Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1520 -0
trainer_state.json
ADDED
@@ -0,0 +1,1520 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 6.0,
|
5 |
+
"eval_steps": 200,
|
6 |
+
"global_step": 210,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.02857142857142857,
|
13 |
+
"grad_norm": 19.810293087197298,
|
14 |
+
"learning_rate": 5e-08,
|
15 |
+
"loss": 1.6319,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.05714285714285714,
|
20 |
+
"grad_norm": 20.10065322014341,
|
21 |
+
"learning_rate": 1e-07,
|
22 |
+
"loss": 1.5294,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.08571428571428572,
|
27 |
+
"grad_norm": 22.139565739974667,
|
28 |
+
"learning_rate": 1.5e-07,
|
29 |
+
"loss": 1.568,
|
30 |
+
"step": 3
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.11428571428571428,
|
34 |
+
"grad_norm": 17.607582690653754,
|
35 |
+
"learning_rate": 2e-07,
|
36 |
+
"loss": 1.6628,
|
37 |
+
"step": 4
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.14285714285714285,
|
41 |
+
"grad_norm": 24.910794234048186,
|
42 |
+
"learning_rate": 2.5e-07,
|
43 |
+
"loss": 1.5738,
|
44 |
+
"step": 5
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.17142857142857143,
|
48 |
+
"grad_norm": 18.006572442480767,
|
49 |
+
"learning_rate": 3e-07,
|
50 |
+
"loss": 1.5898,
|
51 |
+
"step": 6
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.2,
|
55 |
+
"grad_norm": 22.152647510768777,
|
56 |
+
"learning_rate": 3.5e-07,
|
57 |
+
"loss": 1.5891,
|
58 |
+
"step": 7
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.22857142857142856,
|
62 |
+
"grad_norm": 17.044462125735237,
|
63 |
+
"learning_rate": 4e-07,
|
64 |
+
"loss": 1.5277,
|
65 |
+
"step": 8
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.2571428571428571,
|
69 |
+
"grad_norm": 19.26164281649356,
|
70 |
+
"learning_rate": 4.5e-07,
|
71 |
+
"loss": 1.5504,
|
72 |
+
"step": 9
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.2857142857142857,
|
76 |
+
"grad_norm": 20.409183726640773,
|
77 |
+
"learning_rate": 5e-07,
|
78 |
+
"loss": 1.5509,
|
79 |
+
"step": 10
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.3142857142857143,
|
83 |
+
"grad_norm": 16.865224915691606,
|
84 |
+
"learning_rate": 5.5e-07,
|
85 |
+
"loss": 1.5752,
|
86 |
+
"step": 11
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.34285714285714286,
|
90 |
+
"grad_norm": 15.722428651072727,
|
91 |
+
"learning_rate": 6e-07,
|
92 |
+
"loss": 1.6208,
|
93 |
+
"step": 12
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.37142857142857144,
|
97 |
+
"grad_norm": 13.23419740428603,
|
98 |
+
"learning_rate": 6.5e-07,
|
99 |
+
"loss": 1.515,
|
100 |
+
"step": 13
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.4,
|
104 |
+
"grad_norm": 12.45173447148941,
|
105 |
+
"learning_rate": 7e-07,
|
106 |
+
"loss": 1.4752,
|
107 |
+
"step": 14
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.42857142857142855,
|
111 |
+
"grad_norm": 10.684936055938318,
|
112 |
+
"learning_rate": 7.5e-07,
|
113 |
+
"loss": 1.5073,
|
114 |
+
"step": 15
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.45714285714285713,
|
118 |
+
"grad_norm": 8.612999241840305,
|
119 |
+
"learning_rate": 8e-07,
|
120 |
+
"loss": 1.4206,
|
121 |
+
"step": 16
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.4857142857142857,
|
125 |
+
"grad_norm": 9.64409059662565,
|
126 |
+
"learning_rate": 8.499999999999999e-07,
|
127 |
+
"loss": 1.3659,
|
128 |
+
"step": 17
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.5142857142857142,
|
132 |
+
"grad_norm": 9.53721910636901,
|
133 |
+
"learning_rate": 9e-07,
|
134 |
+
"loss": 1.4999,
|
135 |
+
"step": 18
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.5428571428571428,
|
139 |
+
"grad_norm": 15.44618942298652,
|
140 |
+
"learning_rate": 9.499999999999999e-07,
|
141 |
+
"loss": 1.452,
|
142 |
+
"step": 19
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.5714285714285714,
|
146 |
+
"grad_norm": 7.797878397007685,
|
147 |
+
"learning_rate": 1e-06,
|
148 |
+
"loss": 1.4059,
|
149 |
+
"step": 20
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.6,
|
153 |
+
"grad_norm": 8.611751272514192,
|
154 |
+
"learning_rate": 9.999316524962345e-07,
|
155 |
+
"loss": 1.3416,
|
156 |
+
"step": 21
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.6285714285714286,
|
160 |
+
"grad_norm": 8.60172100284345,
|
161 |
+
"learning_rate": 9.99726628670463e-07,
|
162 |
+
"loss": 1.3334,
|
163 |
+
"step": 22
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.6571428571428571,
|
167 |
+
"grad_norm": 7.707986778608113,
|
168 |
+
"learning_rate": 9.993849845741523e-07,
|
169 |
+
"loss": 1.3032,
|
170 |
+
"step": 23
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.6857142857142857,
|
174 |
+
"grad_norm": 7.692312413668571,
|
175 |
+
"learning_rate": 9.989068136093872e-07,
|
176 |
+
"loss": 1.3742,
|
177 |
+
"step": 24
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.7142857142857143,
|
181 |
+
"grad_norm": 6.364211690724886,
|
182 |
+
"learning_rate": 9.982922465033348e-07,
|
183 |
+
"loss": 1.2641,
|
184 |
+
"step": 25
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.7428571428571429,
|
188 |
+
"grad_norm": 6.85599706491363,
|
189 |
+
"learning_rate": 9.975414512725056e-07,
|
190 |
+
"loss": 1.2624,
|
191 |
+
"step": 26
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.7714285714285715,
|
195 |
+
"grad_norm": 6.401247246252304,
|
196 |
+
"learning_rate": 9.966546331768192e-07,
|
197 |
+
"loss": 1.258,
|
198 |
+
"step": 27
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.8,
|
202 |
+
"grad_norm": 9.111986647021705,
|
203 |
+
"learning_rate": 9.956320346634875e-07,
|
204 |
+
"loss": 1.3228,
|
205 |
+
"step": 28
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.8285714285714286,
|
209 |
+
"grad_norm": 6.326533367398195,
|
210 |
+
"learning_rate": 9.944739353007341e-07,
|
211 |
+
"loss": 1.2034,
|
212 |
+
"step": 29
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.8571428571428571,
|
216 |
+
"grad_norm": 6.561308869602955,
|
217 |
+
"learning_rate": 9.931806517013612e-07,
|
218 |
+
"loss": 1.1795,
|
219 |
+
"step": 30
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.8857142857142857,
|
223 |
+
"grad_norm": 5.9273022974036795,
|
224 |
+
"learning_rate": 9.917525374361911e-07,
|
225 |
+
"loss": 1.2288,
|
226 |
+
"step": 31
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.9142857142857143,
|
230 |
+
"grad_norm": 5.348130355740025,
|
231 |
+
"learning_rate": 9.901899829374047e-07,
|
232 |
+
"loss": 1.1764,
|
233 |
+
"step": 32
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.9428571428571428,
|
237 |
+
"grad_norm": 5.274551340001603,
|
238 |
+
"learning_rate": 9.884934153917996e-07,
|
239 |
+
"loss": 1.1402,
|
240 |
+
"step": 33
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.9714285714285714,
|
244 |
+
"grad_norm": 5.681514700875701,
|
245 |
+
"learning_rate": 9.866632986240029e-07,
|
246 |
+
"loss": 1.1604,
|
247 |
+
"step": 34
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.0,
|
251 |
+
"grad_norm": 5.77664764652199,
|
252 |
+
"learning_rate": 9.847001329696652e-07,
|
253 |
+
"loss": 1.1567,
|
254 |
+
"step": 35
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.0285714285714285,
|
258 |
+
"grad_norm": 5.836094968508217,
|
259 |
+
"learning_rate": 9.826044551386742e-07,
|
260 |
+
"loss": 1.084,
|
261 |
+
"step": 36
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 1.0571428571428572,
|
265 |
+
"grad_norm": 5.827344070494953,
|
266 |
+
"learning_rate": 9.803768380684242e-07,
|
267 |
+
"loss": 1.1305,
|
268 |
+
"step": 37
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 1.0857142857142856,
|
272 |
+
"grad_norm": 5.1906563321747115,
|
273 |
+
"learning_rate": 9.780178907671788e-07,
|
274 |
+
"loss": 1.0574,
|
275 |
+
"step": 38
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.1142857142857143,
|
279 |
+
"grad_norm": 5.562551195695788,
|
280 |
+
"learning_rate": 9.755282581475767e-07,
|
281 |
+
"loss": 1.1391,
|
282 |
+
"step": 39
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 1.1428571428571428,
|
286 |
+
"grad_norm": 5.717325403158778,
|
287 |
+
"learning_rate": 9.729086208503173e-07,
|
288 |
+
"loss": 1.1267,
|
289 |
+
"step": 40
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 1.1714285714285715,
|
293 |
+
"grad_norm": 5.969025543565558,
|
294 |
+
"learning_rate": 9.701596950580807e-07,
|
295 |
+
"loss": 1.1055,
|
296 |
+
"step": 41
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.2,
|
300 |
+
"grad_norm": 4.274011328484305,
|
301 |
+
"learning_rate": 9.672822322997304e-07,
|
302 |
+
"loss": 1.0392,
|
303 |
+
"step": 42
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 1.2285714285714286,
|
307 |
+
"grad_norm": 4.4561906748259545,
|
308 |
+
"learning_rate": 9.642770192448535e-07,
|
309 |
+
"loss": 1.0393,
|
310 |
+
"step": 43
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 1.2571428571428571,
|
314 |
+
"grad_norm": 4.618703002376858,
|
315 |
+
"learning_rate": 9.611448774886923e-07,
|
316 |
+
"loss": 1.0795,
|
317 |
+
"step": 44
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 1.2857142857142856,
|
321 |
+
"grad_norm": 4.473713036590507,
|
322 |
+
"learning_rate": 9.578866633275286e-07,
|
323 |
+
"loss": 1.0194,
|
324 |
+
"step": 45
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.3142857142857143,
|
328 |
+
"grad_norm": 4.3150195602071415,
|
329 |
+
"learning_rate": 9.545032675245813e-07,
|
330 |
+
"loss": 1.0409,
|
331 |
+
"step": 46
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.342857142857143,
|
335 |
+
"grad_norm": 4.847108292314838,
|
336 |
+
"learning_rate": 9.509956150664795e-07,
|
337 |
+
"loss": 1.0475,
|
338 |
+
"step": 47
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.3714285714285714,
|
342 |
+
"grad_norm": 4.716522743464805,
|
343 |
+
"learning_rate": 9.473646649103817e-07,
|
344 |
+
"loss": 1.046,
|
345 |
+
"step": 48
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.4,
|
349 |
+
"grad_norm": 4.942648131785751,
|
350 |
+
"learning_rate": 9.436114097218058e-07,
|
351 |
+
"loss": 0.9876,
|
352 |
+
"step": 49
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.4285714285714286,
|
356 |
+
"grad_norm": 5.1753554721234565,
|
357 |
+
"learning_rate": 9.397368756032444e-07,
|
358 |
+
"loss": 1.0394,
|
359 |
+
"step": 50
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.457142857142857,
|
363 |
+
"grad_norm": 4.996816148990504,
|
364 |
+
"learning_rate": 9.357421218136386e-07,
|
365 |
+
"loss": 0.9647,
|
366 |
+
"step": 51
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.4857142857142858,
|
370 |
+
"grad_norm": 5.982295572473842,
|
371 |
+
"learning_rate": 9.316282404787869e-07,
|
372 |
+
"loss": 1.0986,
|
373 |
+
"step": 52
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.5142857142857142,
|
377 |
+
"grad_norm": 4.701714221899932,
|
378 |
+
"learning_rate": 9.273963562927694e-07,
|
379 |
+
"loss": 1.0493,
|
380 |
+
"step": 53
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.5428571428571427,
|
384 |
+
"grad_norm": 4.514923820929648,
|
385 |
+
"learning_rate": 9.230476262104676e-07,
|
386 |
+
"loss": 0.9847,
|
387 |
+
"step": 54
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.5714285714285714,
|
391 |
+
"grad_norm": 4.701919910281847,
|
392 |
+
"learning_rate": 9.185832391312642e-07,
|
393 |
+
"loss": 0.9794,
|
394 |
+
"step": 55
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.6,
|
398 |
+
"grad_norm": 5.175054800986126,
|
399 |
+
"learning_rate": 9.1400441557401e-07,
|
400 |
+
"loss": 1.0607,
|
401 |
+
"step": 56
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.6285714285714286,
|
405 |
+
"grad_norm": 4.080487196055978,
|
406 |
+
"learning_rate": 9.093124073433462e-07,
|
407 |
+
"loss": 0.9856,
|
408 |
+
"step": 57
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.657142857142857,
|
412 |
+
"grad_norm": 4.168373937047882,
|
413 |
+
"learning_rate": 9.045084971874737e-07,
|
414 |
+
"loss": 0.9107,
|
415 |
+
"step": 58
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.6857142857142857,
|
419 |
+
"grad_norm": 4.223886586576581,
|
420 |
+
"learning_rate": 8.995939984474623e-07,
|
421 |
+
"loss": 0.9733,
|
422 |
+
"step": 59
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.7142857142857144,
|
426 |
+
"grad_norm": 4.513527637157093,
|
427 |
+
"learning_rate": 8.945702546981968e-07,
|
428 |
+
"loss": 0.9476,
|
429 |
+
"step": 60
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.7428571428571429,
|
433 |
+
"grad_norm": 6.610164639785705,
|
434 |
+
"learning_rate": 8.894386393810562e-07,
|
435 |
+
"loss": 0.9161,
|
436 |
+
"step": 61
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.7714285714285714,
|
440 |
+
"grad_norm": 4.703651247017542,
|
441 |
+
"learning_rate": 8.842005554284295e-07,
|
442 |
+
"loss": 1.0058,
|
443 |
+
"step": 62
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.8,
|
447 |
+
"grad_norm": 4.843849342547474,
|
448 |
+
"learning_rate": 8.788574348801674e-07,
|
449 |
+
"loss": 0.8815,
|
450 |
+
"step": 63
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.8285714285714287,
|
454 |
+
"grad_norm": 4.850275003883815,
|
455 |
+
"learning_rate": 8.734107384920769e-07,
|
456 |
+
"loss": 0.9945,
|
457 |
+
"step": 64
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.8571428571428572,
|
461 |
+
"grad_norm": 5.453604273607254,
|
462 |
+
"learning_rate": 8.678619553365658e-07,
|
463 |
+
"loss": 0.9822,
|
464 |
+
"step": 65
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.8857142857142857,
|
468 |
+
"grad_norm": 5.3760203289286155,
|
469 |
+
"learning_rate": 8.622126023955445e-07,
|
470 |
+
"loss": 0.9781,
|
471 |
+
"step": 66
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.9142857142857141,
|
475 |
+
"grad_norm": 3.7890108184255062,
|
476 |
+
"learning_rate": 8.564642241456986e-07,
|
477 |
+
"loss": 0.9188,
|
478 |
+
"step": 67
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.9428571428571428,
|
482 |
+
"grad_norm": 4.482132592462384,
|
483 |
+
"learning_rate": 8.506183921362442e-07,
|
484 |
+
"loss": 0.9193,
|
485 |
+
"step": 68
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.9714285714285715,
|
489 |
+
"grad_norm": 5.311797532408616,
|
490 |
+
"learning_rate": 8.446767045592829e-07,
|
491 |
+
"loss": 0.9563,
|
492 |
+
"step": 69
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 2.0,
|
496 |
+
"grad_norm": 4.950363360704343,
|
497 |
+
"learning_rate": 8.386407858128706e-07,
|
498 |
+
"loss": 0.9172,
|
499 |
+
"step": 70
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 2.0285714285714285,
|
503 |
+
"grad_norm": 4.784370125705732,
|
504 |
+
"learning_rate": 8.325122860569241e-07,
|
505 |
+
"loss": 0.8843,
|
506 |
+
"step": 71
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 2.057142857142857,
|
510 |
+
"grad_norm": 3.63025112886302,
|
511 |
+
"learning_rate": 8.262928807620843e-07,
|
512 |
+
"loss": 0.8547,
|
513 |
+
"step": 72
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 2.085714285714286,
|
517 |
+
"grad_norm": 5.218576028144485,
|
518 |
+
"learning_rate": 8.199842702516582e-07,
|
519 |
+
"loss": 0.9414,
|
520 |
+
"step": 73
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 2.1142857142857143,
|
524 |
+
"grad_norm": 4.677814392084542,
|
525 |
+
"learning_rate": 8.135881792367685e-07,
|
526 |
+
"loss": 0.8748,
|
527 |
+
"step": 74
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 2.142857142857143,
|
531 |
+
"grad_norm": 4.400225941517684,
|
532 |
+
"learning_rate": 8.071063563448339e-07,
|
533 |
+
"loss": 0.8646,
|
534 |
+
"step": 75
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 2.1714285714285713,
|
538 |
+
"grad_norm": 4.629375190268374,
|
539 |
+
"learning_rate": 8.005405736415125e-07,
|
540 |
+
"loss": 0.8744,
|
541 |
+
"step": 76
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 2.2,
|
545 |
+
"grad_norm": 4.1365979610935,
|
546 |
+
"learning_rate": 7.938926261462365e-07,
|
547 |
+
"loss": 0.8735,
|
548 |
+
"step": 77
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.2285714285714286,
|
552 |
+
"grad_norm": 4.796843279900597,
|
553 |
+
"learning_rate": 7.871643313414718e-07,
|
554 |
+
"loss": 0.9161,
|
555 |
+
"step": 78
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 2.257142857142857,
|
559 |
+
"grad_norm": 4.297808554492733,
|
560 |
+
"learning_rate": 7.803575286758363e-07,
|
561 |
+
"loss": 0.8119,
|
562 |
+
"step": 79
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 2.2857142857142856,
|
566 |
+
"grad_norm": 4.229241471681201,
|
567 |
+
"learning_rate": 7.734740790612136e-07,
|
568 |
+
"loss": 0.8698,
|
569 |
+
"step": 80
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 2.314285714285714,
|
573 |
+
"grad_norm": 5.474712766868538,
|
574 |
+
"learning_rate": 7.665158643639969e-07,
|
575 |
+
"loss": 0.8673,
|
576 |
+
"step": 81
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 2.342857142857143,
|
580 |
+
"grad_norm": 5.8097182446336495,
|
581 |
+
"learning_rate": 7.594847868906076e-07,
|
582 |
+
"loss": 0.8561,
|
583 |
+
"step": 82
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 2.3714285714285714,
|
587 |
+
"grad_norm": 6.090812187892629,
|
588 |
+
"learning_rate": 7.523827688674219e-07,
|
589 |
+
"loss": 0.9032,
|
590 |
+
"step": 83
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 2.4,
|
594 |
+
"grad_norm": 4.2456911319822614,
|
595 |
+
"learning_rate": 7.452117519152541e-07,
|
596 |
+
"loss": 0.8431,
|
597 |
+
"step": 84
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 2.4285714285714284,
|
601 |
+
"grad_norm": 4.562925662849317,
|
602 |
+
"learning_rate": 7.379736965185368e-07,
|
603 |
+
"loss": 0.844,
|
604 |
+
"step": 85
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 2.4571428571428573,
|
608 |
+
"grad_norm": 4.517660029971984,
|
609 |
+
"learning_rate": 7.306705814893439e-07,
|
610 |
+
"loss": 0.8071,
|
611 |
+
"step": 86
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 2.4857142857142858,
|
615 |
+
"grad_norm": 4.970820782182107,
|
616 |
+
"learning_rate": 7.233044034264033e-07,
|
617 |
+
"loss": 0.8726,
|
618 |
+
"step": 87
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 2.5142857142857142,
|
622 |
+
"grad_norm": 4.817397128446438,
|
623 |
+
"learning_rate": 7.158771761692464e-07,
|
624 |
+
"loss": 0.8651,
|
625 |
+
"step": 88
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 2.5428571428571427,
|
629 |
+
"grad_norm": 4.521311016914525,
|
630 |
+
"learning_rate": 7.083909302476452e-07,
|
631 |
+
"loss": 0.8023,
|
632 |
+
"step": 89
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 2.571428571428571,
|
636 |
+
"grad_norm": 5.57347106069374,
|
637 |
+
"learning_rate": 7.008477123264847e-07,
|
638 |
+
"loss": 0.8107,
|
639 |
+
"step": 90
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 2.6,
|
643 |
+
"grad_norm": 4.958424900327067,
|
644 |
+
"learning_rate": 6.932495846462261e-07,
|
645 |
+
"loss": 0.8321,
|
646 |
+
"step": 91
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 2.6285714285714286,
|
650 |
+
"grad_norm": 4.34206205323412,
|
651 |
+
"learning_rate": 6.855986244591103e-07,
|
652 |
+
"loss": 0.8003,
|
653 |
+
"step": 92
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 2.657142857142857,
|
657 |
+
"grad_norm": 4.417137471491474,
|
658 |
+
"learning_rate": 6.778969234612583e-07,
|
659 |
+
"loss": 0.81,
|
660 |
+
"step": 93
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 2.685714285714286,
|
664 |
+
"grad_norm": 4.714391490425693,
|
665 |
+
"learning_rate": 6.701465872208216e-07,
|
666 |
+
"loss": 0.8362,
|
667 |
+
"step": 94
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 2.7142857142857144,
|
671 |
+
"grad_norm": 3.835031010544762,
|
672 |
+
"learning_rate": 6.623497346023417e-07,
|
673 |
+
"loss": 0.7408,
|
674 |
+
"step": 95
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.742857142857143,
|
678 |
+
"grad_norm": 4.43328902038276,
|
679 |
+
"learning_rate": 6.545084971874736e-07,
|
680 |
+
"loss": 0.803,
|
681 |
+
"step": 96
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.7714285714285714,
|
685 |
+
"grad_norm": 5.573988325671216,
|
686 |
+
"learning_rate": 6.466250186922324e-07,
|
687 |
+
"loss": 0.8499,
|
688 |
+
"step": 97
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.8,
|
692 |
+
"grad_norm": 5.080352906047553,
|
693 |
+
"learning_rate": 6.387014543809223e-07,
|
694 |
+
"loss": 0.8536,
|
695 |
+
"step": 98
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 2.8285714285714287,
|
699 |
+
"grad_norm": 4.433727646568338,
|
700 |
+
"learning_rate": 6.307399704769098e-07,
|
701 |
+
"loss": 0.8167,
|
702 |
+
"step": 99
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 2.857142857142857,
|
706 |
+
"grad_norm": 4.72983530688845,
|
707 |
+
"learning_rate": 6.227427435703995e-07,
|
708 |
+
"loss": 0.8247,
|
709 |
+
"step": 100
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 2.8857142857142857,
|
713 |
+
"grad_norm": 4.031675844503338,
|
714 |
+
"learning_rate": 6.147119600233758e-07,
|
715 |
+
"loss": 0.7923,
|
716 |
+
"step": 101
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 2.914285714285714,
|
720 |
+
"grad_norm": 4.657562682714027,
|
721 |
+
"learning_rate": 6.066498153718734e-07,
|
722 |
+
"loss": 0.8259,
|
723 |
+
"step": 102
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 2.942857142857143,
|
727 |
+
"grad_norm": 3.742914618609364,
|
728 |
+
"learning_rate": 5.985585137257401e-07,
|
729 |
+
"loss": 0.7416,
|
730 |
+
"step": 103
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 2.9714285714285715,
|
734 |
+
"grad_norm": 5.392956655820792,
|
735 |
+
"learning_rate": 5.90440267166055e-07,
|
736 |
+
"loss": 0.86,
|
737 |
+
"step": 104
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 3.0,
|
741 |
+
"grad_norm": 4.370818695376769,
|
742 |
+
"learning_rate": 5.82297295140367e-07,
|
743 |
+
"loss": 0.7756,
|
744 |
+
"step": 105
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 3.0285714285714285,
|
748 |
+
"grad_norm": 4.1080059715561426,
|
749 |
+
"learning_rate": 5.741318238559209e-07,
|
750 |
+
"loss": 0.7417,
|
751 |
+
"step": 106
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 3.057142857142857,
|
755 |
+
"grad_norm": 5.242835158751313,
|
756 |
+
"learning_rate": 5.659460856710345e-07,
|
757 |
+
"loss": 0.7413,
|
758 |
+
"step": 107
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 3.085714285714286,
|
762 |
+
"grad_norm": 4.213242126183361,
|
763 |
+
"learning_rate": 5.577423184847931e-07,
|
764 |
+
"loss": 0.774,
|
765 |
+
"step": 108
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 3.1142857142857143,
|
769 |
+
"grad_norm": 5.14212832829309,
|
770 |
+
"learning_rate": 5.495227651252315e-07,
|
771 |
+
"loss": 0.7596,
|
772 |
+
"step": 109
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 3.142857142857143,
|
776 |
+
"grad_norm": 4.536657005455878,
|
777 |
+
"learning_rate": 5.412896727361662e-07,
|
778 |
+
"loss": 0.7494,
|
779 |
+
"step": 110
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 3.1714285714285713,
|
783 |
+
"grad_norm": 5.180517603140962,
|
784 |
+
"learning_rate": 5.330452921628497e-07,
|
785 |
+
"loss": 0.8146,
|
786 |
+
"step": 111
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 3.2,
|
790 |
+
"grad_norm": 4.499166275543811,
|
791 |
+
"learning_rate": 5.247918773366111e-07,
|
792 |
+
"loss": 0.7437,
|
793 |
+
"step": 112
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 3.2285714285714286,
|
797 |
+
"grad_norm": 4.78892229810655,
|
798 |
+
"learning_rate": 5.165316846586541e-07,
|
799 |
+
"loss": 0.732,
|
800 |
+
"step": 113
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 3.257142857142857,
|
804 |
+
"grad_norm": 4.643947668361775,
|
805 |
+
"learning_rate": 5.082669723831793e-07,
|
806 |
+
"loss": 0.769,
|
807 |
+
"step": 114
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 3.2857142857142856,
|
811 |
+
"grad_norm": 4.79036988016491,
|
812 |
+
"learning_rate": 5e-07,
|
813 |
+
"loss": 0.7586,
|
814 |
+
"step": 115
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 3.314285714285714,
|
818 |
+
"grad_norm": 5.1592671986124845,
|
819 |
+
"learning_rate": 4.917330276168208e-07,
|
820 |
+
"loss": 0.7513,
|
821 |
+
"step": 116
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 3.342857142857143,
|
825 |
+
"grad_norm": 3.944922328543468,
|
826 |
+
"learning_rate": 4.834683153413459e-07,
|
827 |
+
"loss": 0.7184,
|
828 |
+
"step": 117
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 3.3714285714285714,
|
832 |
+
"grad_norm": 4.429646542430914,
|
833 |
+
"learning_rate": 4.752081226633888e-07,
|
834 |
+
"loss": 0.6937,
|
835 |
+
"step": 118
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 3.4,
|
839 |
+
"grad_norm": 5.016602862879482,
|
840 |
+
"learning_rate": 4.669547078371503e-07,
|
841 |
+
"loss": 0.7638,
|
842 |
+
"step": 119
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 3.4285714285714284,
|
846 |
+
"grad_norm": 4.443230989038584,
|
847 |
+
"learning_rate": 4.5871032726383385e-07,
|
848 |
+
"loss": 0.7029,
|
849 |
+
"step": 120
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 3.4571428571428573,
|
853 |
+
"grad_norm": 4.244778886390162,
|
854 |
+
"learning_rate": 4.5047723487476864e-07,
|
855 |
+
"loss": 0.7298,
|
856 |
+
"step": 121
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 3.4857142857142858,
|
860 |
+
"grad_norm": 5.068619137512623,
|
861 |
+
"learning_rate": 4.4225768151520694e-07,
|
862 |
+
"loss": 0.7716,
|
863 |
+
"step": 122
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 3.5142857142857142,
|
867 |
+
"grad_norm": 4.652036232274711,
|
868 |
+
"learning_rate": 4.340539143289655e-07,
|
869 |
+
"loss": 0.7276,
|
870 |
+
"step": 123
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 3.5428571428571427,
|
874 |
+
"grad_norm": 4.95389292196556,
|
875 |
+
"learning_rate": 4.258681761440789e-07,
|
876 |
+
"loss": 0.7345,
|
877 |
+
"step": 124
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 3.571428571428571,
|
881 |
+
"grad_norm": 4.275659541196405,
|
882 |
+
"learning_rate": 4.1770270485963294e-07,
|
883 |
+
"loss": 0.7323,
|
884 |
+
"step": 125
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 3.6,
|
888 |
+
"grad_norm": 4.607417501587952,
|
889 |
+
"learning_rate": 4.095597328339452e-07,
|
890 |
+
"loss": 0.7606,
|
891 |
+
"step": 126
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 3.6285714285714286,
|
895 |
+
"grad_norm": 4.794919375541705,
|
896 |
+
"learning_rate": 4.0144148627425986e-07,
|
897 |
+
"loss": 0.7149,
|
898 |
+
"step": 127
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 3.657142857142857,
|
902 |
+
"grad_norm": 4.102606927748694,
|
903 |
+
"learning_rate": 3.9335018462812664e-07,
|
904 |
+
"loss": 0.6956,
|
905 |
+
"step": 128
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 3.685714285714286,
|
909 |
+
"grad_norm": 4.243593175311764,
|
910 |
+
"learning_rate": 3.8528803997662423e-07,
|
911 |
+
"loss": 0.6894,
|
912 |
+
"step": 129
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 3.7142857142857144,
|
916 |
+
"grad_norm": 4.76702832294717,
|
917 |
+
"learning_rate": 3.772572564296004e-07,
|
918 |
+
"loss": 0.7188,
|
919 |
+
"step": 130
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 3.742857142857143,
|
923 |
+
"grad_norm": 4.651784836564843,
|
924 |
+
"learning_rate": 3.692600295230901e-07,
|
925 |
+
"loss": 0.7059,
|
926 |
+
"step": 131
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 3.7714285714285714,
|
930 |
+
"grad_norm": 4.977499267065962,
|
931 |
+
"learning_rate": 3.612985456190778e-07,
|
932 |
+
"loss": 0.7022,
|
933 |
+
"step": 132
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 3.8,
|
937 |
+
"grad_norm": 4.629091116356692,
|
938 |
+
"learning_rate": 3.533749813077677e-07,
|
939 |
+
"loss": 0.692,
|
940 |
+
"step": 133
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 3.8285714285714287,
|
944 |
+
"grad_norm": 5.180269168283943,
|
945 |
+
"learning_rate": 3.454915028125263e-07,
|
946 |
+
"loss": 0.7245,
|
947 |
+
"step": 134
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 3.857142857142857,
|
951 |
+
"grad_norm": 4.804587126213361,
|
952 |
+
"learning_rate": 3.3765026539765827e-07,
|
953 |
+
"loss": 0.6965,
|
954 |
+
"step": 135
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 3.8857142857142857,
|
958 |
+
"grad_norm": 5.24711927553455,
|
959 |
+
"learning_rate": 3.2985341277917846e-07,
|
960 |
+
"loss": 0.7271,
|
961 |
+
"step": 136
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 3.914285714285714,
|
965 |
+
"grad_norm": 4.306919498942911,
|
966 |
+
"learning_rate": 3.221030765387417e-07,
|
967 |
+
"loss": 0.7287,
|
968 |
+
"step": 137
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 3.942857142857143,
|
972 |
+
"grad_norm": 5.1155454022030815,
|
973 |
+
"learning_rate": 3.1440137554088953e-07,
|
974 |
+
"loss": 0.7441,
|
975 |
+
"step": 138
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 3.9714285714285715,
|
979 |
+
"grad_norm": 4.184043691688485,
|
980 |
+
"learning_rate": 3.06750415353774e-07,
|
981 |
+
"loss": 0.7349,
|
982 |
+
"step": 139
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 4.0,
|
986 |
+
"grad_norm": 4.262350945325998,
|
987 |
+
"learning_rate": 2.9915228767351535e-07,
|
988 |
+
"loss": 0.7377,
|
989 |
+
"step": 140
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 4.0285714285714285,
|
993 |
+
"grad_norm": 4.103707140569048,
|
994 |
+
"learning_rate": 2.916090697523549e-07,
|
995 |
+
"loss": 0.6864,
|
996 |
+
"step": 141
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 4.057142857142857,
|
1000 |
+
"grad_norm": 4.9358404348712765,
|
1001 |
+
"learning_rate": 2.841228238307536e-07,
|
1002 |
+
"loss": 0.7073,
|
1003 |
+
"step": 142
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 4.085714285714285,
|
1007 |
+
"grad_norm": 4.224580170912438,
|
1008 |
+
"learning_rate": 2.7669559657359673e-07,
|
1009 |
+
"loss": 0.6669,
|
1010 |
+
"step": 143
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 4.114285714285714,
|
1014 |
+
"grad_norm": 4.571605759259632,
|
1015 |
+
"learning_rate": 2.6932941851065615e-07,
|
1016 |
+
"loss": 0.6673,
|
1017 |
+
"step": 144
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 4.142857142857143,
|
1021 |
+
"grad_norm": 4.161244011003443,
|
1022 |
+
"learning_rate": 2.620263034814632e-07,
|
1023 |
+
"loss": 0.6644,
|
1024 |
+
"step": 145
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 4.171428571428572,
|
1028 |
+
"grad_norm": 4.128269627797568,
|
1029 |
+
"learning_rate": 2.547882480847461e-07,
|
1030 |
+
"loss": 0.6568,
|
1031 |
+
"step": 146
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 4.2,
|
1035 |
+
"grad_norm": 4.562112299679241,
|
1036 |
+
"learning_rate": 2.476172311325783e-07,
|
1037 |
+
"loss": 0.7048,
|
1038 |
+
"step": 147
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 4.228571428571429,
|
1042 |
+
"grad_norm": 5.336142167342819,
|
1043 |
+
"learning_rate": 2.4051521310939254e-07,
|
1044 |
+
"loss": 0.7082,
|
1045 |
+
"step": 148
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 4.257142857142857,
|
1049 |
+
"grad_norm": 4.796197499964798,
|
1050 |
+
"learning_rate": 2.3348413563600323e-07,
|
1051 |
+
"loss": 0.6497,
|
1052 |
+
"step": 149
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 4.285714285714286,
|
1056 |
+
"grad_norm": 4.825647232190606,
|
1057 |
+
"learning_rate": 2.2652592093878665e-07,
|
1058 |
+
"loss": 0.6851,
|
1059 |
+
"step": 150
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 4.314285714285714,
|
1063 |
+
"grad_norm": 4.462461397938811,
|
1064 |
+
"learning_rate": 2.1964247132416368e-07,
|
1065 |
+
"loss": 0.6974,
|
1066 |
+
"step": 151
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 4.3428571428571425,
|
1070 |
+
"grad_norm": 4.829646877353178,
|
1071 |
+
"learning_rate": 2.128356686585282e-07,
|
1072 |
+
"loss": 0.6439,
|
1073 |
+
"step": 152
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 4.371428571428572,
|
1077 |
+
"grad_norm": 5.043886293339906,
|
1078 |
+
"learning_rate": 2.0610737385376348e-07,
|
1079 |
+
"loss": 0.7215,
|
1080 |
+
"step": 153
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 4.4,
|
1084 |
+
"grad_norm": 4.598743427591961,
|
1085 |
+
"learning_rate": 1.9945942635848745e-07,
|
1086 |
+
"loss": 0.644,
|
1087 |
+
"step": 154
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 4.428571428571429,
|
1091 |
+
"grad_norm": 3.8253679810994012,
|
1092 |
+
"learning_rate": 1.9289364365516607e-07,
|
1093 |
+
"loss": 0.6397,
|
1094 |
+
"step": 155
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 4.457142857142857,
|
1098 |
+
"grad_norm": 4.807293265176188,
|
1099 |
+
"learning_rate": 1.8641182076323148e-07,
|
1100 |
+
"loss": 0.689,
|
1101 |
+
"step": 156
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 4.485714285714286,
|
1105 |
+
"grad_norm": 3.8062822348294363,
|
1106 |
+
"learning_rate": 1.8001572974834168e-07,
|
1107 |
+
"loss": 0.6466,
|
1108 |
+
"step": 157
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 4.514285714285714,
|
1112 |
+
"grad_norm": 4.883594185819761,
|
1113 |
+
"learning_rate": 1.7370711923791564e-07,
|
1114 |
+
"loss": 0.6691,
|
1115 |
+
"step": 158
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 4.542857142857143,
|
1119 |
+
"grad_norm": 5.143361513426633,
|
1120 |
+
"learning_rate": 1.674877139430758e-07,
|
1121 |
+
"loss": 0.6432,
|
1122 |
+
"step": 159
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 4.571428571428571,
|
1126 |
+
"grad_norm": 4.99547680018691,
|
1127 |
+
"learning_rate": 1.6135921418712955e-07,
|
1128 |
+
"loss": 0.6685,
|
1129 |
+
"step": 160
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 4.6,
|
1133 |
+
"grad_norm": 4.6838462077186795,
|
1134 |
+
"learning_rate": 1.553232954407171e-07,
|
1135 |
+
"loss": 0.657,
|
1136 |
+
"step": 161
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 4.628571428571428,
|
1140 |
+
"grad_norm": 4.841965011015589,
|
1141 |
+
"learning_rate": 1.493816078637557e-07,
|
1142 |
+
"loss": 0.6671,
|
1143 |
+
"step": 162
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 4.6571428571428575,
|
1147 |
+
"grad_norm": 5.2138374952108535,
|
1148 |
+
"learning_rate": 1.435357758543015e-07,
|
1149 |
+
"loss": 0.7393,
|
1150 |
+
"step": 163
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 4.685714285714286,
|
1154 |
+
"grad_norm": 4.7664344007331225,
|
1155 |
+
"learning_rate": 1.3778739760445552e-07,
|
1156 |
+
"loss": 0.6904,
|
1157 |
+
"step": 164
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 4.714285714285714,
|
1161 |
+
"grad_norm": 4.757741131293643,
|
1162 |
+
"learning_rate": 1.321380446634342e-07,
|
1163 |
+
"loss": 0.7093,
|
1164 |
+
"step": 165
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 4.742857142857143,
|
1168 |
+
"grad_norm": 5.064771039141748,
|
1169 |
+
"learning_rate": 1.2658926150792322e-07,
|
1170 |
+
"loss": 0.6588,
|
1171 |
+
"step": 166
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 4.771428571428571,
|
1175 |
+
"grad_norm": 4.745894552114562,
|
1176 |
+
"learning_rate": 1.2114256511983274e-07,
|
1177 |
+
"loss": 0.6914,
|
1178 |
+
"step": 167
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 4.8,
|
1182 |
+
"grad_norm": 4.5842953923274425,
|
1183 |
+
"learning_rate": 1.1579944457157059e-07,
|
1184 |
+
"loss": 0.6819,
|
1185 |
+
"step": 168
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 4.828571428571428,
|
1189 |
+
"grad_norm": 6.02007293004415,
|
1190 |
+
"learning_rate": 1.1056136061894384e-07,
|
1191 |
+
"loss": 0.7107,
|
1192 |
+
"step": 169
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 4.857142857142857,
|
1196 |
+
"grad_norm": 4.04792450108683,
|
1197 |
+
"learning_rate": 1.0542974530180327e-07,
|
1198 |
+
"loss": 0.6577,
|
1199 |
+
"step": 170
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 4.885714285714286,
|
1203 |
+
"grad_norm": 4.334397719277511,
|
1204 |
+
"learning_rate": 1.0040600155253764e-07,
|
1205 |
+
"loss": 0.6926,
|
1206 |
+
"step": 171
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 4.914285714285715,
|
1210 |
+
"grad_norm": 4.763783176385905,
|
1211 |
+
"learning_rate": 9.549150281252632e-08,
|
1212 |
+
"loss": 0.6824,
|
1213 |
+
"step": 172
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 4.942857142857143,
|
1217 |
+
"grad_norm": 3.8987128530052075,
|
1218 |
+
"learning_rate": 9.068759265665382e-08,
|
1219 |
+
"loss": 0.6293,
|
1220 |
+
"step": 173
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 4.9714285714285715,
|
1224 |
+
"grad_norm": 5.260545837340523,
|
1225 |
+
"learning_rate": 8.599558442598998e-08,
|
1226 |
+
"loss": 0.7605,
|
1227 |
+
"step": 174
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 5.0,
|
1231 |
+
"grad_norm": 4.28845521901137,
|
1232 |
+
"learning_rate": 8.141676086873573e-08,
|
1233 |
+
"loss": 0.6685,
|
1234 |
+
"step": 175
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 5.0285714285714285,
|
1238 |
+
"grad_norm": 4.317069591639633,
|
1239 |
+
"learning_rate": 7.695237378953224e-08,
|
1240 |
+
"loss": 0.6178,
|
1241 |
+
"step": 176
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 5.057142857142857,
|
1245 |
+
"grad_norm": 4.673363595181425,
|
1246 |
+
"learning_rate": 7.260364370723043e-08,
|
1247 |
+
"loss": 0.6338,
|
1248 |
+
"step": 177
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 5.085714285714285,
|
1252 |
+
"grad_norm": 4.179297774765176,
|
1253 |
+
"learning_rate": 6.837175952121304e-08,
|
1254 |
+
"loss": 0.656,
|
1255 |
+
"step": 178
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 5.114285714285714,
|
1259 |
+
"grad_norm": 4.2740125604739685,
|
1260 |
+
"learning_rate": 6.42578781863613e-08,
|
1261 |
+
"loss": 0.7062,
|
1262 |
+
"step": 179
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 5.142857142857143,
|
1266 |
+
"grad_norm": 3.7179175942036378,
|
1267 |
+
"learning_rate": 6.026312439675551e-08,
|
1268 |
+
"loss": 0.6132,
|
1269 |
+
"step": 180
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 5.171428571428572,
|
1273 |
+
"grad_norm": 4.586691811463999,
|
1274 |
+
"learning_rate": 5.638859027819409e-08,
|
1275 |
+
"loss": 0.6846,
|
1276 |
+
"step": 181
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 5.2,
|
1280 |
+
"grad_norm": 4.74824620606461,
|
1281 |
+
"learning_rate": 5.263533508961826e-08,
|
1282 |
+
"loss": 0.695,
|
1283 |
+
"step": 182
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 5.228571428571429,
|
1287 |
+
"grad_norm": 4.552865468215117,
|
1288 |
+
"learning_rate": 4.9004384933520547e-08,
|
1289 |
+
"loss": 0.6363,
|
1290 |
+
"step": 183
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 5.257142857142857,
|
1294 |
+
"grad_norm": 3.868803646466012,
|
1295 |
+
"learning_rate": 4.549673247541874e-08,
|
1296 |
+
"loss": 0.6197,
|
1297 |
+
"step": 184
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 5.285714285714286,
|
1301 |
+
"grad_norm": 4.257422013394126,
|
1302 |
+
"learning_rate": 4.2113336672471245e-08,
|
1303 |
+
"loss": 0.6325,
|
1304 |
+
"step": 185
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 5.314285714285714,
|
1308 |
+
"grad_norm": 4.138411901026596,
|
1309 |
+
"learning_rate": 3.8855122511307626e-08,
|
1310 |
+
"loss": 0.6074,
|
1311 |
+
"step": 186
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 5.3428571428571425,
|
1315 |
+
"grad_norm": 4.386100060157392,
|
1316 |
+
"learning_rate": 3.572298075514652e-08,
|
1317 |
+
"loss": 0.6298,
|
1318 |
+
"step": 187
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 5.371428571428572,
|
1322 |
+
"grad_norm": 3.899133464905973,
|
1323 |
+
"learning_rate": 3.271776770026963e-08,
|
1324 |
+
"loss": 0.5991,
|
1325 |
+
"step": 188
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 5.4,
|
1329 |
+
"grad_norm": 5.227219263192165,
|
1330 |
+
"learning_rate": 2.9840304941919416e-08,
|
1331 |
+
"loss": 0.6917,
|
1332 |
+
"step": 189
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 5.428571428571429,
|
1336 |
+
"grad_norm": 4.194540185559828,
|
1337 |
+
"learning_rate": 2.7091379149682682e-08,
|
1338 |
+
"loss": 0.6592,
|
1339 |
+
"step": 190
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 5.457142857142857,
|
1343 |
+
"grad_norm": 4.103130224713111,
|
1344 |
+
"learning_rate": 2.4471741852423233e-08,
|
1345 |
+
"loss": 0.6467,
|
1346 |
+
"step": 191
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 5.485714285714286,
|
1350 |
+
"grad_norm": 4.401362210384649,
|
1351 |
+
"learning_rate": 2.1982109232821176e-08,
|
1352 |
+
"loss": 0.6471,
|
1353 |
+
"step": 192
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 5.514285714285714,
|
1357 |
+
"grad_norm": 4.4025758572699365,
|
1358 |
+
"learning_rate": 1.9623161931575926e-08,
|
1359 |
+
"loss": 0.6537,
|
1360 |
+
"step": 193
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 5.542857142857143,
|
1364 |
+
"grad_norm": 4.940644969360135,
|
1365 |
+
"learning_rate": 1.7395544861325718e-08,
|
1366 |
+
"loss": 0.6647,
|
1367 |
+
"step": 194
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 5.571428571428571,
|
1371 |
+
"grad_norm": 4.562635610658606,
|
1372 |
+
"learning_rate": 1.5299867030334813e-08,
|
1373 |
+
"loss": 0.6422,
|
1374 |
+
"step": 195
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 5.6,
|
1378 |
+
"grad_norm": 4.58494040795585,
|
1379 |
+
"learning_rate": 1.3336701375997127e-08,
|
1380 |
+
"loss": 0.7141,
|
1381 |
+
"step": 196
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 5.628571428571428,
|
1385 |
+
"grad_norm": 4.240912508697285,
|
1386 |
+
"learning_rate": 1.1506584608200364e-08,
|
1387 |
+
"loss": 0.6423,
|
1388 |
+
"step": 197
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 5.6571428571428575,
|
1392 |
+
"grad_norm": 4.474353548561071,
|
1393 |
+
"learning_rate": 9.810017062595321e-09,
|
1394 |
+
"loss": 0.6362,
|
1395 |
+
"step": 198
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 5.685714285714286,
|
1399 |
+
"grad_norm": 5.263243471264244,
|
1400 |
+
"learning_rate": 8.247462563808816e-09,
|
1401 |
+
"loss": 0.6436,
|
1402 |
+
"step": 199
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 5.714285714285714,
|
1406 |
+
"grad_norm": 5.629514926131208,
|
1407 |
+
"learning_rate": 6.819348298638839e-09,
|
1408 |
+
"loss": 0.6434,
|
1409 |
+
"step": 200
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 5.714285714285714,
|
1413 |
+
"eval_loss": 1.0673894882202148,
|
1414 |
+
"eval_runtime": 1.3433,
|
1415 |
+
"eval_samples_per_second": 17.122,
|
1416 |
+
"eval_steps_per_second": 0.744,
|
1417 |
+
"step": 200
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 5.742857142857143,
|
1421 |
+
"grad_norm": 4.388069390318575,
|
1422 |
+
"learning_rate": 5.526064699265753e-09,
|
1423 |
+
"loss": 0.6946,
|
1424 |
+
"step": 201
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 5.771428571428571,
|
1428 |
+
"grad_norm": 3.85292435393061,
|
1429 |
+
"learning_rate": 4.367965336512403e-09,
|
1430 |
+
"loss": 0.6113,
|
1431 |
+
"step": 202
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 5.8,
|
1435 |
+
"grad_norm": 4.3104727339441755,
|
1436 |
+
"learning_rate": 3.3453668231809283e-09,
|
1437 |
+
"loss": 0.6493,
|
1438 |
+
"step": 203
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 5.828571428571428,
|
1442 |
+
"grad_norm": 4.21000584494849,
|
1443 |
+
"learning_rate": 2.458548727494292e-09,
|
1444 |
+
"loss": 0.653,
|
1445 |
+
"step": 204
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 5.857142857142857,
|
1449 |
+
"grad_norm": 5.114411845255924,
|
1450 |
+
"learning_rate": 1.7077534966650765e-09,
|
1451 |
+
"loss": 0.6527,
|
1452 |
+
"step": 205
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 5.885714285714286,
|
1456 |
+
"grad_norm": 4.988848182549562,
|
1457 |
+
"learning_rate": 1.0931863906127325e-09,
|
1458 |
+
"loss": 0.6935,
|
1459 |
+
"step": 206
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 5.914285714285715,
|
1463 |
+
"grad_norm": 4.005152725886777,
|
1464 |
+
"learning_rate": 6.150154258476314e-10,
|
1465 |
+
"loss": 0.6559,
|
1466 |
+
"step": 207
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 5.942857142857143,
|
1470 |
+
"grad_norm": 6.395756261520298,
|
1471 |
+
"learning_rate": 2.733713295369755e-10,
|
1472 |
+
"loss": 0.693,
|
1473 |
+
"step": 208
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 5.9714285714285715,
|
1477 |
+
"grad_norm": 5.465002272924693,
|
1478 |
+
"learning_rate": 6.834750376549791e-11,
|
1479 |
+
"loss": 0.7231,
|
1480 |
+
"step": 209
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 6.0,
|
1484 |
+
"grad_norm": 4.273940345648053,
|
1485 |
+
"learning_rate": 0.0,
|
1486 |
+
"loss": 0.6288,
|
1487 |
+
"step": 210
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 6.0,
|
1491 |
+
"step": 210,
|
1492 |
+
"total_flos": 18587454013440.0,
|
1493 |
+
"train_loss": 0.8877790099098569,
|
1494 |
+
"train_runtime": 1031.8496,
|
1495 |
+
"train_samples_per_second": 12.932,
|
1496 |
+
"train_steps_per_second": 0.204
|
1497 |
+
}
|
1498 |
+
],
|
1499 |
+
"logging_steps": 1.0,
|
1500 |
+
"max_steps": 210,
|
1501 |
+
"num_input_tokens_seen": 0,
|
1502 |
+
"num_train_epochs": 6,
|
1503 |
+
"save_steps": 1000,
|
1504 |
+
"stateful_callbacks": {
|
1505 |
+
"TrainerControl": {
|
1506 |
+
"args": {
|
1507 |
+
"should_epoch_stop": false,
|
1508 |
+
"should_evaluate": false,
|
1509 |
+
"should_log": false,
|
1510 |
+
"should_save": true,
|
1511 |
+
"should_training_stop": true
|
1512 |
+
},
|
1513 |
+
"attributes": {}
|
1514 |
+
}
|
1515 |
+
},
|
1516 |
+
"total_flos": 18587454013440.0,
|
1517 |
+
"train_batch_size": 8,
|
1518 |
+
"trial_name": null,
|
1519 |
+
"trial_params": null
|
1520 |
+
}
|