Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1533 -0
trainer_state.json
ADDED
@@ -0,0 +1,1533 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 19.962616822429908,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2136,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.09345794392523364,
|
13 |
+
"grad_norm": 4.17994499206543,
|
14 |
+
"learning_rate": 1.869158878504673e-05,
|
15 |
+
"loss": 0.9424,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.18691588785046728,
|
20 |
+
"grad_norm": 2.603377342224121,
|
21 |
+
"learning_rate": 3.738317757009346e-05,
|
22 |
+
"loss": 0.4804,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.2803738317757009,
|
27 |
+
"grad_norm": 1.72787606716156,
|
28 |
+
"learning_rate": 5.607476635514019e-05,
|
29 |
+
"loss": 0.2668,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.37383177570093457,
|
34 |
+
"grad_norm": 0.5973343849182129,
|
35 |
+
"learning_rate": 7.476635514018692e-05,
|
36 |
+
"loss": 0.1802,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.4672897196261682,
|
41 |
+
"grad_norm": 1.0159034729003906,
|
42 |
+
"learning_rate": 9.345794392523365e-05,
|
43 |
+
"loss": 0.1275,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.5607476635514018,
|
48 |
+
"grad_norm": 0.8789228200912476,
|
49 |
+
"learning_rate": 0.00011214953271028037,
|
50 |
+
"loss": 0.1168,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.6542056074766355,
|
55 |
+
"grad_norm": 1.1317589282989502,
|
56 |
+
"learning_rate": 0.0001308411214953271,
|
57 |
+
"loss": 0.1075,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.7476635514018691,
|
62 |
+
"grad_norm": 1.391225814819336,
|
63 |
+
"learning_rate": 0.00014953271028037384,
|
64 |
+
"loss": 0.1022,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.8411214953271028,
|
69 |
+
"grad_norm": 1.0744414329528809,
|
70 |
+
"learning_rate": 0.00016822429906542056,
|
71 |
+
"loss": 0.0905,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.9345794392523364,
|
76 |
+
"grad_norm": 0.6361038684844971,
|
77 |
+
"learning_rate": 0.0001869158878504673,
|
78 |
+
"loss": 0.0794,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 1.02803738317757,
|
83 |
+
"grad_norm": 0.6731115579605103,
|
84 |
+
"learning_rate": 0.00019999892118398673,
|
85 |
+
"loss": 0.0823,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 1.1214953271028036,
|
90 |
+
"grad_norm": 0.6741369962692261,
|
91 |
+
"learning_rate": 0.0001999797428801708,
|
92 |
+
"loss": 0.0762,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.2149532710280373,
|
97 |
+
"grad_norm": 0.7907876372337341,
|
98 |
+
"learning_rate": 0.00019993659617930244,
|
99 |
+
"loss": 0.0669,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.308411214953271,
|
104 |
+
"grad_norm": 0.7666932940483093,
|
105 |
+
"learning_rate": 0.00019986949142504959,
|
106 |
+
"loss": 0.0739,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.4018691588785046,
|
111 |
+
"grad_norm": 0.7501648664474487,
|
112 |
+
"learning_rate": 0.00019977844470460494,
|
113 |
+
"loss": 0.0688,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.4953271028037383,
|
118 |
+
"grad_norm": 0.45659929513931274,
|
119 |
+
"learning_rate": 0.00019966347784482945,
|
120 |
+
"loss": 0.0654,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.588785046728972,
|
125 |
+
"grad_norm": 0.6907665729522705,
|
126 |
+
"learning_rate": 0.0001995246184070197,
|
127 |
+
"loss": 0.0613,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.6822429906542056,
|
132 |
+
"grad_norm": 0.47750258445739746,
|
133 |
+
"learning_rate": 0.0001993618996803005,
|
134 |
+
"loss": 0.0536,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.7757009345794392,
|
139 |
+
"grad_norm": 0.7647998332977295,
|
140 |
+
"learning_rate": 0.00019917536067364444,
|
141 |
+
"loss": 0.0579,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.8691588785046729,
|
146 |
+
"grad_norm": 0.5567947030067444,
|
147 |
+
"learning_rate": 0.00019896504610652016,
|
148 |
+
"loss": 0.0548,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.9626168224299065,
|
153 |
+
"grad_norm": 0.46288764476776123,
|
154 |
+
"learning_rate": 0.00019873100639817166,
|
155 |
+
"loss": 0.0527,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 2.05607476635514,
|
160 |
+
"grad_norm": 0.4530458450317383,
|
161 |
+
"learning_rate": 0.00019847329765553115,
|
162 |
+
"loss": 0.0465,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 2.149532710280374,
|
167 |
+
"grad_norm": 0.3930870592594147,
|
168 |
+
"learning_rate": 0.00019819198165976838,
|
169 |
+
"loss": 0.0519,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 2.2429906542056073,
|
174 |
+
"grad_norm": 0.41593921184539795,
|
175 |
+
"learning_rate": 0.00019788712585147976,
|
176 |
+
"loss": 0.0507,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 2.336448598130841,
|
181 |
+
"grad_norm": 0.47823095321655273,
|
182 |
+
"learning_rate": 0.00019755880331452045,
|
183 |
+
"loss": 0.043,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 2.4299065420560746,
|
188 |
+
"grad_norm": 0.316738486289978,
|
189 |
+
"learning_rate": 0.00019720709275848408,
|
190 |
+
"loss": 0.0421,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 2.5233644859813085,
|
195 |
+
"grad_norm": 0.33323222398757935,
|
196 |
+
"learning_rate": 0.00019683207849983326,
|
197 |
+
"loss": 0.0451,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.616822429906542,
|
202 |
+
"grad_norm": 0.43355607986450195,
|
203 |
+
"learning_rate": 0.00019643385044168632,
|
204 |
+
"loss": 0.0402,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 2.710280373831776,
|
209 |
+
"grad_norm": 0.40903353691101074,
|
210 |
+
"learning_rate": 0.0001960125040522645,
|
211 |
+
"loss": 0.0433,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.803738317757009,
|
216 |
+
"grad_norm": 0.3322107493877411,
|
217 |
+
"learning_rate": 0.00019556814034200518,
|
218 |
+
"loss": 0.0415,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 2.897196261682243,
|
223 |
+
"grad_norm": 0.3895340859889984,
|
224 |
+
"learning_rate": 0.00019510086583934643,
|
225 |
+
"loss": 0.0419,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 2.9906542056074765,
|
230 |
+
"grad_norm": 0.39366844296455383,
|
231 |
+
"learning_rate": 0.0001946107925651885,
|
232 |
+
"loss": 0.0358,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 3.0841121495327104,
|
237 |
+
"grad_norm": 0.3918449282646179,
|
238 |
+
"learning_rate": 0.00019409803800603904,
|
239 |
+
"loss": 0.041,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 3.177570093457944,
|
244 |
+
"grad_norm": 0.273848295211792,
|
245 |
+
"learning_rate": 0.00019356272508584753,
|
246 |
+
"loss": 0.0447,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 3.2710280373831777,
|
251 |
+
"grad_norm": 0.4000929296016693,
|
252 |
+
"learning_rate": 0.00019300498213653647,
|
253 |
+
"loss": 0.0434,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 3.364485981308411,
|
258 |
+
"grad_norm": 0.2600472867488861,
|
259 |
+
"learning_rate": 0.00019242494286723612,
|
260 |
+
"loss": 0.0429,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 3.457943925233645,
|
265 |
+
"grad_norm": 0.29327982664108276,
|
266 |
+
"learning_rate": 0.00019182274633223,
|
267 |
+
"loss": 0.0362,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 3.5514018691588785,
|
272 |
+
"grad_norm": 0.35170868039131165,
|
273 |
+
"learning_rate": 0.00019119853689761912,
|
274 |
+
"loss": 0.0342,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 3.6448598130841123,
|
279 |
+
"grad_norm": 0.46894872188568115,
|
280 |
+
"learning_rate": 0.00019055246420671257,
|
281 |
+
"loss": 0.0364,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 3.7383177570093458,
|
286 |
+
"grad_norm": 0.4111644923686981,
|
287 |
+
"learning_rate": 0.0001898846831441533,
|
288 |
+
"loss": 0.0336,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 3.831775700934579,
|
293 |
+
"grad_norm": 0.40484148263931274,
|
294 |
+
"learning_rate": 0.0001891953537987873,
|
295 |
+
"loss": 0.0406,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 3.925233644859813,
|
300 |
+
"grad_norm": 0.29574188590049744,
|
301 |
+
"learning_rate": 0.00018848464142528481,
|
302 |
+
"loss": 0.0344,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 4.018691588785047,
|
307 |
+
"grad_norm": 0.30757296085357666,
|
308 |
+
"learning_rate": 0.00018775271640452377,
|
309 |
+
"loss": 0.0365,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 4.11214953271028,
|
314 |
+
"grad_norm": 0.26074057817459106,
|
315 |
+
"learning_rate": 0.00018699975420274358,
|
316 |
+
"loss": 0.0356,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 4.205607476635514,
|
321 |
+
"grad_norm": 0.3508523404598236,
|
322 |
+
"learning_rate": 0.00018622593532948055,
|
323 |
+
"loss": 0.0325,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 4.299065420560748,
|
328 |
+
"grad_norm": 0.2975287437438965,
|
329 |
+
"learning_rate": 0.0001854314452942935,
|
330 |
+
"loss": 0.0321,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 4.392523364485982,
|
335 |
+
"grad_norm": 0.30913522839546204,
|
336 |
+
"learning_rate": 0.0001846164745622914,
|
337 |
+
"loss": 0.0329,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 4.485981308411215,
|
342 |
+
"grad_norm": 0.30218392610549927,
|
343 |
+
"learning_rate": 0.00018378121850847249,
|
344 |
+
"loss": 0.0344,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 4.579439252336448,
|
349 |
+
"grad_norm": 0.4823051691055298,
|
350 |
+
"learning_rate": 0.00018292587737088665,
|
351 |
+
"loss": 0.0331,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 4.672897196261682,
|
356 |
+
"grad_norm": 0.28159743547439575,
|
357 |
+
"learning_rate": 0.00018205065620263168,
|
358 |
+
"loss": 0.0374,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 4.766355140186916,
|
363 |
+
"grad_norm": 0.36674001812934875,
|
364 |
+
"learning_rate": 0.0001811557648226956,
|
365 |
+
"loss": 0.036,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 4.859813084112149,
|
370 |
+
"grad_norm": 0.31007957458496094,
|
371 |
+
"learning_rate": 0.00018024141776565606,
|
372 |
+
"loss": 0.0321,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 4.953271028037383,
|
377 |
+
"grad_norm": 0.2760983109474182,
|
378 |
+
"learning_rate": 0.0001793078342302495,
|
379 |
+
"loss": 0.0326,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 5.046728971962617,
|
384 |
+
"grad_norm": 0.23232242465019226,
|
385 |
+
"learning_rate": 0.00017835523802682194,
|
386 |
+
"loss": 0.032,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 5.140186915887851,
|
391 |
+
"grad_norm": 0.3053811192512512,
|
392 |
+
"learning_rate": 0.0001773838575236747,
|
393 |
+
"loss": 0.0302,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 5.233644859813084,
|
398 |
+
"grad_norm": 0.2696075737476349,
|
399 |
+
"learning_rate": 0.0001763939255923166,
|
400 |
+
"loss": 0.0325,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 5.327102803738318,
|
405 |
+
"grad_norm": 0.28349894285202026,
|
406 |
+
"learning_rate": 0.00017538567955163755,
|
407 |
+
"loss": 0.0358,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 5.420560747663552,
|
412 |
+
"grad_norm": 0.20442534983158112,
|
413 |
+
"learning_rate": 0.00017435936111101518,
|
414 |
+
"loss": 0.0293,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 5.5140186915887845,
|
419 |
+
"grad_norm": 0.271004855632782,
|
420 |
+
"learning_rate": 0.0001733152163123694,
|
421 |
+
"loss": 0.0275,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 5.607476635514018,
|
426 |
+
"grad_norm": 0.14870816469192505,
|
427 |
+
"learning_rate": 0.00017225349547117798,
|
428 |
+
"loss": 0.0315,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 5.700934579439252,
|
433 |
+
"grad_norm": 0.3340059518814087,
|
434 |
+
"learning_rate": 0.00017117445311646802,
|
435 |
+
"loss": 0.029,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 5.794392523364486,
|
440 |
+
"grad_norm": 0.32217147946357727,
|
441 |
+
"learning_rate": 0.00017007834792979695,
|
442 |
+
"loss": 0.0294,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 5.88785046728972,
|
447 |
+
"grad_norm": 0.32910802960395813,
|
448 |
+
"learning_rate": 0.00016896544268323792,
|
449 |
+
"loss": 0.033,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 5.981308411214953,
|
454 |
+
"grad_norm": 0.5195891261100769,
|
455 |
+
"learning_rate": 0.00016783600417638535,
|
456 |
+
"loss": 0.0334,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 6.074766355140187,
|
461 |
+
"grad_norm": 0.31266453862190247,
|
462 |
+
"learning_rate": 0.0001666903031723937,
|
463 |
+
"loss": 0.0311,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 6.168224299065421,
|
468 |
+
"grad_norm": 0.21037207543849945,
|
469 |
+
"learning_rate": 0.00016552861433306726,
|
470 |
+
"loss": 0.0297,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 6.261682242990654,
|
475 |
+
"grad_norm": 0.20664794743061066,
|
476 |
+
"learning_rate": 0.00016435121615301457,
|
477 |
+
"loss": 0.0286,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 6.355140186915888,
|
482 |
+
"grad_norm": 0.2566857635974884,
|
483 |
+
"learning_rate": 0.00016315839089288411,
|
484 |
+
"loss": 0.0229,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 6.4485981308411215,
|
489 |
+
"grad_norm": 0.2402873933315277,
|
490 |
+
"learning_rate": 0.00016195042451169753,
|
491 |
+
"loss": 0.0264,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 6.542056074766355,
|
496 |
+
"grad_norm": 0.36038416624069214,
|
497 |
+
"learning_rate": 0.00016072760659829562,
|
498 |
+
"loss": 0.0252,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 6.635514018691588,
|
503 |
+
"grad_norm": 0.4320838749408722,
|
504 |
+
"learning_rate": 0.00015949023030191475,
|
505 |
+
"loss": 0.0293,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 6.728971962616822,
|
510 |
+
"grad_norm": 0.26101937890052795,
|
511 |
+
"learning_rate": 0.0001582385922619092,
|
512 |
+
"loss": 0.0286,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 6.822429906542056,
|
517 |
+
"grad_norm": 0.32656329870224,
|
518 |
+
"learning_rate": 0.00015697299253663723,
|
519 |
+
"loss": 0.0294,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 6.91588785046729,
|
524 |
+
"grad_norm": 0.2876569926738739,
|
525 |
+
"learning_rate": 0.00015569373453152728,
|
526 |
+
"loss": 0.0309,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 7.009345794392523,
|
531 |
+
"grad_norm": 0.2590012848377228,
|
532 |
+
"learning_rate": 0.00015440112492634192,
|
533 |
+
"loss": 0.0261,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 7.102803738317757,
|
538 |
+
"grad_norm": 0.37834206223487854,
|
539 |
+
"learning_rate": 0.00015309547360165647,
|
540 |
+
"loss": 0.03,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 7.196261682242991,
|
545 |
+
"grad_norm": 0.27962639927864075,
|
546 |
+
"learning_rate": 0.00015177709356457077,
|
547 |
+
"loss": 0.0297,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 7.289719626168225,
|
552 |
+
"grad_norm": 0.3405240476131439,
|
553 |
+
"learning_rate": 0.00015044630087367118,
|
554 |
+
"loss": 0.0314,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 7.383177570093458,
|
559 |
+
"grad_norm": 0.31650567054748535,
|
560 |
+
"learning_rate": 0.0001491034145632609,
|
561 |
+
"loss": 0.0249,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 7.4766355140186915,
|
566 |
+
"grad_norm": 0.3112686276435852,
|
567 |
+
"learning_rate": 0.00014774875656687727,
|
568 |
+
"loss": 0.0257,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 7.570093457943925,
|
573 |
+
"grad_norm": 0.23344942927360535,
|
574 |
+
"learning_rate": 0.0001463826516401137,
|
575 |
+
"loss": 0.0245,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 7.663551401869158,
|
580 |
+
"grad_norm": 0.1889248788356781,
|
581 |
+
"learning_rate": 0.00014500542728276518,
|
582 |
+
"loss": 0.0229,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 7.757009345794392,
|
587 |
+
"grad_norm": 0.24938005208969116,
|
588 |
+
"learning_rate": 0.00014361741366031603,
|
589 |
+
"loss": 0.0246,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 7.850467289719626,
|
594 |
+
"grad_norm": 0.22689329087734222,
|
595 |
+
"learning_rate": 0.0001422189435247884,
|
596 |
+
"loss": 0.0251,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 7.94392523364486,
|
601 |
+
"grad_norm": 0.28100717067718506,
|
602 |
+
"learning_rate": 0.00014081035213497082,
|
603 |
+
"loss": 0.0277,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 8.037383177570094,
|
608 |
+
"grad_norm": 0.2773335874080658,
|
609 |
+
"learning_rate": 0.0001393919771760457,
|
610 |
+
"loss": 0.027,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 8.130841121495328,
|
615 |
+
"grad_norm": 0.2691657841205597,
|
616 |
+
"learning_rate": 0.00013796415867863528,
|
617 |
+
"loss": 0.0239,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 8.22429906542056,
|
622 |
+
"grad_norm": 0.4514601230621338,
|
623 |
+
"learning_rate": 0.00013652723893728532,
|
624 |
+
"loss": 0.029,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 8.317757009345794,
|
629 |
+
"grad_norm": 0.3693794012069702,
|
630 |
+
"learning_rate": 0.00013508156242840566,
|
631 |
+
"loss": 0.0251,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 8.411214953271028,
|
636 |
+
"grad_norm": 0.2953411042690277,
|
637 |
+
"learning_rate": 0.00013362747572768832,
|
638 |
+
"loss": 0.0284,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 8.504672897196262,
|
643 |
+
"grad_norm": 0.25673815608024597,
|
644 |
+
"learning_rate": 0.0001321653274270219,
|
645 |
+
"loss": 0.0251,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 8.598130841121495,
|
650 |
+
"grad_norm": 0.2374838888645172,
|
651 |
+
"learning_rate": 0.00013069546805092276,
|
652 |
+
"loss": 0.0211,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 8.69158878504673,
|
657 |
+
"grad_norm": 0.2258038967847824,
|
658 |
+
"learning_rate": 0.00012921824997250294,
|
659 |
+
"loss": 0.0235,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 8.785046728971963,
|
664 |
+
"grad_norm": 0.35665759444236755,
|
665 |
+
"learning_rate": 0.00012773402732899481,
|
666 |
+
"loss": 0.0262,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 8.878504672897197,
|
671 |
+
"grad_norm": 0.2746776342391968,
|
672 |
+
"learning_rate": 0.00012624315593685282,
|
673 |
+
"loss": 0.0228,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 8.97196261682243,
|
678 |
+
"grad_norm": 0.1950758695602417,
|
679 |
+
"learning_rate": 0.0001247459932064529,
|
680 |
+
"loss": 0.028,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 9.065420560747663,
|
685 |
+
"grad_norm": 0.25761884450912476,
|
686 |
+
"learning_rate": 0.00012324289805640934,
|
687 |
+
"loss": 0.0256,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 9.158878504672897,
|
692 |
+
"grad_norm": 0.24331067502498627,
|
693 |
+
"learning_rate": 0.00012173423082753041,
|
694 |
+
"loss": 0.0226,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 9.25233644859813,
|
699 |
+
"grad_norm": 0.2336030751466751,
|
700 |
+
"learning_rate": 0.00012022035319643305,
|
701 |
+
"loss": 0.024,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 9.345794392523365,
|
706 |
+
"grad_norm": 0.26760831475257874,
|
707 |
+
"learning_rate": 0.00011870162808883703,
|
708 |
+
"loss": 0.0231,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 9.439252336448599,
|
713 |
+
"grad_norm": 0.22399088740348816,
|
714 |
+
"learning_rate": 0.00011717841959255999,
|
715 |
+
"loss": 0.0212,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 9.532710280373832,
|
720 |
+
"grad_norm": 0.13075126707553864,
|
721 |
+
"learning_rate": 0.0001156510928702337,
|
722 |
+
"loss": 0.027,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 9.626168224299064,
|
727 |
+
"grad_norm": 0.19196636974811554,
|
728 |
+
"learning_rate": 0.00011412001407176265,
|
729 |
+
"loss": 0.0213,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 9.719626168224298,
|
734 |
+
"grad_norm": 0.24825584888458252,
|
735 |
+
"learning_rate": 0.00011258555024654612,
|
736 |
+
"loss": 0.02,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 9.813084112149532,
|
741 |
+
"grad_norm": 0.20732861757278442,
|
742 |
+
"learning_rate": 0.00011104806925548435,
|
743 |
+
"loss": 0.0209,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 9.906542056074766,
|
748 |
+
"grad_norm": 0.15755069255828857,
|
749 |
+
"learning_rate": 0.00010950793968279045,
|
750 |
+
"loss": 0.0198,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 10.0,
|
755 |
+
"grad_norm": 0.261300265789032,
|
756 |
+
"learning_rate": 0.00010796553074762867,
|
757 |
+
"loss": 0.0228,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 10.093457943925234,
|
762 |
+
"grad_norm": 0.20064909756183624,
|
763 |
+
"learning_rate": 0.00010642121221560066,
|
764 |
+
"loss": 0.0237,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 10.186915887850468,
|
769 |
+
"grad_norm": 0.2862524092197418,
|
770 |
+
"learning_rate": 0.00010487535431010049,
|
771 |
+
"loss": 0.0189,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 10.280373831775702,
|
776 |
+
"grad_norm": 0.2752748727798462,
|
777 |
+
"learning_rate": 0.0001033283276235602,
|
778 |
+
"loss": 0.0231,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 10.373831775700934,
|
783 |
+
"grad_norm": 0.1559881865978241,
|
784 |
+
"learning_rate": 0.00010178050302860662,
|
785 |
+
"loss": 0.017,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 10.467289719626168,
|
790 |
+
"grad_norm": 0.16235685348510742,
|
791 |
+
"learning_rate": 0.00010023225158915119,
|
792 |
+
"loss": 0.0214,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 10.560747663551401,
|
797 |
+
"grad_norm": 0.26877355575561523,
|
798 |
+
"learning_rate": 9.868394447143394e-05,
|
799 |
+
"loss": 0.0221,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 10.654205607476635,
|
804 |
+
"grad_norm": 0.26213476061820984,
|
805 |
+
"learning_rate": 9.713595285504266e-05,
|
806 |
+
"loss": 0.0218,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 10.74766355140187,
|
811 |
+
"grad_norm": 0.2717665433883667,
|
812 |
+
"learning_rate": 9.558864784392932e-05,
|
813 |
+
"loss": 0.0203,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 10.841121495327103,
|
818 |
+
"grad_norm": 0.29337963461875916,
|
819 |
+
"learning_rate": 9.404240037744418e-05,
|
820 |
+
"loss": 0.0208,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 10.934579439252337,
|
825 |
+
"grad_norm": 0.18261757493019104,
|
826 |
+
"learning_rate": 9.249758114140977e-05,
|
827 |
+
"loss": 0.0237,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 11.02803738317757,
|
832 |
+
"grad_norm": 0.18363697826862335,
|
833 |
+
"learning_rate": 9.095456047925517e-05,
|
834 |
+
"loss": 0.0202,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 11.121495327102803,
|
839 |
+
"grad_norm": 0.147624671459198,
|
840 |
+
"learning_rate": 8.941370830323286e-05,
|
841 |
+
"loss": 0.0177,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 11.214953271028037,
|
846 |
+
"grad_norm": 0.19984780251979828,
|
847 |
+
"learning_rate": 8.787539400573867e-05,
|
848 |
+
"loss": 0.0208,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 11.30841121495327,
|
853 |
+
"grad_norm": 0.2558623254299164,
|
854 |
+
"learning_rate": 8.633998637075634e-05,
|
855 |
+
"loss": 0.0187,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 11.401869158878505,
|
860 |
+
"grad_norm": 0.15088629722595215,
|
861 |
+
"learning_rate": 8.480785348544802e-05,
|
862 |
+
"loss": 0.0191,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 11.495327102803738,
|
867 |
+
"grad_norm": 0.15472617745399475,
|
868 |
+
"learning_rate": 8.32793626519119e-05,
|
869 |
+
"loss": 0.0174,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 11.588785046728972,
|
874 |
+
"grad_norm": 0.1284412145614624,
|
875 |
+
"learning_rate": 8.175488029912783e-05,
|
876 |
+
"loss": 0.0183,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 11.682242990654206,
|
881 |
+
"grad_norm": 0.2205333113670349,
|
882 |
+
"learning_rate": 8.02347718951124e-05,
|
883 |
+
"loss": 0.0179,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 11.77570093457944,
|
888 |
+
"grad_norm": 0.2073797583580017,
|
889 |
+
"learning_rate": 7.871940185930438e-05,
|
890 |
+
"loss": 0.023,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 11.869158878504672,
|
895 |
+
"grad_norm": 0.22515133023262024,
|
896 |
+
"learning_rate": 7.720913347520154e-05,
|
897 |
+
"loss": 0.0178,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 11.962616822429906,
|
902 |
+
"grad_norm": 0.18330000340938568,
|
903 |
+
"learning_rate": 7.57043288032698e-05,
|
904 |
+
"loss": 0.0206,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 12.05607476635514,
|
909 |
+
"grad_norm": 0.19882416725158691,
|
910 |
+
"learning_rate": 7.420534859414542e-05,
|
911 |
+
"loss": 0.0186,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 12.149532710280374,
|
916 |
+
"grad_norm": 0.4009910821914673,
|
917 |
+
"learning_rate": 7.271255220215159e-05,
|
918 |
+
"loss": 0.0203,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 12.242990654205608,
|
923 |
+
"grad_norm": 0.21823973953723907,
|
924 |
+
"learning_rate": 7.122629749914917e-05,
|
925 |
+
"loss": 0.0183,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 12.336448598130842,
|
930 |
+
"grad_norm": 0.17465481162071228,
|
931 |
+
"learning_rate": 6.974694078874344e-05,
|
932 |
+
"loss": 0.0164,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 12.429906542056075,
|
937 |
+
"grad_norm": 0.23976965248584747,
|
938 |
+
"learning_rate": 6.827483672086649e-05,
|
939 |
+
"loss": 0.0228,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 12.523364485981308,
|
944 |
+
"grad_norm": 0.17434823513031006,
|
945 |
+
"learning_rate": 6.681033820675596e-05,
|
946 |
+
"loss": 0.0185,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 12.616822429906541,
|
951 |
+
"grad_norm": 0.21936717629432678,
|
952 |
+
"learning_rate": 6.535379633435082e-05,
|
953 |
+
"loss": 0.019,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 12.710280373831775,
|
958 |
+
"grad_norm": 0.20729029178619385,
|
959 |
+
"learning_rate": 6.390556028412421e-05,
|
960 |
+
"loss": 0.0187,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 12.80373831775701,
|
965 |
+
"grad_norm": 0.20753750205039978,
|
966 |
+
"learning_rate": 6.246597724537342e-05,
|
967 |
+
"loss": 0.0182,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 12.897196261682243,
|
972 |
+
"grad_norm": 0.17846998572349548,
|
973 |
+
"learning_rate": 6.10353923329875e-05,
|
974 |
+
"loss": 0.0158,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 12.990654205607477,
|
979 |
+
"grad_norm": 0.21448595821857452,
|
980 |
+
"learning_rate": 5.961414850471172e-05,
|
981 |
+
"loss": 0.0159,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 13.08411214953271,
|
986 |
+
"grad_norm": 0.21095584332942963,
|
987 |
+
"learning_rate": 5.8202586478929646e-05,
|
988 |
+
"loss": 0.0169,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 13.177570093457945,
|
993 |
+
"grad_norm": 0.2717122435569763,
|
994 |
+
"learning_rate": 5.680104465298186e-05,
|
995 |
+
"loss": 0.0163,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 13.271028037383177,
|
1000 |
+
"grad_norm": 0.15730348229408264,
|
1001 |
+
"learning_rate": 5.540985902204099e-05,
|
1002 |
+
"loss": 0.0154,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 13.36448598130841,
|
1007 |
+
"grad_norm": 0.2034991979598999,
|
1008 |
+
"learning_rate": 5.4029363098562794e-05,
|
1009 |
+
"loss": 0.0181,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 13.457943925233645,
|
1014 |
+
"grad_norm": 0.1877489984035492,
|
1015 |
+
"learning_rate": 5.2659887832332424e-05,
|
1016 |
+
"loss": 0.0149,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 13.551401869158878,
|
1021 |
+
"grad_norm": 0.25242817401885986,
|
1022 |
+
"learning_rate": 5.130176153112485e-05,
|
1023 |
+
"loss": 0.0162,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 13.644859813084112,
|
1028 |
+
"grad_norm": 0.3759627342224121,
|
1029 |
+
"learning_rate": 4.9955309781999096e-05,
|
1030 |
+
"loss": 0.0205,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 13.738317757009346,
|
1035 |
+
"grad_norm": 0.1744040697813034,
|
1036 |
+
"learning_rate": 4.8620855373244056e-05,
|
1037 |
+
"loss": 0.0158,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 13.83177570093458,
|
1042 |
+
"grad_norm": 0.2127053588628769,
|
1043 |
+
"learning_rate": 4.729871821699587e-05,
|
1044 |
+
"loss": 0.0181,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 13.925233644859812,
|
1049 |
+
"grad_norm": 0.16114765405654907,
|
1050 |
+
"learning_rate": 4.5989215272544605e-05,
|
1051 |
+
"loss": 0.0166,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 14.018691588785046,
|
1056 |
+
"grad_norm": 0.16290584206581116,
|
1057 |
+
"learning_rate": 4.469266047034861e-05,
|
1058 |
+
"loss": 0.0157,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 14.11214953271028,
|
1063 |
+
"grad_norm": 0.29187917709350586,
|
1064 |
+
"learning_rate": 4.3409364636775415e-05,
|
1065 |
+
"loss": 0.0172,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 14.205607476635514,
|
1070 |
+
"grad_norm": 0.2521297335624695,
|
1071 |
+
"learning_rate": 4.213963541958631e-05,
|
1072 |
+
"loss": 0.0147,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 14.299065420560748,
|
1077 |
+
"grad_norm": 0.2301267385482788,
|
1078 |
+
"learning_rate": 4.0883777214183306e-05,
|
1079 |
+
"loss": 0.0172,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 14.392523364485982,
|
1084 |
+
"grad_norm": 0.20023106038570404,
|
1085 |
+
"learning_rate": 3.96420910906356e-05,
|
1086 |
+
"loss": 0.0178,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 14.485981308411215,
|
1091 |
+
"grad_norm": 0.20669516921043396,
|
1092 |
+
"learning_rate": 3.8414874721503294e-05,
|
1093 |
+
"loss": 0.016,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 14.57943925233645,
|
1098 |
+
"grad_norm": 0.22793950140476227,
|
1099 |
+
"learning_rate": 3.720242231047568e-05,
|
1100 |
+
"loss": 0.0168,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 14.672897196261681,
|
1105 |
+
"grad_norm": 0.21381746232509613,
|
1106 |
+
"learning_rate": 3.6005024521840916e-05,
|
1107 |
+
"loss": 0.0156,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 14.766355140186915,
|
1112 |
+
"grad_norm": 0.20497943460941315,
|
1113 |
+
"learning_rate": 3.482296841080433e-05,
|
1114 |
+
"loss": 0.0152,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 14.85981308411215,
|
1119 |
+
"grad_norm": 0.2997715473175049,
|
1120 |
+
"learning_rate": 3.365653735467217e-05,
|
1121 |
+
"loss": 0.0159,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 14.953271028037383,
|
1126 |
+
"grad_norm": 0.1698296070098877,
|
1127 |
+
"learning_rate": 3.2506010984916505e-05,
|
1128 |
+
"loss": 0.0153,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 15.046728971962617,
|
1133 |
+
"grad_norm": 0.13616669178009033,
|
1134 |
+
"learning_rate": 3.137166512013882e-05,
|
1135 |
+
"loss": 0.0155,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 15.14018691588785,
|
1140 |
+
"grad_norm": 0.2119993269443512,
|
1141 |
+
"learning_rate": 3.025377169994702e-05,
|
1142 |
+
"loss": 0.013,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 15.233644859813085,
|
1147 |
+
"grad_norm": 0.18577918410301208,
|
1148 |
+
"learning_rate": 2.9152598719762846e-05,
|
1149 |
+
"loss": 0.0135,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 15.327102803738319,
|
1154 |
+
"grad_norm": 0.16220198571681976,
|
1155 |
+
"learning_rate": 2.8068410166574566e-05,
|
1156 |
+
"loss": 0.015,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 15.42056074766355,
|
1161 |
+
"grad_norm": 0.3275705575942993,
|
1162 |
+
"learning_rate": 2.7001465955650774e-05,
|
1163 |
+
"loss": 0.0136,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 15.514018691588785,
|
1168 |
+
"grad_norm": 0.14544951915740967,
|
1169 |
+
"learning_rate": 2.5952021868230135e-05,
|
1170 |
+
"loss": 0.013,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 15.607476635514018,
|
1175 |
+
"grad_norm": 0.1696663796901703,
|
1176 |
+
"learning_rate": 2.492032949020249e-05,
|
1177 |
+
"loss": 0.0135,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 15.700934579439252,
|
1182 |
+
"grad_norm": 0.22859309613704681,
|
1183 |
+
"learning_rate": 2.390663615179528e-05,
|
1184 |
+
"loss": 0.0145,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 15.794392523364486,
|
1189 |
+
"grad_norm": 0.16665244102478027,
|
1190 |
+
"learning_rate": 2.291118486828071e-05,
|
1191 |
+
"loss": 0.0188,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 15.88785046728972,
|
1196 |
+
"grad_norm": 0.2959059774875641,
|
1197 |
+
"learning_rate": 2.193421428171685e-05,
|
1198 |
+
"loss": 0.0176,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 15.981308411214954,
|
1203 |
+
"grad_norm": 0.2710016071796417,
|
1204 |
+
"learning_rate": 2.0975958603737477e-05,
|
1205 |
+
"loss": 0.0159,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 16.074766355140188,
|
1210 |
+
"grad_norm": 0.18647173047065735,
|
1211 |
+
"learning_rate": 2.0036647559403997e-05,
|
1212 |
+
"loss": 0.0137,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 16.16822429906542,
|
1217 |
+
"grad_norm": 0.17502279579639435,
|
1218 |
+
"learning_rate": 1.9116506332132734e-05,
|
1219 |
+
"loss": 0.0149,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 16.261682242990656,
|
1224 |
+
"grad_norm": 0.24166816473007202,
|
1225 |
+
"learning_rate": 1.8215755509711196e-05,
|
1226 |
+
"loss": 0.0152,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 16.35514018691589,
|
1231 |
+
"grad_norm": 0.1839790791273117,
|
1232 |
+
"learning_rate": 1.7334611031416047e-05,
|
1233 |
+
"loss": 0.0144,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 16.44859813084112,
|
1238 |
+
"grad_norm": 0.20211663842201233,
|
1239 |
+
"learning_rate": 1.647328413624537e-05,
|
1240 |
+
"loss": 0.0149,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 16.542056074766354,
|
1245 |
+
"grad_norm": 0.14310133457183838,
|
1246 |
+
"learning_rate": 1.5631981312277886e-05,
|
1247 |
+
"loss": 0.0138,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 16.635514018691588,
|
1252 |
+
"grad_norm": 0.1992831528186798,
|
1253 |
+
"learning_rate": 1.4810904247170953e-05,
|
1254 |
+
"loss": 0.0145,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 16.72897196261682,
|
1259 |
+
"grad_norm": 0.20767998695373535,
|
1260 |
+
"learning_rate": 1.4010249779809503e-05,
|
1261 |
+
"loss": 0.0142,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 16.822429906542055,
|
1266 |
+
"grad_norm": 0.11333203315734863,
|
1267 |
+
"learning_rate": 1.3230209853117448e-05,
|
1268 |
+
"loss": 0.0144,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 16.91588785046729,
|
1273 |
+
"grad_norm": 0.10632842034101486,
|
1274 |
+
"learning_rate": 1.2470971468042558e-05,
|
1275 |
+
"loss": 0.0139,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 17.009345794392523,
|
1280 |
+
"grad_norm": 0.18203343451023102,
|
1281 |
+
"learning_rate": 1.1732716638726382e-05,
|
1282 |
+
"loss": 0.014,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 17.102803738317757,
|
1287 |
+
"grad_norm": 0.21806679666042328,
|
1288 |
+
"learning_rate": 1.1015622348869491e-05,
|
1289 |
+
"loss": 0.0129,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 17.19626168224299,
|
1294 |
+
"grad_norm": 0.19995306432247162,
|
1295 |
+
"learning_rate": 1.0319860509302637e-05,
|
1296 |
+
"loss": 0.0131,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 17.289719626168225,
|
1301 |
+
"grad_norm": 0.0872287005186081,
|
1302 |
+
"learning_rate": 9.64559791677434e-06,
|
1303 |
+
"loss": 0.0153,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 17.38317757009346,
|
1308 |
+
"grad_norm": 0.13513228297233582,
|
1309 |
+
"learning_rate": 8.992996213964067e-06,
|
1310 |
+
"loss": 0.0158,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 17.476635514018692,
|
1315 |
+
"grad_norm": 0.1807233691215515,
|
1316 |
+
"learning_rate": 8.362211850731382e-06,
|
1317 |
+
"loss": 0.0113,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 17.570093457943926,
|
1322 |
+
"grad_norm": 0.19154734909534454,
|
1323 |
+
"learning_rate": 7.75339604660973e-06,
|
1324 |
+
"loss": 0.0161,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 17.66355140186916,
|
1329 |
+
"grad_norm": 0.2096325159072876,
|
1330 |
+
"learning_rate": 7.166694754554193e-06,
|
1331 |
+
"loss": 0.016,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 17.757009345794394,
|
1336 |
+
"grad_norm": 0.16648273169994354,
|
1337 |
+
"learning_rate": 6.602248625951934e-06,
|
1338 |
+
"loss": 0.013,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 17.850467289719624,
|
1343 |
+
"grad_norm": 0.19170570373535156,
|
1344 |
+
"learning_rate": 6.0601929769033115e-06,
|
1345 |
+
"loss": 0.0116,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 17.94392523364486,
|
1350 |
+
"grad_norm": 0.14835980534553528,
|
1351 |
+
"learning_rate": 5.540657755782341e-06,
|
1352 |
+
"loss": 0.0136,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 18.037383177570092,
|
1357 |
+
"grad_norm": 0.10613177716732025,
|
1358 |
+
"learning_rate": 5.043767512083797e-06,
|
1359 |
+
"loss": 0.012,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 18.130841121495326,
|
1364 |
+
"grad_norm": 0.18721149861812592,
|
1365 |
+
"learning_rate": 4.569641366564614e-06,
|
1366 |
+
"loss": 0.0153,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 18.22429906542056,
|
1371 |
+
"grad_norm": 0.17616799473762512,
|
1372 |
+
"learning_rate": 4.1183929826868476e-06,
|
1373 |
+
"loss": 0.0115,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 18.317757009345794,
|
1378 |
+
"grad_norm": 0.2676970064640045,
|
1379 |
+
"learning_rate": 3.6901305393687634e-06,
|
1380 |
+
"loss": 0.0113,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 18.411214953271028,
|
1385 |
+
"grad_norm": 0.10376067459583282,
|
1386 |
+
"learning_rate": 3.284956705050868e-06,
|
1387 |
+
"loss": 0.0122,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 18.50467289719626,
|
1392 |
+
"grad_norm": 0.12165194004774094,
|
1393 |
+
"learning_rate": 2.902968613082957e-06,
|
1394 |
+
"loss": 0.0134,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 18.598130841121495,
|
1399 |
+
"grad_norm": 0.15639284253120422,
|
1400 |
+
"learning_rate": 2.5442578384380245e-06,
|
1401 |
+
"loss": 0.0115,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 18.69158878504673,
|
1406 |
+
"grad_norm": 0.14054954051971436,
|
1407 |
+
"learning_rate": 2.208910375758888e-06,
|
1408 |
+
"loss": 0.0111,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 18.785046728971963,
|
1413 |
+
"grad_norm": 0.13985030353069305,
|
1414 |
+
"learning_rate": 1.8970066187423364e-06,
|
1415 |
+
"loss": 0.0115,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 18.878504672897197,
|
1420 |
+
"grad_norm": 0.12971541285514832,
|
1421 |
+
"learning_rate": 1.6086213408662232e-06,
|
1422 |
+
"loss": 0.0117,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 18.97196261682243,
|
1427 |
+
"grad_norm": 0.17401276528835297,
|
1428 |
+
"learning_rate": 1.343823677463818e-06,
|
1429 |
+
"loss": 0.0139,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 19.065420560747665,
|
1434 |
+
"grad_norm": 0.24669478833675385,
|
1435 |
+
"learning_rate": 1.1026771091498079e-06,
|
1436 |
+
"loss": 0.0143,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 19.1588785046729,
|
1441 |
+
"grad_norm": 0.1944689005613327,
|
1442 |
+
"learning_rate": 8.852394466019487e-07,
|
1443 |
+
"loss": 0.0116,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 19.252336448598133,
|
1448 |
+
"grad_norm": 0.15337271988391876,
|
1449 |
+
"learning_rate": 6.915628167019628e-07,
|
1450 |
+
"loss": 0.0126,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 19.345794392523363,
|
1455 |
+
"grad_norm": 0.22086739540100098,
|
1456 |
+
"learning_rate": 5.216936500389679e-07,
|
1457 |
+
"loss": 0.0123,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 19.439252336448597,
|
1462 |
+
"grad_norm": 0.10731634497642517,
|
1463 |
+
"learning_rate": 3.7567266977868165e-07,
|
1464 |
+
"loss": 0.0117,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 19.53271028037383,
|
1469 |
+
"grad_norm": 0.16525092720985413,
|
1470 |
+
"learning_rate": 2.535348819006078e-07,
|
1471 |
+
"loss": 0.0142,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 19.626168224299064,
|
1476 |
+
"grad_norm": 0.22534893453121185,
|
1477 |
+
"learning_rate": 1.5530956680606114e-07,
|
1478 |
+
"loss": 0.0136,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 19.7196261682243,
|
1483 |
+
"grad_norm": 0.14636439085006714,
|
1484 |
+
"learning_rate": 8.102027229868192e-08,
|
1485 |
+
"loss": 0.0098,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 19.813084112149532,
|
1490 |
+
"grad_norm": 0.1717785745859146,
|
1491 |
+
"learning_rate": 3.0684807939262985e-08,
|
1492 |
+
"loss": 0.0105,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 19.906542056074766,
|
1497 |
+
"grad_norm": 0.15287938714027405,
|
1498 |
+
"learning_rate": 4.315240776175955e-09,
|
1499 |
+
"loss": 0.0135,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 19.962616822429908,
|
1504 |
+
"step": 2136,
|
1505 |
+
"total_flos": 3.092595209923816e+17,
|
1506 |
+
"train_loss": 0.03558622600950217,
|
1507 |
+
"train_runtime": 2455.8794,
|
1508 |
+
"train_samples_per_second": 55.664,
|
1509 |
+
"train_steps_per_second": 0.87
|
1510 |
+
}
|
1511 |
+
],
|
1512 |
+
"logging_steps": 10,
|
1513 |
+
"max_steps": 2136,
|
1514 |
+
"num_input_tokens_seen": 0,
|
1515 |
+
"num_train_epochs": 20,
|
1516 |
+
"save_steps": 10000,
|
1517 |
+
"stateful_callbacks": {
|
1518 |
+
"TrainerControl": {
|
1519 |
+
"args": {
|
1520 |
+
"should_epoch_stop": false,
|
1521 |
+
"should_evaluate": false,
|
1522 |
+
"should_log": false,
|
1523 |
+
"should_save": true,
|
1524 |
+
"should_training_stop": true
|
1525 |
+
},
|
1526 |
+
"attributes": {}
|
1527 |
+
}
|
1528 |
+
},
|
1529 |
+
"total_flos": 3.092595209923816e+17,
|
1530 |
+
"train_batch_size": 64,
|
1531 |
+
"trial_name": null,
|
1532 |
+
"trial_params": null
|
1533 |
+
}
|