8956853292366181,
343
+ "eval_f1_weighted": 0.8979145067423913,
344
+ "eval_loss": 0.7282315492630005,
345
+ "eval_precision_macro": 0.9088936988936989,
346
+ "eval_precision_weighted": 0.905464405964406,
347
+ "eval_recall_macro": 0.8933558197086264,
348
+ "eval_recall_weighted": 0.9,
349
+ "eval_runtime": 0.3802,
350
+ "eval_samples_per_second": 473.411,
351
+ "eval_steps_per_second": 15.78,
352
+ "step": 459
353
+ },
354
+ {
355
+ "epoch": 9.316831683168317,
356
+ "grad_norm": 1.7061387300491333,
357
+ "learning_rate": 7.66884531590414e-06,
358
+ "loss": 0.5522,
359
+ "step": 475
360
+ },
361
+ {
362
+ "epoch": 9.811881188118813,
363
+ "grad_norm": 5.173563003540039,
364
+ "learning_rate": 7.487291212781409e-06,
365
+ "loss": 0.5386,
366
+ "step": 500
367
+ },
368
+ {
369
+ "epoch": 10.0,
370
+ "eval_accuracy": 0.8888888888888888,
371
+ "eval_f1_class_1": 1.0,
372
+ "eval_f1_class_2": 0.7,
373
+ "eval_f1_class_3": 0.972972972972973,
374
+ "eval_f1_class_4": 0.8636363636363636,
375
+ "eval_f1_class_5": 0.9230769230769231,
376
+ "eval_f1_class_6": 0.7096774193548387,
377
+ "eval_f1_class_7": 1.0,
378
+ "eval_f1_class_8": 0.8260869565217391,
379
+ "eval_f1_class_9": 0.9583333333333334,
380
+ "eval_f1_macro": 0.8837537743217968,
381
+ "eval_f1_weighted": 0.887174908396858,
382
+ "eval_loss": 0.7506593465805054,
383
+ "eval_precision_macro": 0.8871215821215821,
384
+ "eval_precision_weighted": 0.8893208273208272,
385
+ "eval_recall_macro": 0.8843523149250703,
386
+ "eval_recall_weighted": 0.8888888888888888,
387
+ "eval_runtime": 0.3848,
388
+ "eval_samples_per_second": 467.763,
389
+ "eval_steps_per_second": 15.592,
390
+ "step": 510
391
+ }
392
+ ],
393
+ "logging_steps": 25,
394
+ "max_steps": 1530,
395
+ "num_input_tokens_seen": 0,
396
+ "num_train_epochs": 30,
397
+ "save_steps": 500,
398
+ "stateful_callbacks": {
399
+ "EarlyStoppingCallback": {
400
+ "args": {
401
+ "early_stopping_patience": 5,
402
+ "early_stopping_threshold": 0.001
403
+ },
404
+ "attributes": {
405
+ "early_stopping_patience_counter": 4
406
+ }
407
+ },
408
+ "TrainerControl": {
409
+ "args": {
410
+ "should_epoch_stop": false,
411
+ "should_evaluate": false,
412
+ "should_log": false,
413
+ "should_save": true,
414
+ "should_training_stop": false
415
+ },
416
+ "attributes": {}
417
+ }
418
+ },
419
+ "total_flos": 2119492753044480.0,
420
+ "train_batch_size": 16,
421
+ "trial_name": null,
422
+ "trial_params": null
423
+ }
telecom_marabert_large_full_pipeline/checkpoint-510/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f988eeda8a1fc14cbf1fe7dca7e292179f39245156399c748eeca11dad745279
3
+ size 5905
telecom_marabert_large_full_pipeline/checkpoint-510/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/checkpoint-561/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": 1,
15
+ "1": 2,
16
+ "2": 3,
17
+ "3": 4,
18
+ "4": 5,
19
+ "5": 6,
20
+ "6": 7,
21
+ "7": 8,
22
+ "8": 9
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "label2id": {
27
+ "1": 0,
28
+ "2": 1,
29
+ "3": 2,
30
+ "4": 3,
31
+ "5": 4,
32
+ "6": 5,
33
+ "7": 6,
34
+ "8": 7,
35
+ "9": 8
36
+ },
37
+ "layer_norm_eps": 1e-12,
38
+ "max_position_embeddings": 512,
39
+ "model_type": "bert",
40
+ "num_attention_heads": 12,
41
+ "num_hidden_layers": 12,
42
+ "pad_token_id": 0,
43
+ "pooler_fc_size": 768,
44
+ "pooler_num_attention_heads": 12,
45
+ "pooler_num_fc_layers": 3,
46
+ "pooler_size_per_head": 128,
47
+ "pooler_type": "first_token_transform",
48
+ "position_embedding_type": "absolute",
49
+ "transformers_version": "4.57.3",
50
+ "type_vocab_size": 2,
51
+ "use_cache": true,
52
+ "vocab_size": 100000
53
+ }
telecom_marabert_large_full_pipeline/checkpoint-561/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13b18b51bcf5c1511522462a4a035eb61913398c7442ddd9f27aca23ef95b09b
3
+ size 651416604
telecom_marabert_large_full_pipeline/checkpoint-561/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5eaf91cc9cea667ac6984583b8ab181719d147e8f9d38b226f37385ce589291e
3
+ size 1302957451
telecom_marabert_large_full_pipeline/checkpoint-561/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:390cc14c18a047ecfec6aaeeb6d72d07d8e9d88e035f384901d1df780a1453f0
3
+ size 14645
telecom_marabert_large_full_pipeline/checkpoint-561/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5d4546df06998d7696f4429bc55482f726f447814ba10911cbdc03caa31e4e
3
+ size 1465
telecom_marabert_large_full_pipeline/checkpoint-561/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
telecom_marabert_large_full_pipeline/checkpoint-561/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/checkpoint-561/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "4": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
telecom_marabert_large_full_pipeline/checkpoint-561/trainer_state.json ADDED
@@ -0,0 +1,461 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 306,
3
+ "best_metric": 0.9036424022837064,
4
+ "best_model_checkpoint": "./telecom_marabert_large_full_pipeline/checkpoint-306",
5
+ "epoch": 11.0,
6
+ "eval_steps": 500,
7
+ "global_step": 561,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.49504950495049505,
14
+ "grad_norm": 1.8976224660873413,
15
+ "learning_rate": 1.5686274509803923e-06,
16
+ "loss": 2.1865,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.9900990099009901,
21
+ "grad_norm": 1.939752459526062,
22
+ "learning_rate": 3.2026143790849674e-06,
23
+ "loss": 2.1373,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 1.0,
28
+ "eval_accuracy": 0.4222222222222222,
29
+ "eval_f1_class_1": 0.8888888888888888,
30
+ "eval_f1_class_2": 0.2222222222222222,
31
+ "eval_f1_class_3": 0.0,
32
+ "eval_f1_class_4": 0.0,
33
+ "eval_f1_class_5": 0.0,
34
+ "eval_f1_class_6": 0.0,
35
+ "eval_f1_class_7": 0.9565217391304348,
36
+ "eval_f1_class_8": 0.43137254901960786,
37
+ "eval_f1_class_9": 0.38016528925619836,
38
+ "eval_f1_macro": 0.3199078542797058,
39
+ "eval_f1_weighted": 0.3570860377261989,
40
+ "eval_loss": 2.050368070602417,
41
+ "eval_precision_macro": 0.32134458721760306,
42
+ "eval_precision_weighted": 0.3565318720080625,
43
+ "eval_recall_macro": 0.375,
44
+ "eval_recall_weighted": 0.4222222222222222,
45
+ "eval_runtime": 0.4171,
46
+ "eval_samples_per_second": 431.574,
47
+ "eval_steps_per_second": 14.386,
48
+ "step": 51
49
+ },
50
+ {
51
+ "epoch": 1.4752475247524752,
52
+ "grad_norm": 2.7972664833068848,
53
+ "learning_rate": 4.836601307189543e-06,
54
+ "loss": 2.0164,
55
+ "step": 75
56
+ },
57
+ {
58
+ "epoch": 1.9702970297029703,
59
+ "grad_norm": 2.76436185836792,
60
+ "learning_rate": 6.470588235294119e-06,
61
+ "loss": 1.8001,
62
+ "step": 100
63
+ },
64
+ {
65
+ "epoch": 2.0,
66
+ "eval_accuracy": 0.8777777777777778,
67
+ "eval_f1_class_1": 1.0,
68
+ "eval_f1_class_2": 0.7567567567567568,
69
+ "eval_f1_class_3": 0.972972972972973,
70
+ "eval_f1_class_4": 0.8837209302325582,
71
+ "eval_f1_class_5": 0.8888888888888888,
72
+ "eval_f1_class_6": 0.6428571428571429,
73
+ "eval_f1_class_7": 0.96,
74
+ "eval_f1_class_8": 0.7755102040816326,
75
+ "eval_f1_class_9": 0.9387755102040817,
76
+ "eval_f1_macro": 0.8688313784437816,
77
+ "eval_f1_weighted": 0.8723569380169602,
78
+ "eval_loss": 1.5676754713058472,
79
+ "eval_precision_macro": 0.8769592612559364,
80
+ "eval_precision_weighted": 0.876579307875983,
81
+ "eval_recall_macro": 0.8712804195002337,
82
+ "eval_recall_weighted": 0.8777777777777778,
83
+ "eval_runtime": 0.3929,
84
+ "eval_samples_per_second": 458.187,
85
+ "eval_steps_per_second": 15.273,
86
+ "step": 102
87
+ },
88
+ {
89
+ "epoch": 2.4554455445544554,
90
+ "grad_norm": 3.865165948867798,
91
+ "learning_rate": 8.104575163398693e-06,
92
+ "loss": 1.5567,
93
+ "step": 125
94
+ },
95
+ {
96
+ "epoch": 2.9504950495049505,
97
+ "grad_norm": 3.0548179149627686,
98
+ "learning_rate": 9.738562091503268e-06,
99
+ "loss": 1.3614,
100
+ "step": 150
101
+ },
102
+ {
103
+ "epoch": 3.0,
104
+ "eval_accuracy": 0.9055555555555556,
105
+ "eval_f1_class_1": 1.0,
106
+ "eval_f1_class_2": 0.8292682926829268,
107
+ "eval_f1_class_3": 0.972972972972973,
108
+ "eval_f1_class_4": 0.926829268292683,
109
+ "eval_f1_class_5": 0.8888888888888888,
110
+ "eval_f1_class_6": 0.6896551724137931,
111
+ "eval_f1_class_7": 0.96,
112
+ "eval_f1_class_8": 0.851063829787234,
113
+ "eval_f1_class_9": 0.9583333333333334,
114
+ "eval_f1_macro": 0.8974457509302035,
115
+ "eval_f1_weighted": 0.9021962765230499,
116
+ "eval_loss": 1.1128010749816895,
117
+ "eval_precision_macro": 0.9019337828033481,
118
+ "eval_precision_weighted": 0.9043258303693086,
119
+ "eval_recall_macro": 0.8991126635089484,
120
+ "eval_recall_weighted": 0.9055555555555556,
121
+ "eval_runtime": 0.3913,
122
+ "eval_samples_per_second": 460.034,
123
+ "eval_steps_per_second": 15.334,
124
+ "step": 153
125
+ },
126
+ {
127
+ "epoch": 3.4356435643564356,
128
+ "grad_norm": 5.272692680358887,
129
+ "learning_rate": 9.847494553376908e-06,
130
+ "loss": 1.1806,
131
+ "step": 175
132
+ },
133
+ {
134
+ "epoch": 3.9306930693069306,
135
+ "grad_norm": 7.215332984924316,
136
+ "learning_rate": 9.665940450254176e-06,
137
+ "loss": 1.0304,
138
+ "step": 200
139
+ },
140
+ {
141
+ "epoch": 4.0,
142
+ "eval_accuracy": 0.9,
143
+ "eval_f1_class_1": 1.0,
144
+ "eval_f1_class_2": 0.7804878048780488,
145
+ "eval_f1_class_3": 1.0,
146
+ "eval_f1_class_4": 0.926829268292683,
147
+ "eval_f1_class_5": 0.88,
148
+ "eval_f1_class_6": 0.64,
149
+ "eval_f1_class_7": 0.9795918367346939,
150
+ "eval_f1_class_8": 0.8301886792452831,
151
+ "eval_f1_class_9": 0.9583333333333334,
152
+ "eval_f1_macro": 0.8883812136093379,
153
+ "eval_f1_weighted": 0.8941263361866708,
154
+ "eval_loss": 0.8936125636100769,
155
+ "eval_precision_macro": 0.9135504469987228,
156
+ "eval_precision_weighted": 0.9092049808429119,
157
+ "eval_recall_macro": 0.8870454164571812,
158
+ "eval_recall_weighted": 0.9,
159
+ "eval_runtime": 0.3826,
160
+ "eval_samples_per_second": 470.424,
161
+ "eval_steps_per_second": 15.681,
162
+ "step": 204
163
+ },
164
+ {
165
+ "epoch": 4.415841584158416,
166
+ "grad_norm": 4.240860939025879,
167
+ "learning_rate": 9.484386347131446e-06,
168
+ "loss": 0.926,
169
+ "step": 225
170
+ },
171
+ {
172
+ "epoch": 4.910891089108911,
173
+ "grad_norm": 7.547500133514404,
174
+ "learning_rate": 9.302832244008714e-06,
175
+ "loss": 0.8455,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 5.0,
180
+ "eval_accuracy": 0.8944444444444445,
181
+ "eval_f1_class_1": 0.975609756097561,
182
+ "eval_f1_class_2": 0.7619047619047619,
183
+ "eval_f1_class_3": 0.9444444444444444,
184
+ "eval_f1_class_4": 0.9,
185
+ "eval_f1_class_5": 0.8888888888888888,
186
+ "eval_f1_class_6": 0.7142857142857143,
187
+ "eval_f1_class_7": 0.96,
188
+ "eval_f1_class_8": 0.875,
189
+ "eval_f1_class_9": 0.9583333333333334,
190
+ "eval_f1_macro": 0.8864963221060782,
191
+ "eval_f1_weighted": 0.8915267453865015,
192
+ "eval_loss": 0.7933213710784912,
193
+ "eval_precision_macro": 0.895996040996041,
194
+ "eval_precision_weighted": 0.8972462352462351,
195
+ "eval_recall_macro": 0.8867832288110926,
196
+ "eval_recall_weighted": 0.8944444444444445,
197
+ "eval_runtime": 0.377,
198
+ "eval_samples_per_second": 477.469,
199
+ "eval_steps_per_second": 15.916,
200
+ "step": 255
201
+ },
202
+ {
203
+ "epoch": 5.396039603960396,
204
+ "grad_norm": 6.939261436462402,
205
+ "learning_rate": 9.121278140885986e-06,
206
+ "loss": 0.7949,
207
+ "step": 275
208
+ },
209
+ {
210
+ "epoch": 5.891089108910891,
211
+ "grad_norm": 4.395618915557861,
212
+ "learning_rate": 8.939724037763254e-06,
213
+ "loss": 0.7405,
214
+ "step": 300
215
+ },
216
+ {
217
+ "epoch": 6.0,
218
+ "eval_accuracy": 0.9055555555555556,
219
+ "eval_f1_class_1": 1.0,
220
+ "eval_f1_class_2": 0.7727272727272727,
221
+ "eval_f1_class_3": 1.0,
222
+ "eval_f1_class_4": 0.95,
223
+ "eval_f1_class_5": 0.9230769230769231,
224
+ "eval_f1_class_6": 0.6666666666666666,
225
+ "eval_f1_class_7": 1.0,
226
+ "eval_f1_class_8": 0.8260869565217391,
227
+ "eval_f1_class_9": 0.9583333333333334,
228
+ "eval_f1_macro": 0.8996545724806596,
229
+ "eval_f1_weighted": 0.9036424022837064,
230
+ "eval_loss": 0.7440095543861389,
231
+ "eval_precision_macro": 0.9038085988085989,
232
+ "eval_precision_weighted": 0.9062825692825692,
233
+ "eval_recall_macro": 0.900330987095693,
234
+ "eval_recall_weighted": 0.9055555555555556,
235
+ "eval_runtime": 0.3904,
236
+ "eval_samples_per_second": 461.066,
237
+ "eval_steps_per_second": 15.369,
238
+ "step": 306
239
+ },
240
+ {
241
+ "epoch": 6.376237623762377,
242
+ "grad_norm": 5.959896087646484,
243
+ "learning_rate": 8.758169934640524e-06,
244
+ "loss": 0.7125,
245
+ "step": 325
246
+ },
247
+ {
248
+ "epoch": 6.871287128712871,
249
+ "grad_norm": 5.87659215927124,
250
+ "learning_rate": 8.576615831517793e-06,
251
+ "loss": 0.673,
252
+ "step": 350
253
+ },
254
+ {
255
+ "epoch": 7.0,
256
+ "eval_accuracy": 0.9055555555555556,
257
+ "eval_f1_class_1": 0.975609756097561,
258
+ "eval_f1_class_2": 0.7368421052631579,
259
+ "eval_f1_class_3": 0.972972972972973,
260
+ "eval_f1_class_4": 0.8888888888888888,
261
+ "eval_f1_class_5": 0.9230769230769231,
262
+ "eval_f1_class_6": 0.7407407407407407,
263
+ "eval_f1_class_7": 1.0,
264
+ "eval_f1_class_8": 0.88,
265
+ "eval_f1_class_9": 0.9583333333333334,
266
+ "eval_f1_macro": 0.8973849689303974,
267
+ "eval_f1_weighted": 0.9014864486152104,
268
+ "eval_loss": 0.7316537499427795,
269
+ "eval_precision_macro": 0.9132654999321667,
270
+ "eval_precision_weighted": 0.9115048161714827,
271
+ "eval_recall_macro": 0.8972608116570964,
272
+ "eval_recall_weighted": 0.9055555555555556,
273
+ "eval_runtime": 0.3857,
274
+ "eval_samples_per_second": 466.734,
275
+ "eval_steps_per_second": 15.558,
276
+ "step": 357
277
+ },
278
+ {
279
+ "epoch": 7.356435643564357,
280
+ "grad_norm": 5.4788947105407715,
281
+ "learning_rate": 8.395061728395062e-06,
282
+ "loss": 0.6209,
283
+ "step": 375
284
+ },
285
+ {
286
+ "epoch": 7.851485148514851,
287
+ "grad_norm": 5.614921569824219,
288
+ "learning_rate": 8.213507625272332e-06,
289
+ "loss": 0.6277,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 8.0,
294
+ "eval_accuracy": 0.8944444444444445,
295
+ "eval_f1_class_1": 0.975609756097561,
296
+ "eval_f1_class_2": 0.7142857142857143,
297
+ "eval_f1_class_3": 0.9444444444444444,
298
+ "eval_f1_class_4": 0.8837209302325582,
299
+ "eval_f1_class_5": 0.8888888888888888,
300
+ "eval_f1_class_6": 0.6923076923076923,
301
+ "eval_f1_class_7": 1.0,
302
+ "eval_f1_class_8": 0.8979591836734694,
303
+ "eval_f1_class_9": 0.9583333333333334,
304
+ "eval_f1_macro": 0.8839499936959624,
305
+ "eval_f1_weighted": 0.8907458103132077,
306
+ "eval_loss": 0.7260087132453918,
307
+ "eval_precision_macro": 0.9019365497626368,
308
+ "eval_precision_weighted": 0.9034921053181922,
309
+ "eval_recall_macro": 0.8848769107283039,
310
+ "eval_recall_weighted": 0.8944444444444445,
311
+ "eval_runtime": 0.3921,
312
+ "eval_samples_per_second": 459.098,
313
+ "eval_steps_per_second": 15.303,
314
+ "step": 408
315
+ },
316
+ {
317
+ "epoch": 8.336633663366337,
318
+ "grad_norm": 0.9043681025505066,
319
+ "learning_rate": 8.031953522149601e-06,
320
+ "loss": 0.5791,
321
+ "step": 425
322
+ },
323
+ {
324
+ "epoch": 8.831683168316832,
325
+ "grad_norm": 8.910099983215332,
326
+ "learning_rate": 7.850399419026871e-06,
327
+ "loss": 0.5777,
328
+ "step": 450
329
+ },
330
+ {
331
+ "epoch": 9.0,
332
+ "eval_accuracy": 0.9,
333
+ "eval_f1_class_1": 1.0,
334
+ "eval_f1_class_2": 0.7317073170731707,
335
+ "eval_f1_class_3": 0.972972972972973,
336
+ "eval_f1_class_4": 0.8888888888888888,
337
+ "eval_f1_class_5": 0.96,
338
+ "eval_f1_class_6": 0.7142857142857143,
339
+ "eval_f1_class_7": 1.0,
340
+ "eval_f1_class_8": 0.8571428571428571,
341
+ "eval_f1_class_9": 0.9361702127659575,
342
+ "eval_f1_macro": 0.8956853292366181,
343
+ "eval_f1_weighted": 0.8979145067423913,
344
+ "eval_loss": 0.7282315492630005,
345
+ "eval_precision_macro": 0.9088936988936989,
346
+ "eval_precision_weighted": 0.905464405964406,
347
+ "eval_recall_macro": 0.8933558197086264,
348
+ "eval_recall_weighted": 0.9,
349
+ "eval_runtime": 0.3802,
350
+ "eval_samples_per_second": 473.411,
351
+ "eval_steps_per_second": 15.78,
352
+ "step": 459
353
+ },
354
+ {
355
+ "epoch": 9.316831683168317,
356
+ "grad_norm": 1.7061387300491333,
357
+ "learning_rate": 7.66884531590414e-06,
358
+ "loss": 0.5522,
359
+ "step": 475
360
+ },
361
+ {
362
+ "epoch": 9.811881188118813,
363
+ "grad_norm": 5.173563003540039,
364
+ "learning_rate": 7.487291212781409e-06,
365
+ "loss": 0.5386,
366
+ "step": 500
367
+ },
368
+ {
369
+ "epoch": 10.0,
370
+ "eval_accuracy": 0.8888888888888888,
371
+ "eval_f1_class_1": 1.0,
372
+ "eval_f1_class_2": 0.7,
373
+ "eval_f1_class_3": 0.972972972972973,
374
+ "eval_f1_class_4": 0.8636363636363636,
375
+ "eval_f1_class_5": 0.9230769230769231,
376
+ "eval_f1_class_6": 0.7096774193548387,
377
+ "eval_f1_class_7": 1.0,
378
+ "eval_f1_class_8": 0.8260869565217391,
379
+ "eval_f1_class_9": 0.9583333333333334,
380
+ "eval_f1_macro": 0.8837537743217968,
381
+ "eval_f1_weighted": 0.887174908396858,
382
+ "eval_loss": 0.7506593465805054,
383
+ "eval_precision_macro": 0.8871215821215821,
384
+ "eval_precision_weighted": 0.8893208273208272,
385
+ "eval_recall_macro": 0.8843523149250703,
386
+ "eval_recall_weighted": 0.8888888888888888,
387
+ "eval_runtime": 0.3848,
388
+ "eval_samples_per_second": 467.763,
389
+ "eval_steps_per_second": 15.592,
390
+ "step": 510
391
+ },
392
+ {
393
+ "epoch": 10.297029702970297,
394
+ "grad_norm": 12.60004711151123,
395
+ "learning_rate": 7.305737109658679e-06,
396
+ "loss": 0.5355,
397
+ "step": 525
398
+ },
399
+ {
400
+ "epoch": 10.792079207920793,
401
+ "grad_norm": 5.767191410064697,
402
+ "learning_rate": 7.124183006535948e-06,
403
+ "loss": 0.5255,
404
+ "step": 550
405
+ },
406
+ {
407
+ "epoch": 11.0,
408
+ "eval_accuracy": 0.8722222222222222,
409
+ "eval_f1_class_1": 0.975609756097561,
410
+ "eval_f1_class_2": 0.7441860465116279,
411
+ "eval_f1_class_3": 0.9444444444444444,
412
+ "eval_f1_class_4": 0.85,
413
+ "eval_f1_class_5": 0.8571428571428571,
414
+ "eval_f1_class_6": 0.6206896551724138,
415
+ "eval_f1_class_7": 0.9795918367346939,
416
+ "eval_f1_class_8": 0.8260869565217391,
417
+ "eval_f1_class_9": 0.9583333333333334,
418
+ "eval_f1_macro": 0.8617872095509633,
419
+ "eval_f1_weighted": 0.8689605526790969,
420
+ "eval_loss": 0.7731823325157166,
421
+ "eval_precision_macro": 0.8657410544367067,
422
+ "eval_precision_weighted": 0.8724329736286257,
423
+ "eval_recall_macro": 0.8654324662838595,
424
+ "eval_recall_weighted": 0.8722222222222222,
425
+ "eval_runtime": 0.377,
426
+ "eval_samples_per_second": 477.489,
427
+ "eval_steps_per_second": 15.916,
428
+ "step": 561
429
+ }
430
+ ],
431
+ "logging_steps": 25,
432
+ "max_steps": 1530,
433
+ "num_input_tokens_seen": 0,
434
+ "num_train_epochs": 30,
435
+ "save_steps": 500,
436
+ "stateful_callbacks": {
437
+ "EarlyStoppingCallback": {
438
+ "args": {
439
+ "early_stopping_patience": 5,
440
+ "early_stopping_threshold": 0.001
441
+ },
442
+ "attributes": {
443
+ "early_stopping_patience_counter": 5
444
+ }
445
+ },
446
+ "TrainerControl": {
447
+ "args": {
448
+ "should_epoch_stop": false,
449
+ "should_evaluate": false,
450
+ "should_log": false,
451
+ "should_save": true,
452
+ "should_training_stop": true
453
+ },
454
+ "attributes": {}
455
+ }
456
+ },
457
+ "total_flos": 2331442028348928.0,
458
+ "train_batch_size": 16,
459
+ "trial_name": null,
460
+ "trial_params": null
461
+ }
telecom_marabert_large_full_pipeline/checkpoint-561/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f988eeda8a1fc14cbf1fe7dca7e292179f39245156399c748eeca11dad745279
3
+ size 5905
telecom_marabert_large_full_pipeline/checkpoint-561/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "1",
15
+ "1": "2",
16
+ "2": "3",
17
+ "3": "4",
18
+ "4": "5",
19
+ "5": "6",
20
+ "6": "7",
21
+ "7": "8",
22
+ "8": "9"
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "label2id": {
27
+ "0": 1,
28
+ "1": 2,
29
+ "2": 3,
30
+ "3": 4,
31
+ "4": 5,
32
+ "5": 6,
33
+ "6": 7,
34
+ "7": 8,
35
+ "8": 9
36
+ },
37
+ "layer_norm_eps": 1e-12,
38
+ "max_position_embeddings": 512,
39
+ "model_type": "bert",
40
+ "num_attention_heads": 12,
41
+ "num_hidden_layers": 12,
42
+ "pad_token_id": 0,
43
+ "pooler_fc_size": 768,
44
+ "pooler_num_attention_heads": 12,
45
+ "pooler_num_fc_layers": 3,
46
+ "pooler_size_per_head": 128,
47
+ "pooler_type": "first_token_transform",
48
+ "position_embedding_type": "absolute",
49
+ "transformers_version": "4.57.3",
50
+ "type_vocab_size": 2,
51
+ "use_cache": true,
52
+ "vocab_size": 100000,
53
+ "num_labels": 9,
54
+ "problem_type": "single_label_classification"
55
+ }
telecom_marabert_large_full_pipeline/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3bf93c2e7faae9340d928e41440193d5bb83b0ea4d4d8db23f0665daa351c9a
3
+ size 651416604
telecom_marabert_large_full_pipeline/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
telecom_marabert_large_full_pipeline/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marabert_large_full_pipeline/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "4": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
telecom_marabert_large_full_pipeline/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f988eeda8a1fc14cbf1fe7dca7e292179f39245156399c748eeca11dad745279
3
+ size 5905
telecom_marabert_large_full_pipeline/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marbertv2_cpt_ft/checkpoint-3500/config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "1",
15
+ "1": "2",
16
+ "2": "3",
17
+ "3": "4",
18
+ "4": "5",
19
+ "5": "6",
20
+ "6": "7",
21
+ "7": "8",
22
+ "8": "9"
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 3072,
26
+ "label2id": {
27
+ "0": 1,
28
+ "1": 2,
29
+ "2": 3,
30
+ "3": 4,
31
+ "4": 5,
32
+ "5": 6,
33
+ "6": 7,
34
+ "7": 8,
35
+ "8": 9
36
+ },
37
+ "layer_norm_eps": 1e-12,
38
+ "max_position_embeddings": 512,
39
+ "model_type": "bert",
40
+ "num_attention_heads": 12,
41
+ "num_hidden_layers": 12,
42
+ "pad_token_id": 0,
43
+ "pooler_fc_size": 768,
44
+ "pooler_num_attention_heads": 12,
45
+ "pooler_num_fc_layers": 3,
46
+ "pooler_size_per_head": 128,
47
+ "pooler_type": "first_token_transform",
48
+ "position_embedding_type": "absolute",
49
+ "problem_type": "single_label_classification",
50
+ "transformers_version": "4.57.3",
51
+ "type_vocab_size": 2,
52
+ "use_cache": true,
53
+ "vocab_size": 100000
54
+ }
telecom_marbertv2_cpt_ft/checkpoint-3500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59f278354228c28cff06335262b0f786a1cdd0ed8f56502f743b33ad6d3efc62
3
+ size 651416604
telecom_marbertv2_cpt_ft/checkpoint-3500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b39f43ac12acf0f23032c62df37bdf1e0593509cf418ad91fb82178960d4da26
3
+ size 1302957451
telecom_marbertv2_cpt_ft/checkpoint-3500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16c4b35aee0bbaeee4ff40ea0ec57a691c91ae4126d7f2611b0621dcea397d18
3
+ size 14645
telecom_marbertv2_cpt_ft/checkpoint-3500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fba3938712732ea9a2d4d892486d311187e6b050bac015163f3186f8c0ea04f4
3
+ size 1465
telecom_marbertv2_cpt_ft/checkpoint-3500/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
telecom_marbertv2_cpt_ft/checkpoint-3500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
telecom_marbertv2_cpt_ft/checkpoint-3500/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "4": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
telecom_marbertv2_cpt_ft/checkpoint-3500/trainer_state.json ADDED
@@ -0,0 +1,524 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 68.62745098039215,
6
+ "eval_steps": 500,
7
+ "global_step": 3500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.9803921568627451,
14
+ "grad_norm": 1.306965947151184,
15
+ "learning_rate": 2.7450980392156867e-06,
16
+ "loss": 2.2062,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 1.9607843137254903,
21
+ "grad_norm": 1.702431559562683,
22
+ "learning_rate": 5.546218487394959e-06,
23
+ "loss": 2.1695,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 2.9411764705882355,
28
+ "grad_norm": 2.5910568237304688,
29
+ "learning_rate": 8.34733893557423e-06,
30
+ "loss": 2.014,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 3.9215686274509802,
35
+ "grad_norm": 3.7203903198242188,
36
+ "learning_rate": 1.1148459383753503e-05,
37
+ "loss": 1.594,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 4.901960784313726,
42
+ "grad_norm": 5.168037414550781,
43
+ "learning_rate": 1.3949579831932774e-05,
44
+ "loss": 1.1392,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 5.882352941176471,
49
+ "grad_norm": 6.241028785705566,
50
+ "learning_rate": 1.6750700280112046e-05,
51
+ "loss": 0.832,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 6.862745098039216,
56
+ "grad_norm": 6.874265193939209,
57
+ "learning_rate": 1.9551820728291318e-05,
58
+ "loss": 0.6115,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 7.8431372549019605,
63
+ "grad_norm": 8.241349220275879,
64
+ "learning_rate": 1.973856209150327e-05,
65
+ "loss": 0.4274,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 8.823529411764707,
70
+ "grad_norm": 11.996331214904785,
71
+ "learning_rate": 1.9427326486150017e-05,
72
+ "loss": 0.2847,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 9.803921568627452,
77
+ "grad_norm": 8.437224388122559,
78
+ "learning_rate": 1.9116090880796766e-05,
79
+ "loss": 0.209,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 10.784313725490197,
84
+ "grad_norm": 4.417978763580322,
85
+ "learning_rate": 1.880485527544351e-05,
86
+ "loss": 0.1514,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 11.764705882352942,
91
+ "grad_norm": 0.3415707051753998,
92
+ "learning_rate": 1.849361967009026e-05,
93
+ "loss": 0.105,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 12.745098039215687,
98
+ "grad_norm": 1.3661662340164185,
99
+ "learning_rate": 1.8182384064737007e-05,
100
+ "loss": 0.0807,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 13.72549019607843,
105
+ "grad_norm": 0.7777091264724731,
106
+ "learning_rate": 1.7871148459383755e-05,
107
+ "loss": 0.0617,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 14.705882352941176,
112
+ "grad_norm": 0.9965397715568542,
113
+ "learning_rate": 1.7559912854030504e-05,
114
+ "loss": 0.0417,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 15.686274509803921,
119
+ "grad_norm": 0.09024341404438019,
120
+ "learning_rate": 1.724867724867725e-05,
121
+ "loss": 0.0555,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 16.666666666666668,
126
+ "grad_norm": 11.35683822631836,
127
+ "learning_rate": 1.6937441643323997e-05,
128
+ "loss": 0.0244,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 17.647058823529413,
133
+ "grad_norm": 0.06510096788406372,
134
+ "learning_rate": 1.6626206037970745e-05,
135
+ "loss": 0.0231,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 18.627450980392158,
140
+ "grad_norm": 0.05045614019036293,
141
+ "learning_rate": 1.631497043261749e-05,
142
+ "loss": 0.0166,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 19.607843137254903,
147
+ "grad_norm": 0.06423764675855637,
148
+ "learning_rate": 1.6003734827264242e-05,
149
+ "loss": 0.02,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 20.58823529411765,
154
+ "grad_norm": 0.037640661001205444,
155
+ "learning_rate": 1.5692499221910987e-05,
156
+ "loss": 0.0128,
157
+ "step": 1050
158
+ },
159
+ {
160
+ "epoch": 21.568627450980394,
161
+ "grad_norm": 0.42402613162994385,
162
+ "learning_rate": 1.5381263616557735e-05,
163
+ "loss": 0.0174,
164
+ "step": 1100
165
+ },
166
+ {
167
+ "epoch": 22.54901960784314,
168
+ "grad_norm": 0.0321834422647953,
169
+ "learning_rate": 1.5070028011204482e-05,
170
+ "loss": 0.0108,
171
+ "step": 1150
172
+ },
173
+ {
174
+ "epoch": 23.529411764705884,
175
+ "grad_norm": 0.03194517642259598,
176
+ "learning_rate": 1.475879240585123e-05,
177
+ "loss": 0.0111,
178
+ "step": 1200
179
+ },
180
+ {
181
+ "epoch": 24.509803921568626,
182
+ "grad_norm": 0.026620032265782356,
183
+ "learning_rate": 1.4447556800497977e-05,
184
+ "loss": 0.0126,
185
+ "step": 1250
186
+ },
187
+ {
188
+ "epoch": 25.49019607843137,
189
+ "grad_norm": 0.02269897237420082,
190
+ "learning_rate": 1.4136321195144727e-05,
191
+ "loss": 0.0095,
192
+ "step": 1300
193
+ },
194
+ {
195
+ "epoch": 26.470588235294116,
196
+ "grad_norm": 0.30052125453948975,
197
+ "learning_rate": 1.3825085589791474e-05,
198
+ "loss": 0.0084,
199
+ "step": 1350
200
+ },
201
+ {
202
+ "epoch": 27.45098039215686,
203
+ "grad_norm": 0.5096309185028076,
204
+ "learning_rate": 1.351384998443822e-05,
205
+ "loss": 0.0052,
206
+ "step": 1400
207
+ },
208
+ {
209
+ "epoch": 28.431372549019606,
210
+ "grad_norm": 0.023257585242390633,
211
+ "learning_rate": 1.3202614379084969e-05,
212
+ "loss": 0.008,
213
+ "step": 1450
214
+ },
215
+ {
216
+ "epoch": 29.41176470588235,
217
+ "grad_norm": 0.017219331115484238,
218
+ "learning_rate": 1.2891378773731715e-05,
219
+ "loss": 0.0058,
220
+ "step": 1500
221
+ },
222
+ {
223
+ "epoch": 30.392156862745097,
224
+ "grad_norm": 0.01873333752155304,
225
+ "learning_rate": 1.2580143168378462e-05,
226
+ "loss": 0.0067,
227
+ "step": 1550
228
+ },
229
+ {
230
+ "epoch": 31.372549019607842,
231
+ "grad_norm": 0.02165828086435795,
232
+ "learning_rate": 1.2268907563025212e-05,
233
+ "loss": 0.0056,
234
+ "step": 1600
235
+ },
236
+ {
237
+ "epoch": 32.35294117647059,
238
+ "grad_norm": 0.01725042052567005,
239
+ "learning_rate": 1.1957671957671959e-05,
240
+ "loss": 0.0073,
241
+ "step": 1650
242
+ },
243
+ {
244
+ "epoch": 33.333333333333336,
245
+ "grad_norm": 0.02131626009941101,
246
+ "learning_rate": 1.1646436352318707e-05,
247
+ "loss": 0.0058,
248
+ "step": 1700
249
+ },
250
+ {
251
+ "epoch": 34.31372549019608,
252
+ "grad_norm": 0.013201112858951092,
253
+ "learning_rate": 1.1335200746965454e-05,
254
+ "loss": 0.0062,
255
+ "step": 1750
256
+ },
257
+ {
258
+ "epoch": 35.294117647058826,
259
+ "grad_norm": 0.016260718926787376,
260
+ "learning_rate": 1.10239651416122e-05,
261
+ "loss": 0.0046,
262
+ "step": 1800
263
+ },
264
+ {
265
+ "epoch": 36.27450980392157,
266
+ "grad_norm": 0.033385492861270905,
267
+ "learning_rate": 1.0712729536258948e-05,
268
+ "loss": 0.006,
269
+ "step": 1850
270
+ },
271
+ {
272
+ "epoch": 37.254901960784316,
273
+ "grad_norm": 0.011549146845936775,
274
+ "learning_rate": 1.0401493930905697e-05,
275
+ "loss": 0.0064,
276
+ "step": 1900
277
+ },
278
+ {
279
+ "epoch": 38.23529411764706,
280
+ "grad_norm": 0.012482107616961002,
281
+ "learning_rate": 1.0090258325552445e-05,
282
+ "loss": 0.0048,
283
+ "step": 1950
284
+ },
285
+ {
286
+ "epoch": 39.21568627450981,
287
+ "grad_norm": 0.010968800634145737,
288
+ "learning_rate": 9.779022720199192e-06,
289
+ "loss": 0.0047,
290
+ "step": 2000
291
+ },
292
+ {
293
+ "epoch": 40.19607843137255,
294
+ "grad_norm": 0.47498399019241333,
295
+ "learning_rate": 9.467787114845938e-06,
296
+ "loss": 0.0073,
297
+ "step": 2050
298
+ },
299
+ {
300
+ "epoch": 41.1764705882353,
301
+ "grad_norm": 0.011364293284714222,
302
+ "learning_rate": 9.156551509492687e-06,
303
+ "loss": 0.0035,
304
+ "step": 2100
305
+ },
306
+ {
307
+ "epoch": 42.15686274509804,
308
+ "grad_norm": 0.009123931638896465,
309
+ "learning_rate": 8.845315904139435e-06,
310
+ "loss": 0.0049,
311
+ "step": 2150
312
+ },
313
+ {
314
+ "epoch": 43.13725490196079,
315
+ "grad_norm": 0.4411880373954773,
316
+ "learning_rate": 8.534080298786182e-06,
317
+ "loss": 0.0057,
318
+ "step": 2200
319
+ },
320
+ {
321
+ "epoch": 44.11764705882353,
322
+ "grad_norm": 0.00864217709749937,
323
+ "learning_rate": 8.22284469343293e-06,
324
+ "loss": 0.0038,
325
+ "step": 2250
326
+ },
327
+ {
328
+ "epoch": 45.09803921568628,
329
+ "grad_norm": 0.009198697283864021,
330
+ "learning_rate": 7.911609088079677e-06,
331
+ "loss": 0.0055,
332
+ "step": 2300
333
+ },
334
+ {
335
+ "epoch": 46.07843137254902,
336
+ "grad_norm": 0.4228628873825073,
337
+ "learning_rate": 7.600373482726424e-06,
338
+ "loss": 0.0044,
339
+ "step": 2350
340
+ },
341
+ {
342
+ "epoch": 47.05882352941177,
343
+ "grad_norm": 0.009074629284441471,
344
+ "learning_rate": 7.2891378773731725e-06,
345
+ "loss": 0.0048,
346
+ "step": 2400
347
+ },
348
+ {
349
+ "epoch": 48.03921568627451,
350
+ "grad_norm": 0.007762749679386616,
351
+ "learning_rate": 6.97790227201992e-06,
352
+ "loss": 0.0042,
353
+ "step": 2450
354
+ },
355
+ {
356
+ "epoch": 49.01960784313726,
357
+ "grad_norm": 0.010210619308054447,
358
+ "learning_rate": 6.666666666666667e-06,
359
+ "loss": 0.0041,
360
+ "step": 2500
361
+ },
362
+ {
363
+ "epoch": 50.0,
364
+ "grad_norm": 0.011346593499183655,
365
+ "learning_rate": 6.355431061313415e-06,
366
+ "loss": 0.0046,
367
+ "step": 2550
368
+ },
369
+ {
370
+ "epoch": 50.98039215686274,
371
+ "grad_norm": 0.007517909165471792,
372
+ "learning_rate": 6.0441954559601625e-06,
373
+ "loss": 0.0048,
374
+ "step": 2600
375
+ },
376
+ {
377
+ "epoch": 51.96078431372549,
378
+ "grad_norm": 0.00800730474293232,
379
+ "learning_rate": 5.732959850606909e-06,
380
+ "loss": 0.0042,
381
+ "step": 2650
382
+ },
383
+ {
384
+ "epoch": 52.94117647058823,
385
+ "grad_norm": 0.008079170249402523,
386
+ "learning_rate": 5.4217242452536574e-06,
387
+ "loss": 0.0043,
388
+ "step": 2700
389
+ },
390
+ {
391
+ "epoch": 53.92156862745098,
392
+ "grad_norm": 0.006971494760364294,
393
+ "learning_rate": 5.110488639900405e-06,
394
+ "loss": 0.0044,
395
+ "step": 2750
396
+ },
397
+ {
398
+ "epoch": 54.90196078431372,
399
+ "grad_norm": 0.4812173843383789,
400
+ "learning_rate": 4.799253034547152e-06,
401
+ "loss": 0.0041,
402
+ "step": 2800
403
+ },
404
+ {
405
+ "epoch": 55.88235294117647,
406
+ "grad_norm": 0.007991774939000607,
407
+ "learning_rate": 4.4880174291939e-06,
408
+ "loss": 0.0034,
409
+ "step": 2850
410
+ },
411
+ {
412
+ "epoch": 56.86274509803921,
413
+ "grad_norm": 0.008251671679317951,
414
+ "learning_rate": 4.176781823840647e-06,
415
+ "loss": 0.004,
416
+ "step": 2900
417
+ },
418
+ {
419
+ "epoch": 57.84313725490196,
420
+ "grad_norm": 0.008690707385540009,
421
+ "learning_rate": 3.865546218487396e-06,
422
+ "loss": 0.004,
423
+ "step": 2950
424
+ },
425
+ {
426
+ "epoch": 58.8235294117647,
427
+ "grad_norm": 0.007442942820489407,
428
+ "learning_rate": 3.5543106131341428e-06,
429
+ "loss": 0.0034,
430
+ "step": 3000
431
+ },
432
+ {
433
+ "epoch": 59.80392156862745,
434
+ "grad_norm": 0.0063288831152021885,
435
+ "learning_rate": 3.2430750077808903e-06,
436
+ "loss": 0.0045,
437
+ "step": 3050
438
+ },
439
+ {
440
+ "epoch": 60.78431372549019,
441
+ "grad_norm": 0.006511453073471785,
442
+ "learning_rate": 2.931839402427638e-06,
443
+ "loss": 0.0075,
444
+ "step": 3100
445
+ },
446
+ {
447
+ "epoch": 61.76470588235294,
448
+ "grad_norm": 0.007184595800936222,
449
+ "learning_rate": 2.6206037970743852e-06,
450
+ "loss": 0.0043,
451
+ "step": 3150
452
+ },
453
+ {
454
+ "epoch": 62.745098039215684,
455
+ "grad_norm": 0.007260785438120365,
456
+ "learning_rate": 2.309368191721133e-06,
457
+ "loss": 0.0026,
458
+ "step": 3200
459
+ },
460
+ {
461
+ "epoch": 63.72549019607843,
462
+ "grad_norm": 0.007063099183142185,
463
+ "learning_rate": 1.9981325863678806e-06,
464
+ "loss": 0.0046,
465
+ "step": 3250
466
+ },
467
+ {
468
+ "epoch": 64.70588235294117,
469
+ "grad_norm": 0.006402399856597185,
470
+ "learning_rate": 1.6868969810146283e-06,
471
+ "loss": 0.0043,
472
+ "step": 3300
473
+ },
474
+ {
475
+ "epoch": 65.68627450980392,
476
+ "grad_norm": 0.007472318597137928,
477
+ "learning_rate": 1.3756613756613758e-06,
478
+ "loss": 0.0045,
479
+ "step": 3350
480
+ },
481
+ {
482
+ "epoch": 66.66666666666667,
483
+ "grad_norm": 0.006515247281640768,
484
+ "learning_rate": 1.0644257703081233e-06,
485
+ "loss": 0.0046,
486
+ "step": 3400
487
+ },
488
+ {
489
+ "epoch": 67.6470588235294,
490
+ "grad_norm": 0.00874197855591774,
491
+ "learning_rate": 7.531901649548709e-07,
492
+ "loss": 0.0034,
493
+ "step": 3450
494
+ },
495
+ {
496
+ "epoch": 68.62745098039215,
497
+ "grad_norm": 0.009945135563611984,
498
+ "learning_rate": 4.419545596016184e-07,
499
+ "loss": 0.0046,
500
+ "step": 3500
501
+ }
502
+ ],
503
+ "logging_steps": 50,
504
+ "max_steps": 3570,
505
+ "num_input_tokens_seen": 0,
506
+ "num_train_epochs": 70,
507
+ "save_steps": 500,
508
+ "stateful_callbacks": {
509
+ "TrainerControl": {
510
+ "args": {
511
+ "should_epoch_stop": false,
512
+ "should_evaluate": false,
513
+ "should_log": false,
514
+ "should_save": true,
515
+ "should_training_stop": false
516
+ },
517
+ "attributes": {}
518
+ }
519
+ },
520
+ "total_flos": 2.909454409554739e+16,
521
+ "train_batch_size": 32,
522
+ "trial_name": null,
523
+ "trial_params": null
524
+ }
telecom_marbertv2_cpt_ft/checkpoint-3500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e7cd97a72ee3db6153e4a7b80b50cb78726745bcba2c134a4831146d5326afd
3
+ size 5841