End of training
Browse files- all_results.json +12 -0
- eval_results.json +8 -0
- train_results.json +7 -0
- trainer_state.json +1524 -0
all_results.json
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 4.98,
|
3 |
+
"eval_accuracy": 0.989844278943805,
|
4 |
+
"eval_loss": 0.029886629432439804,
|
5 |
+
"eval_runtime": 6.2367,
|
6 |
+
"eval_samples_per_second": 473.648,
|
7 |
+
"eval_steps_per_second": 14.912,
|
8 |
+
"train_loss": 0.12655677452755434,
|
9 |
+
"train_runtime": 660.3476,
|
10 |
+
"train_samples_per_second": 201.303,
|
11 |
+
"train_steps_per_second": 1.567
|
12 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 4.98,
|
3 |
+
"eval_accuracy": 0.989844278943805,
|
4 |
+
"eval_loss": 0.029886629432439804,
|
5 |
+
"eval_runtime": 6.2367,
|
6 |
+
"eval_samples_per_second": 473.648,
|
7 |
+
"eval_steps_per_second": 14.912
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 4.98,
|
3 |
+
"train_loss": 0.12655677452755434,
|
4 |
+
"train_runtime": 660.3476,
|
5 |
+
"train_samples_per_second": 201.303,
|
6 |
+
"train_steps_per_second": 1.567
|
7 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1524 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.989844278943805,
|
3 |
+
"best_model_checkpoint": "salon_image_classifier_v1_convnext/checkpoint-1035",
|
4 |
+
"epoch": 4.981949458483754,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1035,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.02,
|
13 |
+
"grad_norm": 6.277402877807617,
|
14 |
+
"learning_rate": 2.403846153846154e-06,
|
15 |
+
"loss": 1.4624,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.05,
|
20 |
+
"grad_norm": 5.703274726867676,
|
21 |
+
"learning_rate": 4.807692307692308e-06,
|
22 |
+
"loss": 1.3971,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.07,
|
27 |
+
"grad_norm": 7.412254810333252,
|
28 |
+
"learning_rate": 7.211538461538461e-06,
|
29 |
+
"loss": 1.3073,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.1,
|
34 |
+
"grad_norm": 5.874199867248535,
|
35 |
+
"learning_rate": 9.615384615384616e-06,
|
36 |
+
"loss": 1.1877,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.12,
|
41 |
+
"grad_norm": 4.8683342933654785,
|
42 |
+
"learning_rate": 1.2019230769230771e-05,
|
43 |
+
"loss": 1.0259,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.14,
|
48 |
+
"grad_norm": 4.822791576385498,
|
49 |
+
"learning_rate": 1.4423076923076923e-05,
|
50 |
+
"loss": 0.8557,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.17,
|
55 |
+
"grad_norm": 4.416042804718018,
|
56 |
+
"learning_rate": 1.682692307692308e-05,
|
57 |
+
"loss": 0.7254,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.19,
|
62 |
+
"grad_norm": 3.968479633331299,
|
63 |
+
"learning_rate": 1.923076923076923e-05,
|
64 |
+
"loss": 0.5629,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.22,
|
69 |
+
"grad_norm": 3.736450433731079,
|
70 |
+
"learning_rate": 2.1634615384615387e-05,
|
71 |
+
"loss": 0.4541,
|
72 |
+
"step": 45
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.24,
|
76 |
+
"grad_norm": 3.005037307739258,
|
77 |
+
"learning_rate": 2.4038461538461542e-05,
|
78 |
+
"loss": 0.3057,
|
79 |
+
"step": 50
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.26,
|
83 |
+
"grad_norm": 4.574312686920166,
|
84 |
+
"learning_rate": 2.6442307692307694e-05,
|
85 |
+
"loss": 0.2917,
|
86 |
+
"step": 55
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.29,
|
90 |
+
"grad_norm": 5.142052173614502,
|
91 |
+
"learning_rate": 2.8846153846153845e-05,
|
92 |
+
"loss": 0.2297,
|
93 |
+
"step": 60
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.31,
|
97 |
+
"grad_norm": 3.3775789737701416,
|
98 |
+
"learning_rate": 3.125e-05,
|
99 |
+
"loss": 0.2004,
|
100 |
+
"step": 65
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.34,
|
104 |
+
"grad_norm": 3.128685712814331,
|
105 |
+
"learning_rate": 3.365384615384616e-05,
|
106 |
+
"loss": 0.1851,
|
107 |
+
"step": 70
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.36,
|
111 |
+
"grad_norm": 5.491213798522949,
|
112 |
+
"learning_rate": 3.605769230769231e-05,
|
113 |
+
"loss": 0.1956,
|
114 |
+
"step": 75
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.39,
|
118 |
+
"grad_norm": 2.380528688430786,
|
119 |
+
"learning_rate": 3.846153846153846e-05,
|
120 |
+
"loss": 0.161,
|
121 |
+
"step": 80
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.41,
|
125 |
+
"grad_norm": 5.325184345245361,
|
126 |
+
"learning_rate": 4.0865384615384615e-05,
|
127 |
+
"loss": 0.1748,
|
128 |
+
"step": 85
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.43,
|
132 |
+
"grad_norm": 3.9706196784973145,
|
133 |
+
"learning_rate": 4.326923076923077e-05,
|
134 |
+
"loss": 0.1487,
|
135 |
+
"step": 90
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.46,
|
139 |
+
"grad_norm": 4.022750377655029,
|
140 |
+
"learning_rate": 4.5673076923076925e-05,
|
141 |
+
"loss": 0.1339,
|
142 |
+
"step": 95
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.48,
|
146 |
+
"grad_norm": 7.611146450042725,
|
147 |
+
"learning_rate": 4.8076923076923084e-05,
|
148 |
+
"loss": 0.1544,
|
149 |
+
"step": 100
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.51,
|
153 |
+
"grad_norm": 9.541460990905762,
|
154 |
+
"learning_rate": 4.9946294307196566e-05,
|
155 |
+
"loss": 0.1224,
|
156 |
+
"step": 105
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.53,
|
160 |
+
"grad_norm": 5.275872230529785,
|
161 |
+
"learning_rate": 4.967776584317938e-05,
|
162 |
+
"loss": 0.1511,
|
163 |
+
"step": 110
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.55,
|
167 |
+
"grad_norm": 2.8402061462402344,
|
168 |
+
"learning_rate": 4.940923737916219e-05,
|
169 |
+
"loss": 0.1081,
|
170 |
+
"step": 115
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.58,
|
174 |
+
"grad_norm": 6.35424280166626,
|
175 |
+
"learning_rate": 4.9140708915145005e-05,
|
176 |
+
"loss": 0.1288,
|
177 |
+
"step": 120
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.6,
|
181 |
+
"grad_norm": 6.0777435302734375,
|
182 |
+
"learning_rate": 4.887218045112782e-05,
|
183 |
+
"loss": 0.1277,
|
184 |
+
"step": 125
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.63,
|
188 |
+
"grad_norm": 3.709442377090454,
|
189 |
+
"learning_rate": 4.860365198711064e-05,
|
190 |
+
"loss": 0.136,
|
191 |
+
"step": 130
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.65,
|
195 |
+
"grad_norm": 6.1887593269348145,
|
196 |
+
"learning_rate": 4.833512352309345e-05,
|
197 |
+
"loss": 0.1758,
|
198 |
+
"step": 135
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.67,
|
202 |
+
"grad_norm": 4.514546871185303,
|
203 |
+
"learning_rate": 4.806659505907626e-05,
|
204 |
+
"loss": 0.156,
|
205 |
+
"step": 140
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.7,
|
209 |
+
"grad_norm": 5.032735824584961,
|
210 |
+
"learning_rate": 4.7798066595059076e-05,
|
211 |
+
"loss": 0.1593,
|
212 |
+
"step": 145
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.72,
|
216 |
+
"grad_norm": 9.502847671508789,
|
217 |
+
"learning_rate": 4.7529538131041896e-05,
|
218 |
+
"loss": 0.2014,
|
219 |
+
"step": 150
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.75,
|
223 |
+
"grad_norm": 2.6441538333892822,
|
224 |
+
"learning_rate": 4.726100966702471e-05,
|
225 |
+
"loss": 0.1331,
|
226 |
+
"step": 155
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.77,
|
230 |
+
"grad_norm": 2.8725085258483887,
|
231 |
+
"learning_rate": 4.699248120300752e-05,
|
232 |
+
"loss": 0.1261,
|
233 |
+
"step": 160
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.79,
|
237 |
+
"grad_norm": 2.6660351753234863,
|
238 |
+
"learning_rate": 4.6723952738990334e-05,
|
239 |
+
"loss": 0.1097,
|
240 |
+
"step": 165
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.82,
|
244 |
+
"grad_norm": 8.11793327331543,
|
245 |
+
"learning_rate": 4.645542427497315e-05,
|
246 |
+
"loss": 0.0933,
|
247 |
+
"step": 170
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.84,
|
251 |
+
"grad_norm": 3.4776053428649902,
|
252 |
+
"learning_rate": 4.618689581095597e-05,
|
253 |
+
"loss": 0.1543,
|
254 |
+
"step": 175
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.87,
|
258 |
+
"grad_norm": 3.1535816192626953,
|
259 |
+
"learning_rate": 4.591836734693878e-05,
|
260 |
+
"loss": 0.1245,
|
261 |
+
"step": 180
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.89,
|
265 |
+
"grad_norm": 4.248240947723389,
|
266 |
+
"learning_rate": 4.564983888292159e-05,
|
267 |
+
"loss": 0.1417,
|
268 |
+
"step": 185
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.91,
|
272 |
+
"grad_norm": 4.858673572540283,
|
273 |
+
"learning_rate": 4.5381310418904406e-05,
|
274 |
+
"loss": 0.1382,
|
275 |
+
"step": 190
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.94,
|
279 |
+
"grad_norm": 3.6954901218414307,
|
280 |
+
"learning_rate": 4.511278195488722e-05,
|
281 |
+
"loss": 0.1133,
|
282 |
+
"step": 195
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.96,
|
286 |
+
"grad_norm": 2.9366323947906494,
|
287 |
+
"learning_rate": 4.484425349087004e-05,
|
288 |
+
"loss": 0.1037,
|
289 |
+
"step": 200
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.99,
|
293 |
+
"grad_norm": 3.420692205429077,
|
294 |
+
"learning_rate": 4.457572502685285e-05,
|
295 |
+
"loss": 0.1269,
|
296 |
+
"step": 205
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.0,
|
300 |
+
"eval_accuracy": 0.9834123222748815,
|
301 |
+
"eval_loss": 0.050108980387449265,
|
302 |
+
"eval_runtime": 6.0814,
|
303 |
+
"eval_samples_per_second": 485.741,
|
304 |
+
"eval_steps_per_second": 15.292,
|
305 |
+
"step": 207
|
306 |
+
},
|
307 |
+
{
|
308 |
+
"epoch": 1.01,
|
309 |
+
"grad_norm": 3.8513619899749756,
|
310 |
+
"learning_rate": 4.4307196562835664e-05,
|
311 |
+
"loss": 0.1179,
|
312 |
+
"step": 210
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"epoch": 1.03,
|
316 |
+
"grad_norm": 3.5678696632385254,
|
317 |
+
"learning_rate": 4.403866809881848e-05,
|
318 |
+
"loss": 0.1067,
|
319 |
+
"step": 215
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"epoch": 1.06,
|
323 |
+
"grad_norm": 3.045541286468506,
|
324 |
+
"learning_rate": 4.3770139634801297e-05,
|
325 |
+
"loss": 0.1014,
|
326 |
+
"step": 220
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 1.08,
|
330 |
+
"grad_norm": 2.9802796840667725,
|
331 |
+
"learning_rate": 4.350161117078411e-05,
|
332 |
+
"loss": 0.0839,
|
333 |
+
"step": 225
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 1.11,
|
337 |
+
"grad_norm": 2.974304437637329,
|
338 |
+
"learning_rate": 4.323308270676692e-05,
|
339 |
+
"loss": 0.096,
|
340 |
+
"step": 230
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 1.13,
|
344 |
+
"grad_norm": 3.204138994216919,
|
345 |
+
"learning_rate": 4.2964554242749735e-05,
|
346 |
+
"loss": 0.088,
|
347 |
+
"step": 235
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 1.16,
|
351 |
+
"grad_norm": 5.182232856750488,
|
352 |
+
"learning_rate": 4.269602577873255e-05,
|
353 |
+
"loss": 0.0926,
|
354 |
+
"step": 240
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 1.18,
|
358 |
+
"grad_norm": 2.830148458480835,
|
359 |
+
"learning_rate": 4.242749731471536e-05,
|
360 |
+
"loss": 0.0915,
|
361 |
+
"step": 245
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 1.2,
|
365 |
+
"grad_norm": 5.067986011505127,
|
366 |
+
"learning_rate": 4.215896885069818e-05,
|
367 |
+
"loss": 0.1051,
|
368 |
+
"step": 250
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 1.23,
|
372 |
+
"grad_norm": 3.488647937774658,
|
373 |
+
"learning_rate": 4.1890440386680994e-05,
|
374 |
+
"loss": 0.0948,
|
375 |
+
"step": 255
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 1.25,
|
379 |
+
"grad_norm": 2.4556047916412354,
|
380 |
+
"learning_rate": 4.1621911922663806e-05,
|
381 |
+
"loss": 0.0915,
|
382 |
+
"step": 260
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 1.28,
|
386 |
+
"grad_norm": 4.81022834777832,
|
387 |
+
"learning_rate": 4.135338345864662e-05,
|
388 |
+
"loss": 0.1068,
|
389 |
+
"step": 265
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 1.3,
|
393 |
+
"grad_norm": 4.263596057891846,
|
394 |
+
"learning_rate": 4.108485499462943e-05,
|
395 |
+
"loss": 0.106,
|
396 |
+
"step": 270
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 1.32,
|
400 |
+
"grad_norm": 2.286785840988159,
|
401 |
+
"learning_rate": 4.0816326530612245e-05,
|
402 |
+
"loss": 0.0784,
|
403 |
+
"step": 275
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 1.35,
|
407 |
+
"grad_norm": 3.0982165336608887,
|
408 |
+
"learning_rate": 4.054779806659506e-05,
|
409 |
+
"loss": 0.0774,
|
410 |
+
"step": 280
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 1.37,
|
414 |
+
"grad_norm": 3.3483657836914062,
|
415 |
+
"learning_rate": 4.027926960257787e-05,
|
416 |
+
"loss": 0.0732,
|
417 |
+
"step": 285
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 1.4,
|
421 |
+
"grad_norm": 3.933288335800171,
|
422 |
+
"learning_rate": 4.0010741138560684e-05,
|
423 |
+
"loss": 0.086,
|
424 |
+
"step": 290
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 1.42,
|
428 |
+
"grad_norm": 3.418642282485962,
|
429 |
+
"learning_rate": 3.97422126745435e-05,
|
430 |
+
"loss": 0.0803,
|
431 |
+
"step": 295
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"epoch": 1.44,
|
435 |
+
"grad_norm": 3.975381374359131,
|
436 |
+
"learning_rate": 3.9473684210526316e-05,
|
437 |
+
"loss": 0.1137,
|
438 |
+
"step": 300
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"epoch": 1.47,
|
442 |
+
"grad_norm": 2.6510424613952637,
|
443 |
+
"learning_rate": 3.920515574650913e-05,
|
444 |
+
"loss": 0.1206,
|
445 |
+
"step": 305
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"epoch": 1.49,
|
449 |
+
"grad_norm": 2.9062960147857666,
|
450 |
+
"learning_rate": 3.893662728249194e-05,
|
451 |
+
"loss": 0.1042,
|
452 |
+
"step": 310
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 1.52,
|
456 |
+
"grad_norm": 2.369260549545288,
|
457 |
+
"learning_rate": 3.866809881847476e-05,
|
458 |
+
"loss": 0.0856,
|
459 |
+
"step": 315
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 1.54,
|
463 |
+
"grad_norm": 5.03189754486084,
|
464 |
+
"learning_rate": 3.8399570354457575e-05,
|
465 |
+
"loss": 0.0831,
|
466 |
+
"step": 320
|
467 |
+
},
|
468 |
+
{
|
469 |
+
"epoch": 1.56,
|
470 |
+
"grad_norm": 4.001619338989258,
|
471 |
+
"learning_rate": 3.813104189044039e-05,
|
472 |
+
"loss": 0.0766,
|
473 |
+
"step": 325
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 1.59,
|
477 |
+
"grad_norm": 3.2500596046447754,
|
478 |
+
"learning_rate": 3.78625134264232e-05,
|
479 |
+
"loss": 0.0881,
|
480 |
+
"step": 330
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 1.61,
|
484 |
+
"grad_norm": 1.553835391998291,
|
485 |
+
"learning_rate": 3.759398496240601e-05,
|
486 |
+
"loss": 0.0714,
|
487 |
+
"step": 335
|
488 |
+
},
|
489 |
+
{
|
490 |
+
"epoch": 1.64,
|
491 |
+
"grad_norm": 4.914017200469971,
|
492 |
+
"learning_rate": 3.732545649838883e-05,
|
493 |
+
"loss": 0.0944,
|
494 |
+
"step": 340
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 1.66,
|
498 |
+
"grad_norm": 2.7451980113983154,
|
499 |
+
"learning_rate": 3.7056928034371646e-05,
|
500 |
+
"loss": 0.0733,
|
501 |
+
"step": 345
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 1.68,
|
505 |
+
"grad_norm": 4.39937162399292,
|
506 |
+
"learning_rate": 3.678839957035446e-05,
|
507 |
+
"loss": 0.1127,
|
508 |
+
"step": 350
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"epoch": 1.71,
|
512 |
+
"grad_norm": 3.2537283897399902,
|
513 |
+
"learning_rate": 3.651987110633727e-05,
|
514 |
+
"loss": 0.1236,
|
515 |
+
"step": 355
|
516 |
+
},
|
517 |
+
{
|
518 |
+
"epoch": 1.73,
|
519 |
+
"grad_norm": 4.104055881500244,
|
520 |
+
"learning_rate": 3.6251342642320084e-05,
|
521 |
+
"loss": 0.114,
|
522 |
+
"step": 360
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"epoch": 1.76,
|
526 |
+
"grad_norm": 2.0853748321533203,
|
527 |
+
"learning_rate": 3.5982814178302904e-05,
|
528 |
+
"loss": 0.0885,
|
529 |
+
"step": 365
|
530 |
+
},
|
531 |
+
{
|
532 |
+
"epoch": 1.78,
|
533 |
+
"grad_norm": 2.1333792209625244,
|
534 |
+
"learning_rate": 3.571428571428572e-05,
|
535 |
+
"loss": 0.0901,
|
536 |
+
"step": 370
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 1.81,
|
540 |
+
"grad_norm": 2.9660935401916504,
|
541 |
+
"learning_rate": 3.544575725026853e-05,
|
542 |
+
"loss": 0.0753,
|
543 |
+
"step": 375
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 1.83,
|
547 |
+
"grad_norm": 1.7010711431503296,
|
548 |
+
"learning_rate": 3.517722878625134e-05,
|
549 |
+
"loss": 0.094,
|
550 |
+
"step": 380
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 1.85,
|
554 |
+
"grad_norm": 3.7029120922088623,
|
555 |
+
"learning_rate": 3.4908700322234156e-05,
|
556 |
+
"loss": 0.1148,
|
557 |
+
"step": 385
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 1.88,
|
561 |
+
"grad_norm": 3.4257864952087402,
|
562 |
+
"learning_rate": 3.4640171858216975e-05,
|
563 |
+
"loss": 0.0879,
|
564 |
+
"step": 390
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 1.9,
|
568 |
+
"grad_norm": 6.967185020446777,
|
569 |
+
"learning_rate": 3.437164339419979e-05,
|
570 |
+
"loss": 0.1001,
|
571 |
+
"step": 395
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 1.93,
|
575 |
+
"grad_norm": 1.414320707321167,
|
576 |
+
"learning_rate": 3.41031149301826e-05,
|
577 |
+
"loss": 0.1218,
|
578 |
+
"step": 400
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 1.95,
|
582 |
+
"grad_norm": 1.8158583641052246,
|
583 |
+
"learning_rate": 3.3834586466165414e-05,
|
584 |
+
"loss": 0.0646,
|
585 |
+
"step": 405
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 1.97,
|
589 |
+
"grad_norm": 2.169985055923462,
|
590 |
+
"learning_rate": 3.3566058002148234e-05,
|
591 |
+
"loss": 0.0776,
|
592 |
+
"step": 410
|
593 |
+
},
|
594 |
+
{
|
595 |
+
"epoch": 2.0,
|
596 |
+
"grad_norm": 2.488377571105957,
|
597 |
+
"learning_rate": 3.3297529538131046e-05,
|
598 |
+
"loss": 0.0572,
|
599 |
+
"step": 415
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 2.0,
|
603 |
+
"eval_accuracy": 0.987136086662153,
|
604 |
+
"eval_loss": 0.03736753389239311,
|
605 |
+
"eval_runtime": 6.0106,
|
606 |
+
"eval_samples_per_second": 491.469,
|
607 |
+
"eval_steps_per_second": 15.473,
|
608 |
+
"step": 415
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 2.02,
|
612 |
+
"grad_norm": 3.571716070175171,
|
613 |
+
"learning_rate": 3.302900107411386e-05,
|
614 |
+
"loss": 0.083,
|
615 |
+
"step": 420
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 2.05,
|
619 |
+
"grad_norm": 2.3017261028289795,
|
620 |
+
"learning_rate": 3.276047261009667e-05,
|
621 |
+
"loss": 0.0581,
|
622 |
+
"step": 425
|
623 |
+
},
|
624 |
+
{
|
625 |
+
"epoch": 2.07,
|
626 |
+
"grad_norm": 4.040441513061523,
|
627 |
+
"learning_rate": 3.2491944146079485e-05,
|
628 |
+
"loss": 0.0762,
|
629 |
+
"step": 430
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 2.09,
|
633 |
+
"grad_norm": 1.5620166063308716,
|
634 |
+
"learning_rate": 3.2223415682062305e-05,
|
635 |
+
"loss": 0.0778,
|
636 |
+
"step": 435
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 2.12,
|
640 |
+
"grad_norm": 2.1466052532196045,
|
641 |
+
"learning_rate": 3.195488721804512e-05,
|
642 |
+
"loss": 0.0876,
|
643 |
+
"step": 440
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"epoch": 2.14,
|
647 |
+
"grad_norm": 1.7395237684249878,
|
648 |
+
"learning_rate": 3.168635875402793e-05,
|
649 |
+
"loss": 0.0996,
|
650 |
+
"step": 445
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 2.17,
|
654 |
+
"grad_norm": 5.349012851715088,
|
655 |
+
"learning_rate": 3.1417830290010743e-05,
|
656 |
+
"loss": 0.072,
|
657 |
+
"step": 450
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 2.19,
|
661 |
+
"grad_norm": 4.39583683013916,
|
662 |
+
"learning_rate": 3.1149301825993556e-05,
|
663 |
+
"loss": 0.0997,
|
664 |
+
"step": 455
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 2.21,
|
668 |
+
"grad_norm": 3.541804313659668,
|
669 |
+
"learning_rate": 3.0880773361976376e-05,
|
670 |
+
"loss": 0.0712,
|
671 |
+
"step": 460
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 2.24,
|
675 |
+
"grad_norm": 2.4857404232025146,
|
676 |
+
"learning_rate": 3.061224489795919e-05,
|
677 |
+
"loss": 0.0789,
|
678 |
+
"step": 465
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 2.26,
|
682 |
+
"grad_norm": 3.7146120071411133,
|
683 |
+
"learning_rate": 3.0343716433942e-05,
|
684 |
+
"loss": 0.088,
|
685 |
+
"step": 470
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 2.29,
|
689 |
+
"grad_norm": 2.6535651683807373,
|
690 |
+
"learning_rate": 3.007518796992481e-05,
|
691 |
+
"loss": 0.0797,
|
692 |
+
"step": 475
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 2.31,
|
696 |
+
"grad_norm": 1.3184159994125366,
|
697 |
+
"learning_rate": 2.980665950590763e-05,
|
698 |
+
"loss": 0.0846,
|
699 |
+
"step": 480
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 2.33,
|
703 |
+
"grad_norm": 2.7791616916656494,
|
704 |
+
"learning_rate": 2.9538131041890444e-05,
|
705 |
+
"loss": 0.074,
|
706 |
+
"step": 485
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 2.36,
|
710 |
+
"grad_norm": 4.284481048583984,
|
711 |
+
"learning_rate": 2.9269602577873257e-05,
|
712 |
+
"loss": 0.0943,
|
713 |
+
"step": 490
|
714 |
+
},
|
715 |
+
{
|
716 |
+
"epoch": 2.38,
|
717 |
+
"grad_norm": 3.457939386367798,
|
718 |
+
"learning_rate": 2.900107411385607e-05,
|
719 |
+
"loss": 0.117,
|
720 |
+
"step": 495
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 2.41,
|
724 |
+
"grad_norm": 1.2724689245224,
|
725 |
+
"learning_rate": 2.8732545649838882e-05,
|
726 |
+
"loss": 0.0803,
|
727 |
+
"step": 500
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 2.43,
|
731 |
+
"grad_norm": 2.6043756008148193,
|
732 |
+
"learning_rate": 2.8464017185821702e-05,
|
733 |
+
"loss": 0.0672,
|
734 |
+
"step": 505
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 2.45,
|
738 |
+
"grad_norm": 3.096151113510132,
|
739 |
+
"learning_rate": 2.8195488721804515e-05,
|
740 |
+
"loss": 0.0848,
|
741 |
+
"step": 510
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 2.48,
|
745 |
+
"grad_norm": 1.409250259399414,
|
746 |
+
"learning_rate": 2.7926960257787328e-05,
|
747 |
+
"loss": 0.0472,
|
748 |
+
"step": 515
|
749 |
+
},
|
750 |
+
{
|
751 |
+
"epoch": 2.5,
|
752 |
+
"grad_norm": 1.6517168283462524,
|
753 |
+
"learning_rate": 2.765843179377014e-05,
|
754 |
+
"loss": 0.0798,
|
755 |
+
"step": 520
|
756 |
+
},
|
757 |
+
{
|
758 |
+
"epoch": 2.53,
|
759 |
+
"grad_norm": 2.9068679809570312,
|
760 |
+
"learning_rate": 2.7389903329752954e-05,
|
761 |
+
"loss": 0.0658,
|
762 |
+
"step": 525
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 2.55,
|
766 |
+
"grad_norm": 3.5304489135742188,
|
767 |
+
"learning_rate": 2.712137486573577e-05,
|
768 |
+
"loss": 0.0512,
|
769 |
+
"step": 530
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"epoch": 2.58,
|
773 |
+
"grad_norm": 4.021263122558594,
|
774 |
+
"learning_rate": 2.6852846401718583e-05,
|
775 |
+
"loss": 0.0889,
|
776 |
+
"step": 535
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 2.6,
|
780 |
+
"grad_norm": 1.863612413406372,
|
781 |
+
"learning_rate": 2.6584317937701396e-05,
|
782 |
+
"loss": 0.0671,
|
783 |
+
"step": 540
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 2.62,
|
787 |
+
"grad_norm": 2.903055191040039,
|
788 |
+
"learning_rate": 2.6315789473684212e-05,
|
789 |
+
"loss": 0.071,
|
790 |
+
"step": 545
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 2.65,
|
794 |
+
"grad_norm": 4.729198455810547,
|
795 |
+
"learning_rate": 2.6047261009667025e-05,
|
796 |
+
"loss": 0.0753,
|
797 |
+
"step": 550
|
798 |
+
},
|
799 |
+
{
|
800 |
+
"epoch": 2.67,
|
801 |
+
"grad_norm": 4.479395866394043,
|
802 |
+
"learning_rate": 2.577873254564984e-05,
|
803 |
+
"loss": 0.088,
|
804 |
+
"step": 555
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 2.7,
|
808 |
+
"grad_norm": 1.303819179534912,
|
809 |
+
"learning_rate": 2.5510204081632654e-05,
|
810 |
+
"loss": 0.0606,
|
811 |
+
"step": 560
|
812 |
+
},
|
813 |
+
{
|
814 |
+
"epoch": 2.72,
|
815 |
+
"grad_norm": 1.7099323272705078,
|
816 |
+
"learning_rate": 2.5241675617615467e-05,
|
817 |
+
"loss": 0.0563,
|
818 |
+
"step": 565
|
819 |
+
},
|
820 |
+
{
|
821 |
+
"epoch": 2.74,
|
822 |
+
"grad_norm": 6.212317943572998,
|
823 |
+
"learning_rate": 2.4973147153598283e-05,
|
824 |
+
"loss": 0.0948,
|
825 |
+
"step": 570
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 2.77,
|
829 |
+
"grad_norm": 2.4912729263305664,
|
830 |
+
"learning_rate": 2.4704618689581096e-05,
|
831 |
+
"loss": 0.0857,
|
832 |
+
"step": 575
|
833 |
+
},
|
834 |
+
{
|
835 |
+
"epoch": 2.79,
|
836 |
+
"grad_norm": 2.978109359741211,
|
837 |
+
"learning_rate": 2.443609022556391e-05,
|
838 |
+
"loss": 0.0976,
|
839 |
+
"step": 580
|
840 |
+
},
|
841 |
+
{
|
842 |
+
"epoch": 2.82,
|
843 |
+
"grad_norm": 1.6303725242614746,
|
844 |
+
"learning_rate": 2.4167561761546725e-05,
|
845 |
+
"loss": 0.0731,
|
846 |
+
"step": 585
|
847 |
+
},
|
848 |
+
{
|
849 |
+
"epoch": 2.84,
|
850 |
+
"grad_norm": 3.7856266498565674,
|
851 |
+
"learning_rate": 2.3899033297529538e-05,
|
852 |
+
"loss": 0.0833,
|
853 |
+
"step": 590
|
854 |
+
},
|
855 |
+
{
|
856 |
+
"epoch": 2.86,
|
857 |
+
"grad_norm": 3.6364758014678955,
|
858 |
+
"learning_rate": 2.3630504833512354e-05,
|
859 |
+
"loss": 0.0659,
|
860 |
+
"step": 595
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"epoch": 2.89,
|
864 |
+
"grad_norm": 3.412216901779175,
|
865 |
+
"learning_rate": 2.3361976369495167e-05,
|
866 |
+
"loss": 0.1227,
|
867 |
+
"step": 600
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 2.91,
|
871 |
+
"grad_norm": 4.180442810058594,
|
872 |
+
"learning_rate": 2.3093447905477984e-05,
|
873 |
+
"loss": 0.0773,
|
874 |
+
"step": 605
|
875 |
+
},
|
876 |
+
{
|
877 |
+
"epoch": 2.94,
|
878 |
+
"grad_norm": 1.9002240896224976,
|
879 |
+
"learning_rate": 2.2824919441460796e-05,
|
880 |
+
"loss": 0.0656,
|
881 |
+
"step": 610
|
882 |
+
},
|
883 |
+
{
|
884 |
+
"epoch": 2.96,
|
885 |
+
"grad_norm": 3.6314845085144043,
|
886 |
+
"learning_rate": 2.255639097744361e-05,
|
887 |
+
"loss": 0.0768,
|
888 |
+
"step": 615
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"epoch": 2.98,
|
892 |
+
"grad_norm": 2.4498918056488037,
|
893 |
+
"learning_rate": 2.2287862513426426e-05,
|
894 |
+
"loss": 0.0806,
|
895 |
+
"step": 620
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"epoch": 3.0,
|
899 |
+
"eval_accuracy": 0.9874746106973595,
|
900 |
+
"eval_loss": 0.03359290212392807,
|
901 |
+
"eval_runtime": 5.9679,
|
902 |
+
"eval_samples_per_second": 494.983,
|
903 |
+
"eval_steps_per_second": 15.583,
|
904 |
+
"step": 623
|
905 |
+
},
|
906 |
+
{
|
907 |
+
"epoch": 3.01,
|
908 |
+
"grad_norm": 3.5386340618133545,
|
909 |
+
"learning_rate": 2.201933404940924e-05,
|
910 |
+
"loss": 0.0704,
|
911 |
+
"step": 625
|
912 |
+
},
|
913 |
+
{
|
914 |
+
"epoch": 3.03,
|
915 |
+
"grad_norm": 2.9113972187042236,
|
916 |
+
"learning_rate": 2.1750805585392055e-05,
|
917 |
+
"loss": 0.0698,
|
918 |
+
"step": 630
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 3.06,
|
922 |
+
"grad_norm": 4.226195812225342,
|
923 |
+
"learning_rate": 2.1482277121374868e-05,
|
924 |
+
"loss": 0.0728,
|
925 |
+
"step": 635
|
926 |
+
},
|
927 |
+
{
|
928 |
+
"epoch": 3.08,
|
929 |
+
"grad_norm": 2.5996837615966797,
|
930 |
+
"learning_rate": 2.121374865735768e-05,
|
931 |
+
"loss": 0.0778,
|
932 |
+
"step": 640
|
933 |
+
},
|
934 |
+
{
|
935 |
+
"epoch": 3.1,
|
936 |
+
"grad_norm": 3.129270076751709,
|
937 |
+
"learning_rate": 2.0945220193340497e-05,
|
938 |
+
"loss": 0.0803,
|
939 |
+
"step": 645
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 3.13,
|
943 |
+
"grad_norm": 4.704379081726074,
|
944 |
+
"learning_rate": 2.067669172932331e-05,
|
945 |
+
"loss": 0.0605,
|
946 |
+
"step": 650
|
947 |
+
},
|
948 |
+
{
|
949 |
+
"epoch": 3.15,
|
950 |
+
"grad_norm": 3.2946126461029053,
|
951 |
+
"learning_rate": 2.0408163265306123e-05,
|
952 |
+
"loss": 0.0719,
|
953 |
+
"step": 655
|
954 |
+
},
|
955 |
+
{
|
956 |
+
"epoch": 3.18,
|
957 |
+
"grad_norm": 3.0328078269958496,
|
958 |
+
"learning_rate": 2.0139634801288935e-05,
|
959 |
+
"loss": 0.064,
|
960 |
+
"step": 660
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 3.2,
|
964 |
+
"grad_norm": 3.618300199508667,
|
965 |
+
"learning_rate": 1.987110633727175e-05,
|
966 |
+
"loss": 0.0673,
|
967 |
+
"step": 665
|
968 |
+
},
|
969 |
+
{
|
970 |
+
"epoch": 3.23,
|
971 |
+
"grad_norm": 3.6683876514434814,
|
972 |
+
"learning_rate": 1.9602577873254565e-05,
|
973 |
+
"loss": 0.0571,
|
974 |
+
"step": 670
|
975 |
+
},
|
976 |
+
{
|
977 |
+
"epoch": 3.25,
|
978 |
+
"grad_norm": 1.6800568103790283,
|
979 |
+
"learning_rate": 1.933404940923738e-05,
|
980 |
+
"loss": 0.047,
|
981 |
+
"step": 675
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 3.27,
|
985 |
+
"grad_norm": 1.3418190479278564,
|
986 |
+
"learning_rate": 1.9065520945220194e-05,
|
987 |
+
"loss": 0.0633,
|
988 |
+
"step": 680
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 3.3,
|
992 |
+
"grad_norm": 2.5490853786468506,
|
993 |
+
"learning_rate": 1.8796992481203007e-05,
|
994 |
+
"loss": 0.0402,
|
995 |
+
"step": 685
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"epoch": 3.32,
|
999 |
+
"grad_norm": 1.2604730129241943,
|
1000 |
+
"learning_rate": 1.8528464017185823e-05,
|
1001 |
+
"loss": 0.0605,
|
1002 |
+
"step": 690
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 3.35,
|
1006 |
+
"grad_norm": 4.190303325653076,
|
1007 |
+
"learning_rate": 1.8259935553168636e-05,
|
1008 |
+
"loss": 0.0886,
|
1009 |
+
"step": 695
|
1010 |
+
},
|
1011 |
+
{
|
1012 |
+
"epoch": 3.37,
|
1013 |
+
"grad_norm": 7.317651748657227,
|
1014 |
+
"learning_rate": 1.7991407089151452e-05,
|
1015 |
+
"loss": 0.0875,
|
1016 |
+
"step": 700
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 3.39,
|
1020 |
+
"grad_norm": 1.9360538721084595,
|
1021 |
+
"learning_rate": 1.7722878625134265e-05,
|
1022 |
+
"loss": 0.0557,
|
1023 |
+
"step": 705
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 3.42,
|
1027 |
+
"grad_norm": 4.210879802703857,
|
1028 |
+
"learning_rate": 1.7454350161117078e-05,
|
1029 |
+
"loss": 0.0735,
|
1030 |
+
"step": 710
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"epoch": 3.44,
|
1034 |
+
"grad_norm": 5.373249053955078,
|
1035 |
+
"learning_rate": 1.7185821697099894e-05,
|
1036 |
+
"loss": 0.0975,
|
1037 |
+
"step": 715
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 3.47,
|
1041 |
+
"grad_norm": 2.7579221725463867,
|
1042 |
+
"learning_rate": 1.6917293233082707e-05,
|
1043 |
+
"loss": 0.0705,
|
1044 |
+
"step": 720
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 3.49,
|
1048 |
+
"grad_norm": 2.673476457595825,
|
1049 |
+
"learning_rate": 1.6648764769065523e-05,
|
1050 |
+
"loss": 0.0524,
|
1051 |
+
"step": 725
|
1052 |
+
},
|
1053 |
+
{
|
1054 |
+
"epoch": 3.51,
|
1055 |
+
"grad_norm": 4.691239356994629,
|
1056 |
+
"learning_rate": 1.6380236305048336e-05,
|
1057 |
+
"loss": 0.0686,
|
1058 |
+
"step": 730
|
1059 |
+
},
|
1060 |
+
{
|
1061 |
+
"epoch": 3.54,
|
1062 |
+
"grad_norm": 4.022672176361084,
|
1063 |
+
"learning_rate": 1.6111707841031152e-05,
|
1064 |
+
"loss": 0.0632,
|
1065 |
+
"step": 735
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 3.56,
|
1069 |
+
"grad_norm": 2.69425630569458,
|
1070 |
+
"learning_rate": 1.5843179377013965e-05,
|
1071 |
+
"loss": 0.0636,
|
1072 |
+
"step": 740
|
1073 |
+
},
|
1074 |
+
{
|
1075 |
+
"epoch": 3.59,
|
1076 |
+
"grad_norm": 4.5648884773254395,
|
1077 |
+
"learning_rate": 1.5574650912996778e-05,
|
1078 |
+
"loss": 0.0627,
|
1079 |
+
"step": 745
|
1080 |
+
},
|
1081 |
+
{
|
1082 |
+
"epoch": 3.61,
|
1083 |
+
"grad_norm": 3.1452832221984863,
|
1084 |
+
"learning_rate": 1.5306122448979594e-05,
|
1085 |
+
"loss": 0.0763,
|
1086 |
+
"step": 750
|
1087 |
+
},
|
1088 |
+
{
|
1089 |
+
"epoch": 3.63,
|
1090 |
+
"grad_norm": 5.030516147613525,
|
1091 |
+
"learning_rate": 1.5037593984962406e-05,
|
1092 |
+
"loss": 0.0581,
|
1093 |
+
"step": 755
|
1094 |
+
},
|
1095 |
+
{
|
1096 |
+
"epoch": 3.66,
|
1097 |
+
"grad_norm": 3.2779948711395264,
|
1098 |
+
"learning_rate": 1.4769065520945222e-05,
|
1099 |
+
"loss": 0.0549,
|
1100 |
+
"step": 760
|
1101 |
+
},
|
1102 |
+
{
|
1103 |
+
"epoch": 3.68,
|
1104 |
+
"grad_norm": 1.7655272483825684,
|
1105 |
+
"learning_rate": 1.4500537056928035e-05,
|
1106 |
+
"loss": 0.0778,
|
1107 |
+
"step": 765
|
1108 |
+
},
|
1109 |
+
{
|
1110 |
+
"epoch": 3.71,
|
1111 |
+
"grad_norm": 3.9161980152130127,
|
1112 |
+
"learning_rate": 1.4232008592910851e-05,
|
1113 |
+
"loss": 0.0773,
|
1114 |
+
"step": 770
|
1115 |
+
},
|
1116 |
+
{
|
1117 |
+
"epoch": 3.73,
|
1118 |
+
"grad_norm": 3.029256820678711,
|
1119 |
+
"learning_rate": 1.3963480128893664e-05,
|
1120 |
+
"loss": 0.056,
|
1121 |
+
"step": 775
|
1122 |
+
},
|
1123 |
+
{
|
1124 |
+
"epoch": 3.75,
|
1125 |
+
"grad_norm": 4.829240322113037,
|
1126 |
+
"learning_rate": 1.3694951664876477e-05,
|
1127 |
+
"loss": 0.0722,
|
1128 |
+
"step": 780
|
1129 |
+
},
|
1130 |
+
{
|
1131 |
+
"epoch": 3.78,
|
1132 |
+
"grad_norm": 2.5825934410095215,
|
1133 |
+
"learning_rate": 1.3426423200859291e-05,
|
1134 |
+
"loss": 0.0372,
|
1135 |
+
"step": 785
|
1136 |
+
},
|
1137 |
+
{
|
1138 |
+
"epoch": 3.8,
|
1139 |
+
"grad_norm": 1.706741452217102,
|
1140 |
+
"learning_rate": 1.3157894736842106e-05,
|
1141 |
+
"loss": 0.0369,
|
1142 |
+
"step": 790
|
1143 |
+
},
|
1144 |
+
{
|
1145 |
+
"epoch": 3.83,
|
1146 |
+
"grad_norm": 2.689628839492798,
|
1147 |
+
"learning_rate": 1.288936627282492e-05,
|
1148 |
+
"loss": 0.0676,
|
1149 |
+
"step": 795
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 3.85,
|
1153 |
+
"grad_norm": 3.5289785861968994,
|
1154 |
+
"learning_rate": 1.2620837808807733e-05,
|
1155 |
+
"loss": 0.0802,
|
1156 |
+
"step": 800
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 3.87,
|
1160 |
+
"grad_norm": 3.4799952507019043,
|
1161 |
+
"learning_rate": 1.2352309344790548e-05,
|
1162 |
+
"loss": 0.0589,
|
1163 |
+
"step": 805
|
1164 |
+
},
|
1165 |
+
{
|
1166 |
+
"epoch": 3.9,
|
1167 |
+
"grad_norm": 1.8066169023513794,
|
1168 |
+
"learning_rate": 1.2083780880773363e-05,
|
1169 |
+
"loss": 0.0662,
|
1170 |
+
"step": 810
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"epoch": 3.92,
|
1174 |
+
"grad_norm": 2.680537700653076,
|
1175 |
+
"learning_rate": 1.1815252416756177e-05,
|
1176 |
+
"loss": 0.0482,
|
1177 |
+
"step": 815
|
1178 |
+
},
|
1179 |
+
{
|
1180 |
+
"epoch": 3.95,
|
1181 |
+
"grad_norm": 4.0623555183410645,
|
1182 |
+
"learning_rate": 1.1546723952738992e-05,
|
1183 |
+
"loss": 0.0786,
|
1184 |
+
"step": 820
|
1185 |
+
},
|
1186 |
+
{
|
1187 |
+
"epoch": 3.97,
|
1188 |
+
"grad_norm": 2.9296209812164307,
|
1189 |
+
"learning_rate": 1.1278195488721805e-05,
|
1190 |
+
"loss": 0.0406,
|
1191 |
+
"step": 825
|
1192 |
+
},
|
1193 |
+
{
|
1194 |
+
"epoch": 4.0,
|
1195 |
+
"grad_norm": 3.4645802974700928,
|
1196 |
+
"learning_rate": 1.100966702470462e-05,
|
1197 |
+
"loss": 0.0589,
|
1198 |
+
"step": 830
|
1199 |
+
},
|
1200 |
+
{
|
1201 |
+
"epoch": 4.0,
|
1202 |
+
"eval_accuracy": 0.988490182802979,
|
1203 |
+
"eval_loss": 0.03382693976163864,
|
1204 |
+
"eval_runtime": 6.0045,
|
1205 |
+
"eval_samples_per_second": 491.966,
|
1206 |
+
"eval_steps_per_second": 15.488,
|
1207 |
+
"step": 831
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 4.02,
|
1211 |
+
"grad_norm": 3.1075704097747803,
|
1212 |
+
"learning_rate": 1.0741138560687434e-05,
|
1213 |
+
"loss": 0.0602,
|
1214 |
+
"step": 835
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 4.04,
|
1218 |
+
"grad_norm": 0.9231489300727844,
|
1219 |
+
"learning_rate": 1.0472610096670248e-05,
|
1220 |
+
"loss": 0.0525,
|
1221 |
+
"step": 840
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 4.07,
|
1225 |
+
"grad_norm": 1.0160259008407593,
|
1226 |
+
"learning_rate": 1.0204081632653061e-05,
|
1227 |
+
"loss": 0.0409,
|
1228 |
+
"step": 845
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 4.09,
|
1232 |
+
"grad_norm": 1.9544317722320557,
|
1233 |
+
"learning_rate": 9.935553168635876e-06,
|
1234 |
+
"loss": 0.0504,
|
1235 |
+
"step": 850
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 4.12,
|
1239 |
+
"grad_norm": 2.4267749786376953,
|
1240 |
+
"learning_rate": 9.66702470461869e-06,
|
1241 |
+
"loss": 0.059,
|
1242 |
+
"step": 855
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 4.14,
|
1246 |
+
"grad_norm": 2.689138889312744,
|
1247 |
+
"learning_rate": 9.398496240601503e-06,
|
1248 |
+
"loss": 0.0625,
|
1249 |
+
"step": 860
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 4.16,
|
1253 |
+
"grad_norm": 1.5462526082992554,
|
1254 |
+
"learning_rate": 9.129967776584318e-06,
|
1255 |
+
"loss": 0.0492,
|
1256 |
+
"step": 865
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 4.19,
|
1260 |
+
"grad_norm": 4.044441223144531,
|
1261 |
+
"learning_rate": 8.861439312567132e-06,
|
1262 |
+
"loss": 0.05,
|
1263 |
+
"step": 870
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 4.21,
|
1267 |
+
"grad_norm": 5.436426162719727,
|
1268 |
+
"learning_rate": 8.592910848549947e-06,
|
1269 |
+
"loss": 0.0455,
|
1270 |
+
"step": 875
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 4.24,
|
1274 |
+
"grad_norm": 2.8710386753082275,
|
1275 |
+
"learning_rate": 8.324382384532762e-06,
|
1276 |
+
"loss": 0.0645,
|
1277 |
+
"step": 880
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 4.26,
|
1281 |
+
"grad_norm": 3.6887547969818115,
|
1282 |
+
"learning_rate": 8.055853920515576e-06,
|
1283 |
+
"loss": 0.0623,
|
1284 |
+
"step": 885
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 4.28,
|
1288 |
+
"grad_norm": 3.6889421939849854,
|
1289 |
+
"learning_rate": 7.787325456498389e-06,
|
1290 |
+
"loss": 0.0569,
|
1291 |
+
"step": 890
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 4.31,
|
1295 |
+
"grad_norm": 3.734835624694824,
|
1296 |
+
"learning_rate": 7.518796992481203e-06,
|
1297 |
+
"loss": 0.0677,
|
1298 |
+
"step": 895
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 4.33,
|
1302 |
+
"grad_norm": 2.4793076515197754,
|
1303 |
+
"learning_rate": 7.250268528464017e-06,
|
1304 |
+
"loss": 0.0445,
|
1305 |
+
"step": 900
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 4.36,
|
1309 |
+
"grad_norm": 5.3167500495910645,
|
1310 |
+
"learning_rate": 6.981740064446832e-06,
|
1311 |
+
"loss": 0.0857,
|
1312 |
+
"step": 905
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 4.38,
|
1316 |
+
"grad_norm": 2.9194352626800537,
|
1317 |
+
"learning_rate": 6.713211600429646e-06,
|
1318 |
+
"loss": 0.0704,
|
1319 |
+
"step": 910
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 4.4,
|
1323 |
+
"grad_norm": 2.78092885017395,
|
1324 |
+
"learning_rate": 6.44468313641246e-06,
|
1325 |
+
"loss": 0.062,
|
1326 |
+
"step": 915
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 4.43,
|
1330 |
+
"grad_norm": 2.4215569496154785,
|
1331 |
+
"learning_rate": 6.176154672395274e-06,
|
1332 |
+
"loss": 0.063,
|
1333 |
+
"step": 920
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 4.45,
|
1337 |
+
"grad_norm": 2.3797125816345215,
|
1338 |
+
"learning_rate": 5.907626208378089e-06,
|
1339 |
+
"loss": 0.052,
|
1340 |
+
"step": 925
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 4.48,
|
1344 |
+
"grad_norm": 0.7256618142127991,
|
1345 |
+
"learning_rate": 5.639097744360902e-06,
|
1346 |
+
"loss": 0.0468,
|
1347 |
+
"step": 930
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 4.5,
|
1351 |
+
"grad_norm": 1.283785104751587,
|
1352 |
+
"learning_rate": 5.370569280343717e-06,
|
1353 |
+
"loss": 0.0577,
|
1354 |
+
"step": 935
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 4.52,
|
1358 |
+
"grad_norm": 4.501251220703125,
|
1359 |
+
"learning_rate": 5.102040816326531e-06,
|
1360 |
+
"loss": 0.0595,
|
1361 |
+
"step": 940
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 4.55,
|
1365 |
+
"grad_norm": 2.7582945823669434,
|
1366 |
+
"learning_rate": 4.833512352309345e-06,
|
1367 |
+
"loss": 0.0693,
|
1368 |
+
"step": 945
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 4.57,
|
1372 |
+
"grad_norm": 5.989513397216797,
|
1373 |
+
"learning_rate": 4.564983888292159e-06,
|
1374 |
+
"loss": 0.0746,
|
1375 |
+
"step": 950
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 4.6,
|
1379 |
+
"grad_norm": 3.373969554901123,
|
1380 |
+
"learning_rate": 4.2964554242749735e-06,
|
1381 |
+
"loss": 0.0529,
|
1382 |
+
"step": 955
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 4.62,
|
1386 |
+
"grad_norm": 2.1218202114105225,
|
1387 |
+
"learning_rate": 4.027926960257788e-06,
|
1388 |
+
"loss": 0.0368,
|
1389 |
+
"step": 960
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 4.65,
|
1393 |
+
"grad_norm": 1.9887917041778564,
|
1394 |
+
"learning_rate": 3.7593984962406014e-06,
|
1395 |
+
"loss": 0.0354,
|
1396 |
+
"step": 965
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 4.67,
|
1400 |
+
"grad_norm": 1.6485974788665771,
|
1401 |
+
"learning_rate": 3.490870032223416e-06,
|
1402 |
+
"loss": 0.0653,
|
1403 |
+
"step": 970
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 4.69,
|
1407 |
+
"grad_norm": 2.1341357231140137,
|
1408 |
+
"learning_rate": 3.22234156820623e-06,
|
1409 |
+
"loss": 0.0421,
|
1410 |
+
"step": 975
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 4.72,
|
1414 |
+
"grad_norm": 5.573878765106201,
|
1415 |
+
"learning_rate": 2.9538131041890443e-06,
|
1416 |
+
"loss": 0.0741,
|
1417 |
+
"step": 980
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 4.74,
|
1421 |
+
"grad_norm": 3.1527225971221924,
|
1422 |
+
"learning_rate": 2.6852846401718585e-06,
|
1423 |
+
"loss": 0.0327,
|
1424 |
+
"step": 985
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 4.77,
|
1428 |
+
"grad_norm": 3.773310661315918,
|
1429 |
+
"learning_rate": 2.4167561761546726e-06,
|
1430 |
+
"loss": 0.0472,
|
1431 |
+
"step": 990
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 4.79,
|
1435 |
+
"grad_norm": 3.0205063819885254,
|
1436 |
+
"learning_rate": 2.1482277121374868e-06,
|
1437 |
+
"loss": 0.0723,
|
1438 |
+
"step": 995
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 4.81,
|
1442 |
+
"grad_norm": 3.098895788192749,
|
1443 |
+
"learning_rate": 1.8796992481203007e-06,
|
1444 |
+
"loss": 0.0743,
|
1445 |
+
"step": 1000
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 4.84,
|
1449 |
+
"grad_norm": 1.734661340713501,
|
1450 |
+
"learning_rate": 1.611170784103115e-06,
|
1451 |
+
"loss": 0.0337,
|
1452 |
+
"step": 1005
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 4.86,
|
1456 |
+
"grad_norm": 1.7586333751678467,
|
1457 |
+
"learning_rate": 1.3426423200859292e-06,
|
1458 |
+
"loss": 0.062,
|
1459 |
+
"step": 1010
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 4.89,
|
1463 |
+
"grad_norm": 2.720187187194824,
|
1464 |
+
"learning_rate": 1.0741138560687434e-06,
|
1465 |
+
"loss": 0.0791,
|
1466 |
+
"step": 1015
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 4.91,
|
1470 |
+
"grad_norm": 1.8995425701141357,
|
1471 |
+
"learning_rate": 8.055853920515575e-07,
|
1472 |
+
"loss": 0.061,
|
1473 |
+
"step": 1020
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 4.93,
|
1477 |
+
"grad_norm": 4.119653224945068,
|
1478 |
+
"learning_rate": 5.370569280343717e-07,
|
1479 |
+
"loss": 0.0486,
|
1480 |
+
"step": 1025
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 4.96,
|
1484 |
+
"grad_norm": 3.3452024459838867,
|
1485 |
+
"learning_rate": 2.6852846401718585e-07,
|
1486 |
+
"loss": 0.0572,
|
1487 |
+
"step": 1030
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 4.98,
|
1491 |
+
"grad_norm": 1.4279868602752686,
|
1492 |
+
"learning_rate": 0.0,
|
1493 |
+
"loss": 0.0326,
|
1494 |
+
"step": 1035
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 4.98,
|
1498 |
+
"eval_accuracy": 0.989844278943805,
|
1499 |
+
"eval_loss": 0.029886629432439804,
|
1500 |
+
"eval_runtime": 5.9932,
|
1501 |
+
"eval_samples_per_second": 492.888,
|
1502 |
+
"eval_steps_per_second": 15.517,
|
1503 |
+
"step": 1035
|
1504 |
+
},
|
1505 |
+
{
|
1506 |
+
"epoch": 4.98,
|
1507 |
+
"step": 1035,
|
1508 |
+
"total_flos": 3.2925088680822374e+18,
|
1509 |
+
"train_loss": 0.12655677452755434,
|
1510 |
+
"train_runtime": 660.3476,
|
1511 |
+
"train_samples_per_second": 201.303,
|
1512 |
+
"train_steps_per_second": 1.567
|
1513 |
+
}
|
1514 |
+
],
|
1515 |
+
"logging_steps": 5,
|
1516 |
+
"max_steps": 1035,
|
1517 |
+
"num_input_tokens_seen": 0,
|
1518 |
+
"num_train_epochs": 5,
|
1519 |
+
"save_steps": 500,
|
1520 |
+
"total_flos": 3.2925088680822374e+18,
|
1521 |
+
"train_batch_size": 32,
|
1522 |
+
"trial_name": null,
|
1523 |
+
"trial_params": null
|
1524 |
+
}
|