Jeremiah Zhou
commited on
Commit
•
3a00009
1
Parent(s):
d22a896
End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- runs/Jun14_16-36-04_pikachu/events.out.tfevents.1655196606.pikachu.4132163.2 +3 -0
- train_results.json +8 -0
- trainer_state.json +109 -0
all_results.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 8.0,
|
3 |
+
"eval_accuracy": 0.7581227436823105,
|
4 |
+
"eval_loss": 0.6365415453910828,
|
5 |
+
"eval_runtime": 0.5623,
|
6 |
+
"eval_samples": 277,
|
7 |
+
"eval_samples_per_second": 492.657,
|
8 |
+
"eval_steps_per_second": 62.249,
|
9 |
+
"train_loss": 0.42912168074876833,
|
10 |
+
"train_runtime": 173.8183,
|
11 |
+
"train_samples": 2490,
|
12 |
+
"train_samples_per_second": 143.253,
|
13 |
+
"train_steps_per_second": 8.975
|
14 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 8.0,
|
3 |
+
"eval_accuracy": 0.7581227436823105,
|
4 |
+
"eval_loss": 0.6365415453910828,
|
5 |
+
"eval_runtime": 0.5623,
|
6 |
+
"eval_samples": 277,
|
7 |
+
"eval_samples_per_second": 492.657,
|
8 |
+
"eval_steps_per_second": 62.249
|
9 |
+
}
|
runs/Jun14_16-36-04_pikachu/events.out.tfevents.1655196606.pikachu.4132163.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:355c39cf8b2852054ef924f142aaf8e66032cace0d87bc6336916efe279d5408
|
3 |
+
size 363
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 8.0,
|
3 |
+
"train_loss": 0.42912168074876833,
|
4 |
+
"train_runtime": 173.8183,
|
5 |
+
"train_samples": 2490,
|
6 |
+
"train_samples_per_second": 143.253,
|
7 |
+
"train_steps_per_second": 8.975
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.7581227436823105,
|
3 |
+
"best_model_checkpoint": "./fine-tune/roberta-base/rte/checkpoint-780",
|
4 |
+
"epoch": 8.0,
|
5 |
+
"global_step": 1248,
|
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": 0.4729241877256318,
|
13 |
+
"eval_loss": 0.7072076797485352,
|
14 |
+
"eval_runtime": 0.4902,
|
15 |
+
"eval_samples_per_second": 565.102,
|
16 |
+
"eval_steps_per_second": 71.403,
|
17 |
+
"step": 156
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 2.0,
|
21 |
+
"eval_accuracy": 0.5270758122743683,
|
22 |
+
"eval_loss": 0.6958328485488892,
|
23 |
+
"eval_runtime": 0.689,
|
24 |
+
"eval_samples_per_second": 402.039,
|
25 |
+
"eval_steps_per_second": 50.799,
|
26 |
+
"step": 312
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 3.0,
|
30 |
+
"eval_accuracy": 0.6462093862815884,
|
31 |
+
"eval_loss": 0.6193148493766785,
|
32 |
+
"eval_runtime": 0.5121,
|
33 |
+
"eval_samples_per_second": 540.927,
|
34 |
+
"eval_steps_per_second": 68.348,
|
35 |
+
"step": 468
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 3.21,
|
39 |
+
"learning_rate": 1.446111869031378e-05,
|
40 |
+
"loss": 0.6759,
|
41 |
+
"step": 500
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 4.0,
|
45 |
+
"eval_accuracy": 0.7075812274368231,
|
46 |
+
"eval_loss": 0.6046352386474609,
|
47 |
+
"eval_runtime": 0.5274,
|
48 |
+
"eval_samples_per_second": 525.205,
|
49 |
+
"eval_steps_per_second": 66.362,
|
50 |
+
"step": 624
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 5.0,
|
54 |
+
"eval_accuracy": 0.7581227436823105,
|
55 |
+
"eval_loss": 0.6365415453910828,
|
56 |
+
"eval_runtime": 0.681,
|
57 |
+
"eval_samples_per_second": 406.732,
|
58 |
+
"eval_steps_per_second": 51.392,
|
59 |
+
"step": 780
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 6.0,
|
63 |
+
"eval_accuracy": 0.7545126353790613,
|
64 |
+
"eval_loss": 0.897487223148346,
|
65 |
+
"eval_runtime": 0.6584,
|
66 |
+
"eval_samples_per_second": 420.704,
|
67 |
+
"eval_steps_per_second": 53.157,
|
68 |
+
"step": 936
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 6.41,
|
72 |
+
"learning_rate": 7.639836289222374e-06,
|
73 |
+
"loss": 0.3194,
|
74 |
+
"step": 1000
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 7.0,
|
78 |
+
"eval_accuracy": 0.7581227436823105,
|
79 |
+
"eval_loss": 1.2031357288360596,
|
80 |
+
"eval_runtime": 0.5817,
|
81 |
+
"eval_samples_per_second": 476.173,
|
82 |
+
"eval_steps_per_second": 60.166,
|
83 |
+
"step": 1092
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 8.0,
|
87 |
+
"eval_accuracy": 0.7581227436823105,
|
88 |
+
"eval_loss": 1.2942094802856445,
|
89 |
+
"eval_runtime": 0.6818,
|
90 |
+
"eval_samples_per_second": 406.254,
|
91 |
+
"eval_steps_per_second": 51.332,
|
92 |
+
"step": 1248
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 8.0,
|
96 |
+
"step": 1248,
|
97 |
+
"total_flos": 1310293055692800.0,
|
98 |
+
"train_loss": 0.42912168074876833,
|
99 |
+
"train_runtime": 173.8183,
|
100 |
+
"train_samples_per_second": 143.253,
|
101 |
+
"train_steps_per_second": 8.975
|
102 |
+
}
|
103 |
+
],
|
104 |
+
"max_steps": 1560,
|
105 |
+
"num_train_epochs": 10,
|
106 |
+
"total_flos": 1310293055692800.0,
|
107 |
+
"trial_name": null,
|
108 |
+
"trial_params": null
|
109 |
+
}
|