Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1668 -0
trainer_state.json
ADDED
@@ -0,0 +1,1668 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.48741263151168823,
|
3 |
+
"best_model_checkpoint": "/home/ray/default/save/checkpoint-1000",
|
4 |
+
"epoch": 1.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1154,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.004332755632582322,
|
13 |
+
"grad_norm": 0.10313185304403305,
|
14 |
+
"learning_rate": 8.333333333333333e-07,
|
15 |
+
"loss": 1.0706,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.008665511265164644,
|
20 |
+
"grad_norm": 0.10810094326734543,
|
21 |
+
"learning_rate": 1.6666666666666667e-06,
|
22 |
+
"loss": 1.0694,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.012998266897746967,
|
27 |
+
"grad_norm": 0.09731286019086838,
|
28 |
+
"learning_rate": 2.5e-06,
|
29 |
+
"loss": 1.0706,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.01733102253032929,
|
34 |
+
"grad_norm": 0.11459868401288986,
|
35 |
+
"learning_rate": 3.3333333333333333e-06,
|
36 |
+
"loss": 1.0772,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.021663778162911613,
|
41 |
+
"grad_norm": 0.10845116525888443,
|
42 |
+
"learning_rate": 4.166666666666667e-06,
|
43 |
+
"loss": 1.0809,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.025996533795493933,
|
48 |
+
"grad_norm": 0.12091381102800369,
|
49 |
+
"learning_rate": 5e-06,
|
50 |
+
"loss": 1.0737,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.030329289428076257,
|
55 |
+
"grad_norm": 0.12315661460161209,
|
56 |
+
"learning_rate": 5.833333333333334e-06,
|
57 |
+
"loss": 1.0675,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.03466204506065858,
|
62 |
+
"grad_norm": 0.13054250180721283,
|
63 |
+
"learning_rate": 6.666666666666667e-06,
|
64 |
+
"loss": 1.0638,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.0389948006932409,
|
69 |
+
"grad_norm": 0.13414187729358673,
|
70 |
+
"learning_rate": 7.5e-06,
|
71 |
+
"loss": 1.0588,
|
72 |
+
"step": 45
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.043327556325823226,
|
76 |
+
"grad_norm": 0.12992985546588898,
|
77 |
+
"learning_rate": 8.333333333333334e-06,
|
78 |
+
"loss": 1.0512,
|
79 |
+
"step": 50
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.047660311958405546,
|
83 |
+
"grad_norm": 0.11500700563192368,
|
84 |
+
"learning_rate": 9.166666666666666e-06,
|
85 |
+
"loss": 1.0545,
|
86 |
+
"step": 55
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.05199306759098787,
|
90 |
+
"grad_norm": 0.10958714783191681,
|
91 |
+
"learning_rate": 1e-05,
|
92 |
+
"loss": 1.0357,
|
93 |
+
"step": 60
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.05632582322357019,
|
97 |
+
"grad_norm": 0.10853663086891174,
|
98 |
+
"learning_rate": 1.0833333333333334e-05,
|
99 |
+
"loss": 1.0268,
|
100 |
+
"step": 65
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.060658578856152515,
|
104 |
+
"grad_norm": 0.12498235702514648,
|
105 |
+
"learning_rate": 1.1666666666666668e-05,
|
106 |
+
"loss": 1.0135,
|
107 |
+
"step": 70
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.06499133448873484,
|
111 |
+
"grad_norm": 0.11716682463884354,
|
112 |
+
"learning_rate": 1.25e-05,
|
113 |
+
"loss": 1.0014,
|
114 |
+
"step": 75
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.06932409012131716,
|
118 |
+
"grad_norm": 0.09530466794967651,
|
119 |
+
"learning_rate": 1.3333333333333333e-05,
|
120 |
+
"loss": 0.9824,
|
121 |
+
"step": 80
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.07365684575389948,
|
125 |
+
"grad_norm": 0.09298353642225266,
|
126 |
+
"learning_rate": 1.4166666666666668e-05,
|
127 |
+
"loss": 0.9676,
|
128 |
+
"step": 85
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.0779896013864818,
|
132 |
+
"grad_norm": 0.09433547407388687,
|
133 |
+
"learning_rate": 1.5e-05,
|
134 |
+
"loss": 0.9463,
|
135 |
+
"step": 90
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.08232235701906412,
|
139 |
+
"grad_norm": 0.09255563467741013,
|
140 |
+
"learning_rate": 1.5833333333333333e-05,
|
141 |
+
"loss": 0.9341,
|
142 |
+
"step": 95
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.08665511265164645,
|
146 |
+
"grad_norm": 0.09951213002204895,
|
147 |
+
"learning_rate": 1.6666666666666667e-05,
|
148 |
+
"loss": 0.9209,
|
149 |
+
"step": 100
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.09098786828422877,
|
153 |
+
"grad_norm": 0.10951012372970581,
|
154 |
+
"learning_rate": 1.75e-05,
|
155 |
+
"loss": 0.8927,
|
156 |
+
"step": 105
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.09532062391681109,
|
160 |
+
"grad_norm": 0.110866479575634,
|
161 |
+
"learning_rate": 1.8333333333333333e-05,
|
162 |
+
"loss": 0.8755,
|
163 |
+
"step": 110
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.09965337954939342,
|
167 |
+
"grad_norm": 0.11648208647966385,
|
168 |
+
"learning_rate": 1.9166666666666667e-05,
|
169 |
+
"loss": 0.8453,
|
170 |
+
"step": 115
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.10398613518197573,
|
174 |
+
"grad_norm": 0.12016862630844116,
|
175 |
+
"learning_rate": 2e-05,
|
176 |
+
"loss": 0.8081,
|
177 |
+
"step": 120
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.10831889081455806,
|
181 |
+
"grad_norm": 0.12387488037347794,
|
182 |
+
"learning_rate": 2.0833333333333336e-05,
|
183 |
+
"loss": 0.7784,
|
184 |
+
"step": 125
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.11265164644714037,
|
188 |
+
"grad_norm": 0.12779255211353302,
|
189 |
+
"learning_rate": 2.1666666666666667e-05,
|
190 |
+
"loss": 0.7353,
|
191 |
+
"step": 130
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.1169844020797227,
|
195 |
+
"grad_norm": 0.12649372220039368,
|
196 |
+
"learning_rate": 2.25e-05,
|
197 |
+
"loss": 0.7085,
|
198 |
+
"step": 135
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.12131715771230503,
|
202 |
+
"grad_norm": 0.1445430964231491,
|
203 |
+
"learning_rate": 2.3333333333333336e-05,
|
204 |
+
"loss": 0.6753,
|
205 |
+
"step": 140
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.12564991334488734,
|
209 |
+
"grad_norm": 0.1329505294561386,
|
210 |
+
"learning_rate": 2.4166666666666667e-05,
|
211 |
+
"loss": 0.6448,
|
212 |
+
"step": 145
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.12998266897746968,
|
216 |
+
"grad_norm": 0.13544394075870514,
|
217 |
+
"learning_rate": 2.5e-05,
|
218 |
+
"loss": 0.621,
|
219 |
+
"step": 150
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.134315424610052,
|
223 |
+
"grad_norm": 0.1351090520620346,
|
224 |
+
"learning_rate": 2.5833333333333336e-05,
|
225 |
+
"loss": 0.5997,
|
226 |
+
"step": 155
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.1386481802426343,
|
230 |
+
"grad_norm": 0.127303346991539,
|
231 |
+
"learning_rate": 2.6666666666666667e-05,
|
232 |
+
"loss": 0.5791,
|
233 |
+
"step": 160
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.14298093587521662,
|
237 |
+
"grad_norm": 0.1476767510175705,
|
238 |
+
"learning_rate": 2.7500000000000004e-05,
|
239 |
+
"loss": 0.57,
|
240 |
+
"step": 165
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.14731369150779897,
|
244 |
+
"grad_norm": 0.13856437802314758,
|
245 |
+
"learning_rate": 2.8333333333333335e-05,
|
246 |
+
"loss": 0.5645,
|
247 |
+
"step": 170
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.15164644714038128,
|
251 |
+
"grad_norm": 0.1533508449792862,
|
252 |
+
"learning_rate": 2.916666666666667e-05,
|
253 |
+
"loss": 0.5583,
|
254 |
+
"step": 175
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.1559792027729636,
|
258 |
+
"grad_norm": 0.13325001299381256,
|
259 |
+
"learning_rate": 3e-05,
|
260 |
+
"loss": 0.555,
|
261 |
+
"step": 180
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.16031195840554593,
|
265 |
+
"grad_norm": 0.13416974246501923,
|
266 |
+
"learning_rate": 3.0833333333333335e-05,
|
267 |
+
"loss": 0.5439,
|
268 |
+
"step": 185
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.16464471403812825,
|
272 |
+
"grad_norm": 0.1278882771730423,
|
273 |
+
"learning_rate": 3.1666666666666666e-05,
|
274 |
+
"loss": 0.537,
|
275 |
+
"step": 190
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.16897746967071056,
|
279 |
+
"grad_norm": 0.14047101140022278,
|
280 |
+
"learning_rate": 3.2500000000000004e-05,
|
281 |
+
"loss": 0.5381,
|
282 |
+
"step": 195
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.1733102253032929,
|
286 |
+
"grad_norm": 0.13340455293655396,
|
287 |
+
"learning_rate": 3.3333333333333335e-05,
|
288 |
+
"loss": 0.5344,
|
289 |
+
"step": 200
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.17764298093587522,
|
293 |
+
"grad_norm": 0.13049094378948212,
|
294 |
+
"learning_rate": 3.4166666666666666e-05,
|
295 |
+
"loss": 0.5291,
|
296 |
+
"step": 205
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.18197573656845753,
|
300 |
+
"grad_norm": 0.16296444833278656,
|
301 |
+
"learning_rate": 3.5e-05,
|
302 |
+
"loss": 0.5342,
|
303 |
+
"step": 210
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.18630849220103987,
|
307 |
+
"grad_norm": 0.1682613343000412,
|
308 |
+
"learning_rate": 3.5833333333333335e-05,
|
309 |
+
"loss": 0.5285,
|
310 |
+
"step": 215
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.19064124783362218,
|
314 |
+
"grad_norm": 0.1439386010169983,
|
315 |
+
"learning_rate": 3.6666666666666666e-05,
|
316 |
+
"loss": 0.5268,
|
317 |
+
"step": 220
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.1949740034662045,
|
321 |
+
"grad_norm": 0.15248768031597137,
|
322 |
+
"learning_rate": 3.7500000000000003e-05,
|
323 |
+
"loss": 0.5252,
|
324 |
+
"step": 225
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.19930675909878684,
|
328 |
+
"grad_norm": 0.1604214906692505,
|
329 |
+
"learning_rate": 3.8333333333333334e-05,
|
330 |
+
"loss": 0.5201,
|
331 |
+
"step": 230
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.20363951473136915,
|
335 |
+
"grad_norm": 0.16192543506622314,
|
336 |
+
"learning_rate": 3.9166666666666665e-05,
|
337 |
+
"loss": 0.5222,
|
338 |
+
"step": 235
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.20797227036395147,
|
342 |
+
"grad_norm": 0.14945088326931,
|
343 |
+
"learning_rate": 4e-05,
|
344 |
+
"loss": 0.5158,
|
345 |
+
"step": 240
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.2123050259965338,
|
349 |
+
"grad_norm": 0.16204427182674408,
|
350 |
+
"learning_rate": 4.0833333333333334e-05,
|
351 |
+
"loss": 0.5127,
|
352 |
+
"step": 245
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.21663778162911612,
|
356 |
+
"grad_norm": 0.1618306040763855,
|
357 |
+
"learning_rate": 4.166666666666667e-05,
|
358 |
+
"loss": 0.5181,
|
359 |
+
"step": 250
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.22097053726169844,
|
363 |
+
"grad_norm": 0.14193885028362274,
|
364 |
+
"learning_rate": 4.25e-05,
|
365 |
+
"loss": 0.5164,
|
366 |
+
"step": 255
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.22530329289428075,
|
370 |
+
"grad_norm": 0.12552691996097565,
|
371 |
+
"learning_rate": 4.3333333333333334e-05,
|
372 |
+
"loss": 0.5149,
|
373 |
+
"step": 260
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.2296360485268631,
|
377 |
+
"grad_norm": 0.14471225440502167,
|
378 |
+
"learning_rate": 4.4166666666666665e-05,
|
379 |
+
"loss": 0.5137,
|
380 |
+
"step": 265
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.2339688041594454,
|
384 |
+
"grad_norm": 0.13988590240478516,
|
385 |
+
"learning_rate": 4.5e-05,
|
386 |
+
"loss": 0.5066,
|
387 |
+
"step": 270
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.23830155979202772,
|
391 |
+
"grad_norm": 0.13964875042438507,
|
392 |
+
"learning_rate": 4.5833333333333334e-05,
|
393 |
+
"loss": 0.5116,
|
394 |
+
"step": 275
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.24263431542461006,
|
398 |
+
"grad_norm": 0.12847208976745605,
|
399 |
+
"learning_rate": 4.666666666666667e-05,
|
400 |
+
"loss": 0.5095,
|
401 |
+
"step": 280
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.24696707105719237,
|
405 |
+
"grad_norm": 0.13142219185829163,
|
406 |
+
"learning_rate": 4.75e-05,
|
407 |
+
"loss": 0.5077,
|
408 |
+
"step": 285
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.2512998266897747,
|
412 |
+
"grad_norm": 0.19098567962646484,
|
413 |
+
"learning_rate": 4.8333333333333334e-05,
|
414 |
+
"loss": 0.5099,
|
415 |
+
"step": 290
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.255632582322357,
|
419 |
+
"grad_norm": 0.1430283784866333,
|
420 |
+
"learning_rate": 4.9166666666666665e-05,
|
421 |
+
"loss": 0.5036,
|
422 |
+
"step": 295
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.25996533795493937,
|
426 |
+
"grad_norm": 0.12594453990459442,
|
427 |
+
"learning_rate": 5e-05,
|
428 |
+
"loss": 0.5085,
|
429 |
+
"step": 300
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.26429809358752165,
|
433 |
+
"grad_norm": 0.14377984404563904,
|
434 |
+
"learning_rate": 4.999577115486055e-05,
|
435 |
+
"loss": 0.4978,
|
436 |
+
"step": 305
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.268630849220104,
|
440 |
+
"grad_norm": 0.12468158453702927,
|
441 |
+
"learning_rate": 4.998308605009268e-05,
|
442 |
+
"loss": 0.5095,
|
443 |
+
"step": 310
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.2729636048526863,
|
447 |
+
"grad_norm": 0.12854167819023132,
|
448 |
+
"learning_rate": 4.996194897716389e-05,
|
449 |
+
"loss": 0.5037,
|
450 |
+
"step": 315
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.2772963604852686,
|
454 |
+
"grad_norm": 0.13529527187347412,
|
455 |
+
"learning_rate": 4.993236708690683e-05,
|
456 |
+
"loss": 0.5058,
|
457 |
+
"step": 320
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.28162911611785096,
|
461 |
+
"grad_norm": 0.1377994418144226,
|
462 |
+
"learning_rate": 4.9894350387100126e-05,
|
463 |
+
"loss": 0.4998,
|
464 |
+
"step": 325
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.28596187175043325,
|
468 |
+
"grad_norm": 0.14942322671413422,
|
469 |
+
"learning_rate": 4.984791173908267e-05,
|
470 |
+
"loss": 0.5007,
|
471 |
+
"step": 330
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.2902946273830156,
|
475 |
+
"grad_norm": 0.1366725116968155,
|
476 |
+
"learning_rate": 4.9793066853402536e-05,
|
477 |
+
"loss": 0.5038,
|
478 |
+
"step": 335
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.29462738301559793,
|
482 |
+
"grad_norm": 0.13133087754249573,
|
483 |
+
"learning_rate": 4.9729834284501995e-05,
|
484 |
+
"loss": 0.5062,
|
485 |
+
"step": 340
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.2989601386481802,
|
489 |
+
"grad_norm": 0.11725670844316483,
|
490 |
+
"learning_rate": 4.965823542444037e-05,
|
491 |
+
"loss": 0.5025,
|
492 |
+
"step": 345
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.30329289428076256,
|
496 |
+
"grad_norm": 0.10628046840429306,
|
497 |
+
"learning_rate": 4.9578294495656965e-05,
|
498 |
+
"loss": 0.4999,
|
499 |
+
"step": 350
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.3076256499133449,
|
503 |
+
"grad_norm": 0.13826170563697815,
|
504 |
+
"learning_rate": 4.949003854277644e-05,
|
505 |
+
"loss": 0.4978,
|
506 |
+
"step": 355
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.3119584055459272,
|
510 |
+
"grad_norm": 0.1305851936340332,
|
511 |
+
"learning_rate": 4.9393497423459376e-05,
|
512 |
+
"loss": 0.4997,
|
513 |
+
"step": 360
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.31629116117850953,
|
517 |
+
"grad_norm": 0.11465763300657272,
|
518 |
+
"learning_rate": 4.928870379830124e-05,
|
519 |
+
"loss": 0.5037,
|
520 |
+
"step": 365
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.32062391681109187,
|
524 |
+
"grad_norm": 0.15975706279277802,
|
525 |
+
"learning_rate": 4.9175693119783013e-05,
|
526 |
+
"loss": 0.4982,
|
527 |
+
"step": 370
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.32495667244367415,
|
531 |
+
"grad_norm": 0.15360799431800842,
|
532 |
+
"learning_rate": 4.905450362027738e-05,
|
533 |
+
"loss": 0.5013,
|
534 |
+
"step": 375
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.3292894280762565,
|
538 |
+
"grad_norm": 0.14006198942661285,
|
539 |
+
"learning_rate": 4.8925176299114416e-05,
|
540 |
+
"loss": 0.5008,
|
541 |
+
"step": 380
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.33362218370883884,
|
545 |
+
"grad_norm": 0.14255651831626892,
|
546 |
+
"learning_rate": 4.878775490871121e-05,
|
547 |
+
"loss": 0.4975,
|
548 |
+
"step": 385
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.3379549393414211,
|
552 |
+
"grad_norm": 0.12012791633605957,
|
553 |
+
"learning_rate": 4.864228593977006e-05,
|
554 |
+
"loss": 0.5047,
|
555 |
+
"step": 390
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.34228769497400346,
|
559 |
+
"grad_norm": 0.11433300375938416,
|
560 |
+
"learning_rate": 4.848881860555035e-05,
|
561 |
+
"loss": 0.4986,
|
562 |
+
"step": 395
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.3466204506065858,
|
566 |
+
"grad_norm": 0.11102011054754257,
|
567 |
+
"learning_rate": 4.832740482521931e-05,
|
568 |
+
"loss": 0.4981,
|
569 |
+
"step": 400
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.3509532062391681,
|
573 |
+
"grad_norm": 0.12340573221445084,
|
574 |
+
"learning_rate": 4.815809920628738e-05,
|
575 |
+
"loss": 0.4984,
|
576 |
+
"step": 405
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.35528596187175043,
|
580 |
+
"grad_norm": 0.1148650050163269,
|
581 |
+
"learning_rate": 4.7980959026134044e-05,
|
582 |
+
"loss": 0.4942,
|
583 |
+
"step": 410
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.3596187175043328,
|
587 |
+
"grad_norm": 0.10840712487697601,
|
588 |
+
"learning_rate": 4.7796044212630486e-05,
|
589 |
+
"loss": 0.4903,
|
590 |
+
"step": 415
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.36395147313691506,
|
594 |
+
"grad_norm": 0.11093516647815704,
|
595 |
+
"learning_rate": 4.7603417323865547e-05,
|
596 |
+
"loss": 0.4957,
|
597 |
+
"step": 420
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.3682842287694974,
|
601 |
+
"grad_norm": 0.1237047016620636,
|
602 |
+
"learning_rate": 4.74031435269818e-05,
|
603 |
+
"loss": 0.499,
|
604 |
+
"step": 425
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.37261698440207974,
|
608 |
+
"grad_norm": 0.11613244563341141,
|
609 |
+
"learning_rate": 4.7195290576129034e-05,
|
610 |
+
"loss": 0.4959,
|
611 |
+
"step": 430
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.37694974003466203,
|
615 |
+
"grad_norm": 0.10890854150056839,
|
616 |
+
"learning_rate": 4.697992878954255e-05,
|
617 |
+
"loss": 0.4944,
|
618 |
+
"step": 435
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.38128249566724437,
|
622 |
+
"grad_norm": 0.11364572495222092,
|
623 |
+
"learning_rate": 4.6757131025753886e-05,
|
624 |
+
"loss": 0.4909,
|
625 |
+
"step": 440
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.3856152512998267,
|
629 |
+
"grad_norm": 0.12619757652282715,
|
630 |
+
"learning_rate": 4.652697265894228e-05,
|
631 |
+
"loss": 0.4966,
|
632 |
+
"step": 445
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.389948006932409,
|
636 |
+
"grad_norm": 0.11669816076755524,
|
637 |
+
"learning_rate": 4.628953155343499e-05,
|
638 |
+
"loss": 0.4956,
|
639 |
+
"step": 450
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.39428076256499134,
|
643 |
+
"grad_norm": 0.13808482885360718,
|
644 |
+
"learning_rate": 4.604488803736523e-05,
|
645 |
+
"loss": 0.4973,
|
646 |
+
"step": 455
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.3986135181975737,
|
650 |
+
"grad_norm": 0.11171045899391174,
|
651 |
+
"learning_rate": 4.579312487549649e-05,
|
652 |
+
"loss": 0.4903,
|
653 |
+
"step": 460
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.40294627383015597,
|
657 |
+
"grad_norm": 0.11149395257234573,
|
658 |
+
"learning_rate": 4.553432724122265e-05,
|
659 |
+
"loss": 0.4999,
|
660 |
+
"step": 465
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.4072790294627383,
|
664 |
+
"grad_norm": 0.11308333277702332,
|
665 |
+
"learning_rate": 4.526858268775313e-05,
|
666 |
+
"loss": 0.4967,
|
667 |
+
"step": 470
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.41161178509532065,
|
671 |
+
"grad_norm": 0.10610105097293854,
|
672 |
+
"learning_rate": 4.499598111849299e-05,
|
673 |
+
"loss": 0.4936,
|
674 |
+
"step": 475
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.41594454072790293,
|
678 |
+
"grad_norm": 0.11356962472200394,
|
679 |
+
"learning_rate": 4.471661475662792e-05,
|
680 |
+
"loss": 0.493,
|
681 |
+
"step": 480
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.4202772963604853,
|
685 |
+
"grad_norm": 0.10310888290405273,
|
686 |
+
"learning_rate": 4.443057811392445e-05,
|
687 |
+
"loss": 0.5002,
|
688 |
+
"step": 485
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.4246100519930676,
|
692 |
+
"grad_norm": 0.11358631402254105,
|
693 |
+
"learning_rate": 4.413796795875586e-05,
|
694 |
+
"loss": 0.4983,
|
695 |
+
"step": 490
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.4289428076256499,
|
699 |
+
"grad_norm": 0.12575574219226837,
|
700 |
+
"learning_rate": 4.383888328336476e-05,
|
701 |
+
"loss": 0.4949,
|
702 |
+
"step": 495
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.43327556325823224,
|
706 |
+
"grad_norm": 0.09878399968147278,
|
707 |
+
"learning_rate": 4.3533425270373216e-05,
|
708 |
+
"loss": 0.4953,
|
709 |
+
"step": 500
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.43327556325823224,
|
713 |
+
"eval_loss": 0.4938061535358429,
|
714 |
+
"eval_runtime": 140.4236,
|
715 |
+
"eval_samples_per_second": 2.685,
|
716 |
+
"eval_steps_per_second": 0.677,
|
717 |
+
"step": 500
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.4376083188908146,
|
721 |
+
"grad_norm": 0.11693672835826874,
|
722 |
+
"learning_rate": 4.3221697258551906e-05,
|
723 |
+
"loss": 0.4934,
|
724 |
+
"step": 505
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 0.44194107452339687,
|
728 |
+
"grad_norm": 0.11737816035747528,
|
729 |
+
"learning_rate": 4.2903804707859835e-05,
|
730 |
+
"loss": 0.4934,
|
731 |
+
"step": 510
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 0.4462738301559792,
|
735 |
+
"grad_norm": 0.11542918533086777,
|
736 |
+
"learning_rate": 4.257985516376644e-05,
|
737 |
+
"loss": 0.4927,
|
738 |
+
"step": 515
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.4506065857885615,
|
742 |
+
"grad_norm": 0.10959002375602722,
|
743 |
+
"learning_rate": 4.224995822086812e-05,
|
744 |
+
"loss": 0.4935,
|
745 |
+
"step": 520
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.45493934142114384,
|
749 |
+
"grad_norm": 0.10852424800395966,
|
750 |
+
"learning_rate": 4.191422548581154e-05,
|
751 |
+
"loss": 0.4947,
|
752 |
+
"step": 525
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.4592720970537262,
|
756 |
+
"grad_norm": 0.11614954471588135,
|
757 |
+
"learning_rate": 4.157277053953631e-05,
|
758 |
+
"loss": 0.4935,
|
759 |
+
"step": 530
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.46360485268630847,
|
763 |
+
"grad_norm": 0.11209428310394287,
|
764 |
+
"learning_rate": 4.1225708898849594e-05,
|
765 |
+
"loss": 0.4975,
|
766 |
+
"step": 535
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.4679376083188908,
|
770 |
+
"grad_norm": 0.11714442819356918,
|
771 |
+
"learning_rate": 4.0873157977346e-05,
|
772 |
+
"loss": 0.4923,
|
773 |
+
"step": 540
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.47227036395147315,
|
777 |
+
"grad_norm": 0.1179489716887474,
|
778 |
+
"learning_rate": 4.051523704568557e-05,
|
779 |
+
"loss": 0.4939,
|
780 |
+
"step": 545
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.47660311958405543,
|
784 |
+
"grad_norm": 0.11872310936450958,
|
785 |
+
"learning_rate": 4.0152067191243696e-05,
|
786 |
+
"loss": 0.4927,
|
787 |
+
"step": 550
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.4809358752166378,
|
791 |
+
"grad_norm": 0.1182030588388443,
|
792 |
+
"learning_rate": 3.978377127714628e-05,
|
793 |
+
"loss": 0.4908,
|
794 |
+
"step": 555
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.4852686308492201,
|
798 |
+
"grad_norm": 0.11203285306692123,
|
799 |
+
"learning_rate": 3.941047390070419e-05,
|
800 |
+
"loss": 0.4898,
|
801 |
+
"step": 560
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.4896013864818024,
|
805 |
+
"grad_norm": 0.10341402143239975,
|
806 |
+
"learning_rate": 3.903230135126113e-05,
|
807 |
+
"loss": 0.4899,
|
808 |
+
"step": 565
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.49393414211438474,
|
812 |
+
"grad_norm": 0.11410869657993317,
|
813 |
+
"learning_rate": 3.864938156746891e-05,
|
814 |
+
"loss": 0.4914,
|
815 |
+
"step": 570
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 0.4982668977469671,
|
819 |
+
"grad_norm": 0.11390075832605362,
|
820 |
+
"learning_rate": 3.8261844094004815e-05,
|
821 |
+
"loss": 0.4922,
|
822 |
+
"step": 575
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 0.5025996533795494,
|
826 |
+
"grad_norm": 0.10876569896936417,
|
827 |
+
"learning_rate": 3.7869820037745776e-05,
|
828 |
+
"loss": 0.4964,
|
829 |
+
"step": 580
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 0.5069324090121318,
|
833 |
+
"grad_norm": 0.11602895706892014,
|
834 |
+
"learning_rate": 3.747344202341373e-05,
|
835 |
+
"loss": 0.4945,
|
836 |
+
"step": 585
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 0.511265164644714,
|
840 |
+
"grad_norm": 0.12035666406154633,
|
841 |
+
"learning_rate": 3.707284414870786e-05,
|
842 |
+
"loss": 0.499,
|
843 |
+
"step": 590
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 0.5155979202772963,
|
847 |
+
"grad_norm": 0.1848023533821106,
|
848 |
+
"learning_rate": 3.666816193893817e-05,
|
849 |
+
"loss": 0.4947,
|
850 |
+
"step": 595
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 0.5199306759098787,
|
854 |
+
"grad_norm": 0.12017575651407242,
|
855 |
+
"learning_rate": 3.6259532301176335e-05,
|
856 |
+
"loss": 0.4905,
|
857 |
+
"step": 600
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 0.524263431542461,
|
861 |
+
"grad_norm": 0.10070477426052094,
|
862 |
+
"learning_rate": 3.5847093477938956e-05,
|
863 |
+
"loss": 0.4974,
|
864 |
+
"step": 605
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 0.5285961871750433,
|
868 |
+
"grad_norm": 0.11581992357969284,
|
869 |
+
"learning_rate": 3.543098500041906e-05,
|
870 |
+
"loss": 0.4927,
|
871 |
+
"step": 610
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 0.5329289428076257,
|
875 |
+
"grad_norm": 0.09777580201625824,
|
876 |
+
"learning_rate": 3.501134764128167e-05,
|
877 |
+
"loss": 0.4907,
|
878 |
+
"step": 615
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 0.537261698440208,
|
882 |
+
"grad_norm": 0.10121456533670425,
|
883 |
+
"learning_rate": 3.458832336703929e-05,
|
884 |
+
"loss": 0.491,
|
885 |
+
"step": 620
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 0.5415944540727903,
|
889 |
+
"grad_norm": 0.10312589257955551,
|
890 |
+
"learning_rate": 3.416205529002363e-05,
|
891 |
+
"loss": 0.4942,
|
892 |
+
"step": 625
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 0.5459272097053726,
|
896 |
+
"grad_norm": 0.10299069434404373,
|
897 |
+
"learning_rate": 3.37326876199695e-05,
|
898 |
+
"loss": 0.493,
|
899 |
+
"step": 630
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 0.550259965337955,
|
903 |
+
"grad_norm": 0.11771389842033386,
|
904 |
+
"learning_rate": 3.3300365615227685e-05,
|
905 |
+
"loss": 0.4942,
|
906 |
+
"step": 635
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 0.5545927209705372,
|
910 |
+
"grad_norm": 0.1121087372303009,
|
911 |
+
"learning_rate": 3.286523553362287e-05,
|
912 |
+
"loss": 0.4946,
|
913 |
+
"step": 640
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 0.5589254766031195,
|
917 |
+
"grad_norm": 0.10604743659496307,
|
918 |
+
"learning_rate": 3.242744458297348e-05,
|
919 |
+
"loss": 0.4907,
|
920 |
+
"step": 645
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 0.5632582322357019,
|
924 |
+
"grad_norm": 0.10975582152605057,
|
925 |
+
"learning_rate": 3.1987140871290236e-05,
|
926 |
+
"loss": 0.4889,
|
927 |
+
"step": 650
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 0.5675909878682842,
|
931 |
+
"grad_norm": 0.10636895149946213,
|
932 |
+
"learning_rate": 3.154447335667001e-05,
|
933 |
+
"loss": 0.4935,
|
934 |
+
"step": 655
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 0.5719237435008665,
|
938 |
+
"grad_norm": 0.10076680779457092,
|
939 |
+
"learning_rate": 3.1099591796902215e-05,
|
940 |
+
"loss": 0.4925,
|
941 |
+
"step": 660
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 0.5762564991334489,
|
945 |
+
"grad_norm": 0.10885344445705414,
|
946 |
+
"learning_rate": 3.0652646698804585e-05,
|
947 |
+
"loss": 0.4892,
|
948 |
+
"step": 665
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 0.5805892547660312,
|
952 |
+
"grad_norm": 0.11372784525156021,
|
953 |
+
"learning_rate": 3.0203789267305567e-05,
|
954 |
+
"loss": 0.4918,
|
955 |
+
"step": 670
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 0.5849220103986135,
|
959 |
+
"grad_norm": 0.11742879450321198,
|
960 |
+
"learning_rate": 2.975317135429056e-05,
|
961 |
+
"loss": 0.4885,
|
962 |
+
"step": 675
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 0.5892547660311959,
|
966 |
+
"grad_norm": 0.09908663481473923,
|
967 |
+
"learning_rate": 2.930094540722927e-05,
|
968 |
+
"loss": 0.4889,
|
969 |
+
"step": 680
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.5935875216637782,
|
973 |
+
"grad_norm": 0.11032566428184509,
|
974 |
+
"learning_rate": 2.884726441760155e-05,
|
975 |
+
"loss": 0.4936,
|
976 |
+
"step": 685
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 0.5979202772963604,
|
980 |
+
"grad_norm": 0.11033419519662857,
|
981 |
+
"learning_rate": 2.8392281869139213e-05,
|
982 |
+
"loss": 0.4925,
|
983 |
+
"step": 690
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 0.6022530329289428,
|
987 |
+
"grad_norm": 0.12115441262722015,
|
988 |
+
"learning_rate": 2.7936151685901373e-05,
|
989 |
+
"loss": 0.49,
|
990 |
+
"step": 695
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 0.6065857885615251,
|
994 |
+
"grad_norm": 0.13125091791152954,
|
995 |
+
"learning_rate": 2.747902818020067e-05,
|
996 |
+
"loss": 0.4919,
|
997 |
+
"step": 700
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 0.6109185441941074,
|
1001 |
+
"grad_norm": 0.11565785109996796,
|
1002 |
+
"learning_rate": 2.7021066000398344e-05,
|
1003 |
+
"loss": 0.4928,
|
1004 |
+
"step": 705
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 0.6152512998266898,
|
1008 |
+
"grad_norm": 0.12640348076820374,
|
1009 |
+
"learning_rate": 2.6562420078585433e-05,
|
1010 |
+
"loss": 0.489,
|
1011 |
+
"step": 710
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 0.6195840554592721,
|
1015 |
+
"grad_norm": 0.12297876179218292,
|
1016 |
+
"learning_rate": 2.6103245578168106e-05,
|
1017 |
+
"loss": 0.4919,
|
1018 |
+
"step": 715
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 0.6239168110918544,
|
1022 |
+
"grad_norm": 0.10511300712823868,
|
1023 |
+
"learning_rate": 2.564369784137472e-05,
|
1024 |
+
"loss": 0.4883,
|
1025 |
+
"step": 720
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 0.6282495667244368,
|
1029 |
+
"grad_norm": 0.10803347080945969,
|
1030 |
+
"learning_rate": 2.5183932336702297e-05,
|
1031 |
+
"loss": 0.4909,
|
1032 |
+
"step": 725
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 0.6325823223570191,
|
1036 |
+
"grad_norm": 0.11415662616491318,
|
1037 |
+
"learning_rate": 2.4724104606320445e-05,
|
1038 |
+
"loss": 0.4878,
|
1039 |
+
"step": 730
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 0.6369150779896013,
|
1043 |
+
"grad_norm": 0.10404873639345169,
|
1044 |
+
"learning_rate": 2.426437021345015e-05,
|
1045 |
+
"loss": 0.4901,
|
1046 |
+
"step": 735
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 0.6412478336221837,
|
1050 |
+
"grad_norm": 0.11284226924180984,
|
1051 |
+
"learning_rate": 2.3804884689735642e-05,
|
1052 |
+
"loss": 0.4906,
|
1053 |
+
"step": 740
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 0.645580589254766,
|
1057 |
+
"grad_norm": 0.1138242855668068,
|
1058 |
+
"learning_rate": 2.3345803482626797e-05,
|
1059 |
+
"loss": 0.494,
|
1060 |
+
"step": 745
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 0.6499133448873483,
|
1064 |
+
"grad_norm": 0.18223117291927338,
|
1065 |
+
"learning_rate": 2.288728190279014e-05,
|
1066 |
+
"loss": 0.4886,
|
1067 |
+
"step": 750
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 0.6542461005199307,
|
1071 |
+
"grad_norm": 0.1119702160358429,
|
1072 |
+
"learning_rate": 2.2429475071565987e-05,
|
1073 |
+
"loss": 0.4908,
|
1074 |
+
"step": 755
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 0.658578856152513,
|
1078 |
+
"grad_norm": 0.11668991297483444,
|
1079 |
+
"learning_rate": 2.1972537868489797e-05,
|
1080 |
+
"loss": 0.4896,
|
1081 |
+
"step": 760
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 0.6629116117850953,
|
1085 |
+
"grad_norm": 0.10587523132562637,
|
1086 |
+
"learning_rate": 2.151662487889518e-05,
|
1087 |
+
"loss": 0.4885,
|
1088 |
+
"step": 765
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 0.6672443674176777,
|
1092 |
+
"grad_norm": 0.10824766755104065,
|
1093 |
+
"learning_rate": 2.1061890341616558e-05,
|
1094 |
+
"loss": 0.4873,
|
1095 |
+
"step": 770
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 0.67157712305026,
|
1099 |
+
"grad_norm": 0.11111487448215485,
|
1100 |
+
"learning_rate": 2.060848809680893e-05,
|
1101 |
+
"loss": 0.4853,
|
1102 |
+
"step": 775
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 0.6759098786828422,
|
1106 |
+
"grad_norm": 0.10754924267530441,
|
1107 |
+
"learning_rate": 2.0156571533902627e-05,
|
1108 |
+
"loss": 0.4849,
|
1109 |
+
"step": 780
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 0.6802426343154246,
|
1113 |
+
"grad_norm": 0.10407605022192001,
|
1114 |
+
"learning_rate": 1.97062935397105e-05,
|
1115 |
+
"loss": 0.488,
|
1116 |
+
"step": 785
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 0.6845753899480069,
|
1120 |
+
"grad_norm": 0.1077013909816742,
|
1121 |
+
"learning_rate": 1.9257806446705116e-05,
|
1122 |
+
"loss": 0.4867,
|
1123 |
+
"step": 790
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 0.6889081455805892,
|
1127 |
+
"grad_norm": 0.11459195613861084,
|
1128 |
+
"learning_rate": 1.8811261981483548e-05,
|
1129 |
+
"loss": 0.4911,
|
1130 |
+
"step": 795
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 0.6932409012131716,
|
1134 |
+
"grad_norm": 0.12345458567142487,
|
1135 |
+
"learning_rate": 1.8366811213437092e-05,
|
1136 |
+
"loss": 0.4888,
|
1137 |
+
"step": 800
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 0.6975736568457539,
|
1141 |
+
"grad_norm": 0.10806908458471298,
|
1142 |
+
"learning_rate": 1.7924604503643367e-05,
|
1143 |
+
"loss": 0.4903,
|
1144 |
+
"step": 805
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 0.7019064124783362,
|
1148 |
+
"grad_norm": 0.12911449372768402,
|
1149 |
+
"learning_rate": 1.7484791453998006e-05,
|
1150 |
+
"loss": 0.4874,
|
1151 |
+
"step": 810
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 0.7062391681109186,
|
1155 |
+
"grad_norm": 0.11117111146450043,
|
1156 |
+
"learning_rate": 1.7047520856603183e-05,
|
1157 |
+
"loss": 0.4895,
|
1158 |
+
"step": 815
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 0.7105719237435009,
|
1162 |
+
"grad_norm": 0.10993503779172897,
|
1163 |
+
"learning_rate": 1.6612940643430138e-05,
|
1164 |
+
"loss": 0.4909,
|
1165 |
+
"step": 820
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 0.7149046793760832,
|
1169 |
+
"grad_norm": 0.10971593856811523,
|
1170 |
+
"learning_rate": 1.618119783627263e-05,
|
1171 |
+
"loss": 0.4898,
|
1172 |
+
"step": 825
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 0.7192374350086655,
|
1176 |
+
"grad_norm": 0.12124701589345932,
|
1177 |
+
"learning_rate": 1.5752438497008405e-05,
|
1178 |
+
"loss": 0.4886,
|
1179 |
+
"step": 830
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 0.7235701906412478,
|
1183 |
+
"grad_norm": 0.10207706689834595,
|
1184 |
+
"learning_rate": 1.5326807678185267e-05,
|
1185 |
+
"loss": 0.4893,
|
1186 |
+
"step": 835
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 0.7279029462738301,
|
1190 |
+
"grad_norm": 0.10885774344205856,
|
1191 |
+
"learning_rate": 1.490444937394879e-05,
|
1192 |
+
"loss": 0.4891,
|
1193 |
+
"step": 840
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 0.7322357019064125,
|
1197 |
+
"grad_norm": 0.10901923477649689,
|
1198 |
+
"learning_rate": 1.4485506471327914e-05,
|
1199 |
+
"loss": 0.4968,
|
1200 |
+
"step": 845
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 0.7365684575389948,
|
1204 |
+
"grad_norm": 0.11790922284126282,
|
1205 |
+
"learning_rate": 1.407012070189524e-05,
|
1206 |
+
"loss": 0.4861,
|
1207 |
+
"step": 850
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 0.7409012131715771,
|
1211 |
+
"grad_norm": 0.11575620621442795,
|
1212 |
+
"learning_rate": 1.3658432593818149e-05,
|
1213 |
+
"loss": 0.4888,
|
1214 |
+
"step": 855
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 0.7452339688041595,
|
1218 |
+
"grad_norm": 0.11284902691841125,
|
1219 |
+
"learning_rate": 1.325058142431701e-05,
|
1220 |
+
"loss": 0.4879,
|
1221 |
+
"step": 860
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 0.7495667244367418,
|
1225 |
+
"grad_norm": 0.11264630407094955,
|
1226 |
+
"learning_rate": 1.2846705172546675e-05,
|
1227 |
+
"loss": 0.4909,
|
1228 |
+
"step": 865
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 0.7538994800693241,
|
1232 |
+
"grad_norm": 0.11645620316267014,
|
1233 |
+
"learning_rate": 1.2446940472917099e-05,
|
1234 |
+
"loss": 0.485,
|
1235 |
+
"step": 870
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 0.7582322357019065,
|
1239 |
+
"grad_norm": 0.10355979204177856,
|
1240 |
+
"learning_rate": 1.2051422568868833e-05,
|
1241 |
+
"loss": 0.4891,
|
1242 |
+
"step": 875
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 0.7625649913344887,
|
1246 |
+
"grad_norm": 0.10600101947784424,
|
1247 |
+
"learning_rate": 1.1660285267119167e-05,
|
1248 |
+
"loss": 0.4903,
|
1249 |
+
"step": 880
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 0.766897746967071,
|
1253 |
+
"grad_norm": 0.11851372569799423,
|
1254 |
+
"learning_rate": 1.1273660892394297e-05,
|
1255 |
+
"loss": 0.4895,
|
1256 |
+
"step": 885
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 0.7712305025996534,
|
1260 |
+
"grad_norm": 0.11533911526203156,
|
1261 |
+
"learning_rate": 1.0891680242662835e-05,
|
1262 |
+
"loss": 0.4843,
|
1263 |
+
"step": 890
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 0.7755632582322357,
|
1267 |
+
"grad_norm": 0.09746929258108139,
|
1268 |
+
"learning_rate": 1.051447254488591e-05,
|
1269 |
+
"loss": 0.4894,
|
1270 |
+
"step": 895
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 0.779896013864818,
|
1274 |
+
"grad_norm": 0.1193188726902008,
|
1275 |
+
"learning_rate": 1.0142165411298662e-05,
|
1276 |
+
"loss": 0.4884,
|
1277 |
+
"step": 900
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 0.7842287694974004,
|
1281 |
+
"grad_norm": 0.11370085179805756,
|
1282 |
+
"learning_rate": 9.774884796238085e-06,
|
1283 |
+
"loss": 0.49,
|
1284 |
+
"step": 905
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 0.7885615251299827,
|
1288 |
+
"grad_norm": 0.1011599749326706,
|
1289 |
+
"learning_rate": 9.412754953531663e-06,
|
1290 |
+
"loss": 0.4884,
|
1291 |
+
"step": 910
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 0.792894280762565,
|
1295 |
+
"grad_norm": 0.11001604050397873,
|
1296 |
+
"learning_rate": 9.055898394461423e-06,
|
1297 |
+
"loss": 0.4883,
|
1298 |
+
"step": 915
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 0.7972270363951474,
|
1302 |
+
"grad_norm": 0.11630392074584961,
|
1303 |
+
"learning_rate": 8.704435846317386e-06,
|
1304 |
+
"loss": 0.4869,
|
1305 |
+
"step": 920
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 0.8015597920277296,
|
1309 |
+
"grad_norm": 0.10578346997499466,
|
1310 |
+
"learning_rate": 8.358486211554637e-06,
|
1311 |
+
"loss": 0.4887,
|
1312 |
+
"step": 925
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 0.8058925476603119,
|
1316 |
+
"grad_norm": 0.11845128238201141,
|
1317 |
+
"learning_rate": 8.018166527567672e-06,
|
1318 |
+
"loss": 0.4852,
|
1319 |
+
"step": 930
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 0.8102253032928943,
|
1323 |
+
"grad_norm": 0.10032304376363754,
|
1324 |
+
"learning_rate": 7.683591927095824e-06,
|
1325 |
+
"loss": 0.4865,
|
1326 |
+
"step": 935
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 0.8145580589254766,
|
1330 |
+
"grad_norm": 0.10029744356870651,
|
1331 |
+
"learning_rate": 7.354875599272928e-06,
|
1332 |
+
"loss": 0.4902,
|
1333 |
+
"step": 940
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 0.8188908145580589,
|
1337 |
+
"grad_norm": 0.10729606449604034,
|
1338 |
+
"learning_rate": 7.0321287513346074e-06,
|
1339 |
+
"loss": 0.4879,
|
1340 |
+
"step": 945
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 0.8232235701906413,
|
1344 |
+
"grad_norm": 0.11077585816383362,
|
1345 |
+
"learning_rate": 6.715460570995988e-06,
|
1346 |
+
"loss": 0.4905,
|
1347 |
+
"step": 950
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 0.8275563258232236,
|
1351 |
+
"grad_norm": 0.112645223736763,
|
1352 |
+
"learning_rate": 6.404978189512681e-06,
|
1353 |
+
"loss": 0.492,
|
1354 |
+
"step": 955
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 0.8318890814558059,
|
1358 |
+
"grad_norm": 0.11094575375318527,
|
1359 |
+
"learning_rate": 6.100786645437481e-06,
|
1360 |
+
"loss": 0.4855,
|
1361 |
+
"step": 960
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 0.8362218370883883,
|
1365 |
+
"grad_norm": 0.10790159553289413,
|
1366 |
+
"learning_rate": 5.8029888490850005e-06,
|
1367 |
+
"loss": 0.4942,
|
1368 |
+
"step": 965
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 0.8405545927209706,
|
1372 |
+
"grad_norm": 0.11253953725099564,
|
1373 |
+
"learning_rate": 5.511685547716328e-06,
|
1374 |
+
"loss": 0.4829,
|
1375 |
+
"step": 970
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 0.8448873483535528,
|
1379 |
+
"grad_norm": 0.11395300179719925,
|
1380 |
+
"learning_rate": 5.226975291455477e-06,
|
1381 |
+
"loss": 0.4856,
|
1382 |
+
"step": 975
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 0.8492201039861352,
|
1386 |
+
"grad_norm": 0.1128508597612381,
|
1387 |
+
"learning_rate": 4.9489543999491045e-06,
|
1388 |
+
"loss": 0.4852,
|
1389 |
+
"step": 980
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 0.8535528596187175,
|
1393 |
+
"grad_norm": 0.11124628782272339,
|
1394 |
+
"learning_rate": 4.67771692978087e-06,
|
1395 |
+
"loss": 0.4885,
|
1396 |
+
"step": 985
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 0.8578856152512998,
|
1400 |
+
"grad_norm": 0.10993985086679459,
|
1401 |
+
"learning_rate": 4.413354642651369e-06,
|
1402 |
+
"loss": 0.4867,
|
1403 |
+
"step": 990
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 0.8622183708838822,
|
1407 |
+
"grad_norm": 0.1036507710814476,
|
1408 |
+
"learning_rate": 4.1559569743344405e-06,
|
1409 |
+
"loss": 0.4839,
|
1410 |
+
"step": 995
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 0.8665511265164645,
|
1414 |
+
"grad_norm": 0.10577788203954697,
|
1415 |
+
"learning_rate": 3.90561100442036e-06,
|
1416 |
+
"loss": 0.4909,
|
1417 |
+
"step": 1000
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 0.8665511265164645,
|
1421 |
+
"eval_loss": 0.48741263151168823,
|
1422 |
+
"eval_runtime": 139.1562,
|
1423 |
+
"eval_samples_per_second": 2.709,
|
1424 |
+
"eval_steps_per_second": 0.683,
|
1425 |
+
"step": 1000
|
1426 |
+
},
|
1427 |
+
{
|
1428 |
+
"epoch": 0.8708838821490468,
|
1429 |
+
"grad_norm": 0.10687436163425446,
|
1430 |
+
"learning_rate": 3.662401426856177e-06,
|
1431 |
+
"loss": 0.4912,
|
1432 |
+
"step": 1005
|
1433 |
+
},
|
1434 |
+
{
|
1435 |
+
"epoch": 0.8752166377816292,
|
1436 |
+
"grad_norm": 0.10468524694442749,
|
1437 |
+
"learning_rate": 3.4264105212930915e-06,
|
1438 |
+
"loss": 0.4882,
|
1439 |
+
"step": 1010
|
1440 |
+
},
|
1441 |
+
{
|
1442 |
+
"epoch": 0.8795493934142115,
|
1443 |
+
"grad_norm": 0.11409106850624084,
|
1444 |
+
"learning_rate": 3.197718125250618e-06,
|
1445 |
+
"loss": 0.4834,
|
1446 |
+
"step": 1015
|
1447 |
+
},
|
1448 |
+
{
|
1449 |
+
"epoch": 0.8838821490467937,
|
1450 |
+
"grad_norm": 0.1069907397031784,
|
1451 |
+
"learning_rate": 2.9764016071069434e-06,
|
1452 |
+
"loss": 0.4881,
|
1453 |
+
"step": 1020
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"epoch": 0.8882149046793761,
|
1457 |
+
"grad_norm": 0.10976656526327133,
|
1458 |
+
"learning_rate": 2.7625358399246376e-06,
|
1459 |
+
"loss": 0.4852,
|
1460 |
+
"step": 1025
|
1461 |
+
},
|
1462 |
+
{
|
1463 |
+
"epoch": 0.8925476603119584,
|
1464 |
+
"grad_norm": 0.11071466654539108,
|
1465 |
+
"learning_rate": 2.5561931761205082e-06,
|
1466 |
+
"loss": 0.4879,
|
1467 |
+
"step": 1030
|
1468 |
+
},
|
1469 |
+
{
|
1470 |
+
"epoch": 0.8968804159445407,
|
1471 |
+
"grad_norm": 0.10794007033109665,
|
1472 |
+
"learning_rate": 2.3574434229882145e-06,
|
1473 |
+
"loss": 0.4865,
|
1474 |
+
"step": 1035
|
1475 |
+
},
|
1476 |
+
{
|
1477 |
+
"epoch": 0.901213171577123,
|
1478 |
+
"grad_norm": 0.11333485692739487,
|
1479 |
+
"learning_rate": 2.166353819081968e-06,
|
1480 |
+
"loss": 0.4855,
|
1481 |
+
"step": 1040
|
1482 |
+
},
|
1483 |
+
{
|
1484 |
+
"epoch": 0.9055459272097054,
|
1485 |
+
"grad_norm": 0.10130605846643448,
|
1486 |
+
"learning_rate": 1.982989011469172e-06,
|
1487 |
+
"loss": 0.4868,
|
1488 |
+
"step": 1045
|
1489 |
+
},
|
1490 |
+
{
|
1491 |
+
"epoch": 0.9098786828422877,
|
1492 |
+
"grad_norm": 0.10454142093658447,
|
1493 |
+
"learning_rate": 1.8074110338598682e-06,
|
1494 |
+
"loss": 0.487,
|
1495 |
+
"step": 1050
|
1496 |
+
},
|
1497 |
+
{
|
1498 |
+
"epoch": 0.91421143847487,
|
1499 |
+
"grad_norm": 0.10157745331525803,
|
1500 |
+
"learning_rate": 1.6396792856202298e-06,
|
1501 |
+
"loss": 0.4881,
|
1502 |
+
"step": 1055
|
1503 |
+
},
|
1504 |
+
{
|
1505 |
+
"epoch": 0.9185441941074524,
|
1506 |
+
"grad_norm": 0.11562693864107132,
|
1507 |
+
"learning_rate": 1.479850511677322e-06,
|
1508 |
+
"loss": 0.4877,
|
1509 |
+
"step": 1060
|
1510 |
+
},
|
1511 |
+
{
|
1512 |
+
"epoch": 0.9228769497400346,
|
1513 |
+
"grad_norm": 0.12777337431907654,
|
1514 |
+
"learning_rate": 1.3279787833218537e-06,
|
1515 |
+
"loss": 0.4886,
|
1516 |
+
"step": 1065
|
1517 |
+
},
|
1518 |
+
{
|
1519 |
+
"epoch": 0.9272097053726169,
|
1520 |
+
"grad_norm": 0.10074026137590408,
|
1521 |
+
"learning_rate": 1.1841154799154374e-06,
|
1522 |
+
"loss": 0.4856,
|
1523 |
+
"step": 1070
|
1524 |
+
},
|
1525 |
+
{
|
1526 |
+
"epoch": 0.9315424610051993,
|
1527 |
+
"grad_norm": 0.11187005788087845,
|
1528 |
+
"learning_rate": 1.0483092715085879e-06,
|
1529 |
+
"loss": 0.4872,
|
1530 |
+
"step": 1075
|
1531 |
+
},
|
1532 |
+
{
|
1533 |
+
"epoch": 0.9358752166377816,
|
1534 |
+
"grad_norm": 0.10271576046943665,
|
1535 |
+
"learning_rate": 9.206061023752516e-07,
|
1536 |
+
"loss": 0.4867,
|
1537 |
+
"step": 1080
|
1538 |
+
},
|
1539 |
+
{
|
1540 |
+
"epoch": 0.9402079722703639,
|
1541 |
+
"grad_norm": 0.11686773598194122,
|
1542 |
+
"learning_rate": 8.010491754695177e-07,
|
1543 |
+
"loss": 0.4868,
|
1544 |
+
"step": 1085
|
1545 |
+
},
|
1546 |
+
{
|
1547 |
+
"epoch": 0.9445407279029463,
|
1548 |
+
"grad_norm": 0.12068697810173035,
|
1549 |
+
"learning_rate": 6.896789378097179e-07,
|
1550 |
+
"loss": 0.4911,
|
1551 |
+
"step": 1090
|
1552 |
+
},
|
1553 |
+
{
|
1554 |
+
"epoch": 0.9488734835355286,
|
1555 |
+
"grad_norm": 0.10469616949558258,
|
1556 |
+
"learning_rate": 5.865330667949115e-07,
|
1557 |
+
"loss": 0.485,
|
1558 |
+
"step": 1095
|
1559 |
+
},
|
1560 |
+
{
|
1561 |
+
"epoch": 0.9532062391681109,
|
1562 |
+
"grad_norm": 0.11302938312292099,
|
1563 |
+
"learning_rate": 4.916464574583251e-07,
|
1564 |
+
"loss": 0.4864,
|
1565 |
+
"step": 1100
|
1566 |
+
},
|
1567 |
+
{
|
1568 |
+
"epoch": 0.9575389948006933,
|
1569 |
+
"grad_norm": 0.11449886858463287,
|
1570 |
+
"learning_rate": 4.050512106620913e-07,
|
1571 |
+
"loss": 0.4873,
|
1572 |
+
"step": 1105
|
1573 |
+
},
|
1574 |
+
{
|
1575 |
+
"epoch": 0.9618717504332756,
|
1576 |
+
"grad_norm": 0.1062023788690567,
|
1577 |
+
"learning_rate": 3.267766222372931e-07,
|
1578 |
+
"loss": 0.4885,
|
1579 |
+
"step": 1110
|
1580 |
+
},
|
1581 |
+
{
|
1582 |
+
"epoch": 0.9662045060658578,
|
1583 |
+
"grad_norm": 0.11473377794027328,
|
1584 |
+
"learning_rate": 2.568491730729539e-07,
|
1585 |
+
"loss": 0.4869,
|
1586 |
+
"step": 1115
|
1587 |
+
},
|
1588 |
+
{
|
1589 |
+
"epoch": 0.9705372616984402,
|
1590 |
+
"grad_norm": 0.10777969658374786,
|
1591 |
+
"learning_rate": 1.95292520157353e-07,
|
1592 |
+
"loss": 0.489,
|
1593 |
+
"step": 1120
|
1594 |
+
},
|
1595 |
+
{
|
1596 |
+
"epoch": 0.9748700173310225,
|
1597 |
+
"grad_norm": 0.10438723862171173,
|
1598 |
+
"learning_rate": 1.4212748857468926e-07,
|
1599 |
+
"loss": 0.4871,
|
1600 |
+
"step": 1125
|
1601 |
+
},
|
1602 |
+
{
|
1603 |
+
"epoch": 0.9792027729636048,
|
1604 |
+
"grad_norm": 0.10945618152618408,
|
1605 |
+
"learning_rate": 9.737206445979696e-08,
|
1606 |
+
"loss": 0.4875,
|
1607 |
+
"step": 1130
|
1608 |
+
},
|
1609 |
+
{
|
1610 |
+
"epoch": 0.9835355285961872,
|
1611 |
+
"grad_norm": 0.10462047904729843,
|
1612 |
+
"learning_rate": 6.104138891329659e-08,
|
1613 |
+
"loss": 0.4895,
|
1614 |
+
"step": 1135
|
1615 |
+
},
|
1616 |
+
{
|
1617 |
+
"epoch": 0.9878682842287695,
|
1618 |
+
"grad_norm": 0.1181010752916336,
|
1619 |
+
"learning_rate": 3.314775287923677e-08,
|
1620 |
+
"loss": 0.4857,
|
1621 |
+
"step": 1140
|
1622 |
+
},
|
1623 |
+
{
|
1624 |
+
"epoch": 0.9922010398613518,
|
1625 |
+
"grad_norm": 0.10257267206907272,
|
1626 |
+
"learning_rate": 1.3700592986998217e-08,
|
1627 |
+
"loss": 0.4866,
|
1628 |
+
"step": 1145
|
1629 |
+
},
|
1630 |
+
{
|
1631 |
+
"epoch": 0.9965337954939342,
|
1632 |
+
"grad_norm": 0.10488082468509674,
|
1633 |
+
"learning_rate": 2.7064883587807345e-09,
|
1634 |
+
"loss": 0.4893,
|
1635 |
+
"step": 1150
|
1636 |
+
},
|
1637 |
+
{
|
1638 |
+
"epoch": 1.0,
|
1639 |
+
"step": 1154,
|
1640 |
+
"total_flos": 8.066906520991826e+17,
|
1641 |
+
"train_loss": 0.5547944433354423,
|
1642 |
+
"train_runtime": 22221.3478,
|
1643 |
+
"train_samples_per_second": 0.831,
|
1644 |
+
"train_steps_per_second": 0.052
|
1645 |
+
}
|
1646 |
+
],
|
1647 |
+
"logging_steps": 5,
|
1648 |
+
"max_steps": 1154,
|
1649 |
+
"num_input_tokens_seen": 0,
|
1650 |
+
"num_train_epochs": 1,
|
1651 |
+
"save_steps": 500,
|
1652 |
+
"stateful_callbacks": {
|
1653 |
+
"TrainerControl": {
|
1654 |
+
"args": {
|
1655 |
+
"should_epoch_stop": false,
|
1656 |
+
"should_evaluate": false,
|
1657 |
+
"should_log": false,
|
1658 |
+
"should_save": true,
|
1659 |
+
"should_training_stop": true
|
1660 |
+
},
|
1661 |
+
"attributes": {}
|
1662 |
+
}
|
1663 |
+
},
|
1664 |
+
"total_flos": 8.066906520991826e+17,
|
1665 |
+
"train_batch_size": 4,
|
1666 |
+
"trial_name": null,
|
1667 |
+
"trial_params": null
|
1668 |
+
}
|