End of training
Browse files- all_results.json +7 -0
- best_results.json +5 -0
- train_results.json +7 -0
- trainer_state.json +1147 -0
all_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 60.0,
|
3 |
+
"train_loss": 0.3189461204740736,
|
4 |
+
"train_runtime": 8512.0872,
|
5 |
+
"train_samples_per_second": 17.552,
|
6 |
+
"train_steps_per_second": 2.199
|
7 |
+
}
|
best_results.json
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_epoch": 58,
|
3 |
+
"best_eval_accuracy": 0.776173285198556,
|
4 |
+
"epoch": 60.0
|
5 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 60.0,
|
3 |
+
"train_loss": 0.3189461204740736,
|
4 |
+
"train_runtime": 8512.0872,
|
5 |
+
"train_samples_per_second": 17.552,
|
6 |
+
"train_steps_per_second": 2.199
|
7 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 60.0,
|
5 |
+
"global_step": 18720,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 1.0,
|
12 |
+
"eval_accuracy": 0.51985559566787,
|
13 |
+
"eval_loss": 0.5319512486457825,
|
14 |
+
"eval_runtime": 9.1619,
|
15 |
+
"eval_samples_per_second": 30.234,
|
16 |
+
"eval_steps_per_second": 3.82,
|
17 |
+
"step": 312
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"best_epoch": 0,
|
21 |
+
"best_eval_accuracy": 0.51985559566787,
|
22 |
+
"epoch": 1.0,
|
23 |
+
"step": 312
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 1.6,
|
27 |
+
"learning_rate": 0.004866452991452991,
|
28 |
+
"loss": 0.6084,
|
29 |
+
"step": 500
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"epoch": 2.0,
|
33 |
+
"eval_accuracy": 0.5306859205776173,
|
34 |
+
"eval_loss": 0.505980372428894,
|
35 |
+
"eval_runtime": 8.9739,
|
36 |
+
"eval_samples_per_second": 30.867,
|
37 |
+
"eval_steps_per_second": 3.9,
|
38 |
+
"step": 624
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"best_epoch": 1,
|
42 |
+
"best_eval_accuracy": 0.5306859205776173,
|
43 |
+
"epoch": 2.0,
|
44 |
+
"step": 624
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 3.0,
|
48 |
+
"eval_accuracy": 0.4729241877256318,
|
49 |
+
"eval_loss": 0.476485937833786,
|
50 |
+
"eval_runtime": 9.1085,
|
51 |
+
"eval_samples_per_second": 30.411,
|
52 |
+
"eval_steps_per_second": 3.843,
|
53 |
+
"step": 936
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"best_epoch": 1,
|
57 |
+
"best_eval_accuracy": 0.5306859205776173,
|
58 |
+
"epoch": 3.0,
|
59 |
+
"step": 936
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 3.21,
|
63 |
+
"learning_rate": 0.004732905982905983,
|
64 |
+
"loss": 0.4786,
|
65 |
+
"step": 1000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 4.0,
|
69 |
+
"eval_accuracy": 0.4729241877256318,
|
70 |
+
"eval_loss": 0.38619130849838257,
|
71 |
+
"eval_runtime": 9.0544,
|
72 |
+
"eval_samples_per_second": 30.593,
|
73 |
+
"eval_steps_per_second": 3.866,
|
74 |
+
"step": 1248
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"best_epoch": 1,
|
78 |
+
"best_eval_accuracy": 0.5306859205776173,
|
79 |
+
"epoch": 4.0,
|
80 |
+
"step": 1248
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 4.81,
|
84 |
+
"learning_rate": 0.004599358974358974,
|
85 |
+
"loss": 0.5253,
|
86 |
+
"step": 1500
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 5.0,
|
90 |
+
"eval_accuracy": 0.5342960288808665,
|
91 |
+
"eval_loss": 0.5090705156326294,
|
92 |
+
"eval_runtime": 9.0148,
|
93 |
+
"eval_samples_per_second": 30.727,
|
94 |
+
"eval_steps_per_second": 3.882,
|
95 |
+
"step": 1560
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"best_epoch": 4,
|
99 |
+
"best_eval_accuracy": 0.5342960288808665,
|
100 |
+
"epoch": 5.0,
|
101 |
+
"step": 1560
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 6.0,
|
105 |
+
"eval_accuracy": 0.4981949458483754,
|
106 |
+
"eval_loss": 0.376774400472641,
|
107 |
+
"eval_runtime": 9.0141,
|
108 |
+
"eval_samples_per_second": 30.73,
|
109 |
+
"eval_steps_per_second": 3.883,
|
110 |
+
"step": 1872
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"best_epoch": 4,
|
114 |
+
"best_eval_accuracy": 0.5342960288808665,
|
115 |
+
"epoch": 6.0,
|
116 |
+
"step": 1872
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 6.41,
|
120 |
+
"learning_rate": 0.004465811965811966,
|
121 |
+
"loss": 0.5144,
|
122 |
+
"step": 2000
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 7.0,
|
126 |
+
"eval_accuracy": 0.5270758122743683,
|
127 |
+
"eval_loss": 0.44060468673706055,
|
128 |
+
"eval_runtime": 9.0181,
|
129 |
+
"eval_samples_per_second": 30.716,
|
130 |
+
"eval_steps_per_second": 3.881,
|
131 |
+
"step": 2184
|
132 |
+
},
|
133 |
+
{
|
134 |
+
"best_epoch": 4,
|
135 |
+
"best_eval_accuracy": 0.5342960288808665,
|
136 |
+
"epoch": 7.0,
|
137 |
+
"step": 2184
|
138 |
+
},
|
139 |
+
{
|
140 |
+
"epoch": 8.0,
|
141 |
+
"eval_accuracy": 0.631768953068592,
|
142 |
+
"eval_loss": 0.3461342751979828,
|
143 |
+
"eval_runtime": 8.9803,
|
144 |
+
"eval_samples_per_second": 30.845,
|
145 |
+
"eval_steps_per_second": 3.897,
|
146 |
+
"step": 2496
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"best_epoch": 7,
|
150 |
+
"best_eval_accuracy": 0.631768953068592,
|
151 |
+
"epoch": 8.0,
|
152 |
+
"step": 2496
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 8.01,
|
156 |
+
"learning_rate": 0.004332264957264957,
|
157 |
+
"loss": 0.4407,
|
158 |
+
"step": 2500
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 9.0,
|
162 |
+
"eval_accuracy": 0.6534296028880866,
|
163 |
+
"eval_loss": 0.34802812337875366,
|
164 |
+
"eval_runtime": 8.976,
|
165 |
+
"eval_samples_per_second": 30.86,
|
166 |
+
"eval_steps_per_second": 3.899,
|
167 |
+
"step": 2808
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"best_epoch": 8,
|
171 |
+
"best_eval_accuracy": 0.6534296028880866,
|
172 |
+
"epoch": 9.0,
|
173 |
+
"step": 2808
|
174 |
+
},
|
175 |
+
{
|
176 |
+
"epoch": 9.62,
|
177 |
+
"learning_rate": 0.004198717948717949,
|
178 |
+
"loss": 0.4002,
|
179 |
+
"step": 3000
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 10.0,
|
183 |
+
"eval_accuracy": 0.6642599277978339,
|
184 |
+
"eval_loss": 0.362854540348053,
|
185 |
+
"eval_runtime": 9.0085,
|
186 |
+
"eval_samples_per_second": 30.749,
|
187 |
+
"eval_steps_per_second": 3.885,
|
188 |
+
"step": 3120
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"best_epoch": 9,
|
192 |
+
"best_eval_accuracy": 0.6642599277978339,
|
193 |
+
"epoch": 10.0,
|
194 |
+
"step": 3120
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 11.0,
|
198 |
+
"eval_accuracy": 0.555956678700361,
|
199 |
+
"eval_loss": 0.39493829011917114,
|
200 |
+
"eval_runtime": 9.036,
|
201 |
+
"eval_samples_per_second": 30.655,
|
202 |
+
"eval_steps_per_second": 3.873,
|
203 |
+
"step": 3432
|
204 |
+
},
|
205 |
+
{
|
206 |
+
"best_epoch": 9,
|
207 |
+
"best_eval_accuracy": 0.6642599277978339,
|
208 |
+
"epoch": 11.0,
|
209 |
+
"step": 3432
|
210 |
+
},
|
211 |
+
{
|
212 |
+
"epoch": 11.22,
|
213 |
+
"learning_rate": 0.00406517094017094,
|
214 |
+
"loss": 0.3576,
|
215 |
+
"step": 3500
|
216 |
+
},
|
217 |
+
{
|
218 |
+
"epoch": 12.0,
|
219 |
+
"eval_accuracy": 0.7075812274368231,
|
220 |
+
"eval_loss": 0.336588591337204,
|
221 |
+
"eval_runtime": 9.0067,
|
222 |
+
"eval_samples_per_second": 30.755,
|
223 |
+
"eval_steps_per_second": 3.886,
|
224 |
+
"step": 3744
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"best_epoch": 11,
|
228 |
+
"best_eval_accuracy": 0.7075812274368231,
|
229 |
+
"epoch": 12.0,
|
230 |
+
"step": 3744
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 12.82,
|
234 |
+
"learning_rate": 0.003931623931623931,
|
235 |
+
"loss": 0.346,
|
236 |
+
"step": 4000
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 13.0,
|
240 |
+
"eval_accuracy": 0.703971119133574,
|
241 |
+
"eval_loss": 0.3301502466201782,
|
242 |
+
"eval_runtime": 9.0221,
|
243 |
+
"eval_samples_per_second": 30.702,
|
244 |
+
"eval_steps_per_second": 3.879,
|
245 |
+
"step": 4056
|
246 |
+
},
|
247 |
+
{
|
248 |
+
"best_epoch": 11,
|
249 |
+
"best_eval_accuracy": 0.7075812274368231,
|
250 |
+
"epoch": 13.0,
|
251 |
+
"step": 4056
|
252 |
+
},
|
253 |
+
{
|
254 |
+
"epoch": 14.0,
|
255 |
+
"eval_accuracy": 0.7184115523465704,
|
256 |
+
"eval_loss": 0.3293074369430542,
|
257 |
+
"eval_runtime": 9.0321,
|
258 |
+
"eval_samples_per_second": 30.668,
|
259 |
+
"eval_steps_per_second": 3.875,
|
260 |
+
"step": 4368
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"best_epoch": 13,
|
264 |
+
"best_eval_accuracy": 0.7184115523465704,
|
265 |
+
"epoch": 14.0,
|
266 |
+
"step": 4368
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 14.42,
|
270 |
+
"learning_rate": 0.003798076923076923,
|
271 |
+
"loss": 0.337,
|
272 |
+
"step": 4500
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 15.0,
|
276 |
+
"eval_accuracy": 0.7292418772563177,
|
277 |
+
"eval_loss": 0.33005112409591675,
|
278 |
+
"eval_runtime": 9.0187,
|
279 |
+
"eval_samples_per_second": 30.714,
|
280 |
+
"eval_steps_per_second": 3.881,
|
281 |
+
"step": 4680
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"best_epoch": 14,
|
285 |
+
"best_eval_accuracy": 0.7292418772563177,
|
286 |
+
"epoch": 15.0,
|
287 |
+
"step": 4680
|
288 |
+
},
|
289 |
+
{
|
290 |
+
"epoch": 16.0,
|
291 |
+
"eval_accuracy": 0.7328519855595668,
|
292 |
+
"eval_loss": 0.33984842896461487,
|
293 |
+
"eval_runtime": 9.0474,
|
294 |
+
"eval_samples_per_second": 30.616,
|
295 |
+
"eval_steps_per_second": 3.869,
|
296 |
+
"step": 4992
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"best_epoch": 15,
|
300 |
+
"best_eval_accuracy": 0.7328519855595668,
|
301 |
+
"epoch": 16.0,
|
302 |
+
"step": 4992
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 16.03,
|
306 |
+
"learning_rate": 0.003664529914529914,
|
307 |
+
"loss": 0.3323,
|
308 |
+
"step": 5000
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 17.0,
|
312 |
+
"eval_accuracy": 0.7256317689530686,
|
313 |
+
"eval_loss": 0.35551536083221436,
|
314 |
+
"eval_runtime": 9.0226,
|
315 |
+
"eval_samples_per_second": 30.701,
|
316 |
+
"eval_steps_per_second": 3.879,
|
317 |
+
"step": 5304
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"best_epoch": 15,
|
321 |
+
"best_eval_accuracy": 0.7328519855595668,
|
322 |
+
"epoch": 17.0,
|
323 |
+
"step": 5304
|
324 |
+
},
|
325 |
+
{
|
326 |
+
"epoch": 17.63,
|
327 |
+
"learning_rate": 0.003530982905982906,
|
328 |
+
"loss": 0.3245,
|
329 |
+
"step": 5500
|
330 |
+
},
|
331 |
+
{
|
332 |
+
"epoch": 18.0,
|
333 |
+
"eval_accuracy": 0.703971119133574,
|
334 |
+
"eval_loss": 0.32570627331733704,
|
335 |
+
"eval_runtime": 9.0171,
|
336 |
+
"eval_samples_per_second": 30.719,
|
337 |
+
"eval_steps_per_second": 3.882,
|
338 |
+
"step": 5616
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"best_epoch": 15,
|
342 |
+
"best_eval_accuracy": 0.7328519855595668,
|
343 |
+
"epoch": 18.0,
|
344 |
+
"step": 5616
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 19.0,
|
348 |
+
"eval_accuracy": 0.7292418772563177,
|
349 |
+
"eval_loss": 0.3257431089878082,
|
350 |
+
"eval_runtime": 9.0276,
|
351 |
+
"eval_samples_per_second": 30.684,
|
352 |
+
"eval_steps_per_second": 3.877,
|
353 |
+
"step": 5928
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"best_epoch": 15,
|
357 |
+
"best_eval_accuracy": 0.7328519855595668,
|
358 |
+
"epoch": 19.0,
|
359 |
+
"step": 5928
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 19.23,
|
363 |
+
"learning_rate": 0.0033974358974358976,
|
364 |
+
"loss": 0.3243,
|
365 |
+
"step": 6000
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"epoch": 20.0,
|
369 |
+
"eval_accuracy": 0.7220216606498195,
|
370 |
+
"eval_loss": 0.3507075309753418,
|
371 |
+
"eval_runtime": 8.9982,
|
372 |
+
"eval_samples_per_second": 30.784,
|
373 |
+
"eval_steps_per_second": 3.89,
|
374 |
+
"step": 6240
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"best_epoch": 15,
|
378 |
+
"best_eval_accuracy": 0.7328519855595668,
|
379 |
+
"epoch": 20.0,
|
380 |
+
"step": 6240
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 20.83,
|
384 |
+
"learning_rate": 0.003263888888888889,
|
385 |
+
"loss": 0.3144,
|
386 |
+
"step": 6500
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 21.0,
|
390 |
+
"eval_accuracy": 0.7184115523465704,
|
391 |
+
"eval_loss": 0.40468713641166687,
|
392 |
+
"eval_runtime": 9.0484,
|
393 |
+
"eval_samples_per_second": 30.613,
|
394 |
+
"eval_steps_per_second": 3.868,
|
395 |
+
"step": 6552
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"best_epoch": 15,
|
399 |
+
"best_eval_accuracy": 0.7328519855595668,
|
400 |
+
"epoch": 21.0,
|
401 |
+
"step": 6552
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 22.0,
|
405 |
+
"eval_accuracy": 0.7220216606498195,
|
406 |
+
"eval_loss": 0.36198413372039795,
|
407 |
+
"eval_runtime": 9.0445,
|
408 |
+
"eval_samples_per_second": 30.626,
|
409 |
+
"eval_steps_per_second": 3.87,
|
410 |
+
"step": 6864
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"best_epoch": 15,
|
414 |
+
"best_eval_accuracy": 0.7328519855595668,
|
415 |
+
"epoch": 22.0,
|
416 |
+
"step": 6864
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 22.44,
|
420 |
+
"learning_rate": 0.0031303418803418806,
|
421 |
+
"loss": 0.3135,
|
422 |
+
"step": 7000
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 23.0,
|
426 |
+
"eval_accuracy": 0.7148014440433214,
|
427 |
+
"eval_loss": 0.3739970028400421,
|
428 |
+
"eval_runtime": 9.0931,
|
429 |
+
"eval_samples_per_second": 30.463,
|
430 |
+
"eval_steps_per_second": 3.849,
|
431 |
+
"step": 7176
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"best_epoch": 15,
|
435 |
+
"best_eval_accuracy": 0.7328519855595668,
|
436 |
+
"epoch": 23.0,
|
437 |
+
"step": 7176
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 24.0,
|
441 |
+
"eval_accuracy": 0.7436823104693141,
|
442 |
+
"eval_loss": 0.33150357007980347,
|
443 |
+
"eval_runtime": 9.0913,
|
444 |
+
"eval_samples_per_second": 30.469,
|
445 |
+
"eval_steps_per_second": 3.85,
|
446 |
+
"step": 7488
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"best_epoch": 23,
|
450 |
+
"best_eval_accuracy": 0.7436823104693141,
|
451 |
+
"epoch": 24.0,
|
452 |
+
"step": 7488
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 24.04,
|
456 |
+
"learning_rate": 0.0029967948717948716,
|
457 |
+
"loss": 0.3063,
|
458 |
+
"step": 7500
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 25.0,
|
462 |
+
"eval_accuracy": 0.7436823104693141,
|
463 |
+
"eval_loss": 0.3291415572166443,
|
464 |
+
"eval_runtime": 9.0712,
|
465 |
+
"eval_samples_per_second": 30.536,
|
466 |
+
"eval_steps_per_second": 3.858,
|
467 |
+
"step": 7800
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"best_epoch": 23,
|
471 |
+
"best_eval_accuracy": 0.7436823104693141,
|
472 |
+
"epoch": 25.0,
|
473 |
+
"step": 7800
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 25.64,
|
477 |
+
"learning_rate": 0.002863247863247863,
|
478 |
+
"loss": 0.2986,
|
479 |
+
"step": 8000
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 26.0,
|
483 |
+
"eval_accuracy": 0.7292418772563177,
|
484 |
+
"eval_loss": 0.36258506774902344,
|
485 |
+
"eval_runtime": 9.1485,
|
486 |
+
"eval_samples_per_second": 30.278,
|
487 |
+
"eval_steps_per_second": 3.826,
|
488 |
+
"step": 8112
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"best_epoch": 23,
|
492 |
+
"best_eval_accuracy": 0.7436823104693141,
|
493 |
+
"epoch": 26.0,
|
494 |
+
"step": 8112
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 27.0,
|
498 |
+
"eval_accuracy": 0.740072202166065,
|
499 |
+
"eval_loss": 0.3281010687351227,
|
500 |
+
"eval_runtime": 9.0555,
|
501 |
+
"eval_samples_per_second": 30.589,
|
502 |
+
"eval_steps_per_second": 3.865,
|
503 |
+
"step": 8424
|
504 |
+
},
|
505 |
+
{
|
506 |
+
"best_epoch": 23,
|
507 |
+
"best_eval_accuracy": 0.7436823104693141,
|
508 |
+
"epoch": 27.0,
|
509 |
+
"step": 8424
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 27.24,
|
513 |
+
"learning_rate": 0.0027297008547008546,
|
514 |
+
"loss": 0.2956,
|
515 |
+
"step": 8500
|
516 |
+
},
|
517 |
+
{
|
518 |
+
"epoch": 28.0,
|
519 |
+
"eval_accuracy": 0.740072202166065,
|
520 |
+
"eval_loss": 0.33756041526794434,
|
521 |
+
"eval_runtime": 9.1719,
|
522 |
+
"eval_samples_per_second": 30.201,
|
523 |
+
"eval_steps_per_second": 3.816,
|
524 |
+
"step": 8736
|
525 |
+
},
|
526 |
+
{
|
527 |
+
"best_epoch": 23,
|
528 |
+
"best_eval_accuracy": 0.7436823104693141,
|
529 |
+
"epoch": 28.0,
|
530 |
+
"step": 8736
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 28.85,
|
534 |
+
"learning_rate": 0.0025961538461538466,
|
535 |
+
"loss": 0.2927,
|
536 |
+
"step": 9000
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 29.0,
|
540 |
+
"eval_accuracy": 0.7545126353790613,
|
541 |
+
"eval_loss": 0.3309880197048187,
|
542 |
+
"eval_runtime": 9.0985,
|
543 |
+
"eval_samples_per_second": 30.445,
|
544 |
+
"eval_steps_per_second": 3.847,
|
545 |
+
"step": 9048
|
546 |
+
},
|
547 |
+
{
|
548 |
+
"best_epoch": 28,
|
549 |
+
"best_eval_accuracy": 0.7545126353790613,
|
550 |
+
"epoch": 29.0,
|
551 |
+
"step": 9048
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 30.0,
|
555 |
+
"eval_accuracy": 0.7436823104693141,
|
556 |
+
"eval_loss": 0.34706664085388184,
|
557 |
+
"eval_runtime": 8.8071,
|
558 |
+
"eval_samples_per_second": 31.452,
|
559 |
+
"eval_steps_per_second": 3.974,
|
560 |
+
"step": 9360
|
561 |
+
},
|
562 |
+
{
|
563 |
+
"best_epoch": 28,
|
564 |
+
"best_eval_accuracy": 0.7545126353790613,
|
565 |
+
"epoch": 30.0,
|
566 |
+
"step": 9360
|
567 |
+
},
|
568 |
+
{
|
569 |
+
"epoch": 30.45,
|
570 |
+
"learning_rate": 0.0024626068376068376,
|
571 |
+
"loss": 0.2853,
|
572 |
+
"step": 9500
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 31.0,
|
576 |
+
"eval_accuracy": 0.7581227436823105,
|
577 |
+
"eval_loss": 0.3204677700996399,
|
578 |
+
"eval_runtime": 9.1308,
|
579 |
+
"eval_samples_per_second": 30.337,
|
580 |
+
"eval_steps_per_second": 3.833,
|
581 |
+
"step": 9672
|
582 |
+
},
|
583 |
+
{
|
584 |
+
"best_epoch": 30,
|
585 |
+
"best_eval_accuracy": 0.7581227436823105,
|
586 |
+
"epoch": 31.0,
|
587 |
+
"step": 9672
|
588 |
+
},
|
589 |
+
{
|
590 |
+
"epoch": 32.0,
|
591 |
+
"eval_accuracy": 0.7509025270758123,
|
592 |
+
"eval_loss": 0.32714536786079407,
|
593 |
+
"eval_runtime": 9.0779,
|
594 |
+
"eval_samples_per_second": 30.514,
|
595 |
+
"eval_steps_per_second": 3.855,
|
596 |
+
"step": 9984
|
597 |
+
},
|
598 |
+
{
|
599 |
+
"best_epoch": 30,
|
600 |
+
"best_eval_accuracy": 0.7581227436823105,
|
601 |
+
"epoch": 32.0,
|
602 |
+
"step": 9984
|
603 |
+
},
|
604 |
+
{
|
605 |
+
"epoch": 32.05,
|
606 |
+
"learning_rate": 0.002329059829059829,
|
607 |
+
"loss": 0.2861,
|
608 |
+
"step": 10000
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 33.0,
|
612 |
+
"eval_accuracy": 0.7509025270758123,
|
613 |
+
"eval_loss": 0.3423185646533966,
|
614 |
+
"eval_runtime": 9.1078,
|
615 |
+
"eval_samples_per_second": 30.414,
|
616 |
+
"eval_steps_per_second": 3.843,
|
617 |
+
"step": 10296
|
618 |
+
},
|
619 |
+
{
|
620 |
+
"best_epoch": 30,
|
621 |
+
"best_eval_accuracy": 0.7581227436823105,
|
622 |
+
"epoch": 33.0,
|
623 |
+
"step": 10296
|
624 |
+
},
|
625 |
+
{
|
626 |
+
"epoch": 33.65,
|
627 |
+
"learning_rate": 0.0021955128205128206,
|
628 |
+
"loss": 0.2782,
|
629 |
+
"step": 10500
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 34.0,
|
633 |
+
"eval_accuracy": 0.7472924187725631,
|
634 |
+
"eval_loss": 0.3327659070491791,
|
635 |
+
"eval_runtime": 9.1866,
|
636 |
+
"eval_samples_per_second": 30.153,
|
637 |
+
"eval_steps_per_second": 3.81,
|
638 |
+
"step": 10608
|
639 |
+
},
|
640 |
+
{
|
641 |
+
"best_epoch": 30,
|
642 |
+
"best_eval_accuracy": 0.7581227436823105,
|
643 |
+
"epoch": 34.0,
|
644 |
+
"step": 10608
|
645 |
+
},
|
646 |
+
{
|
647 |
+
"epoch": 35.0,
|
648 |
+
"eval_accuracy": 0.7617328519855595,
|
649 |
+
"eval_loss": 0.32888853549957275,
|
650 |
+
"eval_runtime": 9.2412,
|
651 |
+
"eval_samples_per_second": 29.974,
|
652 |
+
"eval_steps_per_second": 3.787,
|
653 |
+
"step": 10920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"best_epoch": 34,
|
657 |
+
"best_eval_accuracy": 0.7617328519855595,
|
658 |
+
"epoch": 35.0,
|
659 |
+
"step": 10920
|
660 |
+
},
|
661 |
+
{
|
662 |
+
"epoch": 35.26,
|
663 |
+
"learning_rate": 0.002061965811965812,
|
664 |
+
"loss": 0.2756,
|
665 |
+
"step": 11000
|
666 |
+
},
|
667 |
+
{
|
668 |
+
"epoch": 36.0,
|
669 |
+
"eval_accuracy": 0.7581227436823105,
|
670 |
+
"eval_loss": 0.330919086933136,
|
671 |
+
"eval_runtime": 9.2586,
|
672 |
+
"eval_samples_per_second": 29.918,
|
673 |
+
"eval_steps_per_second": 3.78,
|
674 |
+
"step": 11232
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"best_epoch": 34,
|
678 |
+
"best_eval_accuracy": 0.7617328519855595,
|
679 |
+
"epoch": 36.0,
|
680 |
+
"step": 11232
|
681 |
+
},
|
682 |
+
{
|
683 |
+
"epoch": 36.86,
|
684 |
+
"learning_rate": 0.0019284188034188036,
|
685 |
+
"loss": 0.2758,
|
686 |
+
"step": 11500
|
687 |
+
},
|
688 |
+
{
|
689 |
+
"epoch": 37.0,
|
690 |
+
"eval_accuracy": 0.7364620938628159,
|
691 |
+
"eval_loss": 0.37405356764793396,
|
692 |
+
"eval_runtime": 9.2,
|
693 |
+
"eval_samples_per_second": 30.109,
|
694 |
+
"eval_steps_per_second": 3.804,
|
695 |
+
"step": 11544
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"best_epoch": 34,
|
699 |
+
"best_eval_accuracy": 0.7617328519855595,
|
700 |
+
"epoch": 37.0,
|
701 |
+
"step": 11544
|
702 |
+
},
|
703 |
+
{
|
704 |
+
"epoch": 38.0,
|
705 |
+
"eval_accuracy": 0.7472924187725631,
|
706 |
+
"eval_loss": 0.33260834217071533,
|
707 |
+
"eval_runtime": 4.3573,
|
708 |
+
"eval_samples_per_second": 63.571,
|
709 |
+
"eval_steps_per_second": 8.032,
|
710 |
+
"step": 11856
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"best_epoch": 34,
|
714 |
+
"best_eval_accuracy": 0.7617328519855595,
|
715 |
+
"epoch": 38.0,
|
716 |
+
"step": 11856
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 38.46,
|
720 |
+
"learning_rate": 0.0017948717948717949,
|
721 |
+
"loss": 0.2714,
|
722 |
+
"step": 12000
|
723 |
+
},
|
724 |
+
{
|
725 |
+
"epoch": 39.0,
|
726 |
+
"eval_accuracy": 0.7184115523465704,
|
727 |
+
"eval_loss": 0.36113840341567993,
|
728 |
+
"eval_runtime": 10.6137,
|
729 |
+
"eval_samples_per_second": 26.098,
|
730 |
+
"eval_steps_per_second": 3.298,
|
731 |
+
"step": 12168
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"best_epoch": 34,
|
735 |
+
"best_eval_accuracy": 0.7617328519855595,
|
736 |
+
"epoch": 39.0,
|
737 |
+
"step": 12168
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 40.0,
|
741 |
+
"eval_accuracy": 0.7472924187725631,
|
742 |
+
"eval_loss": 0.33517616987228394,
|
743 |
+
"eval_runtime": 10.5938,
|
744 |
+
"eval_samples_per_second": 26.147,
|
745 |
+
"eval_steps_per_second": 3.304,
|
746 |
+
"step": 12480
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"best_epoch": 34,
|
750 |
+
"best_eval_accuracy": 0.7617328519855595,
|
751 |
+
"epoch": 40.0,
|
752 |
+
"step": 12480
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 40.06,
|
756 |
+
"learning_rate": 0.0016613247863247864,
|
757 |
+
"loss": 0.2687,
|
758 |
+
"step": 12500
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 41.0,
|
762 |
+
"eval_accuracy": 0.7436823104693141,
|
763 |
+
"eval_loss": 0.34048405289649963,
|
764 |
+
"eval_runtime": 10.5168,
|
765 |
+
"eval_samples_per_second": 26.339,
|
766 |
+
"eval_steps_per_second": 3.328,
|
767 |
+
"step": 12792
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"best_epoch": 34,
|
771 |
+
"best_eval_accuracy": 0.7617328519855595,
|
772 |
+
"epoch": 41.0,
|
773 |
+
"step": 12792
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 41.67,
|
777 |
+
"learning_rate": 0.0015277777777777779,
|
778 |
+
"loss": 0.2685,
|
779 |
+
"step": 13000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 42.0,
|
783 |
+
"eval_accuracy": 0.7364620938628159,
|
784 |
+
"eval_loss": 0.34080803394317627,
|
785 |
+
"eval_runtime": 10.8178,
|
786 |
+
"eval_samples_per_second": 25.606,
|
787 |
+
"eval_steps_per_second": 3.235,
|
788 |
+
"step": 13104
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"best_epoch": 34,
|
792 |
+
"best_eval_accuracy": 0.7617328519855595,
|
793 |
+
"epoch": 42.0,
|
794 |
+
"step": 13104
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 43.0,
|
798 |
+
"eval_accuracy": 0.7472924187725631,
|
799 |
+
"eval_loss": 0.34136372804641724,
|
800 |
+
"eval_runtime": 10.56,
|
801 |
+
"eval_samples_per_second": 26.231,
|
802 |
+
"eval_steps_per_second": 3.314,
|
803 |
+
"step": 13416
|
804 |
+
},
|
805 |
+
{
|
806 |
+
"best_epoch": 34,
|
807 |
+
"best_eval_accuracy": 0.7617328519855595,
|
808 |
+
"epoch": 43.0,
|
809 |
+
"step": 13416
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 43.27,
|
813 |
+
"learning_rate": 0.0013942307692307694,
|
814 |
+
"loss": 0.2649,
|
815 |
+
"step": 13500
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 44.0,
|
819 |
+
"eval_accuracy": 0.7545126353790613,
|
820 |
+
"eval_loss": 0.3369286060333252,
|
821 |
+
"eval_runtime": 10.5795,
|
822 |
+
"eval_samples_per_second": 26.183,
|
823 |
+
"eval_steps_per_second": 3.308,
|
824 |
+
"step": 13728
|
825 |
+
},
|
826 |
+
{
|
827 |
+
"best_epoch": 34,
|
828 |
+
"best_eval_accuracy": 0.7617328519855595,
|
829 |
+
"epoch": 44.0,
|
830 |
+
"step": 13728
|
831 |
+
},
|
832 |
+
{
|
833 |
+
"epoch": 44.87,
|
834 |
+
"learning_rate": 0.0012606837606837606,
|
835 |
+
"loss": 0.2615,
|
836 |
+
"step": 14000
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 45.0,
|
840 |
+
"eval_accuracy": 0.7545126353790613,
|
841 |
+
"eval_loss": 0.33710670471191406,
|
842 |
+
"eval_runtime": 10.8841,
|
843 |
+
"eval_samples_per_second": 25.45,
|
844 |
+
"eval_steps_per_second": 3.216,
|
845 |
+
"step": 14040
|
846 |
+
},
|
847 |
+
{
|
848 |
+
"best_epoch": 34,
|
849 |
+
"best_eval_accuracy": 0.7617328519855595,
|
850 |
+
"epoch": 45.0,
|
851 |
+
"step": 14040
|
852 |
+
},
|
853 |
+
{
|
854 |
+
"epoch": 46.0,
|
855 |
+
"eval_accuracy": 0.7509025270758123,
|
856 |
+
"eval_loss": 0.34283646941185,
|
857 |
+
"eval_runtime": 10.5807,
|
858 |
+
"eval_samples_per_second": 26.18,
|
859 |
+
"eval_steps_per_second": 3.308,
|
860 |
+
"step": 14352
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"best_epoch": 34,
|
864 |
+
"best_eval_accuracy": 0.7617328519855595,
|
865 |
+
"epoch": 46.0,
|
866 |
+
"step": 14352
|
867 |
+
},
|
868 |
+
{
|
869 |
+
"epoch": 46.47,
|
870 |
+
"learning_rate": 0.0011271367521367521,
|
871 |
+
"loss": 0.2602,
|
872 |
+
"step": 14500
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 47.0,
|
876 |
+
"eval_accuracy": 0.7545126353790613,
|
877 |
+
"eval_loss": 0.32864174246788025,
|
878 |
+
"eval_runtime": 10.9225,
|
879 |
+
"eval_samples_per_second": 25.361,
|
880 |
+
"eval_steps_per_second": 3.204,
|
881 |
+
"step": 14664
|
882 |
+
},
|
883 |
+
{
|
884 |
+
"best_epoch": 34,
|
885 |
+
"best_eval_accuracy": 0.7617328519855595,
|
886 |
+
"epoch": 47.0,
|
887 |
+
"step": 14664
|
888 |
+
},
|
889 |
+
{
|
890 |
+
"epoch": 48.0,
|
891 |
+
"eval_accuracy": 0.7581227436823105,
|
892 |
+
"eval_loss": 0.33157122135162354,
|
893 |
+
"eval_runtime": 10.8742,
|
894 |
+
"eval_samples_per_second": 25.473,
|
895 |
+
"eval_steps_per_second": 3.219,
|
896 |
+
"step": 14976
|
897 |
+
},
|
898 |
+
{
|
899 |
+
"best_epoch": 34,
|
900 |
+
"best_eval_accuracy": 0.7617328519855595,
|
901 |
+
"epoch": 48.0,
|
902 |
+
"step": 14976
|
903 |
+
},
|
904 |
+
{
|
905 |
+
"epoch": 48.08,
|
906 |
+
"learning_rate": 0.0009935897435897436,
|
907 |
+
"loss": 0.2595,
|
908 |
+
"step": 15000
|
909 |
+
},
|
910 |
+
{
|
911 |
+
"epoch": 49.0,
|
912 |
+
"eval_accuracy": 0.7545126353790613,
|
913 |
+
"eval_loss": 0.3400510847568512,
|
914 |
+
"eval_runtime": 9.388,
|
915 |
+
"eval_samples_per_second": 29.506,
|
916 |
+
"eval_steps_per_second": 3.728,
|
917 |
+
"step": 15288
|
918 |
+
},
|
919 |
+
{
|
920 |
+
"best_epoch": 34,
|
921 |
+
"best_eval_accuracy": 0.7617328519855595,
|
922 |
+
"epoch": 49.0,
|
923 |
+
"step": 15288
|
924 |
+
},
|
925 |
+
{
|
926 |
+
"epoch": 49.68,
|
927 |
+
"learning_rate": 0.0008600427350427351,
|
928 |
+
"loss": 0.2551,
|
929 |
+
"step": 15500
|
930 |
+
},
|
931 |
+
{
|
932 |
+
"epoch": 50.0,
|
933 |
+
"eval_accuracy": 0.7653429602888087,
|
934 |
+
"eval_loss": 0.3362467288970947,
|
935 |
+
"eval_runtime": 9.2771,
|
936 |
+
"eval_samples_per_second": 29.859,
|
937 |
+
"eval_steps_per_second": 3.773,
|
938 |
+
"step": 15600
|
939 |
+
},
|
940 |
+
{
|
941 |
+
"best_epoch": 49,
|
942 |
+
"best_eval_accuracy": 0.7653429602888087,
|
943 |
+
"epoch": 50.0,
|
944 |
+
"step": 15600
|
945 |
+
},
|
946 |
+
{
|
947 |
+
"epoch": 51.0,
|
948 |
+
"eval_accuracy": 0.7653429602888087,
|
949 |
+
"eval_loss": 0.34337136149406433,
|
950 |
+
"eval_runtime": 9.1926,
|
951 |
+
"eval_samples_per_second": 30.133,
|
952 |
+
"eval_steps_per_second": 3.807,
|
953 |
+
"step": 15912
|
954 |
+
},
|
955 |
+
{
|
956 |
+
"best_epoch": 49,
|
957 |
+
"best_eval_accuracy": 0.7653429602888087,
|
958 |
+
"epoch": 51.0,
|
959 |
+
"step": 15912
|
960 |
+
},
|
961 |
+
{
|
962 |
+
"epoch": 51.28,
|
963 |
+
"learning_rate": 0.0007264957264957266,
|
964 |
+
"loss": 0.2574,
|
965 |
+
"step": 16000
|
966 |
+
},
|
967 |
+
{
|
968 |
+
"epoch": 52.0,
|
969 |
+
"eval_accuracy": 0.7725631768953068,
|
970 |
+
"eval_loss": 0.33018356561660767,
|
971 |
+
"eval_runtime": 9.1431,
|
972 |
+
"eval_samples_per_second": 30.296,
|
973 |
+
"eval_steps_per_second": 3.828,
|
974 |
+
"step": 16224
|
975 |
+
},
|
976 |
+
{
|
977 |
+
"best_epoch": 51,
|
978 |
+
"best_eval_accuracy": 0.7725631768953068,
|
979 |
+
"epoch": 52.0,
|
980 |
+
"step": 16224
|
981 |
+
},
|
982 |
+
{
|
983 |
+
"epoch": 52.88,
|
984 |
+
"learning_rate": 0.000592948717948718,
|
985 |
+
"loss": 0.2515,
|
986 |
+
"step": 16500
|
987 |
+
},
|
988 |
+
{
|
989 |
+
"epoch": 53.0,
|
990 |
+
"eval_accuracy": 0.7472924187725631,
|
991 |
+
"eval_loss": 0.3463674485683441,
|
992 |
+
"eval_runtime": 9.139,
|
993 |
+
"eval_samples_per_second": 30.31,
|
994 |
+
"eval_steps_per_second": 3.83,
|
995 |
+
"step": 16536
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"best_epoch": 51,
|
999 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1000 |
+
"epoch": 53.0,
|
1001 |
+
"step": 16536
|
1002 |
+
},
|
1003 |
+
{
|
1004 |
+
"epoch": 54.0,
|
1005 |
+
"eval_accuracy": 0.7689530685920578,
|
1006 |
+
"eval_loss": 0.3337368667125702,
|
1007 |
+
"eval_runtime": 9.3533,
|
1008 |
+
"eval_samples_per_second": 29.615,
|
1009 |
+
"eval_steps_per_second": 3.742,
|
1010 |
+
"step": 16848
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"best_epoch": 51,
|
1014 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1015 |
+
"epoch": 54.0,
|
1016 |
+
"step": 16848
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 54.49,
|
1020 |
+
"learning_rate": 0.00045940170940170943,
|
1021 |
+
"loss": 0.252,
|
1022 |
+
"step": 17000
|
1023 |
+
},
|
1024 |
+
{
|
1025 |
+
"epoch": 55.0,
|
1026 |
+
"eval_accuracy": 0.7689530685920578,
|
1027 |
+
"eval_loss": 0.33643844723701477,
|
1028 |
+
"eval_runtime": 9.1243,
|
1029 |
+
"eval_samples_per_second": 30.359,
|
1030 |
+
"eval_steps_per_second": 3.836,
|
1031 |
+
"step": 17160
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"best_epoch": 51,
|
1035 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1036 |
+
"epoch": 55.0,
|
1037 |
+
"step": 17160
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 56.0,
|
1041 |
+
"eval_accuracy": 0.7509025270758123,
|
1042 |
+
"eval_loss": 0.3417775630950928,
|
1043 |
+
"eval_runtime": 8.8741,
|
1044 |
+
"eval_samples_per_second": 31.214,
|
1045 |
+
"eval_steps_per_second": 3.944,
|
1046 |
+
"step": 17472
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"best_epoch": 51,
|
1050 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1051 |
+
"epoch": 56.0,
|
1052 |
+
"step": 17472
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 56.09,
|
1056 |
+
"learning_rate": 0.00032585470085470087,
|
1057 |
+
"loss": 0.2497,
|
1058 |
+
"step": 17500
|
1059 |
+
},
|
1060 |
+
{
|
1061 |
+
"epoch": 57.0,
|
1062 |
+
"eval_accuracy": 0.7581227436823105,
|
1063 |
+
"eval_loss": 0.340709924697876,
|
1064 |
+
"eval_runtime": 8.9435,
|
1065 |
+
"eval_samples_per_second": 30.972,
|
1066 |
+
"eval_steps_per_second": 3.913,
|
1067 |
+
"step": 17784
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"best_epoch": 51,
|
1071 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1072 |
+
"epoch": 57.0,
|
1073 |
+
"step": 17784
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 57.69,
|
1077 |
+
"learning_rate": 0.00019230769230769233,
|
1078 |
+
"loss": 0.2503,
|
1079 |
+
"step": 18000
|
1080 |
+
},
|
1081 |
+
{
|
1082 |
+
"epoch": 58.0,
|
1083 |
+
"eval_accuracy": 0.7545126353790613,
|
1084 |
+
"eval_loss": 0.3418755829334259,
|
1085 |
+
"eval_runtime": 8.9319,
|
1086 |
+
"eval_samples_per_second": 31.012,
|
1087 |
+
"eval_steps_per_second": 3.919,
|
1088 |
+
"step": 18096
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"best_epoch": 51,
|
1092 |
+
"best_eval_accuracy": 0.7725631768953068,
|
1093 |
+
"epoch": 58.0,
|
1094 |
+
"step": 18096
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 59.0,
|
1098 |
+
"eval_accuracy": 0.776173285198556,
|
1099 |
+
"eval_loss": 0.33760103583335876,
|
1100 |
+
"eval_runtime": 9.2259,
|
1101 |
+
"eval_samples_per_second": 30.024,
|
1102 |
+
"eval_steps_per_second": 3.794,
|
1103 |
+
"step": 18408
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"best_epoch": 58,
|
1107 |
+
"best_eval_accuracy": 0.776173285198556,
|
1108 |
+
"epoch": 59.0,
|
1109 |
+
"step": 18408
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 59.29,
|
1113 |
+
"learning_rate": 5.876068376068376e-05,
|
1114 |
+
"loss": 0.2504,
|
1115 |
+
"step": 18500
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 60.0,
|
1119 |
+
"eval_accuracy": 0.7725631768953068,
|
1120 |
+
"eval_loss": 0.3379180431365967,
|
1121 |
+
"eval_runtime": 9.2108,
|
1122 |
+
"eval_samples_per_second": 30.073,
|
1123 |
+
"eval_steps_per_second": 3.8,
|
1124 |
+
"step": 18720
|
1125 |
+
},
|
1126 |
+
{
|
1127 |
+
"best_epoch": 58,
|
1128 |
+
"best_eval_accuracy": 0.776173285198556,
|
1129 |
+
"epoch": 60.0,
|
1130 |
+
"step": 18720
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 60.0,
|
1134 |
+
"step": 18720,
|
1135 |
+
"total_flos": 6.96152728406016e+16,
|
1136 |
+
"train_loss": 0.3189461204740736,
|
1137 |
+
"train_runtime": 8512.0872,
|
1138 |
+
"train_samples_per_second": 17.552,
|
1139 |
+
"train_steps_per_second": 2.199
|
1140 |
+
}
|
1141 |
+
],
|
1142 |
+
"max_steps": 18720,
|
1143 |
+
"num_train_epochs": 60,
|
1144 |
+
"total_flos": 6.96152728406016e+16,
|
1145 |
+
"trial_name": null,
|
1146 |
+
"trial_params": null
|
1147 |
+
}
|