Go4miii commited on
Commit
c92e6c3
1 Parent(s): 0a2afe7

commit from root

Browse files
Baichuan-13B-Chat-lora-Computing/README.md ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+ ### Framework versions
7
+
8
+
9
+ - PEFT 0.4.0
Baichuan-13B-Chat-lora-Computing/adapter_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "/root/ysb/LLaMA-Efficient-Tuning/model/Baichuan-13B-Chat",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "W_pack"
18
+ ],
19
+ "task_type": "CAUSAL_LM"
20
+ }
Baichuan-13B-Chat-lora-Computing/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c0817e6cbd8c34921834a91f5b62499ac09bc227ab9f920b5785d7aaa0221c9
3
+ size 26241825
Baichuan-13B-Chat-lora-Computing/all_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.1560091979059696,
4
+ "train_runtime": 21225.5051,
5
+ "train_samples_per_second": 3.081,
6
+ "train_steps_per_second": 0.193
7
+ }
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/README.md ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+ ### Framework versions
7
+
8
+
9
+ - PEFT 0.4.0
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/adapter_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "/root/ysb/LLaMA-Efficient-Tuning/model/Baichuan-13B-Chat",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "W_pack"
18
+ ],
19
+ "task_type": "CAUSAL_LM"
20
+ }
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d4832deecb2c24a2c1c7edeb37f9521dbec68f6747f4e405057b623f76241ea
3
+ size 26241825
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/finetuning_args.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "finetuning_type": "lora",
3
+ "lora_alpha": 32.0,
4
+ "lora_dropout": 0.1,
5
+ "lora_rank": 8,
6
+ "lora_target": [
7
+ "W_pack"
8
+ ],
9
+ "name_module_trainable": "mlp",
10
+ "num_hidden_layers": 32,
11
+ "num_layer_trainable": 3
12
+ }
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b2a47c54ec38a264ce2ff0b28263a91b1c330bfaa26e35dde59cfe38bd8c1e8
3
+ size 52496005
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5aaa4814eb13e4f5969c12488df154b94108220258379c62547b466b92abced
3
+ size 14575
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e269fabb4b8a314d2e031cc6507cc95f5d640507558990cf7de0e9f273dfb51
3
+ size 627
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,2419 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9787129924149743,
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": 0.0,
13
+ "learning_rate": 4.999926141863722e-05,
14
+ "loss": 1.0318,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999704571818908e-05,
20
+ "loss": 0.8189,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 4.999335302957357e-05,
26
+ "loss": 0.6632,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4.9988183570978795e-05,
32
+ "loss": 0.5815,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 4.998153764784998e-05,
38
+ "loss": 0.5221,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.9973415652871555e-05,
44
+ "loss": 0.4635,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.02,
49
+ "learning_rate": 4.9963818065943846e-05,
50
+ "loss": 0.4266,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 4.9952745454154736e-05,
56
+ "loss": 0.4073,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.02,
61
+ "learning_rate": 4.9940198471746226e-05,
62
+ "loss": 0.373,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.9926177860075696e-05,
68
+ "loss": 0.3172,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.03,
73
+ "learning_rate": 4.991068444757214e-05,
74
+ "loss": 0.2985,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.03,
79
+ "learning_rate": 4.989371914968722e-05,
80
+ "loss": 0.2965,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.03,
85
+ "learning_rate": 4.987528296884116e-05,
86
+ "loss": 0.29,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.03,
91
+ "learning_rate": 4.985537699436353e-05,
92
+ "loss": 0.2725,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.04,
97
+ "learning_rate": 4.983400240242887e-05,
98
+ "loss": 0.236,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.04,
103
+ "learning_rate": 4.9811160455987196e-05,
104
+ "loss": 0.2672,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.04,
109
+ "learning_rate": 4.978685250468939e-05,
110
+ "loss": 0.2448,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.04,
115
+ "learning_rate": 4.9761079984807426e-05,
116
+ "loss": 0.2553,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.05,
121
+ "learning_rate": 4.973384441914953e-05,
122
+ "loss": 0.2133,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.05,
127
+ "learning_rate": 4.970514741697021e-05,
128
+ "loss": 0.2351,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.05,
133
+ "learning_rate": 4.967499067387514e-05,
134
+ "loss": 0.1805,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.05,
139
+ "learning_rate": 4.964337597172099e-05,
140
+ "loss": 0.2253,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.06,
145
+ "learning_rate": 4.961030517851014e-05,
146
+ "loss": 0.1853,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.06,
151
+ "learning_rate": 4.957578024828033e-05,
152
+ "loss": 0.2271,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.06,
157
+ "learning_rate": 4.953980322098914e-05,
158
+ "loss": 0.2112,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.06,
163
+ "learning_rate": 4.9502376222393525e-05,
164
+ "loss": 0.22,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.07,
169
+ "learning_rate": 4.9463501463924175e-05,
170
+ "loss": 0.1825,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.07,
175
+ "learning_rate": 4.942318124255487e-05,
176
+ "loss": 0.242,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.07,
181
+ "learning_rate": 4.9381417940666715e-05,
182
+ "loss": 0.2179,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.07,
187
+ "learning_rate": 4.9338214025907424e-05,
188
+ "loss": 0.183,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.08,
193
+ "learning_rate": 4.9293572051045514e-05,
194
+ "loss": 0.1886,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.08,
199
+ "learning_rate": 4.924749465381942e-05,
200
+ "loss": 0.2062,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.08,
205
+ "learning_rate": 4.91999845567817e-05,
206
+ "loss": 0.1647,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.08,
211
+ "learning_rate": 4.915104456713812e-05,
212
+ "loss": 0.1828,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.09,
217
+ "learning_rate": 4.910067757658182e-05,
218
+ "loss": 0.1781,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.09,
223
+ "learning_rate": 4.904888656112246e-05,
224
+ "loss": 0.2016,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.09,
229
+ "learning_rate": 4.899567458091031e-05,
230
+ "loss": 0.1761,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.09,
235
+ "learning_rate": 4.894104478005554e-05,
236
+ "loss": 0.1729,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.1,
241
+ "learning_rate": 4.888500038644238e-05,
242
+ "loss": 0.1894,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.1,
247
+ "learning_rate": 4.8827544711538386e-05,
248
+ "loss": 0.2072,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.1,
253
+ "learning_rate": 4.876868115019881e-05,
254
+ "loss": 0.1724,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.1,
259
+ "learning_rate": 4.8708413180466e-05,
260
+ "loss": 0.2249,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.11,
265
+ "learning_rate": 4.8646744363363904e-05,
266
+ "loss": 0.2007,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.11,
271
+ "learning_rate": 4.858367834268763e-05,
272
+ "loss": 0.1877,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.11,
277
+ "learning_rate": 4.851921884478817e-05,
278
+ "loss": 0.1827,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.11,
283
+ "learning_rate": 4.845336967835226e-05,
284
+ "loss": 0.1819,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.11,
289
+ "learning_rate": 4.8386134734177225e-05,
290
+ "loss": 0.1856,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.12,
295
+ "learning_rate": 4.831751798494123e-05,
296
+ "loss": 0.1685,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.12,
301
+ "learning_rate": 4.824752348496844e-05,
302
+ "loss": 0.1679,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.12,
307
+ "learning_rate": 4.8176155369989505e-05,
308
+ "loss": 0.1788,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.12,
313
+ "learning_rate": 4.8103417856897206e-05,
314
+ "loss": 0.1872,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.13,
319
+ "learning_rate": 4.8029315243497244e-05,
320
+ "loss": 0.1697,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.13,
325
+ "learning_rate": 4.7953851908254376e-05,
326
+ "loss": 0.162,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.13,
331
+ "learning_rate": 4.787703231003363e-05,
332
+ "loss": 0.1718,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.13,
337
+ "learning_rate": 4.77988609878369e-05,
338
+ "loss": 0.1866,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.14,
343
+ "learning_rate": 4.771934256053471e-05,
344
+ "loss": 0.1609,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.14,
349
+ "learning_rate": 4.763848172659333e-05,
350
+ "loss": 0.1629,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.14,
355
+ "learning_rate": 4.755628326379716e-05,
356
+ "loss": 0.1947,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.14,
361
+ "learning_rate": 4.747275202896642e-05,
362
+ "loss": 0.1813,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.15,
367
+ "learning_rate": 4.738789295767016e-05,
368
+ "loss": 0.1672,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.15,
373
+ "learning_rate": 4.730171106393466e-05,
374
+ "loss": 0.1597,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.15,
379
+ "learning_rate": 4.7214211439947174e-05,
380
+ "loss": 0.1855,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.15,
385
+ "learning_rate": 4.712539925575502e-05,
386
+ "loss": 0.1729,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.16,
391
+ "learning_rate": 4.7035279758960104e-05,
392
+ "loss": 0.1771,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.16,
397
+ "learning_rate": 4.6943858274408915e-05,
398
+ "loss": 0.1891,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.16,
403
+ "learning_rate": 4.68511402038778e-05,
404
+ "loss": 0.1442,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.16,
409
+ "learning_rate": 4.6757131025753886e-05,
410
+ "loss": 0.1666,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.17,
415
+ "learning_rate": 4.666183629471132e-05,
416
+ "loss": 0.1672,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.17,
421
+ "learning_rate": 4.656526164138307e-05,
422
+ "loss": 0.1525,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.17,
427
+ "learning_rate": 4.646741277202829e-05,
428
+ "loss": 0.1758,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.17,
433
+ "learning_rate": 4.6368295468195064e-05,
434
+ "loss": 0.1708,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.18,
439
+ "learning_rate": 4.626791558637887e-05,
440
+ "loss": 0.1682,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.18,
445
+ "learning_rate": 4.6166279057676494e-05,
446
+ "loss": 0.1515,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.18,
451
+ "learning_rate": 4.6063391887435605e-05,
452
+ "loss": 0.1687,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.18,
457
+ "learning_rate": 4.595926015489991e-05,
458
+ "loss": 0.1641,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.19,
463
+ "learning_rate": 4.585389001284999e-05,
464
+ "loss": 0.1536,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.19,
469
+ "learning_rate": 4.574728768723965e-05,
470
+ "loss": 0.1683,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 0.19,
475
+ "learning_rate": 4.563945947682821e-05,
476
+ "loss": 0.1464,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 0.19,
481
+ "learning_rate": 4.5530411752808166e-05,
482
+ "loss": 0.1764,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 0.2,
487
+ "learning_rate": 4.542015095842885e-05,
488
+ "loss": 0.19,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 0.2,
493
+ "learning_rate": 4.5308683608615706e-05,
494
+ "loss": 0.1819,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 0.2,
499
+ "learning_rate": 4.519601628958529e-05,
500
+ "loss": 0.1512,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 0.2,
505
+ "learning_rate": 4.508215565845615e-05,
506
+ "loss": 0.1689,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 0.21,
511
+ "learning_rate": 4.496710844285551e-05,
512
+ "loss": 0.1739,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 0.21,
517
+ "learning_rate": 4.485088144052172e-05,
518
+ "loss": 0.1535,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 0.21,
523
+ "learning_rate": 4.4733481518902584e-05,
524
+ "loss": 0.142,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 0.21,
529
+ "learning_rate": 4.4614915614749645e-05,
530
+ "loss": 0.1584,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 0.22,
535
+ "learning_rate": 4.449519073370827e-05,
536
+ "loss": 0.1483,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 0.22,
541
+ "learning_rate": 4.43743139499037e-05,
542
+ "loss": 0.1587,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 0.22,
547
+ "learning_rate": 4.425229240552313e-05,
548
+ "loss": 0.1497,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 0.22,
553
+ "learning_rate": 4.4129133310393646e-05,
554
+ "loss": 0.1368,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 0.23,
559
+ "learning_rate": 4.400484394155622e-05,
560
+ "loss": 0.1565,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 0.23,
565
+ "learning_rate": 4.3879431642835784e-05,
566
+ "loss": 0.1377,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 0.23,
571
+ "learning_rate": 4.3752903824407234e-05,
572
+ "loss": 0.1755,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 0.23,
577
+ "learning_rate": 4.3625267962357676e-05,
578
+ "loss": 0.165,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 0.23,
583
+ "learning_rate": 4.3496531598244616e-05,
584
+ "loss": 0.1397,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 0.24,
589
+ "learning_rate": 4.33667023386504e-05,
590
+ "loss": 0.1518,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 0.24,
595
+ "learning_rate": 4.323578785473272e-05,
596
+ "loss": 0.1332,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 0.24,
601
+ "learning_rate": 4.3103795881771454e-05,
602
+ "loss": 0.1541,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 0.24,
607
+ "learning_rate": 4.2970734218711474e-05,
608
+ "loss": 0.1508,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 0.25,
613
+ "learning_rate": 4.283661072770193e-05,
614
+ "loss": 0.1465,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 0.25,
619
+ "learning_rate": 4.270143333363171e-05,
620
+ "loss": 0.1505,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 0.25,
625
+ "learning_rate": 4.2565210023661104e-05,
626
+ "loss": 0.1674,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 0.25,
631
+ "learning_rate": 4.242794884674996e-05,
632
+ "loss": 0.1352,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 0.26,
637
+ "learning_rate": 4.2289657913182045e-05,
638
+ "loss": 0.1667,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 0.26,
643
+ "learning_rate": 4.2150345394085845e-05,
644
+ "loss": 0.1495,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 0.26,
649
+ "learning_rate": 4.201001952095178e-05,
650
+ "loss": 0.1538,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 0.26,
655
+ "learning_rate": 4.186868858514582e-05,
656
+ "loss": 0.1425,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 0.27,
661
+ "learning_rate": 4.172636093741959e-05,
662
+ "loss": 0.1357,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 0.27,
667
+ "learning_rate": 4.158304498741691e-05,
668
+ "loss": 0.1493,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 0.27,
673
+ "learning_rate": 4.143874920317696e-05,
674
+ "loss": 0.1439,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 0.27,
679
+ "learning_rate": 4.1293482110633905e-05,
680
+ "loss": 0.1449,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 0.28,
685
+ "learning_rate": 4.114725229311311e-05,
686
+ "loss": 0.1451,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 0.28,
691
+ "learning_rate": 4.100006839082402e-05,
692
+ "loss": 0.1559,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 0.28,
697
+ "learning_rate": 4.085193910034959e-05,
698
+ "loss": 0.1452,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 0.28,
703
+ "learning_rate": 4.0702873174132483e-05,
704
+ "loss": 0.1584,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 0.29,
709
+ "learning_rate": 4.0552879419957904e-05,
710
+ "loss": 0.1341,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 0.29,
715
+ "learning_rate": 4.0401966700433144e-05,
716
+ "loss": 0.1433,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 0.29,
721
+ "learning_rate": 4.025014393246399e-05,
722
+ "loss": 0.1583,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 0.29,
727
+ "learning_rate": 4.0097420086727766e-05,
728
+ "loss": 0.1508,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 0.3,
733
+ "learning_rate": 3.9943804187143384e-05,
734
+ "loss": 0.1577,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 0.3,
739
+ "learning_rate": 3.978930531033807e-05,
740
+ "loss": 0.1431,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 0.3,
745
+ "learning_rate": 3.963393258511109e-05,
746
+ "loss": 0.161,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 0.3,
751
+ "learning_rate": 3.9477695191894395e-05,
752
+ "loss": 0.1396,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 0.31,
757
+ "learning_rate": 3.9320602362210116e-05,
758
+ "loss": 0.1477,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 0.31,
763
+ "learning_rate": 3.9162663378125166e-05,
764
+ "loss": 0.1461,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 0.31,
769
+ "learning_rate": 3.900388757170274e-05,
770
+ "loss": 0.152,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 0.31,
775
+ "learning_rate": 3.8844284324450956e-05,
776
+ "loss": 0.1563,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 0.32,
781
+ "learning_rate": 3.8683863066768534e-05,
782
+ "loss": 0.1441,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 0.32,
787
+ "learning_rate": 3.8522633277387555e-05,
788
+ "loss": 0.1448,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 0.32,
793
+ "learning_rate": 3.836060448281342e-05,
794
+ "loss": 0.1659,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 0.32,
799
+ "learning_rate": 3.819778625676197e-05,
800
+ "loss": 0.1464,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 0.33,
805
+ "learning_rate": 3.803418821959379e-05,
806
+ "loss": 0.147,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 0.33,
811
+ "learning_rate": 3.7869820037745776e-05,
812
+ "loss": 0.1479,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 0.33,
817
+ "learning_rate": 3.770469142315997e-05,
818
+ "loss": 0.1352,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 0.33,
823
+ "learning_rate": 3.7538812132709775e-05,
824
+ "loss": 0.1638,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 0.34,
829
+ "learning_rate": 3.737219196762337e-05,
830
+ "loss": 0.149,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 0.34,
835
+ "learning_rate": 3.7204840772904645e-05,
836
+ "loss": 0.155,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 0.34,
841
+ "learning_rate": 3.703676843675147e-05,
842
+ "loss": 0.159,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 0.34,
847
+ "learning_rate": 3.686798488997146e-05,
848
+ "loss": 0.1455,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 0.34,
853
+ "learning_rate": 3.669850010539518e-05,
854
+ "loss": 0.1439,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 0.35,
859
+ "learning_rate": 3.652832409728686e-05,
860
+ "loss": 0.1344,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 0.35,
865
+ "learning_rate": 3.635746692075276e-05,
866
+ "loss": 0.1519,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 0.35,
871
+ "learning_rate": 3.6185938671146975e-05,
872
+ "loss": 0.1469,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 0.35,
877
+ "learning_rate": 3.601374948347498e-05,
878
+ "loss": 0.1343,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 0.36,
883
+ "learning_rate": 3.584090953179475e-05,
884
+ "loss": 0.143,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 0.36,
889
+ "learning_rate": 3.566742902861566e-05,
890
+ "loss": 0.137,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 0.36,
895
+ "learning_rate": 3.549331822429503e-05,
896
+ "loss": 0.1485,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 0.36,
901
+ "learning_rate": 3.531858740643247e-05,
902
+ "loss": 0.1182,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 0.37,
907
+ "learning_rate": 3.5143246899262014e-05,
908
+ "loss": 0.1528,
909
+ "step": 1500
910
+ },
911
+ {
912
+ "epoch": 0.37,
913
+ "learning_rate": 3.4967307063042134e-05,
914
+ "loss": 0.1648,
915
+ "step": 1510
916
+ },
917
+ {
918
+ "epoch": 0.37,
919
+ "learning_rate": 3.479077829344354e-05,
920
+ "loss": 0.1213,
921
+ "step": 1520
922
+ },
923
+ {
924
+ "epoch": 0.37,
925
+ "learning_rate": 3.461367102093498e-05,
926
+ "loss": 0.1412,
927
+ "step": 1530
928
+ },
929
+ {
930
+ "epoch": 0.38,
931
+ "learning_rate": 3.443599571016689e-05,
932
+ "loss": 0.1397,
933
+ "step": 1540
934
+ },
935
+ {
936
+ "epoch": 0.38,
937
+ "learning_rate": 3.425776285935315e-05,
938
+ "loss": 0.1527,
939
+ "step": 1550
940
+ },
941
+ {
942
+ "epoch": 0.38,
943
+ "learning_rate": 3.407898299965069e-05,
944
+ "loss": 0.137,
945
+ "step": 1560
946
+ },
947
+ {
948
+ "epoch": 0.38,
949
+ "learning_rate": 3.3899666694537307e-05,
950
+ "loss": 0.121,
951
+ "step": 1570
952
+ },
953
+ {
954
+ "epoch": 0.39,
955
+ "learning_rate": 3.3719824539187484e-05,
956
+ "loss": 0.1401,
957
+ "step": 1580
958
+ },
959
+ {
960
+ "epoch": 0.39,
961
+ "learning_rate": 3.353946715984635e-05,
962
+ "loss": 0.1163,
963
+ "step": 1590
964
+ },
965
+ {
966
+ "epoch": 0.39,
967
+ "learning_rate": 3.3358605213201835e-05,
968
+ "loss": 0.1562,
969
+ "step": 1600
970
+ },
971
+ {
972
+ "epoch": 0.39,
973
+ "learning_rate": 3.3177249385754964e-05,
974
+ "loss": 0.1324,
975
+ "step": 1610
976
+ },
977
+ {
978
+ "epoch": 0.4,
979
+ "learning_rate": 3.299541039318849e-05,
980
+ "loss": 0.1586,
981
+ "step": 1620
982
+ },
983
+ {
984
+ "epoch": 0.4,
985
+ "learning_rate": 3.281309897973368e-05,
986
+ "loss": 0.1626,
987
+ "step": 1630
988
+ },
989
+ {
990
+ "epoch": 0.4,
991
+ "learning_rate": 3.263032591753552e-05,
992
+ "loss": 0.1366,
993
+ "step": 1640
994
+ },
995
+ {
996
+ "epoch": 0.4,
997
+ "learning_rate": 3.2447102006016186e-05,
998
+ "loss": 0.1336,
999
+ "step": 1650
1000
+ },
1001
+ {
1002
+ "epoch": 0.41,
1003
+ "learning_rate": 3.226343807123697e-05,
1004
+ "loss": 0.1376,
1005
+ "step": 1660
1006
+ },
1007
+ {
1008
+ "epoch": 0.41,
1009
+ "learning_rate": 3.2079344965258645e-05,
1010
+ "loss": 0.132,
1011
+ "step": 1670
1012
+ },
1013
+ {
1014
+ "epoch": 0.41,
1015
+ "learning_rate": 3.189483356550015e-05,
1016
+ "loss": 0.1316,
1017
+ "step": 1680
1018
+ },
1019
+ {
1020
+ "epoch": 0.41,
1021
+ "learning_rate": 3.1709914774095975e-05,
1022
+ "loss": 0.1448,
1023
+ "step": 1690
1024
+ },
1025
+ {
1026
+ "epoch": 0.42,
1027
+ "learning_rate": 3.152459951725196e-05,
1028
+ "loss": 0.132,
1029
+ "step": 1700
1030
+ },
1031
+ {
1032
+ "epoch": 0.42,
1033
+ "learning_rate": 3.133889874459971e-05,
1034
+ "loss": 0.133,
1035
+ "step": 1710
1036
+ },
1037
+ {
1038
+ "epoch": 0.42,
1039
+ "learning_rate": 3.115282342854959e-05,
1040
+ "loss": 0.1558,
1041
+ "step": 1720
1042
+ },
1043
+ {
1044
+ "epoch": 0.42,
1045
+ "learning_rate": 3.096638456364245e-05,
1046
+ "loss": 0.152,
1047
+ "step": 1730
1048
+ },
1049
+ {
1050
+ "epoch": 0.43,
1051
+ "learning_rate": 3.0779593165899954e-05,
1052
+ "loss": 0.1293,
1053
+ "step": 1740
1054
+ },
1055
+ {
1056
+ "epoch": 0.43,
1057
+ "learning_rate": 3.059246027217372e-05,
1058
+ "loss": 0.1361,
1059
+ "step": 1750
1060
+ },
1061
+ {
1062
+ "epoch": 0.43,
1063
+ "learning_rate": 3.0404996939493156e-05,
1064
+ "loss": 0.1462,
1065
+ "step": 1760
1066
+ },
1067
+ {
1068
+ "epoch": 0.43,
1069
+ "learning_rate": 3.021721424441216e-05,
1070
+ "loss": 0.1671,
1071
+ "step": 1770
1072
+ },
1073
+ {
1074
+ "epoch": 0.44,
1075
+ "learning_rate": 3.0029123282354638e-05,
1076
+ "loss": 0.1269,
1077
+ "step": 1780
1078
+ },
1079
+ {
1080
+ "epoch": 0.44,
1081
+ "learning_rate": 2.984073516695892e-05,
1082
+ "loss": 0.1571,
1083
+ "step": 1790
1084
+ },
1085
+ {
1086
+ "epoch": 0.44,
1087
+ "learning_rate": 2.9652061029421087e-05,
1088
+ "loss": 0.139,
1089
+ "step": 1800
1090
+ },
1091
+ {
1092
+ "epoch": 0.44,
1093
+ "learning_rate": 2.9463112017837262e-05,
1094
+ "loss": 0.1234,
1095
+ "step": 1810
1096
+ },
1097
+ {
1098
+ "epoch": 0.45,
1099
+ "learning_rate": 2.927389929654494e-05,
1100
+ "loss": 0.1511,
1101
+ "step": 1820
1102
+ },
1103
+ {
1104
+ "epoch": 0.45,
1105
+ "learning_rate": 2.9084434045463255e-05,
1106
+ "loss": 0.1347,
1107
+ "step": 1830
1108
+ },
1109
+ {
1110
+ "epoch": 0.45,
1111
+ "learning_rate": 2.8894727459432504e-05,
1112
+ "loss": 0.1464,
1113
+ "step": 1840
1114
+ },
1115
+ {
1116
+ "epoch": 0.45,
1117
+ "learning_rate": 2.870479074755257e-05,
1118
+ "loss": 0.1446,
1119
+ "step": 1850
1120
+ },
1121
+ {
1122
+ "epoch": 0.46,
1123
+ "learning_rate": 2.8514635132520716e-05,
1124
+ "loss": 0.1371,
1125
+ "step": 1860
1126
+ },
1127
+ {
1128
+ "epoch": 0.46,
1129
+ "learning_rate": 2.8324271849968394e-05,
1130
+ "loss": 0.1345,
1131
+ "step": 1870
1132
+ },
1133
+ {
1134
+ "epoch": 0.46,
1135
+ "learning_rate": 2.8133712147797404e-05,
1136
+ "loss": 0.1396,
1137
+ "step": 1880
1138
+ },
1139
+ {
1140
+ "epoch": 0.46,
1141
+ "learning_rate": 2.794296728551533e-05,
1142
+ "loss": 0.1225,
1143
+ "step": 1890
1144
+ },
1145
+ {
1146
+ "epoch": 0.46,
1147
+ "learning_rate": 2.7752048533570186e-05,
1148
+ "loss": 0.1352,
1149
+ "step": 1900
1150
+ },
1151
+ {
1152
+ "epoch": 0.47,
1153
+ "learning_rate": 2.756096717268453e-05,
1154
+ "loss": 0.146,
1155
+ "step": 1910
1156
+ },
1157
+ {
1158
+ "epoch": 0.47,
1159
+ "learning_rate": 2.7369734493188924e-05,
1160
+ "loss": 0.1426,
1161
+ "step": 1920
1162
+ },
1163
+ {
1164
+ "epoch": 0.47,
1165
+ "learning_rate": 2.71783617943548e-05,
1166
+ "loss": 0.1494,
1167
+ "step": 1930
1168
+ },
1169
+ {
1170
+ "epoch": 0.47,
1171
+ "learning_rate": 2.6986860383726865e-05,
1172
+ "loss": 0.1098,
1173
+ "step": 1940
1174
+ },
1175
+ {
1176
+ "epoch": 0.48,
1177
+ "learning_rate": 2.679524157645494e-05,
1178
+ "loss": 0.1436,
1179
+ "step": 1950
1180
+ },
1181
+ {
1182
+ "epoch": 0.48,
1183
+ "learning_rate": 2.660351669462541e-05,
1184
+ "loss": 0.1269,
1185
+ "step": 1960
1186
+ },
1187
+ {
1188
+ "epoch": 0.48,
1189
+ "learning_rate": 2.6411697066592228e-05,
1190
+ "loss": 0.1283,
1191
+ "step": 1970
1192
+ },
1193
+ {
1194
+ "epoch": 0.48,
1195
+ "learning_rate": 2.62197940263076e-05,
1196
+ "loss": 0.1379,
1197
+ "step": 1980
1198
+ },
1199
+ {
1200
+ "epoch": 0.49,
1201
+ "learning_rate": 2.6027818912652218e-05,
1202
+ "loss": 0.135,
1203
+ "step": 1990
1204
+ },
1205
+ {
1206
+ "epoch": 0.49,
1207
+ "learning_rate": 2.5835783068765378e-05,
1208
+ "loss": 0.1365,
1209
+ "step": 2000
1210
+ },
1211
+ {
1212
+ "epoch": 0.49,
1213
+ "learning_rate": 2.564369784137472e-05,
1214
+ "loss": 0.1235,
1215
+ "step": 2010
1216
+ },
1217
+ {
1218
+ "epoch": 0.49,
1219
+ "learning_rate": 2.5451574580125738e-05,
1220
+ "loss": 0.1514,
1221
+ "step": 2020
1222
+ },
1223
+ {
1224
+ "epoch": 0.5,
1225
+ "learning_rate": 2.5259424636911256e-05,
1226
+ "loss": 0.1358,
1227
+ "step": 2030
1228
+ },
1229
+ {
1230
+ "epoch": 0.5,
1231
+ "learning_rate": 2.5067259365200624e-05,
1232
+ "loss": 0.1444,
1233
+ "step": 2040
1234
+ },
1235
+ {
1236
+ "epoch": 0.5,
1237
+ "learning_rate": 2.4875090119368905e-05,
1238
+ "loss": 0.1254,
1239
+ "step": 2050
1240
+ },
1241
+ {
1242
+ "epoch": 0.5,
1243
+ "learning_rate": 2.468292825402597e-05,
1244
+ "loss": 0.1247,
1245
+ "step": 2060
1246
+ },
1247
+ {
1248
+ "epoch": 0.51,
1249
+ "learning_rate": 2.449078512334562e-05,
1250
+ "loss": 0.1293,
1251
+ "step": 2070
1252
+ },
1253
+ {
1254
+ "epoch": 0.51,
1255
+ "learning_rate": 2.429867208039467e-05,
1256
+ "loss": 0.126,
1257
+ "step": 2080
1258
+ },
1259
+ {
1260
+ "epoch": 0.51,
1261
+ "learning_rate": 2.4106600476462173e-05,
1262
+ "loss": 0.1351,
1263
+ "step": 2090
1264
+ },
1265
+ {
1266
+ "epoch": 0.51,
1267
+ "learning_rate": 2.3914581660388683e-05,
1268
+ "loss": 0.1386,
1269
+ "step": 2100
1270
+ },
1271
+ {
1272
+ "epoch": 0.52,
1273
+ "learning_rate": 2.3722626977895706e-05,
1274
+ "loss": 0.1294,
1275
+ "step": 2110
1276
+ },
1277
+ {
1278
+ "epoch": 0.52,
1279
+ "learning_rate": 2.3530747770915328e-05,
1280
+ "loss": 0.1188,
1281
+ "step": 2120
1282
+ },
1283
+ {
1284
+ "epoch": 0.52,
1285
+ "learning_rate": 2.3338955376920034e-05,
1286
+ "loss": 0.1283,
1287
+ "step": 2130
1288
+ },
1289
+ {
1290
+ "epoch": 0.52,
1291
+ "learning_rate": 2.3147261128252858e-05,
1292
+ "loss": 0.1239,
1293
+ "step": 2140
1294
+ },
1295
+ {
1296
+ "epoch": 0.53,
1297
+ "learning_rate": 2.295567635145774e-05,
1298
+ "loss": 0.1258,
1299
+ "step": 2150
1300
+ },
1301
+ {
1302
+ "epoch": 0.53,
1303
+ "learning_rate": 2.276421236661032e-05,
1304
+ "loss": 0.145,
1305
+ "step": 2160
1306
+ },
1307
+ {
1308
+ "epoch": 0.53,
1309
+ "learning_rate": 2.2572880486649086e-05,
1310
+ "loss": 0.1232,
1311
+ "step": 2170
1312
+ },
1313
+ {
1314
+ "epoch": 0.53,
1315
+ "learning_rate": 2.2381692016706862e-05,
1316
+ "loss": 0.14,
1317
+ "step": 2180
1318
+ },
1319
+ {
1320
+ "epoch": 0.54,
1321
+ "learning_rate": 2.219065825344292e-05,
1322
+ "loss": 0.1594,
1323
+ "step": 2190
1324
+ },
1325
+ {
1326
+ "epoch": 0.54,
1327
+ "learning_rate": 2.1999790484375433e-05,
1328
+ "loss": 0.1475,
1329
+ "step": 2200
1330
+ },
1331
+ {
1332
+ "epoch": 0.54,
1333
+ "learning_rate": 2.1809099987214567e-05,
1334
+ "loss": 0.1176,
1335
+ "step": 2210
1336
+ },
1337
+ {
1338
+ "epoch": 0.54,
1339
+ "learning_rate": 2.1618598029196082e-05,
1340
+ "loss": 0.1317,
1341
+ "step": 2220
1342
+ },
1343
+ {
1344
+ "epoch": 0.55,
1345
+ "learning_rate": 2.142829586641566e-05,
1346
+ "loss": 0.1318,
1347
+ "step": 2230
1348
+ },
1349
+ {
1350
+ "epoch": 0.55,
1351
+ "learning_rate": 2.1238204743163764e-05,
1352
+ "loss": 0.1453,
1353
+ "step": 2240
1354
+ },
1355
+ {
1356
+ "epoch": 0.55,
1357
+ "learning_rate": 2.104833589126124e-05,
1358
+ "loss": 0.1297,
1359
+ "step": 2250
1360
+ },
1361
+ {
1362
+ "epoch": 0.55,
1363
+ "learning_rate": 2.085870052939572e-05,
1364
+ "loss": 0.1393,
1365
+ "step": 2260
1366
+ },
1367
+ {
1368
+ "epoch": 0.56,
1369
+ "learning_rate": 2.0669309862458756e-05,
1370
+ "loss": 0.1364,
1371
+ "step": 2270
1372
+ },
1373
+ {
1374
+ "epoch": 0.56,
1375
+ "learning_rate": 2.0480175080883663e-05,
1376
+ "loss": 0.1246,
1377
+ "step": 2280
1378
+ },
1379
+ {
1380
+ "epoch": 0.56,
1381
+ "learning_rate": 2.0291307359984435e-05,
1382
+ "loss": 0.1437,
1383
+ "step": 2290
1384
+ },
1385
+ {
1386
+ "epoch": 0.56,
1387
+ "learning_rate": 2.0102717859295365e-05,
1388
+ "loss": 0.1268,
1389
+ "step": 2300
1390
+ },
1391
+ {
1392
+ "epoch": 0.57,
1393
+ "learning_rate": 1.9914417721911687e-05,
1394
+ "loss": 0.1299,
1395
+ "step": 2310
1396
+ },
1397
+ {
1398
+ "epoch": 0.57,
1399
+ "learning_rate": 1.9726418073831167e-05,
1400
+ "loss": 0.1334,
1401
+ "step": 2320
1402
+ },
1403
+ {
1404
+ "epoch": 0.57,
1405
+ "learning_rate": 1.9538730023296708e-05,
1406
+ "loss": 0.138,
1407
+ "step": 2330
1408
+ },
1409
+ {
1410
+ "epoch": 0.57,
1411
+ "learning_rate": 1.9351364660140004e-05,
1412
+ "loss": 0.1247,
1413
+ "step": 2340
1414
+ },
1415
+ {
1416
+ "epoch": 0.57,
1417
+ "learning_rate": 1.9164333055126265e-05,
1418
+ "loss": 0.1218,
1419
+ "step": 2350
1420
+ },
1421
+ {
1422
+ "epoch": 0.58,
1423
+ "learning_rate": 1.897764625930012e-05,
1424
+ "loss": 0.1391,
1425
+ "step": 2360
1426
+ },
1427
+ {
1428
+ "epoch": 0.58,
1429
+ "learning_rate": 1.879131530333261e-05,
1430
+ "loss": 0.1405,
1431
+ "step": 2370
1432
+ },
1433
+ {
1434
+ "epoch": 0.58,
1435
+ "learning_rate": 1.8605351196869453e-05,
1436
+ "loss": 0.1454,
1437
+ "step": 2380
1438
+ },
1439
+ {
1440
+ "epoch": 0.58,
1441
+ "learning_rate": 1.8419764927880488e-05,
1442
+ "loss": 0.1404,
1443
+ "step": 2390
1444
+ },
1445
+ {
1446
+ "epoch": 0.59,
1447
+ "learning_rate": 1.8234567462010482e-05,
1448
+ "loss": 0.1277,
1449
+ "step": 2400
1450
+ },
1451
+ {
1452
+ "epoch": 0.59,
1453
+ "learning_rate": 1.8049769741931178e-05,
1454
+ "loss": 0.1298,
1455
+ "step": 2410
1456
+ },
1457
+ {
1458
+ "epoch": 0.59,
1459
+ "learning_rate": 1.786538268669472e-05,
1460
+ "loss": 0.1378,
1461
+ "step": 2420
1462
+ },
1463
+ {
1464
+ "epoch": 0.59,
1465
+ "learning_rate": 1.768141719108852e-05,
1466
+ "loss": 0.1336,
1467
+ "step": 2430
1468
+ },
1469
+ {
1470
+ "epoch": 0.6,
1471
+ "learning_rate": 1.749788412499149e-05,
1472
+ "loss": 0.131,
1473
+ "step": 2440
1474
+ },
1475
+ {
1476
+ "epoch": 0.6,
1477
+ "learning_rate": 1.7314794332731786e-05,
1478
+ "loss": 0.1187,
1479
+ "step": 2450
1480
+ },
1481
+ {
1482
+ "epoch": 0.6,
1483
+ "learning_rate": 1.7132158632446085e-05,
1484
+ "loss": 0.1291,
1485
+ "step": 2460
1486
+ },
1487
+ {
1488
+ "epoch": 0.6,
1489
+ "learning_rate": 1.6949987815440332e-05,
1490
+ "loss": 0.1277,
1491
+ "step": 2470
1492
+ },
1493
+ {
1494
+ "epoch": 0.61,
1495
+ "learning_rate": 1.676829264555216e-05,
1496
+ "loss": 0.1419,
1497
+ "step": 2480
1498
+ },
1499
+ {
1500
+ "epoch": 0.61,
1501
+ "learning_rate": 1.658708385851485e-05,
1502
+ "loss": 0.128,
1503
+ "step": 2490
1504
+ },
1505
+ {
1506
+ "epoch": 0.61,
1507
+ "learning_rate": 1.6406372161323035e-05,
1508
+ "loss": 0.1213,
1509
+ "step": 2500
1510
+ },
1511
+ {
1512
+ "epoch": 0.61,
1513
+ "learning_rate": 1.6226168231600058e-05,
1514
+ "loss": 0.1436,
1515
+ "step": 2510
1516
+ },
1517
+ {
1518
+ "epoch": 0.62,
1519
+ "learning_rate": 1.604648271696701e-05,
1520
+ "loss": 0.1297,
1521
+ "step": 2520
1522
+ },
1523
+ {
1524
+ "epoch": 0.62,
1525
+ "learning_rate": 1.5867326234413692e-05,
1526
+ "loss": 0.1481,
1527
+ "step": 2530
1528
+ },
1529
+ {
1530
+ "epoch": 0.62,
1531
+ "learning_rate": 1.5688709369671222e-05,
1532
+ "loss": 0.1237,
1533
+ "step": 2540
1534
+ },
1535
+ {
1536
+ "epoch": 0.62,
1537
+ "learning_rate": 1.5510642676586606e-05,
1538
+ "loss": 0.1342,
1539
+ "step": 2550
1540
+ },
1541
+ {
1542
+ "epoch": 0.63,
1543
+ "learning_rate": 1.5333136676499078e-05,
1544
+ "loss": 0.1265,
1545
+ "step": 2560
1546
+ },
1547
+ {
1548
+ "epoch": 0.63,
1549
+ "learning_rate": 1.5156201857618532e-05,
1550
+ "loss": 0.1294,
1551
+ "step": 2570
1552
+ },
1553
+ {
1554
+ "epoch": 0.63,
1555
+ "learning_rate": 1.4979848674405755e-05,
1556
+ "loss": 0.1218,
1557
+ "step": 2580
1558
+ },
1559
+ {
1560
+ "epoch": 0.63,
1561
+ "learning_rate": 1.480408754695467e-05,
1562
+ "loss": 0.1377,
1563
+ "step": 2590
1564
+ },
1565
+ {
1566
+ "epoch": 0.64,
1567
+ "learning_rate": 1.4628928860376733e-05,
1568
+ "loss": 0.1367,
1569
+ "step": 2600
1570
+ },
1571
+ {
1572
+ "epoch": 0.64,
1573
+ "learning_rate": 1.4454382964187263e-05,
1574
+ "loss": 0.1338,
1575
+ "step": 2610
1576
+ },
1577
+ {
1578
+ "epoch": 0.64,
1579
+ "learning_rate": 1.4280460171693924e-05,
1580
+ "loss": 0.1357,
1581
+ "step": 2620
1582
+ },
1583
+ {
1584
+ "epoch": 0.64,
1585
+ "learning_rate": 1.4107170759387364e-05,
1586
+ "loss": 0.1308,
1587
+ "step": 2630
1588
+ },
1589
+ {
1590
+ "epoch": 0.65,
1591
+ "learning_rate": 1.3934524966334012e-05,
1592
+ "loss": 0.1466,
1593
+ "step": 2640
1594
+ },
1595
+ {
1596
+ "epoch": 0.65,
1597
+ "learning_rate": 1.3762532993571079e-05,
1598
+ "loss": 0.1371,
1599
+ "step": 2650
1600
+ },
1601
+ {
1602
+ "epoch": 0.65,
1603
+ "learning_rate": 1.35912050035038e-05,
1604
+ "loss": 0.1357,
1605
+ "step": 2660
1606
+ },
1607
+ {
1608
+ "epoch": 0.65,
1609
+ "learning_rate": 1.342055111930502e-05,
1610
+ "loss": 0.1208,
1611
+ "step": 2670
1612
+ },
1613
+ {
1614
+ "epoch": 0.66,
1615
+ "learning_rate": 1.325058142431701e-05,
1616
+ "loss": 0.1168,
1617
+ "step": 2680
1618
+ },
1619
+ {
1620
+ "epoch": 0.66,
1621
+ "learning_rate": 1.3081305961455658e-05,
1622
+ "loss": 0.121,
1623
+ "step": 2690
1624
+ },
1625
+ {
1626
+ "epoch": 0.66,
1627
+ "learning_rate": 1.291273473261716e-05,
1628
+ "loss": 0.1215,
1629
+ "step": 2700
1630
+ },
1631
+ {
1632
+ "epoch": 0.66,
1633
+ "learning_rate": 1.2744877698086943e-05,
1634
+ "loss": 0.1302,
1635
+ "step": 2710
1636
+ },
1637
+ {
1638
+ "epoch": 0.67,
1639
+ "learning_rate": 1.2577744775951194e-05,
1640
+ "loss": 0.1381,
1641
+ "step": 2720
1642
+ },
1643
+ {
1644
+ "epoch": 0.67,
1645
+ "learning_rate": 1.2411345841510808e-05,
1646
+ "loss": 0.1398,
1647
+ "step": 2730
1648
+ },
1649
+ {
1650
+ "epoch": 0.67,
1651
+ "learning_rate": 1.2245690726697939e-05,
1652
+ "loss": 0.1279,
1653
+ "step": 2740
1654
+ },
1655
+ {
1656
+ "epoch": 0.67,
1657
+ "learning_rate": 1.2080789219495028e-05,
1658
+ "loss": 0.1419,
1659
+ "step": 2750
1660
+ },
1661
+ {
1662
+ "epoch": 0.68,
1663
+ "learning_rate": 1.1916651063356455e-05,
1664
+ "loss": 0.1466,
1665
+ "step": 2760
1666
+ },
1667
+ {
1668
+ "epoch": 0.68,
1669
+ "learning_rate": 1.1753285956632867e-05,
1670
+ "loss": 0.1237,
1671
+ "step": 2770
1672
+ },
1673
+ {
1674
+ "epoch": 0.68,
1675
+ "learning_rate": 1.1590703551998111e-05,
1676
+ "loss": 0.1248,
1677
+ "step": 2780
1678
+ },
1679
+ {
1680
+ "epoch": 0.68,
1681
+ "learning_rate": 1.1428913455878932e-05,
1682
+ "loss": 0.1256,
1683
+ "step": 2790
1684
+ },
1685
+ {
1686
+ "epoch": 0.69,
1687
+ "learning_rate": 1.1267925227887275e-05,
1688
+ "loss": 0.1272,
1689
+ "step": 2800
1690
+ },
1691
+ {
1692
+ "epoch": 0.69,
1693
+ "learning_rate": 1.1107748380255536e-05,
1694
+ "loss": 0.1325,
1695
+ "step": 2810
1696
+ },
1697
+ {
1698
+ "epoch": 0.69,
1699
+ "learning_rate": 1.0948392377274477e-05,
1700
+ "loss": 0.1099,
1701
+ "step": 2820
1702
+ },
1703
+ {
1704
+ "epoch": 0.69,
1705
+ "learning_rate": 1.0789866634733988e-05,
1706
+ "loss": 0.1418,
1707
+ "step": 2830
1708
+ },
1709
+ {
1710
+ "epoch": 0.69,
1711
+ "learning_rate": 1.0632180519366796e-05,
1712
+ "loss": 0.1202,
1713
+ "step": 2840
1714
+ },
1715
+ {
1716
+ "epoch": 0.7,
1717
+ "learning_rate": 1.0475343348294978e-05,
1718
+ "loss": 0.1304,
1719
+ "step": 2850
1720
+ },
1721
+ {
1722
+ "epoch": 0.7,
1723
+ "learning_rate": 1.0319364388479455e-05,
1724
+ "loss": 0.15,
1725
+ "step": 2860
1726
+ },
1727
+ {
1728
+ "epoch": 0.7,
1729
+ "learning_rate": 1.0164252856172445e-05,
1730
+ "loss": 0.1306,
1731
+ "step": 2870
1732
+ },
1733
+ {
1734
+ "epoch": 0.7,
1735
+ "learning_rate": 1.0010017916372908e-05,
1736
+ "loss": 0.1394,
1737
+ "step": 2880
1738
+ },
1739
+ {
1740
+ "epoch": 0.71,
1741
+ "learning_rate": 9.856668682285003e-06,
1742
+ "loss": 0.1545,
1743
+ "step": 2890
1744
+ },
1745
+ {
1746
+ "epoch": 0.71,
1747
+ "learning_rate": 9.704214214779625e-06,
1748
+ "loss": 0.1261,
1749
+ "step": 2900
1750
+ },
1751
+ {
1752
+ "epoch": 0.71,
1753
+ "learning_rate": 9.552663521859048e-06,
1754
+ "loss": 0.1222,
1755
+ "step": 2910
1756
+ },
1757
+ {
1758
+ "epoch": 0.71,
1759
+ "learning_rate": 9.402025558124658e-06,
1760
+ "loss": 0.1253,
1761
+ "step": 2920
1762
+ },
1763
+ {
1764
+ "epoch": 0.72,
1765
+ "learning_rate": 9.25230922424786e-06,
1766
+ "loss": 0.1418,
1767
+ "step": 2930
1768
+ },
1769
+ {
1770
+ "epoch": 0.72,
1771
+ "learning_rate": 9.103523366444164e-06,
1772
+ "loss": 0.1211,
1773
+ "step": 2940
1774
+ },
1775
+ {
1776
+ "epoch": 0.72,
1777
+ "learning_rate": 8.9556767759505e-06,
1778
+ "loss": 0.1251,
1779
+ "step": 2950
1780
+ },
1781
+ {
1782
+ "epoch": 0.72,
1783
+ "learning_rate": 8.808778188505779e-06,
1784
+ "loss": 0.1331,
1785
+ "step": 2960
1786
+ },
1787
+ {
1788
+ "epoch": 0.73,
1789
+ "learning_rate": 8.6628362838347e-06,
1790
+ "loss": 0.15,
1791
+ "step": 2970
1792
+ },
1793
+ {
1794
+ "epoch": 0.73,
1795
+ "learning_rate": 8.517859685134932e-06,
1796
+ "loss": 0.1152,
1797
+ "step": 2980
1798
+ },
1799
+ {
1800
+ "epoch": 0.73,
1801
+ "learning_rate": 8.373856958567589e-06,
1802
+ "loss": 0.1244,
1803
+ "step": 2990
1804
+ },
1805
+ {
1806
+ "epoch": 0.73,
1807
+ "learning_rate": 8.230836612751069e-06,
1808
+ "loss": 0.1491,
1809
+ "step": 3000
1810
+ },
1811
+ {
1812
+ "epoch": 0.74,
1813
+ "learning_rate": 8.088807098258328e-06,
1814
+ "loss": 0.1387,
1815
+ "step": 3010
1816
+ },
1817
+ {
1818
+ "epoch": 0.74,
1819
+ "learning_rate": 7.94777680711756e-06,
1820
+ "loss": 0.1497,
1821
+ "step": 3020
1822
+ },
1823
+ {
1824
+ "epoch": 0.74,
1825
+ "learning_rate": 7.807754072316323e-06,
1826
+ "loss": 0.1296,
1827
+ "step": 3030
1828
+ },
1829
+ {
1830
+ "epoch": 0.74,
1831
+ "learning_rate": 7.668747167309209e-06,
1832
+ "loss": 0.1278,
1833
+ "step": 3040
1834
+ },
1835
+ {
1836
+ "epoch": 0.75,
1837
+ "learning_rate": 7.530764305528959e-06,
1838
+ "loss": 0.1334,
1839
+ "step": 3050
1840
+ },
1841
+ {
1842
+ "epoch": 0.75,
1843
+ "learning_rate": 7.393813639901195e-06,
1844
+ "loss": 0.1314,
1845
+ "step": 3060
1846
+ },
1847
+ {
1848
+ "epoch": 0.75,
1849
+ "learning_rate": 7.257903262362637e-06,
1850
+ "loss": 0.1251,
1851
+ "step": 3070
1852
+ },
1853
+ {
1854
+ "epoch": 0.75,
1855
+ "learning_rate": 7.123041203383032e-06,
1856
+ "loss": 0.1127,
1857
+ "step": 3080
1858
+ },
1859
+ {
1860
+ "epoch": 0.76,
1861
+ "learning_rate": 6.989235431490676e-06,
1862
+ "loss": 0.1363,
1863
+ "step": 3090
1864
+ },
1865
+ {
1866
+ "epoch": 0.76,
1867
+ "learning_rate": 6.856493852801485e-06,
1868
+ "loss": 0.1184,
1869
+ "step": 3100
1870
+ },
1871
+ {
1872
+ "epoch": 0.76,
1873
+ "learning_rate": 6.724824310551961e-06,
1874
+ "loss": 0.1475,
1875
+ "step": 3110
1876
+ },
1877
+ {
1878
+ "epoch": 0.76,
1879
+ "learning_rate": 6.594234584635694e-06,
1880
+ "loss": 0.1336,
1881
+ "step": 3120
1882
+ },
1883
+ {
1884
+ "epoch": 0.77,
1885
+ "learning_rate": 6.464732391143713e-06,
1886
+ "loss": 0.127,
1887
+ "step": 3130
1888
+ },
1889
+ {
1890
+ "epoch": 0.77,
1891
+ "learning_rate": 6.336325381908529e-06,
1892
+ "loss": 0.1315,
1893
+ "step": 3140
1894
+ },
1895
+ {
1896
+ "epoch": 0.77,
1897
+ "learning_rate": 6.209021144052054e-06,
1898
+ "loss": 0.1257,
1899
+ "step": 3150
1900
+ },
1901
+ {
1902
+ "epoch": 0.77,
1903
+ "learning_rate": 6.082827199537295e-06,
1904
+ "loss": 0.1356,
1905
+ "step": 3160
1906
+ },
1907
+ {
1908
+ "epoch": 0.78,
1909
+ "learning_rate": 5.9577510047238825e-06,
1910
+ "loss": 0.1306,
1911
+ "step": 3170
1912
+ },
1913
+ {
1914
+ "epoch": 0.78,
1915
+ "learning_rate": 5.833799949927537e-06,
1916
+ "loss": 0.1163,
1917
+ "step": 3180
1918
+ },
1919
+ {
1920
+ "epoch": 0.78,
1921
+ "learning_rate": 5.710981358983384e-06,
1922
+ "loss": 0.1333,
1923
+ "step": 3190
1924
+ },
1925
+ {
1926
+ "epoch": 0.78,
1927
+ "learning_rate": 5.589302488813186e-06,
1928
+ "loss": 0.139,
1929
+ "step": 3200
1930
+ },
1931
+ {
1932
+ "epoch": 0.79,
1933
+ "learning_rate": 5.468770528996614e-06,
1934
+ "loss": 0.1261,
1935
+ "step": 3210
1936
+ },
1937
+ {
1938
+ "epoch": 0.79,
1939
+ "learning_rate": 5.349392601346398e-06,
1940
+ "loss": 0.1274,
1941
+ "step": 3220
1942
+ },
1943
+ {
1944
+ "epoch": 0.79,
1945
+ "learning_rate": 5.231175759487536e-06,
1946
+ "loss": 0.1136,
1947
+ "step": 3230
1948
+ },
1949
+ {
1950
+ "epoch": 0.79,
1951
+ "learning_rate": 5.114126988440523e-06,
1952
+ "loss": 0.1289,
1953
+ "step": 3240
1954
+ },
1955
+ {
1956
+ "epoch": 0.8,
1957
+ "learning_rate": 4.998253204208625e-06,
1958
+ "loss": 0.1462,
1959
+ "step": 3250
1960
+ },
1961
+ {
1962
+ "epoch": 0.8,
1963
+ "learning_rate": 4.8835612533692455e-06,
1964
+ "loss": 0.1209,
1965
+ "step": 3260
1966
+ },
1967
+ {
1968
+ "epoch": 0.8,
1969
+ "learning_rate": 4.770057912669362e-06,
1970
+ "loss": 0.1266,
1971
+ "step": 3270
1972
+ },
1973
+ {
1974
+ "epoch": 0.8,
1975
+ "learning_rate": 4.65774988862514e-06,
1976
+ "loss": 0.1487,
1977
+ "step": 3280
1978
+ },
1979
+ {
1980
+ "epoch": 0.8,
1981
+ "learning_rate": 4.546643817125662e-06,
1982
+ "loss": 0.1101,
1983
+ "step": 3290
1984
+ },
1985
+ {
1986
+ "epoch": 0.81,
1987
+ "learning_rate": 4.436746263040819e-06,
1988
+ "loss": 0.1183,
1989
+ "step": 3300
1990
+ },
1991
+ {
1992
+ "epoch": 0.81,
1993
+ "learning_rate": 4.328063719833433e-06,
1994
+ "loss": 0.1133,
1995
+ "step": 3310
1996
+ },
1997
+ {
1998
+ "epoch": 0.81,
1999
+ "learning_rate": 4.220602609175575e-06,
2000
+ "loss": 0.1159,
2001
+ "step": 3320
2002
+ },
2003
+ {
2004
+ "epoch": 0.81,
2005
+ "learning_rate": 4.114369280569136e-06,
2006
+ "loss": 0.1288,
2007
+ "step": 3330
2008
+ },
2009
+ {
2010
+ "epoch": 0.82,
2011
+ "learning_rate": 4.009370010970634e-06,
2012
+ "loss": 0.1222,
2013
+ "step": 3340
2014
+ },
2015
+ {
2016
+ "epoch": 0.82,
2017
+ "learning_rate": 3.90561100442036e-06,
2018
+ "loss": 0.1208,
2019
+ "step": 3350
2020
+ },
2021
+ {
2022
+ "epoch": 0.82,
2023
+ "learning_rate": 3.8030983916757966e-06,
2024
+ "loss": 0.1096,
2025
+ "step": 3360
2026
+ },
2027
+ {
2028
+ "epoch": 0.82,
2029
+ "learning_rate": 3.7018382298493635e-06,
2030
+ "loss": 0.1157,
2031
+ "step": 3370
2032
+ },
2033
+ {
2034
+ "epoch": 0.83,
2035
+ "learning_rate": 3.6018365020505124e-06,
2036
+ "loss": 0.1264,
2037
+ "step": 3380
2038
+ },
2039
+ {
2040
+ "epoch": 0.83,
2041
+ "learning_rate": 3.503099117032249e-06,
2042
+ "loss": 0.1315,
2043
+ "step": 3390
2044
+ },
2045
+ {
2046
+ "epoch": 0.83,
2047
+ "learning_rate": 3.405631908841966e-06,
2048
+ "loss": 0.1317,
2049
+ "step": 3400
2050
+ },
2051
+ {
2052
+ "epoch": 0.83,
2053
+ "learning_rate": 3.3094406364767267e-06,
2054
+ "loss": 0.1382,
2055
+ "step": 3410
2056
+ },
2057
+ {
2058
+ "epoch": 0.84,
2059
+ "learning_rate": 3.214530983543021e-06,
2060
+ "loss": 0.1366,
2061
+ "step": 3420
2062
+ },
2063
+ {
2064
+ "epoch": 0.84,
2065
+ "learning_rate": 3.120908557920918e-06,
2066
+ "loss": 0.1308,
2067
+ "step": 3430
2068
+ },
2069
+ {
2070
+ "epoch": 0.84,
2071
+ "learning_rate": 3.028578891432701e-06,
2072
+ "loss": 0.1448,
2073
+ "step": 3440
2074
+ },
2075
+ {
2076
+ "epoch": 0.84,
2077
+ "learning_rate": 2.9375474395160528e-06,
2078
+ "loss": 0.1292,
2079
+ "step": 3450
2080
+ },
2081
+ {
2082
+ "epoch": 0.85,
2083
+ "learning_rate": 2.8478195809016674e-06,
2084
+ "loss": 0.13,
2085
+ "step": 3460
2086
+ },
2087
+ {
2088
+ "epoch": 0.85,
2089
+ "learning_rate": 2.7594006172954906e-06,
2090
+ "loss": 0.1333,
2091
+ "step": 3470
2092
+ },
2093
+ {
2094
+ "epoch": 0.85,
2095
+ "learning_rate": 2.6722957730653973e-06,
2096
+ "loss": 0.1527,
2097
+ "step": 3480
2098
+ },
2099
+ {
2100
+ "epoch": 0.85,
2101
+ "learning_rate": 2.5865101949325567e-06,
2102
+ "loss": 0.1154,
2103
+ "step": 3490
2104
+ },
2105
+ {
2106
+ "epoch": 0.86,
2107
+ "learning_rate": 2.50204895166731e-06,
2108
+ "loss": 0.1247,
2109
+ "step": 3500
2110
+ },
2111
+ {
2112
+ "epoch": 0.86,
2113
+ "learning_rate": 2.418917033789661e-06,
2114
+ "loss": 0.1211,
2115
+ "step": 3510
2116
+ },
2117
+ {
2118
+ "epoch": 0.86,
2119
+ "learning_rate": 2.3371193532744305e-06,
2120
+ "loss": 0.1208,
2121
+ "step": 3520
2122
+ },
2123
+ {
2124
+ "epoch": 0.86,
2125
+ "learning_rate": 2.2566607432610047e-06,
2126
+ "loss": 0.1208,
2127
+ "step": 3530
2128
+ },
2129
+ {
2130
+ "epoch": 0.87,
2131
+ "learning_rate": 2.177545957767771e-06,
2132
+ "loss": 0.1251,
2133
+ "step": 3540
2134
+ },
2135
+ {
2136
+ "epoch": 0.87,
2137
+ "learning_rate": 2.0997796714112144e-06,
2138
+ "loss": 0.1175,
2139
+ "step": 3550
2140
+ },
2141
+ {
2142
+ "epoch": 0.87,
2143
+ "learning_rate": 2.023366479129718e-06,
2144
+ "loss": 0.1207,
2145
+ "step": 3560
2146
+ },
2147
+ {
2148
+ "epoch": 0.87,
2149
+ "learning_rate": 1.9483108959120587e-06,
2150
+ "loss": 0.1198,
2151
+ "step": 3570
2152
+ },
2153
+ {
2154
+ "epoch": 0.88,
2155
+ "learning_rate": 1.874617356530628e-06,
2156
+ "loss": 0.113,
2157
+ "step": 3580
2158
+ },
2159
+ {
2160
+ "epoch": 0.88,
2161
+ "learning_rate": 1.8022902152794036e-06,
2162
+ "loss": 0.1261,
2163
+ "step": 3590
2164
+ },
2165
+ {
2166
+ "epoch": 0.88,
2167
+ "learning_rate": 1.7313337457166707e-06,
2168
+ "loss": 0.1297,
2169
+ "step": 3600
2170
+ },
2171
+ {
2172
+ "epoch": 0.88,
2173
+ "learning_rate": 1.661752140412512e-06,
2174
+ "loss": 0.1197,
2175
+ "step": 3610
2176
+ },
2177
+ {
2178
+ "epoch": 0.89,
2179
+ "learning_rate": 1.5935495107010728e-06,
2180
+ "loss": 0.133,
2181
+ "step": 3620
2182
+ },
2183
+ {
2184
+ "epoch": 0.89,
2185
+ "learning_rate": 1.5267298864376527e-06,
2186
+ "loss": 0.1245,
2187
+ "step": 3630
2188
+ },
2189
+ {
2190
+ "epoch": 0.89,
2191
+ "learning_rate": 1.461297215760582e-06,
2192
+ "loss": 0.1337,
2193
+ "step": 3640
2194
+ },
2195
+ {
2196
+ "epoch": 0.89,
2197
+ "learning_rate": 1.3972553648579468e-06,
2198
+ "loss": 0.1437,
2199
+ "step": 3650
2200
+ },
2201
+ {
2202
+ "epoch": 0.9,
2203
+ "learning_rate": 1.3346081177391472e-06,
2204
+ "loss": 0.1467,
2205
+ "step": 3660
2206
+ },
2207
+ {
2208
+ "epoch": 0.9,
2209
+ "learning_rate": 1.2733591760113184e-06,
2210
+ "loss": 0.1298,
2211
+ "step": 3670
2212
+ },
2213
+ {
2214
+ "epoch": 0.9,
2215
+ "learning_rate": 1.2135121586606014e-06,
2216
+ "loss": 0.1239,
2217
+ "step": 3680
2218
+ },
2219
+ {
2220
+ "epoch": 0.9,
2221
+ "learning_rate": 1.1550706018383345e-06,
2222
+ "loss": 0.1252,
2223
+ "step": 3690
2224
+ },
2225
+ {
2226
+ "epoch": 0.91,
2227
+ "learning_rate": 1.09803795865209e-06,
2228
+ "loss": 0.1475,
2229
+ "step": 3700
2230
+ },
2231
+ {
2232
+ "epoch": 0.91,
2233
+ "learning_rate": 1.0424175989616563e-06,
2234
+ "loss": 0.1365,
2235
+ "step": 3710
2236
+ },
2237
+ {
2238
+ "epoch": 0.91,
2239
+ "learning_rate": 9.882128091799114e-07,
2240
+ "loss": 0.127,
2241
+ "step": 3720
2242
+ },
2243
+ {
2244
+ "epoch": 0.91,
2245
+ "learning_rate": 9.354267920786614e-07,
2246
+ "loss": 0.1281,
2247
+ "step": 3730
2248
+ },
2249
+ {
2250
+ "epoch": 0.92,
2251
+ "learning_rate": 8.840626665993807e-07,
2252
+ "loss": 0.1219,
2253
+ "step": 3740
2254
+ },
2255
+ {
2256
+ "epoch": 0.92,
2257
+ "learning_rate": 8.341234676689319e-07,
2258
+ "loss": 0.1227,
2259
+ "step": 3750
2260
+ },
2261
+ {
2262
+ "epoch": 0.92,
2263
+ "learning_rate": 7.856121460202398e-07,
2264
+ "loss": 0.1568,
2265
+ "step": 3760
2266
+ },
2267
+ {
2268
+ "epoch": 0.92,
2269
+ "learning_rate": 7.385315680179583e-07,
2270
+ "loss": 0.1202,
2271
+ "step": 3770
2272
+ },
2273
+ {
2274
+ "epoch": 0.92,
2275
+ "learning_rate": 6.928845154890757e-07,
2276
+ "loss": 0.1374,
2277
+ "step": 3780
2278
+ },
2279
+ {
2280
+ "epoch": 0.93,
2281
+ "learning_rate": 6.486736855585762e-07,
2282
+ "loss": 0.1235,
2283
+ "step": 3790
2284
+ },
2285
+ {
2286
+ "epoch": 0.93,
2287
+ "learning_rate": 6.059016904900628e-07,
2288
+ "loss": 0.1252,
2289
+ "step": 3800
2290
+ },
2291
+ {
2292
+ "epoch": 0.93,
2293
+ "learning_rate": 5.645710575314073e-07,
2294
+ "loss": 0.1261,
2295
+ "step": 3810
2296
+ },
2297
+ {
2298
+ "epoch": 0.93,
2299
+ "learning_rate": 5.246842287654235e-07,
2300
+ "loss": 0.1338,
2301
+ "step": 3820
2302
+ },
2303
+ {
2304
+ "epoch": 0.94,
2305
+ "learning_rate": 4.862435609655824e-07,
2306
+ "loss": 0.1345,
2307
+ "step": 3830
2308
+ },
2309
+ {
2310
+ "epoch": 0.94,
2311
+ "learning_rate": 4.492513254567482e-07,
2312
+ "loss": 0.1164,
2313
+ "step": 3840
2314
+ },
2315
+ {
2316
+ "epoch": 0.94,
2317
+ "learning_rate": 4.137097079809776e-07,
2318
+ "loss": 0.1212,
2319
+ "step": 3850
2320
+ },
2321
+ {
2322
+ "epoch": 0.94,
2323
+ "learning_rate": 3.796208085683733e-07,
2324
+ "loss": 0.1289,
2325
+ "step": 3860
2326
+ },
2327
+ {
2328
+ "epoch": 0.95,
2329
+ "learning_rate": 3.4698664141299686e-07,
2330
+ "loss": 0.1412,
2331
+ "step": 3870
2332
+ },
2333
+ {
2334
+ "epoch": 0.95,
2335
+ "learning_rate": 3.1580913475386107e-07,
2336
+ "loss": 0.1293,
2337
+ "step": 3880
2338
+ },
2339
+ {
2340
+ "epoch": 0.95,
2341
+ "learning_rate": 2.8609013076099365e-07,
2342
+ "loss": 0.1191,
2343
+ "step": 3890
2344
+ },
2345
+ {
2346
+ "epoch": 0.95,
2347
+ "learning_rate": 2.5783138542659345e-07,
2348
+ "loss": 0.1247,
2349
+ "step": 3900
2350
+ },
2351
+ {
2352
+ "epoch": 0.96,
2353
+ "learning_rate": 2.3103456846126913e-07,
2354
+ "loss": 0.1175,
2355
+ "step": 3910
2356
+ },
2357
+ {
2358
+ "epoch": 0.96,
2359
+ "learning_rate": 2.057012631953903e-07,
2360
+ "loss": 0.1317,
2361
+ "step": 3920
2362
+ },
2363
+ {
2364
+ "epoch": 0.96,
2365
+ "learning_rate": 1.8183296648552616e-07,
2366
+ "loss": 0.1421,
2367
+ "step": 3930
2368
+ },
2369
+ {
2370
+ "epoch": 0.96,
2371
+ "learning_rate": 1.5943108862600809e-07,
2372
+ "loss": 0.1365,
2373
+ "step": 3940
2374
+ },
2375
+ {
2376
+ "epoch": 0.97,
2377
+ "learning_rate": 1.384969532655933e-07,
2378
+ "loss": 0.1261,
2379
+ "step": 3950
2380
+ },
2381
+ {
2382
+ "epoch": 0.97,
2383
+ "learning_rate": 1.190317973292554e-07,
2384
+ "loss": 0.1054,
2385
+ "step": 3960
2386
+ },
2387
+ {
2388
+ "epoch": 0.97,
2389
+ "learning_rate": 1.0103677094510932e-07,
2390
+ "loss": 0.1251,
2391
+ "step": 3970
2392
+ },
2393
+ {
2394
+ "epoch": 0.97,
2395
+ "learning_rate": 8.451293737644638e-08,
2396
+ "loss": 0.1446,
2397
+ "step": 3980
2398
+ },
2399
+ {
2400
+ "epoch": 0.98,
2401
+ "learning_rate": 6.946127295890392e-08,
2402
+ "loss": 0.1271,
2403
+ "step": 3990
2404
+ },
2405
+ {
2406
+ "epoch": 0.98,
2407
+ "learning_rate": 5.5882667042789214e-08,
2408
+ "loss": 0.1372,
2409
+ "step": 4000
2410
+ }
2411
+ ],
2412
+ "logging_steps": 10,
2413
+ "max_steps": 4087,
2414
+ "num_train_epochs": 1,
2415
+ "save_steps": 1000,
2416
+ "total_flos": 2.4557146854199296e+18,
2417
+ "trial_name": null,
2418
+ "trial_params": null
2419
+ }
Baichuan-13B-Chat-lora-Computing/checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fe319e7156520da103d89f67f7026a8a887d2ca0d81076b6f0b7a8365ab7db8
3
+ size 3513
Baichuan-13B-Chat-lora-Computing/finetuning_args.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "finetuning_type": "lora",
3
+ "lora_alpha": 32.0,
4
+ "lora_dropout": 0.1,
5
+ "lora_rank": 8,
6
+ "lora_target": [
7
+ "W_pack"
8
+ ],
9
+ "name_module_trainable": "mlp",
10
+ "num_hidden_layers": 32,
11
+ "num_layer_trainable": 3
12
+ }
Baichuan-13B-Chat-lora-Computing/runs/Oct14_22-56-58_64d98f3565d5/events.out.tfevents.1697324346.64d98f3565d5.1992391.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:158678988d541382e81eaebd07cb3c161f08e89a82f94e96ca1f9fd854fdb941
3
+ size 68890
Baichuan-13B-Chat-lora-Computing/train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.1560091979059696,
4
+ "train_runtime": 21225.5051,
5
+ "train_samples_per_second": 3.081,
6
+ "train_steps_per_second": 0.193
7
+ }
Baichuan-13B-Chat-lora-Computing/trainer_log.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
Baichuan-13B-Chat-lora-Computing/trainer_state.json ADDED
@@ -0,0 +1,2476 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 4087,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4.999926141863722e-05,
14
+ "loss": 1.0318,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999704571818908e-05,
20
+ "loss": 0.8189,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 4.999335302957357e-05,
26
+ "loss": 0.6632,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 4.9988183570978795e-05,
32
+ "loss": 0.5815,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 4.998153764784998e-05,
38
+ "loss": 0.5221,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 4.9973415652871555e-05,
44
+ "loss": 0.4635,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.02,
49
+ "learning_rate": 4.9963818065943846e-05,
50
+ "loss": 0.4266,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 4.9952745454154736e-05,
56
+ "loss": 0.4073,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.02,
61
+ "learning_rate": 4.9940198471746226e-05,
62
+ "loss": 0.373,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 4.9926177860075696e-05,
68
+ "loss": 0.3172,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.03,
73
+ "learning_rate": 4.991068444757214e-05,
74
+ "loss": 0.2985,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.03,
79
+ "learning_rate": 4.989371914968722e-05,
80
+ "loss": 0.2965,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.03,
85
+ "learning_rate": 4.987528296884116e-05,
86
+ "loss": 0.29,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.03,
91
+ "learning_rate": 4.985537699436353e-05,
92
+ "loss": 0.2725,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.04,
97
+ "learning_rate": 4.983400240242887e-05,
98
+ "loss": 0.236,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.04,
103
+ "learning_rate": 4.9811160455987196e-05,
104
+ "loss": 0.2672,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.04,
109
+ "learning_rate": 4.978685250468939e-05,
110
+ "loss": 0.2448,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.04,
115
+ "learning_rate": 4.9761079984807426e-05,
116
+ "loss": 0.2553,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.05,
121
+ "learning_rate": 4.973384441914953e-05,
122
+ "loss": 0.2133,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.05,
127
+ "learning_rate": 4.970514741697021e-05,
128
+ "loss": 0.2351,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.05,
133
+ "learning_rate": 4.967499067387514e-05,
134
+ "loss": 0.1805,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.05,
139
+ "learning_rate": 4.964337597172099e-05,
140
+ "loss": 0.2253,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.06,
145
+ "learning_rate": 4.961030517851014e-05,
146
+ "loss": 0.1853,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.06,
151
+ "learning_rate": 4.957578024828033e-05,
152
+ "loss": 0.2271,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.06,
157
+ "learning_rate": 4.953980322098914e-05,
158
+ "loss": 0.2112,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.06,
163
+ "learning_rate": 4.9502376222393525e-05,
164
+ "loss": 0.22,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.07,
169
+ "learning_rate": 4.9463501463924175e-05,
170
+ "loss": 0.1825,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.07,
175
+ "learning_rate": 4.942318124255487e-05,
176
+ "loss": 0.242,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.07,
181
+ "learning_rate": 4.9381417940666715e-05,
182
+ "loss": 0.2179,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.07,
187
+ "learning_rate": 4.9338214025907424e-05,
188
+ "loss": 0.183,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.08,
193
+ "learning_rate": 4.9293572051045514e-05,
194
+ "loss": 0.1886,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.08,
199
+ "learning_rate": 4.924749465381942e-05,
200
+ "loss": 0.2062,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.08,
205
+ "learning_rate": 4.91999845567817e-05,
206
+ "loss": 0.1647,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.08,
211
+ "learning_rate": 4.915104456713812e-05,
212
+ "loss": 0.1828,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.09,
217
+ "learning_rate": 4.910067757658182e-05,
218
+ "loss": 0.1781,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.09,
223
+ "learning_rate": 4.904888656112246e-05,
224
+ "loss": 0.2016,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.09,
229
+ "learning_rate": 4.899567458091031e-05,
230
+ "loss": 0.1761,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.09,
235
+ "learning_rate": 4.894104478005554e-05,
236
+ "loss": 0.1729,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.1,
241
+ "learning_rate": 4.888500038644238e-05,
242
+ "loss": 0.1894,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.1,
247
+ "learning_rate": 4.8827544711538386e-05,
248
+ "loss": 0.2072,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.1,
253
+ "learning_rate": 4.876868115019881e-05,
254
+ "loss": 0.1724,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.1,
259
+ "learning_rate": 4.8708413180466e-05,
260
+ "loss": 0.2249,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.11,
265
+ "learning_rate": 4.8646744363363904e-05,
266
+ "loss": 0.2007,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.11,
271
+ "learning_rate": 4.858367834268763e-05,
272
+ "loss": 0.1877,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.11,
277
+ "learning_rate": 4.851921884478817e-05,
278
+ "loss": 0.1827,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.11,
283
+ "learning_rate": 4.845336967835226e-05,
284
+ "loss": 0.1819,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.11,
289
+ "learning_rate": 4.8386134734177225e-05,
290
+ "loss": 0.1856,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.12,
295
+ "learning_rate": 4.831751798494123e-05,
296
+ "loss": 0.1685,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.12,
301
+ "learning_rate": 4.824752348496844e-05,
302
+ "loss": 0.1679,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.12,
307
+ "learning_rate": 4.8176155369989505e-05,
308
+ "loss": 0.1788,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.12,
313
+ "learning_rate": 4.8103417856897206e-05,
314
+ "loss": 0.1872,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.13,
319
+ "learning_rate": 4.8029315243497244e-05,
320
+ "loss": 0.1697,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.13,
325
+ "learning_rate": 4.7953851908254376e-05,
326
+ "loss": 0.162,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.13,
331
+ "learning_rate": 4.787703231003363e-05,
332
+ "loss": 0.1718,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.13,
337
+ "learning_rate": 4.77988609878369e-05,
338
+ "loss": 0.1866,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.14,
343
+ "learning_rate": 4.771934256053471e-05,
344
+ "loss": 0.1609,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.14,
349
+ "learning_rate": 4.763848172659333e-05,
350
+ "loss": 0.1629,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.14,
355
+ "learning_rate": 4.755628326379716e-05,
356
+ "loss": 0.1947,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.14,
361
+ "learning_rate": 4.747275202896642e-05,
362
+ "loss": 0.1813,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.15,
367
+ "learning_rate": 4.738789295767016e-05,
368
+ "loss": 0.1672,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.15,
373
+ "learning_rate": 4.730171106393466e-05,
374
+ "loss": 0.1597,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.15,
379
+ "learning_rate": 4.7214211439947174e-05,
380
+ "loss": 0.1855,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.15,
385
+ "learning_rate": 4.712539925575502e-05,
386
+ "loss": 0.1729,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.16,
391
+ "learning_rate": 4.7035279758960104e-05,
392
+ "loss": 0.1771,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.16,
397
+ "learning_rate": 4.6943858274408915e-05,
398
+ "loss": 0.1891,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.16,
403
+ "learning_rate": 4.68511402038778e-05,
404
+ "loss": 0.1442,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.16,
409
+ "learning_rate": 4.6757131025753886e-05,
410
+ "loss": 0.1666,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.17,
415
+ "learning_rate": 4.666183629471132e-05,
416
+ "loss": 0.1672,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.17,
421
+ "learning_rate": 4.656526164138307e-05,
422
+ "loss": 0.1525,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.17,
427
+ "learning_rate": 4.646741277202829e-05,
428
+ "loss": 0.1758,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.17,
433
+ "learning_rate": 4.6368295468195064e-05,
434
+ "loss": 0.1708,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.18,
439
+ "learning_rate": 4.626791558637887e-05,
440
+ "loss": 0.1682,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.18,
445
+ "learning_rate": 4.6166279057676494e-05,
446
+ "loss": 0.1515,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.18,
451
+ "learning_rate": 4.6063391887435605e-05,
452
+ "loss": 0.1687,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.18,
457
+ "learning_rate": 4.595926015489991e-05,
458
+ "loss": 0.1641,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.19,
463
+ "learning_rate": 4.585389001284999e-05,
464
+ "loss": 0.1536,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.19,
469
+ "learning_rate": 4.574728768723965e-05,
470
+ "loss": 0.1683,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 0.19,
475
+ "learning_rate": 4.563945947682821e-05,
476
+ "loss": 0.1464,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 0.19,
481
+ "learning_rate": 4.5530411752808166e-05,
482
+ "loss": 0.1764,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 0.2,
487
+ "learning_rate": 4.542015095842885e-05,
488
+ "loss": 0.19,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 0.2,
493
+ "learning_rate": 4.5308683608615706e-05,
494
+ "loss": 0.1819,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 0.2,
499
+ "learning_rate": 4.519601628958529e-05,
500
+ "loss": 0.1512,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 0.2,
505
+ "learning_rate": 4.508215565845615e-05,
506
+ "loss": 0.1689,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 0.21,
511
+ "learning_rate": 4.496710844285551e-05,
512
+ "loss": 0.1739,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 0.21,
517
+ "learning_rate": 4.485088144052172e-05,
518
+ "loss": 0.1535,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 0.21,
523
+ "learning_rate": 4.4733481518902584e-05,
524
+ "loss": 0.142,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 0.21,
529
+ "learning_rate": 4.4614915614749645e-05,
530
+ "loss": 0.1584,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 0.22,
535
+ "learning_rate": 4.449519073370827e-05,
536
+ "loss": 0.1483,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 0.22,
541
+ "learning_rate": 4.43743139499037e-05,
542
+ "loss": 0.1587,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 0.22,
547
+ "learning_rate": 4.425229240552313e-05,
548
+ "loss": 0.1497,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 0.22,
553
+ "learning_rate": 4.4129133310393646e-05,
554
+ "loss": 0.1368,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 0.23,
559
+ "learning_rate": 4.400484394155622e-05,
560
+ "loss": 0.1565,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 0.23,
565
+ "learning_rate": 4.3879431642835784e-05,
566
+ "loss": 0.1377,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 0.23,
571
+ "learning_rate": 4.3752903824407234e-05,
572
+ "loss": 0.1755,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 0.23,
577
+ "learning_rate": 4.3625267962357676e-05,
578
+ "loss": 0.165,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 0.23,
583
+ "learning_rate": 4.3496531598244616e-05,
584
+ "loss": 0.1397,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 0.24,
589
+ "learning_rate": 4.33667023386504e-05,
590
+ "loss": 0.1518,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 0.24,
595
+ "learning_rate": 4.323578785473272e-05,
596
+ "loss": 0.1332,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 0.24,
601
+ "learning_rate": 4.3103795881771454e-05,
602
+ "loss": 0.1541,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 0.24,
607
+ "learning_rate": 4.2970734218711474e-05,
608
+ "loss": 0.1508,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 0.25,
613
+ "learning_rate": 4.283661072770193e-05,
614
+ "loss": 0.1465,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 0.25,
619
+ "learning_rate": 4.270143333363171e-05,
620
+ "loss": 0.1505,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 0.25,
625
+ "learning_rate": 4.2565210023661104e-05,
626
+ "loss": 0.1674,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 0.25,
631
+ "learning_rate": 4.242794884674996e-05,
632
+ "loss": 0.1352,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 0.26,
637
+ "learning_rate": 4.2289657913182045e-05,
638
+ "loss": 0.1667,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 0.26,
643
+ "learning_rate": 4.2150345394085845e-05,
644
+ "loss": 0.1495,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 0.26,
649
+ "learning_rate": 4.201001952095178e-05,
650
+ "loss": 0.1538,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 0.26,
655
+ "learning_rate": 4.186868858514582e-05,
656
+ "loss": 0.1425,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 0.27,
661
+ "learning_rate": 4.172636093741959e-05,
662
+ "loss": 0.1357,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 0.27,
667
+ "learning_rate": 4.158304498741691e-05,
668
+ "loss": 0.1493,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 0.27,
673
+ "learning_rate": 4.143874920317696e-05,
674
+ "loss": 0.1439,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 0.27,
679
+ "learning_rate": 4.1293482110633905e-05,
680
+ "loss": 0.1449,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 0.28,
685
+ "learning_rate": 4.114725229311311e-05,
686
+ "loss": 0.1451,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 0.28,
691
+ "learning_rate": 4.100006839082402e-05,
692
+ "loss": 0.1559,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 0.28,
697
+ "learning_rate": 4.085193910034959e-05,
698
+ "loss": 0.1452,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 0.28,
703
+ "learning_rate": 4.0702873174132483e-05,
704
+ "loss": 0.1584,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 0.29,
709
+ "learning_rate": 4.0552879419957904e-05,
710
+ "loss": 0.1341,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 0.29,
715
+ "learning_rate": 4.0401966700433144e-05,
716
+ "loss": 0.1433,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 0.29,
721
+ "learning_rate": 4.025014393246399e-05,
722
+ "loss": 0.1583,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 0.29,
727
+ "learning_rate": 4.0097420086727766e-05,
728
+ "loss": 0.1508,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 0.3,
733
+ "learning_rate": 3.9943804187143384e-05,
734
+ "loss": 0.1577,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 0.3,
739
+ "learning_rate": 3.978930531033807e-05,
740
+ "loss": 0.1431,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 0.3,
745
+ "learning_rate": 3.963393258511109e-05,
746
+ "loss": 0.161,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 0.3,
751
+ "learning_rate": 3.9477695191894395e-05,
752
+ "loss": 0.1396,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 0.31,
757
+ "learning_rate": 3.9320602362210116e-05,
758
+ "loss": 0.1477,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 0.31,
763
+ "learning_rate": 3.9162663378125166e-05,
764
+ "loss": 0.1461,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 0.31,
769
+ "learning_rate": 3.900388757170274e-05,
770
+ "loss": 0.152,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 0.31,
775
+ "learning_rate": 3.8844284324450956e-05,
776
+ "loss": 0.1563,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 0.32,
781
+ "learning_rate": 3.8683863066768534e-05,
782
+ "loss": 0.1441,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 0.32,
787
+ "learning_rate": 3.8522633277387555e-05,
788
+ "loss": 0.1448,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 0.32,
793
+ "learning_rate": 3.836060448281342e-05,
794
+ "loss": 0.1659,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 0.32,
799
+ "learning_rate": 3.819778625676197e-05,
800
+ "loss": 0.1464,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 0.33,
805
+ "learning_rate": 3.803418821959379e-05,
806
+ "loss": 0.147,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 0.33,
811
+ "learning_rate": 3.7869820037745776e-05,
812
+ "loss": 0.1479,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 0.33,
817
+ "learning_rate": 3.770469142315997e-05,
818
+ "loss": 0.1352,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 0.33,
823
+ "learning_rate": 3.7538812132709775e-05,
824
+ "loss": 0.1638,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 0.34,
829
+ "learning_rate": 3.737219196762337e-05,
830
+ "loss": 0.149,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 0.34,
835
+ "learning_rate": 3.7204840772904645e-05,
836
+ "loss": 0.155,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 0.34,
841
+ "learning_rate": 3.703676843675147e-05,
842
+ "loss": 0.159,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 0.34,
847
+ "learning_rate": 3.686798488997146e-05,
848
+ "loss": 0.1455,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 0.34,
853
+ "learning_rate": 3.669850010539518e-05,
854
+ "loss": 0.1439,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 0.35,
859
+ "learning_rate": 3.652832409728686e-05,
860
+ "loss": 0.1344,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 0.35,
865
+ "learning_rate": 3.635746692075276e-05,
866
+ "loss": 0.1519,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 0.35,
871
+ "learning_rate": 3.6185938671146975e-05,
872
+ "loss": 0.1469,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 0.35,
877
+ "learning_rate": 3.601374948347498e-05,
878
+ "loss": 0.1343,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 0.36,
883
+ "learning_rate": 3.584090953179475e-05,
884
+ "loss": 0.143,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 0.36,
889
+ "learning_rate": 3.566742902861566e-05,
890
+ "loss": 0.137,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 0.36,
895
+ "learning_rate": 3.549331822429503e-05,
896
+ "loss": 0.1485,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 0.36,
901
+ "learning_rate": 3.531858740643247e-05,
902
+ "loss": 0.1182,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 0.37,
907
+ "learning_rate": 3.5143246899262014e-05,
908
+ "loss": 0.1528,
909
+ "step": 1500
910
+ },
911
+ {
912
+ "epoch": 0.37,
913
+ "learning_rate": 3.4967307063042134e-05,
914
+ "loss": 0.1648,
915
+ "step": 1510
916
+ },
917
+ {
918
+ "epoch": 0.37,
919
+ "learning_rate": 3.479077829344354e-05,
920
+ "loss": 0.1213,
921
+ "step": 1520
922
+ },
923
+ {
924
+ "epoch": 0.37,
925
+ "learning_rate": 3.461367102093498e-05,
926
+ "loss": 0.1412,
927
+ "step": 1530
928
+ },
929
+ {
930
+ "epoch": 0.38,
931
+ "learning_rate": 3.443599571016689e-05,
932
+ "loss": 0.1397,
933
+ "step": 1540
934
+ },
935
+ {
936
+ "epoch": 0.38,
937
+ "learning_rate": 3.425776285935315e-05,
938
+ "loss": 0.1527,
939
+ "step": 1550
940
+ },
941
+ {
942
+ "epoch": 0.38,
943
+ "learning_rate": 3.407898299965069e-05,
944
+ "loss": 0.137,
945
+ "step": 1560
946
+ },
947
+ {
948
+ "epoch": 0.38,
949
+ "learning_rate": 3.3899666694537307e-05,
950
+ "loss": 0.121,
951
+ "step": 1570
952
+ },
953
+ {
954
+ "epoch": 0.39,
955
+ "learning_rate": 3.3719824539187484e-05,
956
+ "loss": 0.1401,
957
+ "step": 1580
958
+ },
959
+ {
960
+ "epoch": 0.39,
961
+ "learning_rate": 3.353946715984635e-05,
962
+ "loss": 0.1163,
963
+ "step": 1590
964
+ },
965
+ {
966
+ "epoch": 0.39,
967
+ "learning_rate": 3.3358605213201835e-05,
968
+ "loss": 0.1562,
969
+ "step": 1600
970
+ },
971
+ {
972
+ "epoch": 0.39,
973
+ "learning_rate": 3.3177249385754964e-05,
974
+ "loss": 0.1324,
975
+ "step": 1610
976
+ },
977
+ {
978
+ "epoch": 0.4,
979
+ "learning_rate": 3.299541039318849e-05,
980
+ "loss": 0.1586,
981
+ "step": 1620
982
+ },
983
+ {
984
+ "epoch": 0.4,
985
+ "learning_rate": 3.281309897973368e-05,
986
+ "loss": 0.1626,
987
+ "step": 1630
988
+ },
989
+ {
990
+ "epoch": 0.4,
991
+ "learning_rate": 3.263032591753552e-05,
992
+ "loss": 0.1366,
993
+ "step": 1640
994
+ },
995
+ {
996
+ "epoch": 0.4,
997
+ "learning_rate": 3.2447102006016186e-05,
998
+ "loss": 0.1336,
999
+ "step": 1650
1000
+ },
1001
+ {
1002
+ "epoch": 0.41,
1003
+ "learning_rate": 3.226343807123697e-05,
1004
+ "loss": 0.1376,
1005
+ "step": 1660
1006
+ },
1007
+ {
1008
+ "epoch": 0.41,
1009
+ "learning_rate": 3.2079344965258645e-05,
1010
+ "loss": 0.132,
1011
+ "step": 1670
1012
+ },
1013
+ {
1014
+ "epoch": 0.41,
1015
+ "learning_rate": 3.189483356550015e-05,
1016
+ "loss": 0.1316,
1017
+ "step": 1680
1018
+ },
1019
+ {
1020
+ "epoch": 0.41,
1021
+ "learning_rate": 3.1709914774095975e-05,
1022
+ "loss": 0.1448,
1023
+ "step": 1690
1024
+ },
1025
+ {
1026
+ "epoch": 0.42,
1027
+ "learning_rate": 3.152459951725196e-05,
1028
+ "loss": 0.132,
1029
+ "step": 1700
1030
+ },
1031
+ {
1032
+ "epoch": 0.42,
1033
+ "learning_rate": 3.133889874459971e-05,
1034
+ "loss": 0.133,
1035
+ "step": 1710
1036
+ },
1037
+ {
1038
+ "epoch": 0.42,
1039
+ "learning_rate": 3.115282342854959e-05,
1040
+ "loss": 0.1558,
1041
+ "step": 1720
1042
+ },
1043
+ {
1044
+ "epoch": 0.42,
1045
+ "learning_rate": 3.096638456364245e-05,
1046
+ "loss": 0.152,
1047
+ "step": 1730
1048
+ },
1049
+ {
1050
+ "epoch": 0.43,
1051
+ "learning_rate": 3.0779593165899954e-05,
1052
+ "loss": 0.1293,
1053
+ "step": 1740
1054
+ },
1055
+ {
1056
+ "epoch": 0.43,
1057
+ "learning_rate": 3.059246027217372e-05,
1058
+ "loss": 0.1361,
1059
+ "step": 1750
1060
+ },
1061
+ {
1062
+ "epoch": 0.43,
1063
+ "learning_rate": 3.0404996939493156e-05,
1064
+ "loss": 0.1462,
1065
+ "step": 1760
1066
+ },
1067
+ {
1068
+ "epoch": 0.43,
1069
+ "learning_rate": 3.021721424441216e-05,
1070
+ "loss": 0.1671,
1071
+ "step": 1770
1072
+ },
1073
+ {
1074
+ "epoch": 0.44,
1075
+ "learning_rate": 3.0029123282354638e-05,
1076
+ "loss": 0.1269,
1077
+ "step": 1780
1078
+ },
1079
+ {
1080
+ "epoch": 0.44,
1081
+ "learning_rate": 2.984073516695892e-05,
1082
+ "loss": 0.1571,
1083
+ "step": 1790
1084
+ },
1085
+ {
1086
+ "epoch": 0.44,
1087
+ "learning_rate": 2.9652061029421087e-05,
1088
+ "loss": 0.139,
1089
+ "step": 1800
1090
+ },
1091
+ {
1092
+ "epoch": 0.44,
1093
+ "learning_rate": 2.9463112017837262e-05,
1094
+ "loss": 0.1234,
1095
+ "step": 1810
1096
+ },
1097
+ {
1098
+ "epoch": 0.45,
1099
+ "learning_rate": 2.927389929654494e-05,
1100
+ "loss": 0.1511,
1101
+ "step": 1820
1102
+ },
1103
+ {
1104
+ "epoch": 0.45,
1105
+ "learning_rate": 2.9084434045463255e-05,
1106
+ "loss": 0.1347,
1107
+ "step": 1830
1108
+ },
1109
+ {
1110
+ "epoch": 0.45,
1111
+ "learning_rate": 2.8894727459432504e-05,
1112
+ "loss": 0.1464,
1113
+ "step": 1840
1114
+ },
1115
+ {
1116
+ "epoch": 0.45,
1117
+ "learning_rate": 2.870479074755257e-05,
1118
+ "loss": 0.1446,
1119
+ "step": 1850
1120
+ },
1121
+ {
1122
+ "epoch": 0.46,
1123
+ "learning_rate": 2.8514635132520716e-05,
1124
+ "loss": 0.1371,
1125
+ "step": 1860
1126
+ },
1127
+ {
1128
+ "epoch": 0.46,
1129
+ "learning_rate": 2.8324271849968394e-05,
1130
+ "loss": 0.1345,
1131
+ "step": 1870
1132
+ },
1133
+ {
1134
+ "epoch": 0.46,
1135
+ "learning_rate": 2.8133712147797404e-05,
1136
+ "loss": 0.1396,
1137
+ "step": 1880
1138
+ },
1139
+ {
1140
+ "epoch": 0.46,
1141
+ "learning_rate": 2.794296728551533e-05,
1142
+ "loss": 0.1225,
1143
+ "step": 1890
1144
+ },
1145
+ {
1146
+ "epoch": 0.46,
1147
+ "learning_rate": 2.7752048533570186e-05,
1148
+ "loss": 0.1352,
1149
+ "step": 1900
1150
+ },
1151
+ {
1152
+ "epoch": 0.47,
1153
+ "learning_rate": 2.756096717268453e-05,
1154
+ "loss": 0.146,
1155
+ "step": 1910
1156
+ },
1157
+ {
1158
+ "epoch": 0.47,
1159
+ "learning_rate": 2.7369734493188924e-05,
1160
+ "loss": 0.1426,
1161
+ "step": 1920
1162
+ },
1163
+ {
1164
+ "epoch": 0.47,
1165
+ "learning_rate": 2.71783617943548e-05,
1166
+ "loss": 0.1494,
1167
+ "step": 1930
1168
+ },
1169
+ {
1170
+ "epoch": 0.47,
1171
+ "learning_rate": 2.6986860383726865e-05,
1172
+ "loss": 0.1098,
1173
+ "step": 1940
1174
+ },
1175
+ {
1176
+ "epoch": 0.48,
1177
+ "learning_rate": 2.679524157645494e-05,
1178
+ "loss": 0.1436,
1179
+ "step": 1950
1180
+ },
1181
+ {
1182
+ "epoch": 0.48,
1183
+ "learning_rate": 2.660351669462541e-05,
1184
+ "loss": 0.1269,
1185
+ "step": 1960
1186
+ },
1187
+ {
1188
+ "epoch": 0.48,
1189
+ "learning_rate": 2.6411697066592228e-05,
1190
+ "loss": 0.1283,
1191
+ "step": 1970
1192
+ },
1193
+ {
1194
+ "epoch": 0.48,
1195
+ "learning_rate": 2.62197940263076e-05,
1196
+ "loss": 0.1379,
1197
+ "step": 1980
1198
+ },
1199
+ {
1200
+ "epoch": 0.49,
1201
+ "learning_rate": 2.6027818912652218e-05,
1202
+ "loss": 0.135,
1203
+ "step": 1990
1204
+ },
1205
+ {
1206
+ "epoch": 0.49,
1207
+ "learning_rate": 2.5835783068765378e-05,
1208
+ "loss": 0.1365,
1209
+ "step": 2000
1210
+ },
1211
+ {
1212
+ "epoch": 0.49,
1213
+ "learning_rate": 2.564369784137472e-05,
1214
+ "loss": 0.1235,
1215
+ "step": 2010
1216
+ },
1217
+ {
1218
+ "epoch": 0.49,
1219
+ "learning_rate": 2.5451574580125738e-05,
1220
+ "loss": 0.1514,
1221
+ "step": 2020
1222
+ },
1223
+ {
1224
+ "epoch": 0.5,
1225
+ "learning_rate": 2.5259424636911256e-05,
1226
+ "loss": 0.1358,
1227
+ "step": 2030
1228
+ },
1229
+ {
1230
+ "epoch": 0.5,
1231
+ "learning_rate": 2.5067259365200624e-05,
1232
+ "loss": 0.1444,
1233
+ "step": 2040
1234
+ },
1235
+ {
1236
+ "epoch": 0.5,
1237
+ "learning_rate": 2.4875090119368905e-05,
1238
+ "loss": 0.1254,
1239
+ "step": 2050
1240
+ },
1241
+ {
1242
+ "epoch": 0.5,
1243
+ "learning_rate": 2.468292825402597e-05,
1244
+ "loss": 0.1247,
1245
+ "step": 2060
1246
+ },
1247
+ {
1248
+ "epoch": 0.51,
1249
+ "learning_rate": 2.449078512334562e-05,
1250
+ "loss": 0.1293,
1251
+ "step": 2070
1252
+ },
1253
+ {
1254
+ "epoch": 0.51,
1255
+ "learning_rate": 2.429867208039467e-05,
1256
+ "loss": 0.126,
1257
+ "step": 2080
1258
+ },
1259
+ {
1260
+ "epoch": 0.51,
1261
+ "learning_rate": 2.4106600476462173e-05,
1262
+ "loss": 0.1351,
1263
+ "step": 2090
1264
+ },
1265
+ {
1266
+ "epoch": 0.51,
1267
+ "learning_rate": 2.3914581660388683e-05,
1268
+ "loss": 0.1386,
1269
+ "step": 2100
1270
+ },
1271
+ {
1272
+ "epoch": 0.52,
1273
+ "learning_rate": 2.3722626977895706e-05,
1274
+ "loss": 0.1294,
1275
+ "step": 2110
1276
+ },
1277
+ {
1278
+ "epoch": 0.52,
1279
+ "learning_rate": 2.3530747770915328e-05,
1280
+ "loss": 0.1188,
1281
+ "step": 2120
1282
+ },
1283
+ {
1284
+ "epoch": 0.52,
1285
+ "learning_rate": 2.3338955376920034e-05,
1286
+ "loss": 0.1283,
1287
+ "step": 2130
1288
+ },
1289
+ {
1290
+ "epoch": 0.52,
1291
+ "learning_rate": 2.3147261128252858e-05,
1292
+ "loss": 0.1239,
1293
+ "step": 2140
1294
+ },
1295
+ {
1296
+ "epoch": 0.53,
1297
+ "learning_rate": 2.295567635145774e-05,
1298
+ "loss": 0.1258,
1299
+ "step": 2150
1300
+ },
1301
+ {
1302
+ "epoch": 0.53,
1303
+ "learning_rate": 2.276421236661032e-05,
1304
+ "loss": 0.145,
1305
+ "step": 2160
1306
+ },
1307
+ {
1308
+ "epoch": 0.53,
1309
+ "learning_rate": 2.2572880486649086e-05,
1310
+ "loss": 0.1232,
1311
+ "step": 2170
1312
+ },
1313
+ {
1314
+ "epoch": 0.53,
1315
+ "learning_rate": 2.2381692016706862e-05,
1316
+ "loss": 0.14,
1317
+ "step": 2180
1318
+ },
1319
+ {
1320
+ "epoch": 0.54,
1321
+ "learning_rate": 2.219065825344292e-05,
1322
+ "loss": 0.1594,
1323
+ "step": 2190
1324
+ },
1325
+ {
1326
+ "epoch": 0.54,
1327
+ "learning_rate": 2.1999790484375433e-05,
1328
+ "loss": 0.1475,
1329
+ "step": 2200
1330
+ },
1331
+ {
1332
+ "epoch": 0.54,
1333
+ "learning_rate": 2.1809099987214567e-05,
1334
+ "loss": 0.1176,
1335
+ "step": 2210
1336
+ },
1337
+ {
1338
+ "epoch": 0.54,
1339
+ "learning_rate": 2.1618598029196082e-05,
1340
+ "loss": 0.1317,
1341
+ "step": 2220
1342
+ },
1343
+ {
1344
+ "epoch": 0.55,
1345
+ "learning_rate": 2.142829586641566e-05,
1346
+ "loss": 0.1318,
1347
+ "step": 2230
1348
+ },
1349
+ {
1350
+ "epoch": 0.55,
1351
+ "learning_rate": 2.1238204743163764e-05,
1352
+ "loss": 0.1453,
1353
+ "step": 2240
1354
+ },
1355
+ {
1356
+ "epoch": 0.55,
1357
+ "learning_rate": 2.104833589126124e-05,
1358
+ "loss": 0.1297,
1359
+ "step": 2250
1360
+ },
1361
+ {
1362
+ "epoch": 0.55,
1363
+ "learning_rate": 2.085870052939572e-05,
1364
+ "loss": 0.1393,
1365
+ "step": 2260
1366
+ },
1367
+ {
1368
+ "epoch": 0.56,
1369
+ "learning_rate": 2.0669309862458756e-05,
1370
+ "loss": 0.1364,
1371
+ "step": 2270
1372
+ },
1373
+ {
1374
+ "epoch": 0.56,
1375
+ "learning_rate": 2.0480175080883663e-05,
1376
+ "loss": 0.1246,
1377
+ "step": 2280
1378
+ },
1379
+ {
1380
+ "epoch": 0.56,
1381
+ "learning_rate": 2.0291307359984435e-05,
1382
+ "loss": 0.1437,
1383
+ "step": 2290
1384
+ },
1385
+ {
1386
+ "epoch": 0.56,
1387
+ "learning_rate": 2.0102717859295365e-05,
1388
+ "loss": 0.1268,
1389
+ "step": 2300
1390
+ },
1391
+ {
1392
+ "epoch": 0.57,
1393
+ "learning_rate": 1.9914417721911687e-05,
1394
+ "loss": 0.1299,
1395
+ "step": 2310
1396
+ },
1397
+ {
1398
+ "epoch": 0.57,
1399
+ "learning_rate": 1.9726418073831167e-05,
1400
+ "loss": 0.1334,
1401
+ "step": 2320
1402
+ },
1403
+ {
1404
+ "epoch": 0.57,
1405
+ "learning_rate": 1.9538730023296708e-05,
1406
+ "loss": 0.138,
1407
+ "step": 2330
1408
+ },
1409
+ {
1410
+ "epoch": 0.57,
1411
+ "learning_rate": 1.9351364660140004e-05,
1412
+ "loss": 0.1247,
1413
+ "step": 2340
1414
+ },
1415
+ {
1416
+ "epoch": 0.57,
1417
+ "learning_rate": 1.9164333055126265e-05,
1418
+ "loss": 0.1218,
1419
+ "step": 2350
1420
+ },
1421
+ {
1422
+ "epoch": 0.58,
1423
+ "learning_rate": 1.897764625930012e-05,
1424
+ "loss": 0.1391,
1425
+ "step": 2360
1426
+ },
1427
+ {
1428
+ "epoch": 0.58,
1429
+ "learning_rate": 1.879131530333261e-05,
1430
+ "loss": 0.1405,
1431
+ "step": 2370
1432
+ },
1433
+ {
1434
+ "epoch": 0.58,
1435
+ "learning_rate": 1.8605351196869453e-05,
1436
+ "loss": 0.1454,
1437
+ "step": 2380
1438
+ },
1439
+ {
1440
+ "epoch": 0.58,
1441
+ "learning_rate": 1.8419764927880488e-05,
1442
+ "loss": 0.1404,
1443
+ "step": 2390
1444
+ },
1445
+ {
1446
+ "epoch": 0.59,
1447
+ "learning_rate": 1.8234567462010482e-05,
1448
+ "loss": 0.1277,
1449
+ "step": 2400
1450
+ },
1451
+ {
1452
+ "epoch": 0.59,
1453
+ "learning_rate": 1.8049769741931178e-05,
1454
+ "loss": 0.1298,
1455
+ "step": 2410
1456
+ },
1457
+ {
1458
+ "epoch": 0.59,
1459
+ "learning_rate": 1.786538268669472e-05,
1460
+ "loss": 0.1378,
1461
+ "step": 2420
1462
+ },
1463
+ {
1464
+ "epoch": 0.59,
1465
+ "learning_rate": 1.768141719108852e-05,
1466
+ "loss": 0.1336,
1467
+ "step": 2430
1468
+ },
1469
+ {
1470
+ "epoch": 0.6,
1471
+ "learning_rate": 1.749788412499149e-05,
1472
+ "loss": 0.131,
1473
+ "step": 2440
1474
+ },
1475
+ {
1476
+ "epoch": 0.6,
1477
+ "learning_rate": 1.7314794332731786e-05,
1478
+ "loss": 0.1187,
1479
+ "step": 2450
1480
+ },
1481
+ {
1482
+ "epoch": 0.6,
1483
+ "learning_rate": 1.7132158632446085e-05,
1484
+ "loss": 0.1291,
1485
+ "step": 2460
1486
+ },
1487
+ {
1488
+ "epoch": 0.6,
1489
+ "learning_rate": 1.6949987815440332e-05,
1490
+ "loss": 0.1277,
1491
+ "step": 2470
1492
+ },
1493
+ {
1494
+ "epoch": 0.61,
1495
+ "learning_rate": 1.676829264555216e-05,
1496
+ "loss": 0.1419,
1497
+ "step": 2480
1498
+ },
1499
+ {
1500
+ "epoch": 0.61,
1501
+ "learning_rate": 1.658708385851485e-05,
1502
+ "loss": 0.128,
1503
+ "step": 2490
1504
+ },
1505
+ {
1506
+ "epoch": 0.61,
1507
+ "learning_rate": 1.6406372161323035e-05,
1508
+ "loss": 0.1213,
1509
+ "step": 2500
1510
+ },
1511
+ {
1512
+ "epoch": 0.61,
1513
+ "learning_rate": 1.6226168231600058e-05,
1514
+ "loss": 0.1436,
1515
+ "step": 2510
1516
+ },
1517
+ {
1518
+ "epoch": 0.62,
1519
+ "learning_rate": 1.604648271696701e-05,
1520
+ "loss": 0.1297,
1521
+ "step": 2520
1522
+ },
1523
+ {
1524
+ "epoch": 0.62,
1525
+ "learning_rate": 1.5867326234413692e-05,
1526
+ "loss": 0.1481,
1527
+ "step": 2530
1528
+ },
1529
+ {
1530
+ "epoch": 0.62,
1531
+ "learning_rate": 1.5688709369671222e-05,
1532
+ "loss": 0.1237,
1533
+ "step": 2540
1534
+ },
1535
+ {
1536
+ "epoch": 0.62,
1537
+ "learning_rate": 1.5510642676586606e-05,
1538
+ "loss": 0.1342,
1539
+ "step": 2550
1540
+ },
1541
+ {
1542
+ "epoch": 0.63,
1543
+ "learning_rate": 1.5333136676499078e-05,
1544
+ "loss": 0.1265,
1545
+ "step": 2560
1546
+ },
1547
+ {
1548
+ "epoch": 0.63,
1549
+ "learning_rate": 1.5156201857618532e-05,
1550
+ "loss": 0.1294,
1551
+ "step": 2570
1552
+ },
1553
+ {
1554
+ "epoch": 0.63,
1555
+ "learning_rate": 1.4979848674405755e-05,
1556
+ "loss": 0.1218,
1557
+ "step": 2580
1558
+ },
1559
+ {
1560
+ "epoch": 0.63,
1561
+ "learning_rate": 1.480408754695467e-05,
1562
+ "loss": 0.1377,
1563
+ "step": 2590
1564
+ },
1565
+ {
1566
+ "epoch": 0.64,
1567
+ "learning_rate": 1.4628928860376733e-05,
1568
+ "loss": 0.1367,
1569
+ "step": 2600
1570
+ },
1571
+ {
1572
+ "epoch": 0.64,
1573
+ "learning_rate": 1.4454382964187263e-05,
1574
+ "loss": 0.1338,
1575
+ "step": 2610
1576
+ },
1577
+ {
1578
+ "epoch": 0.64,
1579
+ "learning_rate": 1.4280460171693924e-05,
1580
+ "loss": 0.1357,
1581
+ "step": 2620
1582
+ },
1583
+ {
1584
+ "epoch": 0.64,
1585
+ "learning_rate": 1.4107170759387364e-05,
1586
+ "loss": 0.1308,
1587
+ "step": 2630
1588
+ },
1589
+ {
1590
+ "epoch": 0.65,
1591
+ "learning_rate": 1.3934524966334012e-05,
1592
+ "loss": 0.1466,
1593
+ "step": 2640
1594
+ },
1595
+ {
1596
+ "epoch": 0.65,
1597
+ "learning_rate": 1.3762532993571079e-05,
1598
+ "loss": 0.1371,
1599
+ "step": 2650
1600
+ },
1601
+ {
1602
+ "epoch": 0.65,
1603
+ "learning_rate": 1.35912050035038e-05,
1604
+ "loss": 0.1357,
1605
+ "step": 2660
1606
+ },
1607
+ {
1608
+ "epoch": 0.65,
1609
+ "learning_rate": 1.342055111930502e-05,
1610
+ "loss": 0.1208,
1611
+ "step": 2670
1612
+ },
1613
+ {
1614
+ "epoch": 0.66,
1615
+ "learning_rate": 1.325058142431701e-05,
1616
+ "loss": 0.1168,
1617
+ "step": 2680
1618
+ },
1619
+ {
1620
+ "epoch": 0.66,
1621
+ "learning_rate": 1.3081305961455658e-05,
1622
+ "loss": 0.121,
1623
+ "step": 2690
1624
+ },
1625
+ {
1626
+ "epoch": 0.66,
1627
+ "learning_rate": 1.291273473261716e-05,
1628
+ "loss": 0.1215,
1629
+ "step": 2700
1630
+ },
1631
+ {
1632
+ "epoch": 0.66,
1633
+ "learning_rate": 1.2744877698086943e-05,
1634
+ "loss": 0.1302,
1635
+ "step": 2710
1636
+ },
1637
+ {
1638
+ "epoch": 0.67,
1639
+ "learning_rate": 1.2577744775951194e-05,
1640
+ "loss": 0.1381,
1641
+ "step": 2720
1642
+ },
1643
+ {
1644
+ "epoch": 0.67,
1645
+ "learning_rate": 1.2411345841510808e-05,
1646
+ "loss": 0.1398,
1647
+ "step": 2730
1648
+ },
1649
+ {
1650
+ "epoch": 0.67,
1651
+ "learning_rate": 1.2245690726697939e-05,
1652
+ "loss": 0.1279,
1653
+ "step": 2740
1654
+ },
1655
+ {
1656
+ "epoch": 0.67,
1657
+ "learning_rate": 1.2080789219495028e-05,
1658
+ "loss": 0.1419,
1659
+ "step": 2750
1660
+ },
1661
+ {
1662
+ "epoch": 0.68,
1663
+ "learning_rate": 1.1916651063356455e-05,
1664
+ "loss": 0.1466,
1665
+ "step": 2760
1666
+ },
1667
+ {
1668
+ "epoch": 0.68,
1669
+ "learning_rate": 1.1753285956632867e-05,
1670
+ "loss": 0.1237,
1671
+ "step": 2770
1672
+ },
1673
+ {
1674
+ "epoch": 0.68,
1675
+ "learning_rate": 1.1590703551998111e-05,
1676
+ "loss": 0.1248,
1677
+ "step": 2780
1678
+ },
1679
+ {
1680
+ "epoch": 0.68,
1681
+ "learning_rate": 1.1428913455878932e-05,
1682
+ "loss": 0.1256,
1683
+ "step": 2790
1684
+ },
1685
+ {
1686
+ "epoch": 0.69,
1687
+ "learning_rate": 1.1267925227887275e-05,
1688
+ "loss": 0.1272,
1689
+ "step": 2800
1690
+ },
1691
+ {
1692
+ "epoch": 0.69,
1693
+ "learning_rate": 1.1107748380255536e-05,
1694
+ "loss": 0.1325,
1695
+ "step": 2810
1696
+ },
1697
+ {
1698
+ "epoch": 0.69,
1699
+ "learning_rate": 1.0948392377274477e-05,
1700
+ "loss": 0.1099,
1701
+ "step": 2820
1702
+ },
1703
+ {
1704
+ "epoch": 0.69,
1705
+ "learning_rate": 1.0789866634733988e-05,
1706
+ "loss": 0.1418,
1707
+ "step": 2830
1708
+ },
1709
+ {
1710
+ "epoch": 0.69,
1711
+ "learning_rate": 1.0632180519366796e-05,
1712
+ "loss": 0.1202,
1713
+ "step": 2840
1714
+ },
1715
+ {
1716
+ "epoch": 0.7,
1717
+ "learning_rate": 1.0475343348294978e-05,
1718
+ "loss": 0.1304,
1719
+ "step": 2850
1720
+ },
1721
+ {
1722
+ "epoch": 0.7,
1723
+ "learning_rate": 1.0319364388479455e-05,
1724
+ "loss": 0.15,
1725
+ "step": 2860
1726
+ },
1727
+ {
1728
+ "epoch": 0.7,
1729
+ "learning_rate": 1.0164252856172445e-05,
1730
+ "loss": 0.1306,
1731
+ "step": 2870
1732
+ },
1733
+ {
1734
+ "epoch": 0.7,
1735
+ "learning_rate": 1.0010017916372908e-05,
1736
+ "loss": 0.1394,
1737
+ "step": 2880
1738
+ },
1739
+ {
1740
+ "epoch": 0.71,
1741
+ "learning_rate": 9.856668682285003e-06,
1742
+ "loss": 0.1545,
1743
+ "step": 2890
1744
+ },
1745
+ {
1746
+ "epoch": 0.71,
1747
+ "learning_rate": 9.704214214779625e-06,
1748
+ "loss": 0.1261,
1749
+ "step": 2900
1750
+ },
1751
+ {
1752
+ "epoch": 0.71,
1753
+ "learning_rate": 9.552663521859048e-06,
1754
+ "loss": 0.1222,
1755
+ "step": 2910
1756
+ },
1757
+ {
1758
+ "epoch": 0.71,
1759
+ "learning_rate": 9.402025558124658e-06,
1760
+ "loss": 0.1253,
1761
+ "step": 2920
1762
+ },
1763
+ {
1764
+ "epoch": 0.72,
1765
+ "learning_rate": 9.25230922424786e-06,
1766
+ "loss": 0.1418,
1767
+ "step": 2930
1768
+ },
1769
+ {
1770
+ "epoch": 0.72,
1771
+ "learning_rate": 9.103523366444164e-06,
1772
+ "loss": 0.1211,
1773
+ "step": 2940
1774
+ },
1775
+ {
1776
+ "epoch": 0.72,
1777
+ "learning_rate": 8.9556767759505e-06,
1778
+ "loss": 0.1251,
1779
+ "step": 2950
1780
+ },
1781
+ {
1782
+ "epoch": 0.72,
1783
+ "learning_rate": 8.808778188505779e-06,
1784
+ "loss": 0.1331,
1785
+ "step": 2960
1786
+ },
1787
+ {
1788
+ "epoch": 0.73,
1789
+ "learning_rate": 8.6628362838347e-06,
1790
+ "loss": 0.15,
1791
+ "step": 2970
1792
+ },
1793
+ {
1794
+ "epoch": 0.73,
1795
+ "learning_rate": 8.517859685134932e-06,
1796
+ "loss": 0.1152,
1797
+ "step": 2980
1798
+ },
1799
+ {
1800
+ "epoch": 0.73,
1801
+ "learning_rate": 8.373856958567589e-06,
1802
+ "loss": 0.1244,
1803
+ "step": 2990
1804
+ },
1805
+ {
1806
+ "epoch": 0.73,
1807
+ "learning_rate": 8.230836612751069e-06,
1808
+ "loss": 0.1491,
1809
+ "step": 3000
1810
+ },
1811
+ {
1812
+ "epoch": 0.74,
1813
+ "learning_rate": 8.088807098258328e-06,
1814
+ "loss": 0.1387,
1815
+ "step": 3010
1816
+ },
1817
+ {
1818
+ "epoch": 0.74,
1819
+ "learning_rate": 7.94777680711756e-06,
1820
+ "loss": 0.1497,
1821
+ "step": 3020
1822
+ },
1823
+ {
1824
+ "epoch": 0.74,
1825
+ "learning_rate": 7.807754072316323e-06,
1826
+ "loss": 0.1296,
1827
+ "step": 3030
1828
+ },
1829
+ {
1830
+ "epoch": 0.74,
1831
+ "learning_rate": 7.668747167309209e-06,
1832
+ "loss": 0.1278,
1833
+ "step": 3040
1834
+ },
1835
+ {
1836
+ "epoch": 0.75,
1837
+ "learning_rate": 7.530764305528959e-06,
1838
+ "loss": 0.1334,
1839
+ "step": 3050
1840
+ },
1841
+ {
1842
+ "epoch": 0.75,
1843
+ "learning_rate": 7.393813639901195e-06,
1844
+ "loss": 0.1314,
1845
+ "step": 3060
1846
+ },
1847
+ {
1848
+ "epoch": 0.75,
1849
+ "learning_rate": 7.257903262362637e-06,
1850
+ "loss": 0.1251,
1851
+ "step": 3070
1852
+ },
1853
+ {
1854
+ "epoch": 0.75,
1855
+ "learning_rate": 7.123041203383032e-06,
1856
+ "loss": 0.1127,
1857
+ "step": 3080
1858
+ },
1859
+ {
1860
+ "epoch": 0.76,
1861
+ "learning_rate": 6.989235431490676e-06,
1862
+ "loss": 0.1363,
1863
+ "step": 3090
1864
+ },
1865
+ {
1866
+ "epoch": 0.76,
1867
+ "learning_rate": 6.856493852801485e-06,
1868
+ "loss": 0.1184,
1869
+ "step": 3100
1870
+ },
1871
+ {
1872
+ "epoch": 0.76,
1873
+ "learning_rate": 6.724824310551961e-06,
1874
+ "loss": 0.1475,
1875
+ "step": 3110
1876
+ },
1877
+ {
1878
+ "epoch": 0.76,
1879
+ "learning_rate": 6.594234584635694e-06,
1880
+ "loss": 0.1336,
1881
+ "step": 3120
1882
+ },
1883
+ {
1884
+ "epoch": 0.77,
1885
+ "learning_rate": 6.464732391143713e-06,
1886
+ "loss": 0.127,
1887
+ "step": 3130
1888
+ },
1889
+ {
1890
+ "epoch": 0.77,
1891
+ "learning_rate": 6.336325381908529e-06,
1892
+ "loss": 0.1315,
1893
+ "step": 3140
1894
+ },
1895
+ {
1896
+ "epoch": 0.77,
1897
+ "learning_rate": 6.209021144052054e-06,
1898
+ "loss": 0.1257,
1899
+ "step": 3150
1900
+ },
1901
+ {
1902
+ "epoch": 0.77,
1903
+ "learning_rate": 6.082827199537295e-06,
1904
+ "loss": 0.1356,
1905
+ "step": 3160
1906
+ },
1907
+ {
1908
+ "epoch": 0.78,
1909
+ "learning_rate": 5.9577510047238825e-06,
1910
+ "loss": 0.1306,
1911
+ "step": 3170
1912
+ },
1913
+ {
1914
+ "epoch": 0.78,
1915
+ "learning_rate": 5.833799949927537e-06,
1916
+ "loss": 0.1163,
1917
+ "step": 3180
1918
+ },
1919
+ {
1920
+ "epoch": 0.78,
1921
+ "learning_rate": 5.710981358983384e-06,
1922
+ "loss": 0.1333,
1923
+ "step": 3190
1924
+ },
1925
+ {
1926
+ "epoch": 0.78,
1927
+ "learning_rate": 5.589302488813186e-06,
1928
+ "loss": 0.139,
1929
+ "step": 3200
1930
+ },
1931
+ {
1932
+ "epoch": 0.79,
1933
+ "learning_rate": 5.468770528996614e-06,
1934
+ "loss": 0.1261,
1935
+ "step": 3210
1936
+ },
1937
+ {
1938
+ "epoch": 0.79,
1939
+ "learning_rate": 5.349392601346398e-06,
1940
+ "loss": 0.1274,
1941
+ "step": 3220
1942
+ },
1943
+ {
1944
+ "epoch": 0.79,
1945
+ "learning_rate": 5.231175759487536e-06,
1946
+ "loss": 0.1136,
1947
+ "step": 3230
1948
+ },
1949
+ {
1950
+ "epoch": 0.79,
1951
+ "learning_rate": 5.114126988440523e-06,
1952
+ "loss": 0.1289,
1953
+ "step": 3240
1954
+ },
1955
+ {
1956
+ "epoch": 0.8,
1957
+ "learning_rate": 4.998253204208625e-06,
1958
+ "loss": 0.1462,
1959
+ "step": 3250
1960
+ },
1961
+ {
1962
+ "epoch": 0.8,
1963
+ "learning_rate": 4.8835612533692455e-06,
1964
+ "loss": 0.1209,
1965
+ "step": 3260
1966
+ },
1967
+ {
1968
+ "epoch": 0.8,
1969
+ "learning_rate": 4.770057912669362e-06,
1970
+ "loss": 0.1266,
1971
+ "step": 3270
1972
+ },
1973
+ {
1974
+ "epoch": 0.8,
1975
+ "learning_rate": 4.65774988862514e-06,
1976
+ "loss": 0.1487,
1977
+ "step": 3280
1978
+ },
1979
+ {
1980
+ "epoch": 0.8,
1981
+ "learning_rate": 4.546643817125662e-06,
1982
+ "loss": 0.1101,
1983
+ "step": 3290
1984
+ },
1985
+ {
1986
+ "epoch": 0.81,
1987
+ "learning_rate": 4.436746263040819e-06,
1988
+ "loss": 0.1183,
1989
+ "step": 3300
1990
+ },
1991
+ {
1992
+ "epoch": 0.81,
1993
+ "learning_rate": 4.328063719833433e-06,
1994
+ "loss": 0.1133,
1995
+ "step": 3310
1996
+ },
1997
+ {
1998
+ "epoch": 0.81,
1999
+ "learning_rate": 4.220602609175575e-06,
2000
+ "loss": 0.1159,
2001
+ "step": 3320
2002
+ },
2003
+ {
2004
+ "epoch": 0.81,
2005
+ "learning_rate": 4.114369280569136e-06,
2006
+ "loss": 0.1288,
2007
+ "step": 3330
2008
+ },
2009
+ {
2010
+ "epoch": 0.82,
2011
+ "learning_rate": 4.009370010970634e-06,
2012
+ "loss": 0.1222,
2013
+ "step": 3340
2014
+ },
2015
+ {
2016
+ "epoch": 0.82,
2017
+ "learning_rate": 3.90561100442036e-06,
2018
+ "loss": 0.1208,
2019
+ "step": 3350
2020
+ },
2021
+ {
2022
+ "epoch": 0.82,
2023
+ "learning_rate": 3.8030983916757966e-06,
2024
+ "loss": 0.1096,
2025
+ "step": 3360
2026
+ },
2027
+ {
2028
+ "epoch": 0.82,
2029
+ "learning_rate": 3.7018382298493635e-06,
2030
+ "loss": 0.1157,
2031
+ "step": 3370
2032
+ },
2033
+ {
2034
+ "epoch": 0.83,
2035
+ "learning_rate": 3.6018365020505124e-06,
2036
+ "loss": 0.1264,
2037
+ "step": 3380
2038
+ },
2039
+ {
2040
+ "epoch": 0.83,
2041
+ "learning_rate": 3.503099117032249e-06,
2042
+ "loss": 0.1315,
2043
+ "step": 3390
2044
+ },
2045
+ {
2046
+ "epoch": 0.83,
2047
+ "learning_rate": 3.405631908841966e-06,
2048
+ "loss": 0.1317,
2049
+ "step": 3400
2050
+ },
2051
+ {
2052
+ "epoch": 0.83,
2053
+ "learning_rate": 3.3094406364767267e-06,
2054
+ "loss": 0.1382,
2055
+ "step": 3410
2056
+ },
2057
+ {
2058
+ "epoch": 0.84,
2059
+ "learning_rate": 3.214530983543021e-06,
2060
+ "loss": 0.1366,
2061
+ "step": 3420
2062
+ },
2063
+ {
2064
+ "epoch": 0.84,
2065
+ "learning_rate": 3.120908557920918e-06,
2066
+ "loss": 0.1308,
2067
+ "step": 3430
2068
+ },
2069
+ {
2070
+ "epoch": 0.84,
2071
+ "learning_rate": 3.028578891432701e-06,
2072
+ "loss": 0.1448,
2073
+ "step": 3440
2074
+ },
2075
+ {
2076
+ "epoch": 0.84,
2077
+ "learning_rate": 2.9375474395160528e-06,
2078
+ "loss": 0.1292,
2079
+ "step": 3450
2080
+ },
2081
+ {
2082
+ "epoch": 0.85,
2083
+ "learning_rate": 2.8478195809016674e-06,
2084
+ "loss": 0.13,
2085
+ "step": 3460
2086
+ },
2087
+ {
2088
+ "epoch": 0.85,
2089
+ "learning_rate": 2.7594006172954906e-06,
2090
+ "loss": 0.1333,
2091
+ "step": 3470
2092
+ },
2093
+ {
2094
+ "epoch": 0.85,
2095
+ "learning_rate": 2.6722957730653973e-06,
2096
+ "loss": 0.1527,
2097
+ "step": 3480
2098
+ },
2099
+ {
2100
+ "epoch": 0.85,
2101
+ "learning_rate": 2.5865101949325567e-06,
2102
+ "loss": 0.1154,
2103
+ "step": 3490
2104
+ },
2105
+ {
2106
+ "epoch": 0.86,
2107
+ "learning_rate": 2.50204895166731e-06,
2108
+ "loss": 0.1247,
2109
+ "step": 3500
2110
+ },
2111
+ {
2112
+ "epoch": 0.86,
2113
+ "learning_rate": 2.418917033789661e-06,
2114
+ "loss": 0.1211,
2115
+ "step": 3510
2116
+ },
2117
+ {
2118
+ "epoch": 0.86,
2119
+ "learning_rate": 2.3371193532744305e-06,
2120
+ "loss": 0.1208,
2121
+ "step": 3520
2122
+ },
2123
+ {
2124
+ "epoch": 0.86,
2125
+ "learning_rate": 2.2566607432610047e-06,
2126
+ "loss": 0.1208,
2127
+ "step": 3530
2128
+ },
2129
+ {
2130
+ "epoch": 0.87,
2131
+ "learning_rate": 2.177545957767771e-06,
2132
+ "loss": 0.1251,
2133
+ "step": 3540
2134
+ },
2135
+ {
2136
+ "epoch": 0.87,
2137
+ "learning_rate": 2.0997796714112144e-06,
2138
+ "loss": 0.1175,
2139
+ "step": 3550
2140
+ },
2141
+ {
2142
+ "epoch": 0.87,
2143
+ "learning_rate": 2.023366479129718e-06,
2144
+ "loss": 0.1207,
2145
+ "step": 3560
2146
+ },
2147
+ {
2148
+ "epoch": 0.87,
2149
+ "learning_rate": 1.9483108959120587e-06,
2150
+ "loss": 0.1198,
2151
+ "step": 3570
2152
+ },
2153
+ {
2154
+ "epoch": 0.88,
2155
+ "learning_rate": 1.874617356530628e-06,
2156
+ "loss": 0.113,
2157
+ "step": 3580
2158
+ },
2159
+ {
2160
+ "epoch": 0.88,
2161
+ "learning_rate": 1.8022902152794036e-06,
2162
+ "loss": 0.1261,
2163
+ "step": 3590
2164
+ },
2165
+ {
2166
+ "epoch": 0.88,
2167
+ "learning_rate": 1.7313337457166707e-06,
2168
+ "loss": 0.1297,
2169
+ "step": 3600
2170
+ },
2171
+ {
2172
+ "epoch": 0.88,
2173
+ "learning_rate": 1.661752140412512e-06,
2174
+ "loss": 0.1197,
2175
+ "step": 3610
2176
+ },
2177
+ {
2178
+ "epoch": 0.89,
2179
+ "learning_rate": 1.5935495107010728e-06,
2180
+ "loss": 0.133,
2181
+ "step": 3620
2182
+ },
2183
+ {
2184
+ "epoch": 0.89,
2185
+ "learning_rate": 1.5267298864376527e-06,
2186
+ "loss": 0.1245,
2187
+ "step": 3630
2188
+ },
2189
+ {
2190
+ "epoch": 0.89,
2191
+ "learning_rate": 1.461297215760582e-06,
2192
+ "loss": 0.1337,
2193
+ "step": 3640
2194
+ },
2195
+ {
2196
+ "epoch": 0.89,
2197
+ "learning_rate": 1.3972553648579468e-06,
2198
+ "loss": 0.1437,
2199
+ "step": 3650
2200
+ },
2201
+ {
2202
+ "epoch": 0.9,
2203
+ "learning_rate": 1.3346081177391472e-06,
2204
+ "loss": 0.1467,
2205
+ "step": 3660
2206
+ },
2207
+ {
2208
+ "epoch": 0.9,
2209
+ "learning_rate": 1.2733591760113184e-06,
2210
+ "loss": 0.1298,
2211
+ "step": 3670
2212
+ },
2213
+ {
2214
+ "epoch": 0.9,
2215
+ "learning_rate": 1.2135121586606014e-06,
2216
+ "loss": 0.1239,
2217
+ "step": 3680
2218
+ },
2219
+ {
2220
+ "epoch": 0.9,
2221
+ "learning_rate": 1.1550706018383345e-06,
2222
+ "loss": 0.1252,
2223
+ "step": 3690
2224
+ },
2225
+ {
2226
+ "epoch": 0.91,
2227
+ "learning_rate": 1.09803795865209e-06,
2228
+ "loss": 0.1475,
2229
+ "step": 3700
2230
+ },
2231
+ {
2232
+ "epoch": 0.91,
2233
+ "learning_rate": 1.0424175989616563e-06,
2234
+ "loss": 0.1365,
2235
+ "step": 3710
2236
+ },
2237
+ {
2238
+ "epoch": 0.91,
2239
+ "learning_rate": 9.882128091799114e-07,
2240
+ "loss": 0.127,
2241
+ "step": 3720
2242
+ },
2243
+ {
2244
+ "epoch": 0.91,
2245
+ "learning_rate": 9.354267920786614e-07,
2246
+ "loss": 0.1281,
2247
+ "step": 3730
2248
+ },
2249
+ {
2250
+ "epoch": 0.92,
2251
+ "learning_rate": 8.840626665993807e-07,
2252
+ "loss": 0.1219,
2253
+ "step": 3740
2254
+ },
2255
+ {
2256
+ "epoch": 0.92,
2257
+ "learning_rate": 8.341234676689319e-07,
2258
+ "loss": 0.1227,
2259
+ "step": 3750
2260
+ },
2261
+ {
2262
+ "epoch": 0.92,
2263
+ "learning_rate": 7.856121460202398e-07,
2264
+ "loss": 0.1568,
2265
+ "step": 3760
2266
+ },
2267
+ {
2268
+ "epoch": 0.92,
2269
+ "learning_rate": 7.385315680179583e-07,
2270
+ "loss": 0.1202,
2271
+ "step": 3770
2272
+ },
2273
+ {
2274
+ "epoch": 0.92,
2275
+ "learning_rate": 6.928845154890757e-07,
2276
+ "loss": 0.1374,
2277
+ "step": 3780
2278
+ },
2279
+ {
2280
+ "epoch": 0.93,
2281
+ "learning_rate": 6.486736855585762e-07,
2282
+ "loss": 0.1235,
2283
+ "step": 3790
2284
+ },
2285
+ {
2286
+ "epoch": 0.93,
2287
+ "learning_rate": 6.059016904900628e-07,
2288
+ "loss": 0.1252,
2289
+ "step": 3800
2290
+ },
2291
+ {
2292
+ "epoch": 0.93,
2293
+ "learning_rate": 5.645710575314073e-07,
2294
+ "loss": 0.1261,
2295
+ "step": 3810
2296
+ },
2297
+ {
2298
+ "epoch": 0.93,
2299
+ "learning_rate": 5.246842287654235e-07,
2300
+ "loss": 0.1338,
2301
+ "step": 3820
2302
+ },
2303
+ {
2304
+ "epoch": 0.94,
2305
+ "learning_rate": 4.862435609655824e-07,
2306
+ "loss": 0.1345,
2307
+ "step": 3830
2308
+ },
2309
+ {
2310
+ "epoch": 0.94,
2311
+ "learning_rate": 4.492513254567482e-07,
2312
+ "loss": 0.1164,
2313
+ "step": 3840
2314
+ },
2315
+ {
2316
+ "epoch": 0.94,
2317
+ "learning_rate": 4.137097079809776e-07,
2318
+ "loss": 0.1212,
2319
+ "step": 3850
2320
+ },
2321
+ {
2322
+ "epoch": 0.94,
2323
+ "learning_rate": 3.796208085683733e-07,
2324
+ "loss": 0.1289,
2325
+ "step": 3860
2326
+ },
2327
+ {
2328
+ "epoch": 0.95,
2329
+ "learning_rate": 3.4698664141299686e-07,
2330
+ "loss": 0.1412,
2331
+ "step": 3870
2332
+ },
2333
+ {
2334
+ "epoch": 0.95,
2335
+ "learning_rate": 3.1580913475386107e-07,
2336
+ "loss": 0.1293,
2337
+ "step": 3880
2338
+ },
2339
+ {
2340
+ "epoch": 0.95,
2341
+ "learning_rate": 2.8609013076099365e-07,
2342
+ "loss": 0.1191,
2343
+ "step": 3890
2344
+ },
2345
+ {
2346
+ "epoch": 0.95,
2347
+ "learning_rate": 2.5783138542659345e-07,
2348
+ "loss": 0.1247,
2349
+ "step": 3900
2350
+ },
2351
+ {
2352
+ "epoch": 0.96,
2353
+ "learning_rate": 2.3103456846126913e-07,
2354
+ "loss": 0.1175,
2355
+ "step": 3910
2356
+ },
2357
+ {
2358
+ "epoch": 0.96,
2359
+ "learning_rate": 2.057012631953903e-07,
2360
+ "loss": 0.1317,
2361
+ "step": 3920
2362
+ },
2363
+ {
2364
+ "epoch": 0.96,
2365
+ "learning_rate": 1.8183296648552616e-07,
2366
+ "loss": 0.1421,
2367
+ "step": 3930
2368
+ },
2369
+ {
2370
+ "epoch": 0.96,
2371
+ "learning_rate": 1.5943108862600809e-07,
2372
+ "loss": 0.1365,
2373
+ "step": 3940
2374
+ },
2375
+ {
2376
+ "epoch": 0.97,
2377
+ "learning_rate": 1.384969532655933e-07,
2378
+ "loss": 0.1261,
2379
+ "step": 3950
2380
+ },
2381
+ {
2382
+ "epoch": 0.97,
2383
+ "learning_rate": 1.190317973292554e-07,
2384
+ "loss": 0.1054,
2385
+ "step": 3960
2386
+ },
2387
+ {
2388
+ "epoch": 0.97,
2389
+ "learning_rate": 1.0103677094510932e-07,
2390
+ "loss": 0.1251,
2391
+ "step": 3970
2392
+ },
2393
+ {
2394
+ "epoch": 0.97,
2395
+ "learning_rate": 8.451293737644638e-08,
2396
+ "loss": 0.1446,
2397
+ "step": 3980
2398
+ },
2399
+ {
2400
+ "epoch": 0.98,
2401
+ "learning_rate": 6.946127295890392e-08,
2402
+ "loss": 0.1271,
2403
+ "step": 3990
2404
+ },
2405
+ {
2406
+ "epoch": 0.98,
2407
+ "learning_rate": 5.5882667042789214e-08,
2408
+ "loss": 0.1372,
2409
+ "step": 4000
2410
+ },
2411
+ {
2412
+ "epoch": 0.98,
2413
+ "learning_rate": 4.377792194052432e-08,
2414
+ "loss": 0.1258,
2415
+ "step": 4010
2416
+ },
2417
+ {
2418
+ "epoch": 0.98,
2419
+ "learning_rate": 3.314775287923677e-08,
2420
+ "loss": 0.1268,
2421
+ "step": 4020
2422
+ },
2423
+ {
2424
+ "epoch": 0.99,
2425
+ "learning_rate": 2.399278795851001e-08,
2426
+ "loss": 0.1372,
2427
+ "step": 4030
2428
+ },
2429
+ {
2430
+ "epoch": 0.99,
2431
+ "learning_rate": 1.631356811325757e-08,
2432
+ "loss": 0.1274,
2433
+ "step": 4040
2434
+ },
2435
+ {
2436
+ "epoch": 0.99,
2437
+ "learning_rate": 1.0110547081776389e-08,
2438
+ "loss": 0.1361,
2439
+ "step": 4050
2440
+ },
2441
+ {
2442
+ "epoch": 0.99,
2443
+ "learning_rate": 5.384091378923817e-09,
2444
+ "loss": 0.1202,
2445
+ "step": 4060
2446
+ },
2447
+ {
2448
+ "epoch": 1.0,
2449
+ "learning_rate": 2.1344802744627246e-09,
2450
+ "loss": 0.1315,
2451
+ "step": 4070
2452
+ },
2453
+ {
2454
+ "epoch": 1.0,
2455
+ "learning_rate": 3.6190577657468914e-10,
2456
+ "loss": 0.1349,
2457
+ "step": 4080
2458
+ },
2459
+ {
2460
+ "epoch": 1.0,
2461
+ "step": 4087,
2462
+ "total_flos": 2.507754096179712e+18,
2463
+ "train_loss": 0.1560091979059696,
2464
+ "train_runtime": 21225.5051,
2465
+ "train_samples_per_second": 3.081,
2466
+ "train_steps_per_second": 0.193
2467
+ }
2468
+ ],
2469
+ "logging_steps": 10,
2470
+ "max_steps": 4087,
2471
+ "num_train_epochs": 1,
2472
+ "save_steps": 1000,
2473
+ "total_flos": 2.507754096179712e+18,
2474
+ "trial_name": null,
2475
+ "trial_params": null
2476
+ }
Baichuan-13B-Chat-lora-Computing/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fe319e7156520da103d89f67f7026a8a887d2ca0d81076b6f0b7a8365ab7db8
3
+ size 3513
Baichuan-13B-Chat-lora-Computing/training_loss.png ADDED