Hugging Face's logo

laurafcamargos
/
distilbert-qasports

Question Answering
Transformers
Safetensors
distilbert
Generated from Trainer
Inference Endpoints
Model card Files Community
laurafcamargos commited on
Commit
8679870
路
verified 路
1 Parent(s): 0a5d782

Adicionando tokenizer

Browse files
This view is limited to 50 files because it contains too many changes. 聽 See raw diff
Files changed (50) hide show
  1. checkpoint-1000/config.json +24 -0
  2. checkpoint-1000/model.safetensors +3 -0
  3. checkpoint-1000/optimizer.pt +3 -0
  4. checkpoint-1000/rng_state.pth +3 -0
  5. checkpoint-1000/scheduler.pt +3 -0
  6. checkpoint-1000/trainer_state.json +148 -0
  7. checkpoint-1000/training_args.bin +3 -0
  8. checkpoint-1500/config.json +24 -0
  9. checkpoint-1500/model.safetensors +3 -0
  10. checkpoint-1500/optimizer.pt +3 -0
  11. checkpoint-1500/rng_state.pth +3 -0
  12. checkpoint-1500/scheduler.pt +3 -0
  13. checkpoint-1500/trainer_state.json +201 -0
  14. checkpoint-1500/training_args.bin +3 -0
  15. checkpoint-2000/config.json +24 -0
  16. checkpoint-2000/model.safetensors +3 -0
  17. checkpoint-2000/optimizer.pt +3 -0
  18. checkpoint-2000/rng_state.pth +3 -0
  19. checkpoint-2000/scheduler.pt +3 -0
  20. checkpoint-2000/trainer_state.json +254 -0
  21. checkpoint-2000/training_args.bin +3 -0
  22. checkpoint-2500/config.json +24 -0
  23. checkpoint-2500/model.safetensors +3 -0
  24. checkpoint-2500/optimizer.pt +3 -0
  25. checkpoint-2500/rng_state.pth +3 -0
  26. checkpoint-2500/scheduler.pt +3 -0
  27. checkpoint-2500/trainer_state.json +307 -0
  28. checkpoint-2500/training_args.bin +3 -0
  29. checkpoint-3000/config.json +24 -0
  30. checkpoint-3000/model.safetensors +3 -0
  31. checkpoint-3000/optimizer.pt +3 -0
  32. checkpoint-3000/rng_state.pth +3 -0
  33. checkpoint-3000/scheduler.pt +3 -0
  34. checkpoint-3000/trainer_state.json +360 -0
  35. checkpoint-3000/training_args.bin +3 -0
  36. checkpoint-3500/config.json +24 -0
  37. checkpoint-3500/model.safetensors +3 -0
  38. checkpoint-3500/optimizer.pt +3 -0
  39. checkpoint-3500/rng_state.pth +3 -0
  40. checkpoint-3500/scheduler.pt +3 -0
  41. checkpoint-3500/trainer_state.json +413 -0
  42. checkpoint-3500/training_args.bin +3 -0
  43. checkpoint-4000/config.json +24 -0
  44. checkpoint-4000/model.safetensors +3 -0
  45. checkpoint-4000/optimizer.pt +3 -0
  46. checkpoint-4000/rng_state.pth +3 -0
  47. checkpoint-4000/scheduler.pt +3 -0
  48. checkpoint-4000/trainer_state.json +466 -0
  49. checkpoint-4000/training_args.bin +3 -0
  50. checkpoint-4500/config.json +24 -0
checkpoint-1000/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:deedcbf2220cb22ccc6eb29fd96171207246b408e2b05896fa4a970bad1025c8
3
+ size 265470032
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9bccbc8d9528e67cec62877dc2b7feefba8a7182649ce2f0e7065aee3e8b5cd6
3
+ size 531000890
checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7205abc132f78134f75870a63b9be517634d2c841c55662c15fcb8834e6b5d6a
3
+ size 14244
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d57f9a5bb738693e11780051ac63d810f94212904d14acac2f786859d2276d3
3
+ size 1064
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,148 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 80.10135733594169,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-1000",
4
+ "epoch": 0.26490066225165565,
5
+ "eval_steps": 500,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ }
117
+ ],
118
+ "logging_steps": 100,
119
+ "max_steps": 188750,
120
+ "num_input_tokens_seen": 0,
121
+ "num_train_epochs": 50,
122
+ "save_steps": 500,
123
+ "stateful_callbacks": {
124
+ "EarlyStoppingCallback": {
125
+ "args": {
126
+ "early_stopping_patience": 5,
127
+ "early_stopping_threshold": 0.0
128
+ },
129
+ "attributes": {
130
+ "early_stopping_patience_counter": 0
131
+ }
132
+ },
133
+ "TrainerControl": {
134
+ "args": {
135
+ "should_epoch_stop": false,
136
+ "should_evaluate": false,
137
+ "should_log": false,
138
+ "should_save": true,
139
+ "should_training_stop": false
140
+ },
141
+ "attributes": {}
142
+ }
143
+ },
144
+ "total_flos": 4180899201024000.0,
145
+ "train_batch_size": 16,
146
+ "trial_name": null,
147
+ "trial_params": null
148
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-1500/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed13557e9240d4a0bed20539f5981b2753464d33782f5dc510138dfe94b67fdf
3
+ size 265470032
checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ffc1b6e25ccfa63c9320689831245df70b0cccd266f3f38c82c086eb2c47411
3
+ size 531000890
checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83da7848191b5b11eb3203ebedf4beefb345a57375d8fd1433caf5d2dc78867c
3
+ size 14244
checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db32f4431a93462e07b12e9fd7e2da7c206873db70d9b7d866353dd730eb9bec
3
+ size 1064
checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,201 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 80.3558190310888,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-1500",
4
+ "epoch": 0.3973509933774834,
5
+ "eval_steps": 500,
6
+ "global_step": 1500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ }
170
+ ],
171
+ "logging_steps": 100,
172
+ "max_steps": 188750,
173
+ "num_input_tokens_seen": 0,
174
+ "num_train_epochs": 50,
175
+ "save_steps": 500,
176
+ "stateful_callbacks": {
177
+ "EarlyStoppingCallback": {
178
+ "args": {
179
+ "early_stopping_patience": 5,
180
+ "early_stopping_threshold": 0.0
181
+ },
182
+ "attributes": {
183
+ "early_stopping_patience_counter": 0
184
+ }
185
+ },
186
+ "TrainerControl": {
187
+ "args": {
188
+ "should_epoch_stop": false,
189
+ "should_evaluate": false,
190
+ "should_log": false,
191
+ "should_save": true,
192
+ "should_training_stop": false
193
+ },
194
+ "attributes": {}
195
+ }
196
+ },
197
+ "total_flos": 6271348801536000.0,
198
+ "train_batch_size": 16,
199
+ "trial_name": null,
200
+ "trial_params": null
201
+ }
checkpoint-1500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-2000/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f88d824cd6987622bac96eed843669fe9794fda90b4b59d1547197a0e9fe5709
3
+ size 265470032
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4432d5cd80440a0dae76370c1ebb1fb8418b33464466ef06354a6544c83edfd2
3
+ size 531000890
checkpoint-2000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb2f2f80ca6a24eb462827f13b12fe1f6f526ef65151bcbc84cfa9f2e1f31250
3
+ size 14244
checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1c241d2e174338e8ca6a41a14062f5a64ddfbc159f5a84518e9420b53cf17e2
3
+ size 1064
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,254 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 81.07514905827573,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-2000",
4
+ "epoch": 0.5298013245033113,
5
+ "eval_steps": 500,
6
+ "global_step": 2000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ },
170
+ {
171
+ "epoch": 0.423841059602649,
172
+ "grad_norm": 8.173700332641602,
173
+ "learning_rate": 9.915337748344372e-06,
174
+ "loss": 0.584,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.4503311258278146,
179
+ "grad_norm": 6.5389723777771,
180
+ "learning_rate": 9.910039735099339e-06,
181
+ "loss": 0.5457,
182
+ "step": 1700
183
+ },
184
+ {
185
+ "epoch": 0.4768211920529801,
186
+ "grad_norm": 9.859861373901367,
187
+ "learning_rate": 9.904741721854306e-06,
188
+ "loss": 0.5417,
189
+ "step": 1800
190
+ },
191
+ {
192
+ "epoch": 0.5033112582781457,
193
+ "grad_norm": 13.78690242767334,
194
+ "learning_rate": 9.899443708609273e-06,
195
+ "loss": 0.5475,
196
+ "step": 1900
197
+ },
198
+ {
199
+ "epoch": 0.5298013245033113,
200
+ "grad_norm": 8.801095008850098,
201
+ "learning_rate": 9.89414569536424e-06,
202
+ "loss": 0.5814,
203
+ "step": 2000
204
+ },
205
+ {
206
+ "epoch": 0.5298013245033113,
207
+ "eval_HasAns_exact": 76.39784588790639,
208
+ "eval_HasAns_f1": 81.07514905827573,
209
+ "eval_HasAns_total": 15041,
210
+ "eval_best_exact": 76.39784588790639,
211
+ "eval_best_exact_thresh": 0.0,
212
+ "eval_best_f1": 81.07514905827573,
213
+ "eval_best_f1_thresh": 0.0,
214
+ "eval_exact": 76.39784588790639,
215
+ "eval_f1": 81.07514905827573,
216
+ "eval_loss": 0.4890206754207611,
217
+ "eval_runtime": 62.2295,
218
+ "eval_samples_per_second": 241.702,
219
+ "eval_steps_per_second": 15.121,
220
+ "eval_total": 15041,
221
+ "step": 2000
222
+ }
223
+ ],
224
+ "logging_steps": 100,
225
+ "max_steps": 188750,
226
+ "num_input_tokens_seen": 0,
227
+ "num_train_epochs": 50,
228
+ "save_steps": 500,
229
+ "stateful_callbacks": {
230
+ "EarlyStoppingCallback": {
231
+ "args": {
232
+ "early_stopping_patience": 5,
233
+ "early_stopping_threshold": 0.0
234
+ },
235
+ "attributes": {
236
+ "early_stopping_patience_counter": 0
237
+ }
238
+ },
239
+ "TrainerControl": {
240
+ "args": {
241
+ "should_epoch_stop": false,
242
+ "should_evaluate": false,
243
+ "should_log": false,
244
+ "should_save": true,
245
+ "should_training_stop": false
246
+ },
247
+ "attributes": {}
248
+ }
249
+ },
250
+ "total_flos": 8361798402048000.0,
251
+ "train_batch_size": 16,
252
+ "trial_name": null,
253
+ "trial_params": null
254
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-2500/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-2500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0798ef4b49da6c52724e153563bdafc17f096e85bd69dd93a5c0780caa0859bd
3
+ size 265470032
checkpoint-2500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:124803248da446f9a6e578b5a7b217a9dd72f142d2c5ff6a45b253bb3e5cfca1
3
+ size 531000890
checkpoint-2500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84238ebbd9b25e70825b74d08d45884067ac11306d95b5b4a5b19e070d7a26c7
3
+ size 14244
checkpoint-2500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:048cf25de4880c9945ff66be59acdbfd8b43d343cd650caaf51861a817320bcc
3
+ size 1064
checkpoint-2500/trainer_state.json ADDED
@@ -0,0 +1,307 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 81.07514905827573,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-2000",
4
+ "epoch": 0.6622516556291391,
5
+ "eval_steps": 500,
6
+ "global_step": 2500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ },
170
+ {
171
+ "epoch": 0.423841059602649,
172
+ "grad_norm": 8.173700332641602,
173
+ "learning_rate": 9.915337748344372e-06,
174
+ "loss": 0.584,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.4503311258278146,
179
+ "grad_norm": 6.5389723777771,
180
+ "learning_rate": 9.910039735099339e-06,
181
+ "loss": 0.5457,
182
+ "step": 1700
183
+ },
184
+ {
185
+ "epoch": 0.4768211920529801,
186
+ "grad_norm": 9.859861373901367,
187
+ "learning_rate": 9.904741721854306e-06,
188
+ "loss": 0.5417,
189
+ "step": 1800
190
+ },
191
+ {
192
+ "epoch": 0.5033112582781457,
193
+ "grad_norm": 13.78690242767334,
194
+ "learning_rate": 9.899443708609273e-06,
195
+ "loss": 0.5475,
196
+ "step": 1900
197
+ },
198
+ {
199
+ "epoch": 0.5298013245033113,
200
+ "grad_norm": 8.801095008850098,
201
+ "learning_rate": 9.89414569536424e-06,
202
+ "loss": 0.5814,
203
+ "step": 2000
204
+ },
205
+ {
206
+ "epoch": 0.5298013245033113,
207
+ "eval_HasAns_exact": 76.39784588790639,
208
+ "eval_HasAns_f1": 81.07514905827573,
209
+ "eval_HasAns_total": 15041,
210
+ "eval_best_exact": 76.39784588790639,
211
+ "eval_best_exact_thresh": 0.0,
212
+ "eval_best_f1": 81.07514905827573,
213
+ "eval_best_f1_thresh": 0.0,
214
+ "eval_exact": 76.39784588790639,
215
+ "eval_f1": 81.07514905827573,
216
+ "eval_loss": 0.4890206754207611,
217
+ "eval_runtime": 62.2295,
218
+ "eval_samples_per_second": 241.702,
219
+ "eval_steps_per_second": 15.121,
220
+ "eval_total": 15041,
221
+ "step": 2000
222
+ },
223
+ {
224
+ "epoch": 0.5562913907284768,
225
+ "grad_norm": 10.412147521972656,
226
+ "learning_rate": 9.888847682119206e-06,
227
+ "loss": 0.5275,
228
+ "step": 2100
229
+ },
230
+ {
231
+ "epoch": 0.5827814569536424,
232
+ "grad_norm": 8.767007827758789,
233
+ "learning_rate": 9.883549668874173e-06,
234
+ "loss": 0.4934,
235
+ "step": 2200
236
+ },
237
+ {
238
+ "epoch": 0.609271523178808,
239
+ "grad_norm": 11.056267738342285,
240
+ "learning_rate": 9.87825165562914e-06,
241
+ "loss": 0.5246,
242
+ "step": 2300
243
+ },
244
+ {
245
+ "epoch": 0.6357615894039735,
246
+ "grad_norm": 9.738068580627441,
247
+ "learning_rate": 9.872953642384107e-06,
248
+ "loss": 0.5127,
249
+ "step": 2400
250
+ },
251
+ {
252
+ "epoch": 0.6622516556291391,
253
+ "grad_norm": 15.755542755126953,
254
+ "learning_rate": 9.867655629139074e-06,
255
+ "loss": 0.5165,
256
+ "step": 2500
257
+ },
258
+ {
259
+ "epoch": 0.6622516556291391,
260
+ "eval_HasAns_exact": 76.21168805265607,
261
+ "eval_HasAns_f1": 80.96151211311172,
262
+ "eval_HasAns_total": 15041,
263
+ "eval_best_exact": 76.21168805265607,
264
+ "eval_best_exact_thresh": 0.0,
265
+ "eval_best_f1": 80.96151211311172,
266
+ "eval_best_f1_thresh": 0.0,
267
+ "eval_exact": 76.21168805265607,
268
+ "eval_f1": 80.96151211311172,
269
+ "eval_loss": 0.4729439914226532,
270
+ "eval_runtime": 62.2287,
271
+ "eval_samples_per_second": 241.705,
272
+ "eval_steps_per_second": 15.122,
273
+ "eval_total": 15041,
274
+ "step": 2500
275
+ }
276
+ ],
277
+ "logging_steps": 100,
278
+ "max_steps": 188750,
279
+ "num_input_tokens_seen": 0,
280
+ "num_train_epochs": 50,
281
+ "save_steps": 500,
282
+ "stateful_callbacks": {
283
+ "EarlyStoppingCallback": {
284
+ "args": {
285
+ "early_stopping_patience": 5,
286
+ "early_stopping_threshold": 0.0
287
+ },
288
+ "attributes": {
289
+ "early_stopping_patience_counter": 1
290
+ }
291
+ },
292
+ "TrainerControl": {
293
+ "args": {
294
+ "should_epoch_stop": false,
295
+ "should_evaluate": false,
296
+ "should_log": false,
297
+ "should_save": true,
298
+ "should_training_stop": false
299
+ },
300
+ "attributes": {}
301
+ }
302
+ },
303
+ "total_flos": 1.045224800256e+16,
304
+ "train_batch_size": 16,
305
+ "trial_name": null,
306
+ "trial_params": null
307
+ }
checkpoint-2500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-3000/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-3000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9f641fd9a2bae1c60ddd82b9a032b9f910132a923e0cb18f2f7c37b0e53ade0
3
+ size 265470032
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99498c5db00200a49309ea8d8c6e752d599fc4373fe2402fa1047e2024cffac7
3
+ size 531000890
checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ac7a288b5703efdb70277bf0a722568746ddf3a96fc14596b4037febae507f4
3
+ size 14244
checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c10d9e37146a122190afa783b54fc268d742be9de213c0fc098c3137d98dc66b
3
+ size 1064
checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,360 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 81.20875033119805,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-3000",
4
+ "epoch": 0.7947019867549668,
5
+ "eval_steps": 500,
6
+ "global_step": 3000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ },
170
+ {
171
+ "epoch": 0.423841059602649,
172
+ "grad_norm": 8.173700332641602,
173
+ "learning_rate": 9.915337748344372e-06,
174
+ "loss": 0.584,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.4503311258278146,
179
+ "grad_norm": 6.5389723777771,
180
+ "learning_rate": 9.910039735099339e-06,
181
+ "loss": 0.5457,
182
+ "step": 1700
183
+ },
184
+ {
185
+ "epoch": 0.4768211920529801,
186
+ "grad_norm": 9.859861373901367,
187
+ "learning_rate": 9.904741721854306e-06,
188
+ "loss": 0.5417,
189
+ "step": 1800
190
+ },
191
+ {
192
+ "epoch": 0.5033112582781457,
193
+ "grad_norm": 13.78690242767334,
194
+ "learning_rate": 9.899443708609273e-06,
195
+ "loss": 0.5475,
196
+ "step": 1900
197
+ },
198
+ {
199
+ "epoch": 0.5298013245033113,
200
+ "grad_norm": 8.801095008850098,
201
+ "learning_rate": 9.89414569536424e-06,
202
+ "loss": 0.5814,
203
+ "step": 2000
204
+ },
205
+ {
206
+ "epoch": 0.5298013245033113,
207
+ "eval_HasAns_exact": 76.39784588790639,
208
+ "eval_HasAns_f1": 81.07514905827573,
209
+ "eval_HasAns_total": 15041,
210
+ "eval_best_exact": 76.39784588790639,
211
+ "eval_best_exact_thresh": 0.0,
212
+ "eval_best_f1": 81.07514905827573,
213
+ "eval_best_f1_thresh": 0.0,
214
+ "eval_exact": 76.39784588790639,
215
+ "eval_f1": 81.07514905827573,
216
+ "eval_loss": 0.4890206754207611,
217
+ "eval_runtime": 62.2295,
218
+ "eval_samples_per_second": 241.702,
219
+ "eval_steps_per_second": 15.121,
220
+ "eval_total": 15041,
221
+ "step": 2000
222
+ },
223
+ {
224
+ "epoch": 0.5562913907284768,
225
+ "grad_norm": 10.412147521972656,
226
+ "learning_rate": 9.888847682119206e-06,
227
+ "loss": 0.5275,
228
+ "step": 2100
229
+ },
230
+ {
231
+ "epoch": 0.5827814569536424,
232
+ "grad_norm": 8.767007827758789,
233
+ "learning_rate": 9.883549668874173e-06,
234
+ "loss": 0.4934,
235
+ "step": 2200
236
+ },
237
+ {
238
+ "epoch": 0.609271523178808,
239
+ "grad_norm": 11.056267738342285,
240
+ "learning_rate": 9.87825165562914e-06,
241
+ "loss": 0.5246,
242
+ "step": 2300
243
+ },
244
+ {
245
+ "epoch": 0.6357615894039735,
246
+ "grad_norm": 9.738068580627441,
247
+ "learning_rate": 9.872953642384107e-06,
248
+ "loss": 0.5127,
249
+ "step": 2400
250
+ },
251
+ {
252
+ "epoch": 0.6622516556291391,
253
+ "grad_norm": 15.755542755126953,
254
+ "learning_rate": 9.867655629139074e-06,
255
+ "loss": 0.5165,
256
+ "step": 2500
257
+ },
258
+ {
259
+ "epoch": 0.6622516556291391,
260
+ "eval_HasAns_exact": 76.21168805265607,
261
+ "eval_HasAns_f1": 80.96151211311172,
262
+ "eval_HasAns_total": 15041,
263
+ "eval_best_exact": 76.21168805265607,
264
+ "eval_best_exact_thresh": 0.0,
265
+ "eval_best_f1": 80.96151211311172,
266
+ "eval_best_f1_thresh": 0.0,
267
+ "eval_exact": 76.21168805265607,
268
+ "eval_f1": 80.96151211311172,
269
+ "eval_loss": 0.4729439914226532,
270
+ "eval_runtime": 62.2287,
271
+ "eval_samples_per_second": 241.705,
272
+ "eval_steps_per_second": 15.122,
273
+ "eval_total": 15041,
274
+ "step": 2500
275
+ },
276
+ {
277
+ "epoch": 0.6887417218543046,
278
+ "grad_norm": 9.808700561523438,
279
+ "learning_rate": 9.86235761589404e-06,
280
+ "loss": 0.5304,
281
+ "step": 2600
282
+ },
283
+ {
284
+ "epoch": 0.7152317880794702,
285
+ "grad_norm": 7.898611068725586,
286
+ "learning_rate": 9.857059602649007e-06,
287
+ "loss": 0.4855,
288
+ "step": 2700
289
+ },
290
+ {
291
+ "epoch": 0.7417218543046358,
292
+ "grad_norm": 11.350944519042969,
293
+ "learning_rate": 9.851761589403974e-06,
294
+ "loss": 0.5395,
295
+ "step": 2800
296
+ },
297
+ {
298
+ "epoch": 0.7682119205298014,
299
+ "grad_norm": 14.699758529663086,
300
+ "learning_rate": 9.846463576158941e-06,
301
+ "loss": 0.4848,
302
+ "step": 2900
303
+ },
304
+ {
305
+ "epoch": 0.7947019867549668,
306
+ "grad_norm": 13.83389663696289,
307
+ "learning_rate": 9.841165562913908e-06,
308
+ "loss": 0.4822,
309
+ "step": 3000
310
+ },
311
+ {
312
+ "epoch": 0.7947019867549668,
313
+ "eval_HasAns_exact": 76.49757329964763,
314
+ "eval_HasAns_f1": 81.20875033119805,
315
+ "eval_HasAns_total": 15041,
316
+ "eval_best_exact": 76.49757329964763,
317
+ "eval_best_exact_thresh": 0.0,
318
+ "eval_best_f1": 81.20875033119805,
319
+ "eval_best_f1_thresh": 0.0,
320
+ "eval_exact": 76.49757329964763,
321
+ "eval_f1": 81.20875033119805,
322
+ "eval_loss": 0.4559445381164551,
323
+ "eval_runtime": 62.2922,
324
+ "eval_samples_per_second": 241.459,
325
+ "eval_steps_per_second": 15.106,
326
+ "eval_total": 15041,
327
+ "step": 3000
328
+ }
329
+ ],
330
+ "logging_steps": 100,
331
+ "max_steps": 188750,
332
+ "num_input_tokens_seen": 0,
333
+ "num_train_epochs": 50,
334
+ "save_steps": 500,
335
+ "stateful_callbacks": {
336
+ "EarlyStoppingCallback": {
337
+ "args": {
338
+ "early_stopping_patience": 5,
339
+ "early_stopping_threshold": 0.0
340
+ },
341
+ "attributes": {
342
+ "early_stopping_patience_counter": 0
343
+ }
344
+ },
345
+ "TrainerControl": {
346
+ "args": {
347
+ "should_epoch_stop": false,
348
+ "should_evaluate": false,
349
+ "should_log": false,
350
+ "should_save": true,
351
+ "should_training_stop": false
352
+ },
353
+ "attributes": {}
354
+ }
355
+ },
356
+ "total_flos": 1.2542697603072e+16,
357
+ "train_batch_size": 16,
358
+ "trial_name": null,
359
+ "trial_params": null
360
+ }
checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-3500/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-3500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:406f4c673c9df45ffa732bc1580883a7ab011119bcdcccb57c1e37514a6ff816
3
+ size 265470032
checkpoint-3500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f41843dc31644d96d49ab5f1e7a42b56d1aa00b56ec97470e1201619b24e042f
3
+ size 531000890
checkpoint-3500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b69bc7665acfd94a70aaf85c453fc865fd45799916ab4ddf488005643959a346
3
+ size 14244
checkpoint-3500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7f7582552700f51cc48ef227b31506b22ce32d930ba46bbd4939294889860d3
3
+ size 1064
checkpoint-3500/trainer_state.json ADDED
@@ -0,0 +1,413 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 81.20875033119805,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-3000",
4
+ "epoch": 0.9271523178807947,
5
+ "eval_steps": 500,
6
+ "global_step": 3500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ },
170
+ {
171
+ "epoch": 0.423841059602649,
172
+ "grad_norm": 8.173700332641602,
173
+ "learning_rate": 9.915337748344372e-06,
174
+ "loss": 0.584,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.4503311258278146,
179
+ "grad_norm": 6.5389723777771,
180
+ "learning_rate": 9.910039735099339e-06,
181
+ "loss": 0.5457,
182
+ "step": 1700
183
+ },
184
+ {
185
+ "epoch": 0.4768211920529801,
186
+ "grad_norm": 9.859861373901367,
187
+ "learning_rate": 9.904741721854306e-06,
188
+ "loss": 0.5417,
189
+ "step": 1800
190
+ },
191
+ {
192
+ "epoch": 0.5033112582781457,
193
+ "grad_norm": 13.78690242767334,
194
+ "learning_rate": 9.899443708609273e-06,
195
+ "loss": 0.5475,
196
+ "step": 1900
197
+ },
198
+ {
199
+ "epoch": 0.5298013245033113,
200
+ "grad_norm": 8.801095008850098,
201
+ "learning_rate": 9.89414569536424e-06,
202
+ "loss": 0.5814,
203
+ "step": 2000
204
+ },
205
+ {
206
+ "epoch": 0.5298013245033113,
207
+ "eval_HasAns_exact": 76.39784588790639,
208
+ "eval_HasAns_f1": 81.07514905827573,
209
+ "eval_HasAns_total": 15041,
210
+ "eval_best_exact": 76.39784588790639,
211
+ "eval_best_exact_thresh": 0.0,
212
+ "eval_best_f1": 81.07514905827573,
213
+ "eval_best_f1_thresh": 0.0,
214
+ "eval_exact": 76.39784588790639,
215
+ "eval_f1": 81.07514905827573,
216
+ "eval_loss": 0.4890206754207611,
217
+ "eval_runtime": 62.2295,
218
+ "eval_samples_per_second": 241.702,
219
+ "eval_steps_per_second": 15.121,
220
+ "eval_total": 15041,
221
+ "step": 2000
222
+ },
223
+ {
224
+ "epoch": 0.5562913907284768,
225
+ "grad_norm": 10.412147521972656,
226
+ "learning_rate": 9.888847682119206e-06,
227
+ "loss": 0.5275,
228
+ "step": 2100
229
+ },
230
+ {
231
+ "epoch": 0.5827814569536424,
232
+ "grad_norm": 8.767007827758789,
233
+ "learning_rate": 9.883549668874173e-06,
234
+ "loss": 0.4934,
235
+ "step": 2200
236
+ },
237
+ {
238
+ "epoch": 0.609271523178808,
239
+ "grad_norm": 11.056267738342285,
240
+ "learning_rate": 9.87825165562914e-06,
241
+ "loss": 0.5246,
242
+ "step": 2300
243
+ },
244
+ {
245
+ "epoch": 0.6357615894039735,
246
+ "grad_norm": 9.738068580627441,
247
+ "learning_rate": 9.872953642384107e-06,
248
+ "loss": 0.5127,
249
+ "step": 2400
250
+ },
251
+ {
252
+ "epoch": 0.6622516556291391,
253
+ "grad_norm": 15.755542755126953,
254
+ "learning_rate": 9.867655629139074e-06,
255
+ "loss": 0.5165,
256
+ "step": 2500
257
+ },
258
+ {
259
+ "epoch": 0.6622516556291391,
260
+ "eval_HasAns_exact": 76.21168805265607,
261
+ "eval_HasAns_f1": 80.96151211311172,
262
+ "eval_HasAns_total": 15041,
263
+ "eval_best_exact": 76.21168805265607,
264
+ "eval_best_exact_thresh": 0.0,
265
+ "eval_best_f1": 80.96151211311172,
266
+ "eval_best_f1_thresh": 0.0,
267
+ "eval_exact": 76.21168805265607,
268
+ "eval_f1": 80.96151211311172,
269
+ "eval_loss": 0.4729439914226532,
270
+ "eval_runtime": 62.2287,
271
+ "eval_samples_per_second": 241.705,
272
+ "eval_steps_per_second": 15.122,
273
+ "eval_total": 15041,
274
+ "step": 2500
275
+ },
276
+ {
277
+ "epoch": 0.6887417218543046,
278
+ "grad_norm": 9.808700561523438,
279
+ "learning_rate": 9.86235761589404e-06,
280
+ "loss": 0.5304,
281
+ "step": 2600
282
+ },
283
+ {
284
+ "epoch": 0.7152317880794702,
285
+ "grad_norm": 7.898611068725586,
286
+ "learning_rate": 9.857059602649007e-06,
287
+ "loss": 0.4855,
288
+ "step": 2700
289
+ },
290
+ {
291
+ "epoch": 0.7417218543046358,
292
+ "grad_norm": 11.350944519042969,
293
+ "learning_rate": 9.851761589403974e-06,
294
+ "loss": 0.5395,
295
+ "step": 2800
296
+ },
297
+ {
298
+ "epoch": 0.7682119205298014,
299
+ "grad_norm": 14.699758529663086,
300
+ "learning_rate": 9.846463576158941e-06,
301
+ "loss": 0.4848,
302
+ "step": 2900
303
+ },
304
+ {
305
+ "epoch": 0.7947019867549668,
306
+ "grad_norm": 13.83389663696289,
307
+ "learning_rate": 9.841165562913908e-06,
308
+ "loss": 0.4822,
309
+ "step": 3000
310
+ },
311
+ {
312
+ "epoch": 0.7947019867549668,
313
+ "eval_HasAns_exact": 76.49757329964763,
314
+ "eval_HasAns_f1": 81.20875033119805,
315
+ "eval_HasAns_total": 15041,
316
+ "eval_best_exact": 76.49757329964763,
317
+ "eval_best_exact_thresh": 0.0,
318
+ "eval_best_f1": 81.20875033119805,
319
+ "eval_best_f1_thresh": 0.0,
320
+ "eval_exact": 76.49757329964763,
321
+ "eval_f1": 81.20875033119805,
322
+ "eval_loss": 0.4559445381164551,
323
+ "eval_runtime": 62.2922,
324
+ "eval_samples_per_second": 241.459,
325
+ "eval_steps_per_second": 15.106,
326
+ "eval_total": 15041,
327
+ "step": 3000
328
+ },
329
+ {
330
+ "epoch": 0.8211920529801324,
331
+ "grad_norm": 8.879410743713379,
332
+ "learning_rate": 9.835867549668875e-06,
333
+ "loss": 0.4936,
334
+ "step": 3100
335
+ },
336
+ {
337
+ "epoch": 0.847682119205298,
338
+ "grad_norm": 8.931514739990234,
339
+ "learning_rate": 9.830569536423842e-06,
340
+ "loss": 0.5157,
341
+ "step": 3200
342
+ },
343
+ {
344
+ "epoch": 0.8741721854304636,
345
+ "grad_norm": 9.465814590454102,
346
+ "learning_rate": 9.825271523178809e-06,
347
+ "loss": 0.4865,
348
+ "step": 3300
349
+ },
350
+ {
351
+ "epoch": 0.9006622516556292,
352
+ "grad_norm": 7.991344451904297,
353
+ "learning_rate": 9.819973509933775e-06,
354
+ "loss": 0.4886,
355
+ "step": 3400
356
+ },
357
+ {
358
+ "epoch": 0.9271523178807947,
359
+ "grad_norm": 10.497336387634277,
360
+ "learning_rate": 9.814675496688742e-06,
361
+ "loss": 0.5015,
362
+ "step": 3500
363
+ },
364
+ {
365
+ "epoch": 0.9271523178807947,
366
+ "eval_HasAns_exact": 76.53081577022805,
367
+ "eval_HasAns_f1": 81.09624697164699,
368
+ "eval_HasAns_total": 15041,
369
+ "eval_best_exact": 76.53081577022805,
370
+ "eval_best_exact_thresh": 0.0,
371
+ "eval_best_f1": 81.09624697164699,
372
+ "eval_best_f1_thresh": 0.0,
373
+ "eval_exact": 76.53081577022805,
374
+ "eval_f1": 81.09624697164699,
375
+ "eval_loss": 0.43427419662475586,
376
+ "eval_runtime": 62.4053,
377
+ "eval_samples_per_second": 241.021,
378
+ "eval_steps_per_second": 15.079,
379
+ "eval_total": 15041,
380
+ "step": 3500
381
+ }
382
+ ],
383
+ "logging_steps": 100,
384
+ "max_steps": 188750,
385
+ "num_input_tokens_seen": 0,
386
+ "num_train_epochs": 50,
387
+ "save_steps": 500,
388
+ "stateful_callbacks": {
389
+ "EarlyStoppingCallback": {
390
+ "args": {
391
+ "early_stopping_patience": 5,
392
+ "early_stopping_threshold": 0.0
393
+ },
394
+ "attributes": {
395
+ "early_stopping_patience_counter": 1
396
+ }
397
+ },
398
+ "TrainerControl": {
399
+ "args": {
400
+ "should_epoch_stop": false,
401
+ "should_evaluate": false,
402
+ "should_log": false,
403
+ "should_save": true,
404
+ "should_training_stop": false
405
+ },
406
+ "attributes": {}
407
+ }
408
+ },
409
+ "total_flos": 1.4633147203584e+16,
410
+ "train_batch_size": 16,
411
+ "trial_name": null,
412
+ "trial_params": null
413
+ }
checkpoint-3500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-4000/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }
checkpoint-4000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddb4143263a3d9813da9fc2785a8047aa77b8a9cb479e8b363dc498bf8cf1bcc
3
+ size 265470032
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0d8752ef8dccf7f87700535d40f55f76214137676dff733cbbc214c9504e4e4
3
+ size 531000890
checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa13bdfaf5935e9a75248e1ecd7043df85ceb54867dd5dfe99d1a5b912176762
3
+ size 14244
checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:573171676354d55a02a52fea42775870f69491a693771f0c3ee0299bac15146e
3
+ size 1064
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,466 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 81.20875033119805,
3
+ "best_model_checkpoint": "distilbert-qasports/checkpoint-3000",
4
+ "epoch": 1.0596026490066226,
5
+ "eval_steps": 500,
6
+ "global_step": 4000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.026490066225165563,
13
+ "grad_norm": 10.109187126159668,
14
+ "learning_rate": 9.994754966887419e-06,
15
+ "loss": 0.8141,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.052980132450331126,
20
+ "grad_norm": 13.866031646728516,
21
+ "learning_rate": 9.989456953642385e-06,
22
+ "loss": 0.7347,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.07947019867549669,
27
+ "grad_norm": 7.728989124298096,
28
+ "learning_rate": 9.984158940397352e-06,
29
+ "loss": 0.6902,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.10596026490066225,
34
+ "grad_norm": 8.372330665588379,
35
+ "learning_rate": 9.978860927152319e-06,
36
+ "loss": 0.6645,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.13245033112582782,
41
+ "grad_norm": 10.106789588928223,
42
+ "learning_rate": 9.973562913907286e-06,
43
+ "loss": 0.6782,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.13245033112582782,
48
+ "eval_HasAns_exact": 74.40994614719766,
49
+ "eval_HasAns_f1": 79.39755830823303,
50
+ "eval_HasAns_total": 15041,
51
+ "eval_best_exact": 74.40994614719766,
52
+ "eval_best_exact_thresh": 0.0,
53
+ "eval_best_f1": 79.39755830823303,
54
+ "eval_best_f1_thresh": 0.0,
55
+ "eval_exact": 74.40994614719766,
56
+ "eval_f1": 79.39755830823303,
57
+ "eval_loss": 0.6026614308357239,
58
+ "eval_runtime": 62.2306,
59
+ "eval_samples_per_second": 241.698,
60
+ "eval_steps_per_second": 15.121,
61
+ "eval_total": 15041,
62
+ "step": 500
63
+ },
64
+ {
65
+ "epoch": 0.15894039735099338,
66
+ "grad_norm": 15.874868392944336,
67
+ "learning_rate": 9.968264900662253e-06,
68
+ "loss": 0.6139,
69
+ "step": 600
70
+ },
71
+ {
72
+ "epoch": 0.18543046357615894,
73
+ "grad_norm": 12.287043571472168,
74
+ "learning_rate": 9.963019867549669e-06,
75
+ "loss": 0.6147,
76
+ "step": 700
77
+ },
78
+ {
79
+ "epoch": 0.2119205298013245,
80
+ "grad_norm": 11.49232006072998,
81
+ "learning_rate": 9.957721854304636e-06,
82
+ "loss": 0.6033,
83
+ "step": 800
84
+ },
85
+ {
86
+ "epoch": 0.23841059602649006,
87
+ "grad_norm": 8.300566673278809,
88
+ "learning_rate": 9.952423841059603e-06,
89
+ "loss": 0.6254,
90
+ "step": 900
91
+ },
92
+ {
93
+ "epoch": 0.26490066225165565,
94
+ "grad_norm": 14.619403839111328,
95
+ "learning_rate": 9.94712582781457e-06,
96
+ "loss": 0.569,
97
+ "step": 1000
98
+ },
99
+ {
100
+ "epoch": 0.26490066225165565,
101
+ "eval_HasAns_exact": 75.10803802938635,
102
+ "eval_HasAns_f1": 80.10135733594169,
103
+ "eval_HasAns_total": 15041,
104
+ "eval_best_exact": 75.10803802938635,
105
+ "eval_best_exact_thresh": 0.0,
106
+ "eval_best_f1": 80.10135733594169,
107
+ "eval_best_f1_thresh": 0.0,
108
+ "eval_exact": 75.10803802938635,
109
+ "eval_f1": 80.10135733594169,
110
+ "eval_loss": 0.5508891344070435,
111
+ "eval_runtime": 62.3805,
112
+ "eval_samples_per_second": 241.117,
113
+ "eval_steps_per_second": 15.085,
114
+ "eval_total": 15041,
115
+ "step": 1000
116
+ },
117
+ {
118
+ "epoch": 0.2913907284768212,
119
+ "grad_norm": 10.70427131652832,
120
+ "learning_rate": 9.941827814569536e-06,
121
+ "loss": 0.6449,
122
+ "step": 1100
123
+ },
124
+ {
125
+ "epoch": 0.31788079470198677,
126
+ "grad_norm": 10.064720153808594,
127
+ "learning_rate": 9.936529801324503e-06,
128
+ "loss": 0.5613,
129
+ "step": 1200
130
+ },
131
+ {
132
+ "epoch": 0.3443708609271523,
133
+ "grad_norm": 8.998592376708984,
134
+ "learning_rate": 9.93123178807947e-06,
135
+ "loss": 0.5896,
136
+ "step": 1300
137
+ },
138
+ {
139
+ "epoch": 0.3708609271523179,
140
+ "grad_norm": 11.581778526306152,
141
+ "learning_rate": 9.925933774834437e-06,
142
+ "loss": 0.5767,
143
+ "step": 1400
144
+ },
145
+ {
146
+ "epoch": 0.3973509933774834,
147
+ "grad_norm": 7.466866493225098,
148
+ "learning_rate": 9.920635761589405e-06,
149
+ "loss": 0.5821,
150
+ "step": 1500
151
+ },
152
+ {
153
+ "epoch": 0.3973509933774834,
154
+ "eval_HasAns_exact": 75.55348713516389,
155
+ "eval_HasAns_f1": 80.3558190310888,
156
+ "eval_HasAns_total": 15041,
157
+ "eval_best_exact": 75.55348713516389,
158
+ "eval_best_exact_thresh": 0.0,
159
+ "eval_best_f1": 80.3558190310888,
160
+ "eval_best_f1_thresh": 0.0,
161
+ "eval_exact": 75.55348713516389,
162
+ "eval_f1": 80.3558190310888,
163
+ "eval_loss": 0.5195178985595703,
164
+ "eval_runtime": 62.2517,
165
+ "eval_samples_per_second": 241.616,
166
+ "eval_steps_per_second": 15.116,
167
+ "eval_total": 15041,
168
+ "step": 1500
169
+ },
170
+ {
171
+ "epoch": 0.423841059602649,
172
+ "grad_norm": 8.173700332641602,
173
+ "learning_rate": 9.915337748344372e-06,
174
+ "loss": 0.584,
175
+ "step": 1600
176
+ },
177
+ {
178
+ "epoch": 0.4503311258278146,
179
+ "grad_norm": 6.5389723777771,
180
+ "learning_rate": 9.910039735099339e-06,
181
+ "loss": 0.5457,
182
+ "step": 1700
183
+ },
184
+ {
185
+ "epoch": 0.4768211920529801,
186
+ "grad_norm": 9.859861373901367,
187
+ "learning_rate": 9.904741721854306e-06,
188
+ "loss": 0.5417,
189
+ "step": 1800
190
+ },
191
+ {
192
+ "epoch": 0.5033112582781457,
193
+ "grad_norm": 13.78690242767334,
194
+ "learning_rate": 9.899443708609273e-06,
195
+ "loss": 0.5475,
196
+ "step": 1900
197
+ },
198
+ {
199
+ "epoch": 0.5298013245033113,
200
+ "grad_norm": 8.801095008850098,
201
+ "learning_rate": 9.89414569536424e-06,
202
+ "loss": 0.5814,
203
+ "step": 2000
204
+ },
205
+ {
206
+ "epoch": 0.5298013245033113,
207
+ "eval_HasAns_exact": 76.39784588790639,
208
+ "eval_HasAns_f1": 81.07514905827573,
209
+ "eval_HasAns_total": 15041,
210
+ "eval_best_exact": 76.39784588790639,
211
+ "eval_best_exact_thresh": 0.0,
212
+ "eval_best_f1": 81.07514905827573,
213
+ "eval_best_f1_thresh": 0.0,
214
+ "eval_exact": 76.39784588790639,
215
+ "eval_f1": 81.07514905827573,
216
+ "eval_loss": 0.4890206754207611,
217
+ "eval_runtime": 62.2295,
218
+ "eval_samples_per_second": 241.702,
219
+ "eval_steps_per_second": 15.121,
220
+ "eval_total": 15041,
221
+ "step": 2000
222
+ },
223
+ {
224
+ "epoch": 0.5562913907284768,
225
+ "grad_norm": 10.412147521972656,
226
+ "learning_rate": 9.888847682119206e-06,
227
+ "loss": 0.5275,
228
+ "step": 2100
229
+ },
230
+ {
231
+ "epoch": 0.5827814569536424,
232
+ "grad_norm": 8.767007827758789,
233
+ "learning_rate": 9.883549668874173e-06,
234
+ "loss": 0.4934,
235
+ "step": 2200
236
+ },
237
+ {
238
+ "epoch": 0.609271523178808,
239
+ "grad_norm": 11.056267738342285,
240
+ "learning_rate": 9.87825165562914e-06,
241
+ "loss": 0.5246,
242
+ "step": 2300
243
+ },
244
+ {
245
+ "epoch": 0.6357615894039735,
246
+ "grad_norm": 9.738068580627441,
247
+ "learning_rate": 9.872953642384107e-06,
248
+ "loss": 0.5127,
249
+ "step": 2400
250
+ },
251
+ {
252
+ "epoch": 0.6622516556291391,
253
+ "grad_norm": 15.755542755126953,
254
+ "learning_rate": 9.867655629139074e-06,
255
+ "loss": 0.5165,
256
+ "step": 2500
257
+ },
258
+ {
259
+ "epoch": 0.6622516556291391,
260
+ "eval_HasAns_exact": 76.21168805265607,
261
+ "eval_HasAns_f1": 80.96151211311172,
262
+ "eval_HasAns_total": 15041,
263
+ "eval_best_exact": 76.21168805265607,
264
+ "eval_best_exact_thresh": 0.0,
265
+ "eval_best_f1": 80.96151211311172,
266
+ "eval_best_f1_thresh": 0.0,
267
+ "eval_exact": 76.21168805265607,
268
+ "eval_f1": 80.96151211311172,
269
+ "eval_loss": 0.4729439914226532,
270
+ "eval_runtime": 62.2287,
271
+ "eval_samples_per_second": 241.705,
272
+ "eval_steps_per_second": 15.122,
273
+ "eval_total": 15041,
274
+ "step": 2500
275
+ },
276
+ {
277
+ "epoch": 0.6887417218543046,
278
+ "grad_norm": 9.808700561523438,
279
+ "learning_rate": 9.86235761589404e-06,
280
+ "loss": 0.5304,
281
+ "step": 2600
282
+ },
283
+ {
284
+ "epoch": 0.7152317880794702,
285
+ "grad_norm": 7.898611068725586,
286
+ "learning_rate": 9.857059602649007e-06,
287
+ "loss": 0.4855,
288
+ "step": 2700
289
+ },
290
+ {
291
+ "epoch": 0.7417218543046358,
292
+ "grad_norm": 11.350944519042969,
293
+ "learning_rate": 9.851761589403974e-06,
294
+ "loss": 0.5395,
295
+ "step": 2800
296
+ },
297
+ {
298
+ "epoch": 0.7682119205298014,
299
+ "grad_norm": 14.699758529663086,
300
+ "learning_rate": 9.846463576158941e-06,
301
+ "loss": 0.4848,
302
+ "step": 2900
303
+ },
304
+ {
305
+ "epoch": 0.7947019867549668,
306
+ "grad_norm": 13.83389663696289,
307
+ "learning_rate": 9.841165562913908e-06,
308
+ "loss": 0.4822,
309
+ "step": 3000
310
+ },
311
+ {
312
+ "epoch": 0.7947019867549668,
313
+ "eval_HasAns_exact": 76.49757329964763,
314
+ "eval_HasAns_f1": 81.20875033119805,
315
+ "eval_HasAns_total": 15041,
316
+ "eval_best_exact": 76.49757329964763,
317
+ "eval_best_exact_thresh": 0.0,
318
+ "eval_best_f1": 81.20875033119805,
319
+ "eval_best_f1_thresh": 0.0,
320
+ "eval_exact": 76.49757329964763,
321
+ "eval_f1": 81.20875033119805,
322
+ "eval_loss": 0.4559445381164551,
323
+ "eval_runtime": 62.2922,
324
+ "eval_samples_per_second": 241.459,
325
+ "eval_steps_per_second": 15.106,
326
+ "eval_total": 15041,
327
+ "step": 3000
328
+ },
329
+ {
330
+ "epoch": 0.8211920529801324,
331
+ "grad_norm": 8.879410743713379,
332
+ "learning_rate": 9.835867549668875e-06,
333
+ "loss": 0.4936,
334
+ "step": 3100
335
+ },
336
+ {
337
+ "epoch": 0.847682119205298,
338
+ "grad_norm": 8.931514739990234,
339
+ "learning_rate": 9.830569536423842e-06,
340
+ "loss": 0.5157,
341
+ "step": 3200
342
+ },
343
+ {
344
+ "epoch": 0.8741721854304636,
345
+ "grad_norm": 9.465814590454102,
346
+ "learning_rate": 9.825271523178809e-06,
347
+ "loss": 0.4865,
348
+ "step": 3300
349
+ },
350
+ {
351
+ "epoch": 0.9006622516556292,
352
+ "grad_norm": 7.991344451904297,
353
+ "learning_rate": 9.819973509933775e-06,
354
+ "loss": 0.4886,
355
+ "step": 3400
356
+ },
357
+ {
358
+ "epoch": 0.9271523178807947,
359
+ "grad_norm": 10.497336387634277,
360
+ "learning_rate": 9.814675496688742e-06,
361
+ "loss": 0.5015,
362
+ "step": 3500
363
+ },
364
+ {
365
+ "epoch": 0.9271523178807947,
366
+ "eval_HasAns_exact": 76.53081577022805,
367
+ "eval_HasAns_f1": 81.09624697164699,
368
+ "eval_HasAns_total": 15041,
369
+ "eval_best_exact": 76.53081577022805,
370
+ "eval_best_exact_thresh": 0.0,
371
+ "eval_best_f1": 81.09624697164699,
372
+ "eval_best_f1_thresh": 0.0,
373
+ "eval_exact": 76.53081577022805,
374
+ "eval_f1": 81.09624697164699,
375
+ "eval_loss": 0.43427419662475586,
376
+ "eval_runtime": 62.4053,
377
+ "eval_samples_per_second": 241.021,
378
+ "eval_steps_per_second": 15.079,
379
+ "eval_total": 15041,
380
+ "step": 3500
381
+ },
382
+ {
383
+ "epoch": 0.9536423841059603,
384
+ "grad_norm": 9.306705474853516,
385
+ "learning_rate": 9.809377483443709e-06,
386
+ "loss": 0.4927,
387
+ "step": 3600
388
+ },
389
+ {
390
+ "epoch": 0.9801324503311258,
391
+ "grad_norm": 8.719966888427734,
392
+ "learning_rate": 9.804079470198676e-06,
393
+ "loss": 0.4859,
394
+ "step": 3700
395
+ },
396
+ {
397
+ "epoch": 1.0066225165562914,
398
+ "grad_norm": 11.576972961425781,
399
+ "learning_rate": 9.798781456953643e-06,
400
+ "loss": 0.4501,
401
+ "step": 3800
402
+ },
403
+ {
404
+ "epoch": 1.033112582781457,
405
+ "grad_norm": 7.64553689956665,
406
+ "learning_rate": 9.79348344370861e-06,
407
+ "loss": 0.4127,
408
+ "step": 3900
409
+ },
410
+ {
411
+ "epoch": 1.0596026490066226,
412
+ "grad_norm": 8.646525382995605,
413
+ "learning_rate": 9.788185430463577e-06,
414
+ "loss": 0.36,
415
+ "step": 4000
416
+ },
417
+ {
418
+ "epoch": 1.0596026490066226,
419
+ "eval_HasAns_exact": 76.53081577022805,
420
+ "eval_HasAns_f1": 81.08277260501552,
421
+ "eval_HasAns_total": 15041,
422
+ "eval_best_exact": 76.53081577022805,
423
+ "eval_best_exact_thresh": 0.0,
424
+ "eval_best_f1": 81.08277260501552,
425
+ "eval_best_f1_thresh": 0.0,
426
+ "eval_exact": 76.53081577022805,
427
+ "eval_f1": 81.08277260501552,
428
+ "eval_loss": 0.43494778871536255,
429
+ "eval_runtime": 62.3029,
430
+ "eval_samples_per_second": 241.417,
431
+ "eval_steps_per_second": 15.104,
432
+ "eval_total": 15041,
433
+ "step": 4000
434
+ }
435
+ ],
436
+ "logging_steps": 100,
437
+ "max_steps": 188750,
438
+ "num_input_tokens_seen": 0,
439
+ "num_train_epochs": 50,
440
+ "save_steps": 500,
441
+ "stateful_callbacks": {
442
+ "EarlyStoppingCallback": {
443
+ "args": {
444
+ "early_stopping_patience": 5,
445
+ "early_stopping_threshold": 0.0
446
+ },
447
+ "attributes": {
448
+ "early_stopping_patience_counter": 2
449
+ }
450
+ },
451
+ "TrainerControl": {
452
+ "args": {
453
+ "should_epoch_stop": false,
454
+ "should_evaluate": false,
455
+ "should_log": false,
456
+ "should_save": true,
457
+ "should_training_stop": false
458
+ },
459
+ "attributes": {}
460
+ }
461
+ },
462
+ "total_flos": 1.6722812885495808e+16,
463
+ "train_batch_size": 16,
464
+ "trial_name": null,
465
+ "trial_params": null
466
+ }
checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb88a2f910cf2cd3f6d22808102adc97823c9e497aa6be1cbe2c9cda9513a4ba
3
+ size 5304
checkpoint-4500/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased-distilled-squad",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForQuestionAnswering"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "initializer_range": 0.02,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "distilbert",
14
+ "n_heads": 12,
15
+ "n_layers": 6,
16
+ "pad_token_id": 0,
17
+ "qa_dropout": 0.1,
18
+ "seq_classif_dropout": 0.2,
19
+ "sinusoidal_pos_embds": false,
20
+ "tie_weights_": true,
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.48.3",
23
+ "vocab_size": 30522
24
+ }