Hugging Face's logo

DylanJHJ
/
bert-base-final-v0

Model card Files Community
DylanJHJ commited on
Commit
7333e13
·
1 Parent(s): 30ed7f3

upload checkpoints

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +17 -0
  2. CHECKPOINTS_DECRIPTIONS.md +0 -0
  3. README.md +3 -3
  4. esnli-zs-highlighter/checkpoint-10000/config.json +27 -0
  5. esnli-zs-highlighter/checkpoint-10000/optimizer.pt +3 -0
  6. esnli-zs-highlighter/checkpoint-10000/pytorch_model.bin +3 -0
  7. esnli-zs-highlighter/checkpoint-10000/rng_state.pth +3 -0
  8. esnli-zs-highlighter/checkpoint-10000/scheduler.pt +3 -0
  9. esnli-zs-highlighter/checkpoint-10000/trainer_state.json +152 -0
  10. esnli-zs-highlighter/checkpoint-10000/training_args.bin +3 -0
  11. esnli-zs-highlighter/checkpoint-15000/config.json +27 -0
  12. esnli-zs-highlighter/checkpoint-15000/optimizer.pt +3 -0
  13. esnli-zs-highlighter/checkpoint-15000/pytorch_model.bin +3 -0
  14. esnli-zs-highlighter/checkpoint-15000/rng_state.pth +3 -0
  15. esnli-zs-highlighter/checkpoint-15000/scheduler.pt +3 -0
  16. esnli-zs-highlighter/checkpoint-15000/trainer_state.json +220 -0
  17. esnli-zs-highlighter/checkpoint-15000/training_args.bin +3 -0
  18. esnli-zs-highlighter/checkpoint-5000/config.json +27 -0
  19. esnli-zs-highlighter/checkpoint-5000/optimizer.pt +3 -0
  20. esnli-zs-highlighter/checkpoint-5000/pytorch_model.bin +3 -0
  21. esnli-zs-highlighter/checkpoint-5000/rng_state.pth +3 -0
  22. esnli-zs-highlighter/checkpoint-5000/scheduler.pt +3 -0
  23. esnli-zs-highlighter/checkpoint-5000/trainer_state.json +84 -0
  24. esnli-zs-highlighter/checkpoint-5000/training_args.bin +3 -0
  25. from-scratch-ll/checkpoint-3000/config.json +27 -0
  26. from-scratch-ll/checkpoint-3000/optimizer.pt +3 -0
  27. from-scratch-ll/checkpoint-3000/pytorch_model.bin +3 -0
  28. from-scratch-ll/checkpoint-3000/rng_state.pth +3 -0
  29. from-scratch-ll/checkpoint-3000/scheduler.pt +3 -0
  30. from-scratch-ll/checkpoint-3000/trainer_state.json +60 -0
  31. from-scratch-ll/checkpoint-3000/training_args.bin +3 -0
  32. from-scratch-ll/checkpoint-6000/config.json +27 -0
  33. from-scratch-ll/checkpoint-6000/optimizer.pt +3 -0
  34. from-scratch-ll/checkpoint-6000/pytorch_model.bin +3 -0
  35. from-scratch-ll/checkpoint-6000/rng_state.pth +3 -0
  36. from-scratch-ll/checkpoint-6000/scheduler.pt +3 -0
  37. from-scratch-ll/checkpoint-6000/trainer_state.json +104 -0
  38. from-scratch-ll/checkpoint-6000/training_args.bin +3 -0
  39. from-scratch-ll/checkpoint-9000/config.json +27 -0
  40. from-scratch-ll/checkpoint-9000/optimizer.pt +3 -0
  41. from-scratch-ll/checkpoint-9000/pytorch_model.bin +3 -0
  42. from-scratch-ll/checkpoint-9000/rng_state.pth +3 -0
  43. from-scratch-ll/checkpoint-9000/scheduler.pt +3 -0
  44. from-scratch-ll/checkpoint-9000/trainer_state.json +148 -0
  45. from-scratch-ll/checkpoint-9000/training_args.bin +3 -0
  46. from-scratch-sl-balance/checkpoint-3000/config.json +27 -0
  47. from-scratch-sl-balance/checkpoint-3000/optimizer.pt +3 -0
  48. from-scratch-sl-balance/checkpoint-3000/pytorch_model.bin +3 -0
  49. from-scratch-sl-balance/checkpoint-3000/rng_state.pth +3 -0
  50. from-scratch-sl-balance/checkpoint-3000/scheduler.pt +3 -0
.gitattributes CHANGED
@@ -33,3 +33,20 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ CHECKPOINTS_DECRIPTIONS.md filter=lfs diff=lfs merge=lfs -text
37
+ README.md filter=lfs diff=lfs merge=lfs -text
38
+ esnli-zs-highlighter filter=lfs diff=lfs merge=lfs -text
39
+ further-finetune-sl-smooth-0.1 filter=lfs diff=lfs merge=lfs -text
40
+ further-finetune-sl-smooth-0.25 filter=lfs diff=lfs merge=lfs -text
41
+ from-scratch filter=lfs diff=lfs merge=lfs -text
42
+ from-scratch-sl-balance filter=lfs diff=lfs merge=lfs -text
43
+ further-finetune-sl-balance filter=lfs diff=lfs merge=lfs -text
44
+ further-finetune-sl-soft filter=lfs diff=lfs merge=lfs -text
45
+ further-finetune-sl-tunned filter=lfs diff=lfs merge=lfs -text
46
+ further-finetune-ll filter=lfs diff=lfs merge=lfs -text
47
+ further-finetune-sl-mod filter=lfs diff=lfs merge=lfs -text
48
+ from-scratch-ll filter=lfs diff=lfs merge=lfs -text
49
+ from-scratch-sl-soft filter=lfs diff=lfs merge=lfs -text
50
+ further-finetune filter=lfs diff=lfs merge=lfs -text
51
+ further-finetune-sl-smooth-0.5 filter=lfs diff=lfs merge=lfs -text
52
+ further-finetune-sl-smooth-2 filter=lfs diff=lfs merge=lfs -text
CHECKPOINTS_DECRIPTIONS.md ADDED
File without changes
README.md CHANGED
@@ -1,3 +1,3 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98b45ea81164d1e1a1dd82255207053b15cd6c69d922a1c5cf3387ce604d4b74
3
+ size 28
esnli-zs-highlighter/checkpoint-10000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
esnli-zs-highlighter/checkpoint-10000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9178293fb30703c9bce1477b5739aed4350941cc5dcfae835bc5d28b59852436
3
+ size 871260901
esnli-zs-highlighter/checkpoint-10000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6aee7f5935b6648d874ff6faae6c05b506a440b728d2ccd185a57d51d5423938
3
+ size 435643185
esnli-zs-highlighter/checkpoint-10000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ae9fa494f9100e37adce915f23854727205f49a2a8d0ea352b87f222b075898
3
+ size 14503
esnli-zs-highlighter/checkpoint-10000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23b8d40a8a5b8586f23fd1ccfcb57ac99e3456f3205969b16746423e888cad02
3
+ size 623
esnli-zs-highlighter/checkpoint-10000/trainer_state.json ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.310272536687631,
5
+ "global_step": 10000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.07,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.4176,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.13,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.3487,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.2,
24
+ "learning_rate": 4.8214285714285716e-05,
25
+ "loss": 0.3403,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.26,
30
+ "learning_rate": 4.642857142857143e-05,
31
+ "loss": 0.3348,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.33,
36
+ "learning_rate": 4.464285714285715e-05,
37
+ "loss": 0.3279,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.39,
42
+ "learning_rate": 4.2857142857142856e-05,
43
+ "loss": 0.3278,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.46,
48
+ "learning_rate": 4.107142857142857e-05,
49
+ "loss": 0.3253,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.52,
54
+ "learning_rate": 3.928571428571429e-05,
55
+ "loss": 0.3211,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.59,
60
+ "learning_rate": 3.7500000000000003e-05,
61
+ "loss": 0.3209,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.66,
66
+ "learning_rate": 3.571428571428572e-05,
67
+ "loss": 0.3251,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.66,
72
+ "eval_loss": 0.3245891034603119,
73
+ "eval_runtime": 22.7214,
74
+ "eval_samples_per_second": 144.269,
75
+ "eval_steps_per_second": 36.089,
76
+ "step": 5000
77
+ },
78
+ {
79
+ "epoch": 0.72,
80
+ "learning_rate": 3.392857142857143e-05,
81
+ "loss": 0.3185,
82
+ "step": 5500
83
+ },
84
+ {
85
+ "epoch": 0.79,
86
+ "learning_rate": 3.2142857142857144e-05,
87
+ "loss": 0.3166,
88
+ "step": 6000
89
+ },
90
+ {
91
+ "epoch": 0.85,
92
+ "learning_rate": 3.0357142857142857e-05,
93
+ "loss": 0.3176,
94
+ "step": 6500
95
+ },
96
+ {
97
+ "epoch": 0.92,
98
+ "learning_rate": 2.857142857142857e-05,
99
+ "loss": 0.3121,
100
+ "step": 7000
101
+ },
102
+ {
103
+ "epoch": 0.98,
104
+ "learning_rate": 2.6785714285714288e-05,
105
+ "loss": 0.3167,
106
+ "step": 7500
107
+ },
108
+ {
109
+ "epoch": 1.05,
110
+ "learning_rate": 2.5e-05,
111
+ "loss": 0.2982,
112
+ "step": 8000
113
+ },
114
+ {
115
+ "epoch": 1.11,
116
+ "learning_rate": 2.3214285714285715e-05,
117
+ "loss": 0.2894,
118
+ "step": 8500
119
+ },
120
+ {
121
+ "epoch": 1.18,
122
+ "learning_rate": 2.1428571428571428e-05,
123
+ "loss": 0.289,
124
+ "step": 9000
125
+ },
126
+ {
127
+ "epoch": 1.24,
128
+ "learning_rate": 1.9642857142857145e-05,
129
+ "loss": 0.2882,
130
+ "step": 9500
131
+ },
132
+ {
133
+ "epoch": 1.31,
134
+ "learning_rate": 1.785714285714286e-05,
135
+ "loss": 0.286,
136
+ "step": 10000
137
+ },
138
+ {
139
+ "epoch": 1.31,
140
+ "eval_loss": 0.3317989408969879,
141
+ "eval_runtime": 22.4167,
142
+ "eval_samples_per_second": 146.23,
143
+ "eval_steps_per_second": 36.58,
144
+ "step": 10000
145
+ }
146
+ ],
147
+ "max_steps": 15000,
148
+ "num_train_epochs": 2,
149
+ "total_flos": 3.159952316520941e+16,
150
+ "trial_name": null,
151
+ "trial_params": null
152
+ }
esnli-zs-highlighter/checkpoint-10000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd63b64122109a9b0f524808fa204698bbcf8f060694881a0cdf278ed40d9426
3
+ size 3311
esnli-zs-highlighter/checkpoint-15000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
esnli-zs-highlighter/checkpoint-15000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6810b7d1291005269822f7a5098f25e21a239f11c5048fba5f6356df83bda065
3
+ size 871260901
esnli-zs-highlighter/checkpoint-15000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:211b56be9089f514b1fb94e698fb9e7536a42a78f025a94b16ad7a78c3113ce6
3
+ size 435643185
esnli-zs-highlighter/checkpoint-15000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c594bf3718517e8e83f44736e280b735b6eed7512710d74da7e4e48262d1e6e9
3
+ size 14503
esnli-zs-highlighter/checkpoint-15000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:085ad382975075e849be4ca66a1a28387318b75408196caf2b382378a2ec04ed
3
+ size 623
esnli-zs-highlighter/checkpoint-15000/trainer_state.json ADDED
@@ -0,0 +1,220 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.9654088050314464,
5
+ "global_step": 15000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.07,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.4176,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.13,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.3487,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.2,
24
+ "learning_rate": 4.8214285714285716e-05,
25
+ "loss": 0.3403,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.26,
30
+ "learning_rate": 4.642857142857143e-05,
31
+ "loss": 0.3348,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.33,
36
+ "learning_rate": 4.464285714285715e-05,
37
+ "loss": 0.3279,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.39,
42
+ "learning_rate": 4.2857142857142856e-05,
43
+ "loss": 0.3278,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.46,
48
+ "learning_rate": 4.107142857142857e-05,
49
+ "loss": 0.3253,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.52,
54
+ "learning_rate": 3.928571428571429e-05,
55
+ "loss": 0.3211,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.59,
60
+ "learning_rate": 3.7500000000000003e-05,
61
+ "loss": 0.3209,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.66,
66
+ "learning_rate": 3.571428571428572e-05,
67
+ "loss": 0.3251,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.66,
72
+ "eval_loss": 0.3245891034603119,
73
+ "eval_runtime": 22.7214,
74
+ "eval_samples_per_second": 144.269,
75
+ "eval_steps_per_second": 36.089,
76
+ "step": 5000
77
+ },
78
+ {
79
+ "epoch": 0.72,
80
+ "learning_rate": 3.392857142857143e-05,
81
+ "loss": 0.3185,
82
+ "step": 5500
83
+ },
84
+ {
85
+ "epoch": 0.79,
86
+ "learning_rate": 3.2142857142857144e-05,
87
+ "loss": 0.3166,
88
+ "step": 6000
89
+ },
90
+ {
91
+ "epoch": 0.85,
92
+ "learning_rate": 3.0357142857142857e-05,
93
+ "loss": 0.3176,
94
+ "step": 6500
95
+ },
96
+ {
97
+ "epoch": 0.92,
98
+ "learning_rate": 2.857142857142857e-05,
99
+ "loss": 0.3121,
100
+ "step": 7000
101
+ },
102
+ {
103
+ "epoch": 0.98,
104
+ "learning_rate": 2.6785714285714288e-05,
105
+ "loss": 0.3167,
106
+ "step": 7500
107
+ },
108
+ {
109
+ "epoch": 1.05,
110
+ "learning_rate": 2.5e-05,
111
+ "loss": 0.2982,
112
+ "step": 8000
113
+ },
114
+ {
115
+ "epoch": 1.11,
116
+ "learning_rate": 2.3214285714285715e-05,
117
+ "loss": 0.2894,
118
+ "step": 8500
119
+ },
120
+ {
121
+ "epoch": 1.18,
122
+ "learning_rate": 2.1428571428571428e-05,
123
+ "loss": 0.289,
124
+ "step": 9000
125
+ },
126
+ {
127
+ "epoch": 1.24,
128
+ "learning_rate": 1.9642857142857145e-05,
129
+ "loss": 0.2882,
130
+ "step": 9500
131
+ },
132
+ {
133
+ "epoch": 1.31,
134
+ "learning_rate": 1.785714285714286e-05,
135
+ "loss": 0.286,
136
+ "step": 10000
137
+ },
138
+ {
139
+ "epoch": 1.31,
140
+ "eval_loss": 0.3317989408969879,
141
+ "eval_runtime": 22.4167,
142
+ "eval_samples_per_second": 146.23,
143
+ "eval_steps_per_second": 36.58,
144
+ "step": 10000
145
+ },
146
+ {
147
+ "epoch": 1.38,
148
+ "learning_rate": 1.6071428571428572e-05,
149
+ "loss": 0.2899,
150
+ "step": 10500
151
+ },
152
+ {
153
+ "epoch": 1.44,
154
+ "learning_rate": 1.4285714285714285e-05,
155
+ "loss": 0.2901,
156
+ "step": 11000
157
+ },
158
+ {
159
+ "epoch": 1.51,
160
+ "learning_rate": 1.25e-05,
161
+ "loss": 0.2877,
162
+ "step": 11500
163
+ },
164
+ {
165
+ "epoch": 1.57,
166
+ "learning_rate": 1.0714285714285714e-05,
167
+ "loss": 0.2877,
168
+ "step": 12000
169
+ },
170
+ {
171
+ "epoch": 1.64,
172
+ "learning_rate": 8.92857142857143e-06,
173
+ "loss": 0.2878,
174
+ "step": 12500
175
+ },
176
+ {
177
+ "epoch": 1.7,
178
+ "learning_rate": 7.142857142857143e-06,
179
+ "loss": 0.2883,
180
+ "step": 13000
181
+ },
182
+ {
183
+ "epoch": 1.77,
184
+ "learning_rate": 5.357142857142857e-06,
185
+ "loss": 0.2872,
186
+ "step": 13500
187
+ },
188
+ {
189
+ "epoch": 1.83,
190
+ "learning_rate": 3.5714285714285714e-06,
191
+ "loss": 0.2847,
192
+ "step": 14000
193
+ },
194
+ {
195
+ "epoch": 1.9,
196
+ "learning_rate": 1.7857142857142857e-06,
197
+ "loss": 0.2868,
198
+ "step": 14500
199
+ },
200
+ {
201
+ "epoch": 1.97,
202
+ "learning_rate": 0.0,
203
+ "loss": 0.284,
204
+ "step": 15000
205
+ },
206
+ {
207
+ "epoch": 1.97,
208
+ "eval_loss": 0.317633718252182,
209
+ "eval_runtime": 22.7238,
210
+ "eval_samples_per_second": 144.254,
211
+ "eval_steps_per_second": 36.085,
212
+ "step": 15000
213
+ }
214
+ ],
215
+ "max_steps": 15000,
216
+ "num_train_epochs": 2,
217
+ "total_flos": 4.739981142408941e+16,
218
+ "trial_name": null,
219
+ "trial_params": null
220
+ }
esnli-zs-highlighter/checkpoint-15000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd63b64122109a9b0f524808fa204698bbcf8f060694881a0cdf278ed40d9426
3
+ size 3311
esnli-zs-highlighter/checkpoint-5000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
esnli-zs-highlighter/checkpoint-5000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:875089a6f40a3752be8e415537a7613d183c7273b69a96a52332dfb71555d7f7
3
+ size 871260901
esnli-zs-highlighter/checkpoint-5000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac02eb656c58c5c1ac713b3f5dd81011ae95137a62654961dca5e350d865f9bb
3
+ size 435643185
esnli-zs-highlighter/checkpoint-5000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35ff5535b37e56eddf1a1a3c0aacc8dedd9b9684e3daffc3202577db835af5ce
3
+ size 14503
esnli-zs-highlighter/checkpoint-5000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa2c955e45e69585032b63891fe8f0c9e66b397fd5c837a8f816cc87cd35d718
3
+ size 623
esnli-zs-highlighter/checkpoint-5000/trainer_state.json ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.6551362683438156,
5
+ "global_step": 5000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.07,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.4176,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.13,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.3487,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.2,
24
+ "learning_rate": 4.8214285714285716e-05,
25
+ "loss": 0.3403,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.26,
30
+ "learning_rate": 4.642857142857143e-05,
31
+ "loss": 0.3348,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.33,
36
+ "learning_rate": 4.464285714285715e-05,
37
+ "loss": 0.3279,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.39,
42
+ "learning_rate": 4.2857142857142856e-05,
43
+ "loss": 0.3278,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.46,
48
+ "learning_rate": 4.107142857142857e-05,
49
+ "loss": 0.3253,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.52,
54
+ "learning_rate": 3.928571428571429e-05,
55
+ "loss": 0.3211,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.59,
60
+ "learning_rate": 3.7500000000000003e-05,
61
+ "loss": 0.3209,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.66,
66
+ "learning_rate": 3.571428571428572e-05,
67
+ "loss": 0.3251,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.66,
72
+ "eval_loss": 0.3245891034603119,
73
+ "eval_runtime": 22.7214,
74
+ "eval_samples_per_second": 144.269,
75
+ "eval_steps_per_second": 36.089,
76
+ "step": 5000
77
+ }
78
+ ],
79
+ "max_steps": 15000,
80
+ "num_train_epochs": 2,
81
+ "total_flos": 1.580028825888e+16,
82
+ "trial_name": null,
83
+ "trial_params": null
84
+ }
esnli-zs-highlighter/checkpoint-5000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd63b64122109a9b0f524808fa204698bbcf8f060694881a0cdf278ed40d9426
3
+ size 3311
from-scratch-ll/checkpoint-3000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
from-scratch-ll/checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7169a56c2cedd2cf2b388b655920ca469102a9e719c63e2f129822571f48c75
3
+ size 871260901
from-scratch-ll/checkpoint-3000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa2834e62a49ddde9f4b5e93313c184acee0b290b8a85b308a361af090144c81
3
+ size 435643185
from-scratch-ll/checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc67bb6793cc9e3a2453616d2e0bb667dbe46c82a25d258f3f82f7c00e57b78d
3
+ size 14503
from-scratch-ll/checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e5d2b076608b10220351f7a090fe60ee9462654745d3c6d779eef696c9c360
3
+ size 623
from-scratch-ll/checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0093770931011385,
5
+ "global_step": 3000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.33,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.2269,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.67,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.0536,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 1.0,
24
+ "learning_rate": 4.722222222222222e-05,
25
+ "loss": 0.0352,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 1.34,
30
+ "learning_rate": 4.4444444444444447e-05,
31
+ "loss": 0.0207,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 1.67,
36
+ "learning_rate": 4.166666666666667e-05,
37
+ "loss": 0.0201,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 2.01,
42
+ "learning_rate": 3.888888888888889e-05,
43
+ "loss": 0.0175,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 2.01,
48
+ "eval_loss": 1.1985770463943481,
49
+ "eval_runtime": 44.1956,
50
+ "eval_samples_per_second": 74.17,
51
+ "eval_steps_per_second": 18.554,
52
+ "step": 3000
53
+ }
54
+ ],
55
+ "max_steps": 10000,
56
+ "num_train_epochs": 7,
57
+ "total_flos": 1.880866314337075e+16,
58
+ "trial_name": null,
59
+ "trial_params": null
60
+ }
from-scratch-ll/checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c12fe8d360bd5886db6b95f7a073c5297fac83419532fd9baa19387974ee3911
3
+ size 3311
from-scratch-ll/checkpoint-6000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
from-scratch-ll/checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44ed68abfe006a9a18304b21ee34f73d946f14ca5d1147897d9c35446efdc341
3
+ size 871260901
from-scratch-ll/checkpoint-6000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:248b8201f8d53d7d3be8db16992c42ffeeddcd014adb164151cd0952f8175bee
3
+ size 435643185
from-scratch-ll/checkpoint-6000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2aa75a157b502a690dfda23707b5ed29e94b28dab58669ce35740a5780b8d9b7
3
+ size 14503
from-scratch-ll/checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d11dc3c90f61aec3ca3b5799bc4183040e0e02a8cef3baadb19eae758969d44
3
+ size 623
from-scratch-ll/checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.018754186202277,
5
+ "global_step": 6000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.33,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.2269,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.67,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.0536,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 1.0,
24
+ "learning_rate": 4.722222222222222e-05,
25
+ "loss": 0.0352,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 1.34,
30
+ "learning_rate": 4.4444444444444447e-05,
31
+ "loss": 0.0207,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 1.67,
36
+ "learning_rate": 4.166666666666667e-05,
37
+ "loss": 0.0201,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 2.01,
42
+ "learning_rate": 3.888888888888889e-05,
43
+ "loss": 0.0175,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 2.01,
48
+ "eval_loss": 1.1985770463943481,
49
+ "eval_runtime": 44.1956,
50
+ "eval_samples_per_second": 74.17,
51
+ "eval_steps_per_second": 18.554,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 2.34,
56
+ "learning_rate": 3.611111111111111e-05,
57
+ "loss": 0.0105,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 2.68,
62
+ "learning_rate": 3.3333333333333335e-05,
63
+ "loss": 0.0095,
64
+ "step": 4000
65
+ },
66
+ {
67
+ "epoch": 3.01,
68
+ "learning_rate": 3.055555555555556e-05,
69
+ "loss": 0.0091,
70
+ "step": 4500
71
+ },
72
+ {
73
+ "epoch": 3.35,
74
+ "learning_rate": 2.777777777777778e-05,
75
+ "loss": 0.0055,
76
+ "step": 5000
77
+ },
78
+ {
79
+ "epoch": 3.68,
80
+ "learning_rate": 2.5e-05,
81
+ "loss": 0.0059,
82
+ "step": 5500
83
+ },
84
+ {
85
+ "epoch": 4.02,
86
+ "learning_rate": 2.2222222222222223e-05,
87
+ "loss": 0.0055,
88
+ "step": 6000
89
+ },
90
+ {
91
+ "epoch": 4.02,
92
+ "eval_loss": 1.4346826076507568,
93
+ "eval_runtime": 43.2685,
94
+ "eval_samples_per_second": 75.76,
95
+ "eval_steps_per_second": 18.951,
96
+ "step": 6000
97
+ }
98
+ ],
99
+ "max_steps": 10000,
100
+ "num_train_epochs": 7,
101
+ "total_flos": 3.76173262867415e+16,
102
+ "trial_name": null,
103
+ "trial_params": null
104
+ }
from-scratch-ll/checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c12fe8d360bd5886db6b95f7a073c5297fac83419532fd9baa19387974ee3911
3
+ size 3311
from-scratch-ll/checkpoint-9000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
from-scratch-ll/checkpoint-9000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a1390e1a6b5e91a2fd5d8e6e68b728a7969aec23b541f7d899bff2eed297fe2
3
+ size 871260901
from-scratch-ll/checkpoint-9000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e1ff8f46a889cd07fda0fb6753f9005a098169b07aca650e24d69bf0c27710c
3
+ size 435643185
from-scratch-ll/checkpoint-9000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e663beca5a03c7cbe63768c5cf1fc8556d27cb20d5fd9d8aed467e33b9f787b4
3
+ size 14503
from-scratch-ll/checkpoint-9000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51dcf4d510d7613409859db8c31ab37e6cb185c4dc0cd816d5ebfdc5c680529c
3
+ size 623
from-scratch-ll/checkpoint-9000/trainer_state.json ADDED
@@ -0,0 +1,148 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 6.028131279303416,
5
+ "global_step": 9000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.33,
12
+ "learning_rate": 2.5e-05,
13
+ "loss": 0.2269,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.67,
18
+ "learning_rate": 5e-05,
19
+ "loss": 0.0536,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 1.0,
24
+ "learning_rate": 4.722222222222222e-05,
25
+ "loss": 0.0352,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 1.34,
30
+ "learning_rate": 4.4444444444444447e-05,
31
+ "loss": 0.0207,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 1.67,
36
+ "learning_rate": 4.166666666666667e-05,
37
+ "loss": 0.0201,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 2.01,
42
+ "learning_rate": 3.888888888888889e-05,
43
+ "loss": 0.0175,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 2.01,
48
+ "eval_loss": 1.1985770463943481,
49
+ "eval_runtime": 44.1956,
50
+ "eval_samples_per_second": 74.17,
51
+ "eval_steps_per_second": 18.554,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 2.34,
56
+ "learning_rate": 3.611111111111111e-05,
57
+ "loss": 0.0105,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 2.68,
62
+ "learning_rate": 3.3333333333333335e-05,
63
+ "loss": 0.0095,
64
+ "step": 4000
65
+ },
66
+ {
67
+ "epoch": 3.01,
68
+ "learning_rate": 3.055555555555556e-05,
69
+ "loss": 0.0091,
70
+ "step": 4500
71
+ },
72
+ {
73
+ "epoch": 3.35,
74
+ "learning_rate": 2.777777777777778e-05,
75
+ "loss": 0.0055,
76
+ "step": 5000
77
+ },
78
+ {
79
+ "epoch": 3.68,
80
+ "learning_rate": 2.5e-05,
81
+ "loss": 0.0059,
82
+ "step": 5500
83
+ },
84
+ {
85
+ "epoch": 4.02,
86
+ "learning_rate": 2.2222222222222223e-05,
87
+ "loss": 0.0055,
88
+ "step": 6000
89
+ },
90
+ {
91
+ "epoch": 4.02,
92
+ "eval_loss": 1.4346826076507568,
93
+ "eval_runtime": 43.2685,
94
+ "eval_samples_per_second": 75.76,
95
+ "eval_steps_per_second": 18.951,
96
+ "step": 6000
97
+ },
98
+ {
99
+ "epoch": 4.35,
100
+ "learning_rate": 1.9444444444444445e-05,
101
+ "loss": 0.0033,
102
+ "step": 6500
103
+ },
104
+ {
105
+ "epoch": 4.69,
106
+ "learning_rate": 1.6666666666666667e-05,
107
+ "loss": 0.0029,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 5.02,
112
+ "learning_rate": 1.388888888888889e-05,
113
+ "loss": 0.0023,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 5.36,
118
+ "learning_rate": 1.1111111111111112e-05,
119
+ "loss": 0.0017,
120
+ "step": 8000
121
+ },
122
+ {
123
+ "epoch": 5.69,
124
+ "learning_rate": 8.333333333333334e-06,
125
+ "loss": 0.0019,
126
+ "step": 8500
127
+ },
128
+ {
129
+ "epoch": 6.03,
130
+ "learning_rate": 5.555555555555556e-06,
131
+ "loss": 0.0015,
132
+ "step": 9000
133
+ },
134
+ {
135
+ "epoch": 6.03,
136
+ "eval_loss": 1.6609243154525757,
137
+ "eval_runtime": 43.4248,
138
+ "eval_samples_per_second": 75.487,
139
+ "eval_steps_per_second": 18.883,
140
+ "step": 9000
141
+ }
142
+ ],
143
+ "max_steps": 10000,
144
+ "num_train_epochs": 7,
145
+ "total_flos": 5.642598943011226e+16,
146
+ "trial_name": null,
147
+ "trial_params": null
148
+ }
from-scratch-ll/checkpoint-9000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c12fe8d360bd5886db6b95f7a073c5297fac83419532fd9baa19387974ee3911
3
+ size 3311
from-scratch-sl-balance/checkpoint-3000/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-uncased",
3
+ "architectures": [
4
+ "BertForHighlightPrediction"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "output_hidden_states": true,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.20.1",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 30522
27
+ }
from-scratch-sl-balance/checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a1d67ea81367c3c4683c20df386765e06a2b8a3c7bbda91cad77b15161505d3
3
+ size 871260901
from-scratch-sl-balance/checkpoint-3000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d9fe480348451983dfa0c525ead2b4fe0735b2838c246a5db834ce303a52c81
3
+ size 435643185
from-scratch-sl-balance/checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc67bb6793cc9e3a2453616d2e0bb667dbe46c82a25d258f3f82f7c00e57b78d
3
+ size 14503
from-scratch-sl-balance/checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e5d2b076608b10220351f7a090fe60ee9462654745d3c6d779eef696c9c360
3
+ size 623