Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1573 -0
trainer_state.json
ADDED
@@ -0,0 +1,1573 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.9730941704035874,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 220,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.017937219730941704,
|
13 |
+
"grad_norm": 1.1350524425506592,
|
14 |
+
"learning_rate": 4e-05,
|
15 |
+
"loss": 2.626,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.03587443946188341,
|
20 |
+
"grad_norm": 1.1098030805587769,
|
21 |
+
"learning_rate": 8e-05,
|
22 |
+
"loss": 2.564,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.053811659192825115,
|
27 |
+
"grad_norm": 1.004073143005371,
|
28 |
+
"learning_rate": 0.00012,
|
29 |
+
"loss": 2.5371,
|
30 |
+
"step": 3
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.07174887892376682,
|
34 |
+
"grad_norm": 1.0235692262649536,
|
35 |
+
"learning_rate": 0.00016,
|
36 |
+
"loss": 2.3217,
|
37 |
+
"step": 4
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.08968609865470852,
|
41 |
+
"grad_norm": 0.9896207451820374,
|
42 |
+
"learning_rate": 0.0002,
|
43 |
+
"loss": 2.1635,
|
44 |
+
"step": 5
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.10762331838565023,
|
48 |
+
"grad_norm": 1.0160284042358398,
|
49 |
+
"learning_rate": 0.00019906976744186048,
|
50 |
+
"loss": 1.9974,
|
51 |
+
"step": 6
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.12556053811659193,
|
55 |
+
"grad_norm": 1.6710572242736816,
|
56 |
+
"learning_rate": 0.00019813953488372096,
|
57 |
+
"loss": 1.721,
|
58 |
+
"step": 7
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.14349775784753363,
|
62 |
+
"grad_norm": 1.414752721786499,
|
63 |
+
"learning_rate": 0.0001972093023255814,
|
64 |
+
"loss": 1.5117,
|
65 |
+
"step": 8
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.16143497757847533,
|
69 |
+
"grad_norm": 1.0044751167297363,
|
70 |
+
"learning_rate": 0.00019627906976744185,
|
71 |
+
"loss": 1.3689,
|
72 |
+
"step": 9
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.17937219730941703,
|
76 |
+
"grad_norm": 0.8621335029602051,
|
77 |
+
"learning_rate": 0.00019534883720930232,
|
78 |
+
"loss": 1.3251,
|
79 |
+
"step": 10
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.19730941704035873,
|
83 |
+
"grad_norm": 0.7370575666427612,
|
84 |
+
"learning_rate": 0.0001944186046511628,
|
85 |
+
"loss": 1.2459,
|
86 |
+
"step": 11
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.21524663677130046,
|
90 |
+
"grad_norm": 0.7463206648826599,
|
91 |
+
"learning_rate": 0.00019348837209302326,
|
92 |
+
"loss": 1.1113,
|
93 |
+
"step": 12
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.23318385650224216,
|
97 |
+
"grad_norm": 0.9221929907798767,
|
98 |
+
"learning_rate": 0.00019255813953488374,
|
99 |
+
"loss": 1.1779,
|
100 |
+
"step": 13
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.25112107623318386,
|
104 |
+
"grad_norm": 0.8653731346130371,
|
105 |
+
"learning_rate": 0.0001916279069767442,
|
106 |
+
"loss": 1.056,
|
107 |
+
"step": 14
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.26905829596412556,
|
111 |
+
"grad_norm": 0.8368218541145325,
|
112 |
+
"learning_rate": 0.00019069767441860466,
|
113 |
+
"loss": 1.0355,
|
114 |
+
"step": 15
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.28699551569506726,
|
118 |
+
"grad_norm": 0.9069833755493164,
|
119 |
+
"learning_rate": 0.00018976744186046513,
|
120 |
+
"loss": 1.076,
|
121 |
+
"step": 16
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.30493273542600896,
|
125 |
+
"grad_norm": 1.010295033454895,
|
126 |
+
"learning_rate": 0.00018883720930232557,
|
127 |
+
"loss": 1.037,
|
128 |
+
"step": 17
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.32286995515695066,
|
132 |
+
"grad_norm": 0.9165616631507874,
|
133 |
+
"learning_rate": 0.00018790697674418605,
|
134 |
+
"loss": 1.0841,
|
135 |
+
"step": 18
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.34080717488789236,
|
139 |
+
"grad_norm": 1.2438362836837769,
|
140 |
+
"learning_rate": 0.00018697674418604652,
|
141 |
+
"loss": 0.928,
|
142 |
+
"step": 19
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.35874439461883406,
|
146 |
+
"grad_norm": 1.1386940479278564,
|
147 |
+
"learning_rate": 0.000186046511627907,
|
148 |
+
"loss": 0.9742,
|
149 |
+
"step": 20
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.37668161434977576,
|
153 |
+
"grad_norm": 1.0614705085754395,
|
154 |
+
"learning_rate": 0.00018511627906976744,
|
155 |
+
"loss": 0.986,
|
156 |
+
"step": 21
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.39461883408071746,
|
160 |
+
"grad_norm": 1.1421048641204834,
|
161 |
+
"learning_rate": 0.0001841860465116279,
|
162 |
+
"loss": 0.9306,
|
163 |
+
"step": 22
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.4125560538116592,
|
167 |
+
"grad_norm": 0.9451465606689453,
|
168 |
+
"learning_rate": 0.00018325581395348838,
|
169 |
+
"loss": 0.9888,
|
170 |
+
"step": 23
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.4304932735426009,
|
174 |
+
"grad_norm": 0.753145158290863,
|
175 |
+
"learning_rate": 0.00018232558139534886,
|
176 |
+
"loss": 0.9598,
|
177 |
+
"step": 24
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.4484304932735426,
|
181 |
+
"grad_norm": 0.6006896495819092,
|
182 |
+
"learning_rate": 0.0001813953488372093,
|
183 |
+
"loss": 0.8188,
|
184 |
+
"step": 25
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.4663677130044843,
|
188 |
+
"grad_norm": 0.6499263048171997,
|
189 |
+
"learning_rate": 0.00018046511627906977,
|
190 |
+
"loss": 0.9449,
|
191 |
+
"step": 26
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.484304932735426,
|
195 |
+
"grad_norm": 0.6340591907501221,
|
196 |
+
"learning_rate": 0.00017953488372093025,
|
197 |
+
"loss": 1.0643,
|
198 |
+
"step": 27
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.5022421524663677,
|
202 |
+
"grad_norm": 0.7478179335594177,
|
203 |
+
"learning_rate": 0.0001786046511627907,
|
204 |
+
"loss": 0.9822,
|
205 |
+
"step": 28
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.5201793721973094,
|
209 |
+
"grad_norm": 0.6700637936592102,
|
210 |
+
"learning_rate": 0.00017767441860465117,
|
211 |
+
"loss": 0.9793,
|
212 |
+
"step": 29
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.5381165919282511,
|
216 |
+
"grad_norm": 0.6026176810264587,
|
217 |
+
"learning_rate": 0.00017674418604651164,
|
218 |
+
"loss": 1.0611,
|
219 |
+
"step": 30
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.5560538116591929,
|
223 |
+
"grad_norm": 0.5661296248435974,
|
224 |
+
"learning_rate": 0.0001758139534883721,
|
225 |
+
"loss": 0.9952,
|
226 |
+
"step": 31
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.5739910313901345,
|
230 |
+
"grad_norm": 0.6180285811424255,
|
231 |
+
"learning_rate": 0.00017488372093023258,
|
232 |
+
"loss": 0.9497,
|
233 |
+
"step": 32
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.5919282511210763,
|
237 |
+
"grad_norm": 0.6067416667938232,
|
238 |
+
"learning_rate": 0.00017395348837209303,
|
239 |
+
"loss": 0.9015,
|
240 |
+
"step": 33
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.6098654708520179,
|
244 |
+
"grad_norm": 0.6353489756584167,
|
245 |
+
"learning_rate": 0.00017302325581395348,
|
246 |
+
"loss": 0.9297,
|
247 |
+
"step": 34
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.6278026905829597,
|
251 |
+
"grad_norm": 0.6017511487007141,
|
252 |
+
"learning_rate": 0.00017209302325581395,
|
253 |
+
"loss": 0.8634,
|
254 |
+
"step": 35
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.6457399103139013,
|
258 |
+
"grad_norm": 0.701750636100769,
|
259 |
+
"learning_rate": 0.00017116279069767442,
|
260 |
+
"loss": 0.9088,
|
261 |
+
"step": 36
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.6636771300448431,
|
265 |
+
"grad_norm": 0.6852689385414124,
|
266 |
+
"learning_rate": 0.0001702325581395349,
|
267 |
+
"loss": 0.9195,
|
268 |
+
"step": 37
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.6816143497757847,
|
272 |
+
"grad_norm": 0.6971113681793213,
|
273 |
+
"learning_rate": 0.00016930232558139537,
|
274 |
+
"loss": 0.8599,
|
275 |
+
"step": 38
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.6995515695067265,
|
279 |
+
"grad_norm": 0.6576591730117798,
|
280 |
+
"learning_rate": 0.00016837209302325584,
|
281 |
+
"loss": 0.764,
|
282 |
+
"step": 39
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.7174887892376681,
|
286 |
+
"grad_norm": 0.8312844038009644,
|
287 |
+
"learning_rate": 0.00016744186046511629,
|
288 |
+
"loss": 0.8577,
|
289 |
+
"step": 40
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.7354260089686099,
|
293 |
+
"grad_norm": 0.7586076259613037,
|
294 |
+
"learning_rate": 0.00016651162790697673,
|
295 |
+
"loss": 0.8069,
|
296 |
+
"step": 41
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.7533632286995515,
|
300 |
+
"grad_norm": 0.6356410384178162,
|
301 |
+
"learning_rate": 0.0001655813953488372,
|
302 |
+
"loss": 0.7753,
|
303 |
+
"step": 42
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.7713004484304933,
|
307 |
+
"grad_norm": 0.6421555280685425,
|
308 |
+
"learning_rate": 0.00016465116279069768,
|
309 |
+
"loss": 0.8733,
|
310 |
+
"step": 43
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.7892376681614349,
|
314 |
+
"grad_norm": 0.8002834916114807,
|
315 |
+
"learning_rate": 0.00016372093023255815,
|
316 |
+
"loss": 1.0058,
|
317 |
+
"step": 44
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.8071748878923767,
|
321 |
+
"grad_norm": 0.6567667126655579,
|
322 |
+
"learning_rate": 0.00016279069767441862,
|
323 |
+
"loss": 0.8771,
|
324 |
+
"step": 45
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.8251121076233184,
|
328 |
+
"grad_norm": 0.5926035642623901,
|
329 |
+
"learning_rate": 0.00016186046511627907,
|
330 |
+
"loss": 0.8556,
|
331 |
+
"step": 46
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.8430493273542601,
|
335 |
+
"grad_norm": 0.613197922706604,
|
336 |
+
"learning_rate": 0.00016093023255813954,
|
337 |
+
"loss": 0.85,
|
338 |
+
"step": 47
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.8609865470852018,
|
342 |
+
"grad_norm": 0.7108270525932312,
|
343 |
+
"learning_rate": 0.00016,
|
344 |
+
"loss": 0.8383,
|
345 |
+
"step": 48
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.8789237668161435,
|
349 |
+
"grad_norm": 0.6039162874221802,
|
350 |
+
"learning_rate": 0.00015906976744186046,
|
351 |
+
"loss": 0.898,
|
352 |
+
"step": 49
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.8968609865470852,
|
356 |
+
"grad_norm": 0.6543579697608948,
|
357 |
+
"learning_rate": 0.00015813953488372093,
|
358 |
+
"loss": 0.8442,
|
359 |
+
"step": 50
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.9147982062780269,
|
363 |
+
"grad_norm": 0.6246331334114075,
|
364 |
+
"learning_rate": 0.0001572093023255814,
|
365 |
+
"loss": 0.7756,
|
366 |
+
"step": 51
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.9327354260089686,
|
370 |
+
"grad_norm": 0.6133050322532654,
|
371 |
+
"learning_rate": 0.00015627906976744188,
|
372 |
+
"loss": 0.8701,
|
373 |
+
"step": 52
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.9506726457399103,
|
377 |
+
"grad_norm": 0.6625930070877075,
|
378 |
+
"learning_rate": 0.00015534883720930232,
|
379 |
+
"loss": 0.9818,
|
380 |
+
"step": 53
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.968609865470852,
|
384 |
+
"grad_norm": 0.6724585294723511,
|
385 |
+
"learning_rate": 0.0001544186046511628,
|
386 |
+
"loss": 0.957,
|
387 |
+
"step": 54
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.9865470852017937,
|
391 |
+
"grad_norm": 0.5864427089691162,
|
392 |
+
"learning_rate": 0.00015348837209302327,
|
393 |
+
"loss": 0.7588,
|
394 |
+
"step": 55
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.0134529147982063,
|
398 |
+
"grad_norm": 1.4326505661010742,
|
399 |
+
"learning_rate": 0.00015255813953488374,
|
400 |
+
"loss": 1.2876,
|
401 |
+
"step": 56
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.031390134529148,
|
405 |
+
"grad_norm": 0.7354760766029358,
|
406 |
+
"learning_rate": 0.0001516279069767442,
|
407 |
+
"loss": 0.6682,
|
408 |
+
"step": 57
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.0493273542600896,
|
412 |
+
"grad_norm": 0.7175352573394775,
|
413 |
+
"learning_rate": 0.00015069767441860466,
|
414 |
+
"loss": 0.6862,
|
415 |
+
"step": 58
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.0672645739910314,
|
419 |
+
"grad_norm": 0.7436112761497498,
|
420 |
+
"learning_rate": 0.0001497674418604651,
|
421 |
+
"loss": 0.7319,
|
422 |
+
"step": 59
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.0852017937219731,
|
426 |
+
"grad_norm": 0.7228017449378967,
|
427 |
+
"learning_rate": 0.00014883720930232558,
|
428 |
+
"loss": 0.842,
|
429 |
+
"step": 60
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.1031390134529149,
|
433 |
+
"grad_norm": 0.7761241793632507,
|
434 |
+
"learning_rate": 0.00014790697674418605,
|
435 |
+
"loss": 0.6666,
|
436 |
+
"step": 61
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.1210762331838564,
|
440 |
+
"grad_norm": 0.8203029632568359,
|
441 |
+
"learning_rate": 0.00014697674418604652,
|
442 |
+
"loss": 0.7403,
|
443 |
+
"step": 62
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.1390134529147982,
|
447 |
+
"grad_norm": 0.7372130751609802,
|
448 |
+
"learning_rate": 0.000146046511627907,
|
449 |
+
"loss": 0.7426,
|
450 |
+
"step": 63
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.15695067264574,
|
454 |
+
"grad_norm": 0.7927672863006592,
|
455 |
+
"learning_rate": 0.00014511627906976747,
|
456 |
+
"loss": 0.7717,
|
457 |
+
"step": 64
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.1748878923766817,
|
461 |
+
"grad_norm": 0.7056854367256165,
|
462 |
+
"learning_rate": 0.00014418604651162791,
|
463 |
+
"loss": 0.6855,
|
464 |
+
"step": 65
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.1928251121076232,
|
468 |
+
"grad_norm": 0.8383380174636841,
|
469 |
+
"learning_rate": 0.00014325581395348836,
|
470 |
+
"loss": 0.7222,
|
471 |
+
"step": 66
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.210762331838565,
|
475 |
+
"grad_norm": 0.6172988414764404,
|
476 |
+
"learning_rate": 0.00014232558139534883,
|
477 |
+
"loss": 0.6318,
|
478 |
+
"step": 67
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.2286995515695067,
|
482 |
+
"grad_norm": 0.7639912962913513,
|
483 |
+
"learning_rate": 0.0001413953488372093,
|
484 |
+
"loss": 0.6523,
|
485 |
+
"step": 68
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.2466367713004485,
|
489 |
+
"grad_norm": 0.7420268654823303,
|
490 |
+
"learning_rate": 0.00014046511627906978,
|
491 |
+
"loss": 0.7358,
|
492 |
+
"step": 69
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.2645739910313902,
|
496 |
+
"grad_norm": 0.781150758266449,
|
497 |
+
"learning_rate": 0.00013953488372093025,
|
498 |
+
"loss": 0.6927,
|
499 |
+
"step": 70
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.2825112107623318,
|
503 |
+
"grad_norm": 0.7435458302497864,
|
504 |
+
"learning_rate": 0.00013860465116279072,
|
505 |
+
"loss": 0.7508,
|
506 |
+
"step": 71
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.3004484304932735,
|
510 |
+
"grad_norm": 0.7637338042259216,
|
511 |
+
"learning_rate": 0.00013767441860465117,
|
512 |
+
"loss": 0.5798,
|
513 |
+
"step": 72
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.3183856502242153,
|
517 |
+
"grad_norm": 0.9575199484825134,
|
518 |
+
"learning_rate": 0.00013674418604651162,
|
519 |
+
"loss": 0.8045,
|
520 |
+
"step": 73
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.336322869955157,
|
524 |
+
"grad_norm": 0.836318850517273,
|
525 |
+
"learning_rate": 0.0001358139534883721,
|
526 |
+
"loss": 0.7916,
|
527 |
+
"step": 74
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.3542600896860986,
|
531 |
+
"grad_norm": 0.7818664908409119,
|
532 |
+
"learning_rate": 0.00013488372093023256,
|
533 |
+
"loss": 0.7037,
|
534 |
+
"step": 75
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.3721973094170403,
|
538 |
+
"grad_norm": 0.7612494826316833,
|
539 |
+
"learning_rate": 0.00013395348837209303,
|
540 |
+
"loss": 0.6847,
|
541 |
+
"step": 76
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 1.390134529147982,
|
545 |
+
"grad_norm": 0.6829874515533447,
|
546 |
+
"learning_rate": 0.0001330232558139535,
|
547 |
+
"loss": 0.6848,
|
548 |
+
"step": 77
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 1.4080717488789238,
|
552 |
+
"grad_norm": 0.6923062801361084,
|
553 |
+
"learning_rate": 0.00013209302325581395,
|
554 |
+
"loss": 0.6928,
|
555 |
+
"step": 78
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.4260089686098656,
|
559 |
+
"grad_norm": 0.6936827898025513,
|
560 |
+
"learning_rate": 0.00013116279069767442,
|
561 |
+
"loss": 0.8813,
|
562 |
+
"step": 79
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 1.4439461883408071,
|
566 |
+
"grad_norm": 0.7367523908615112,
|
567 |
+
"learning_rate": 0.0001302325581395349,
|
568 |
+
"loss": 0.7386,
|
569 |
+
"step": 80
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.4618834080717489,
|
573 |
+
"grad_norm": 0.7084159255027771,
|
574 |
+
"learning_rate": 0.00012930232558139534,
|
575 |
+
"loss": 0.604,
|
576 |
+
"step": 81
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 1.4798206278026906,
|
580 |
+
"grad_norm": 0.817794144153595,
|
581 |
+
"learning_rate": 0.00012837209302325582,
|
582 |
+
"loss": 0.7644,
|
583 |
+
"step": 82
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 1.4977578475336322,
|
587 |
+
"grad_norm": 0.7807640433311462,
|
588 |
+
"learning_rate": 0.0001274418604651163,
|
589 |
+
"loss": 0.8061,
|
590 |
+
"step": 83
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 1.515695067264574,
|
594 |
+
"grad_norm": 0.7616767883300781,
|
595 |
+
"learning_rate": 0.00012651162790697676,
|
596 |
+
"loss": 0.7786,
|
597 |
+
"step": 84
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.5336322869955157,
|
601 |
+
"grad_norm": 0.7925138473510742,
|
602 |
+
"learning_rate": 0.0001255813953488372,
|
603 |
+
"loss": 0.6859,
|
604 |
+
"step": 85
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 1.5515695067264574,
|
608 |
+
"grad_norm": 0.7205699682235718,
|
609 |
+
"learning_rate": 0.00012465116279069768,
|
610 |
+
"loss": 0.7581,
|
611 |
+
"step": 86
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.5695067264573992,
|
615 |
+
"grad_norm": 0.6984810829162598,
|
616 |
+
"learning_rate": 0.00012372093023255815,
|
617 |
+
"loss": 0.6652,
|
618 |
+
"step": 87
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 1.587443946188341,
|
622 |
+
"grad_norm": 0.7267066836357117,
|
623 |
+
"learning_rate": 0.00012279069767441863,
|
624 |
+
"loss": 0.6419,
|
625 |
+
"step": 88
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 1.6053811659192825,
|
629 |
+
"grad_norm": 0.7686505913734436,
|
630 |
+
"learning_rate": 0.00012186046511627907,
|
631 |
+
"loss": 0.7497,
|
632 |
+
"step": 89
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 1.6233183856502242,
|
636 |
+
"grad_norm": 0.755163311958313,
|
637 |
+
"learning_rate": 0.00012093023255813953,
|
638 |
+
"loss": 0.7757,
|
639 |
+
"step": 90
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 1.6412556053811658,
|
643 |
+
"grad_norm": 0.7927355766296387,
|
644 |
+
"learning_rate": 0.00012,
|
645 |
+
"loss": 0.7754,
|
646 |
+
"step": 91
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 1.6591928251121075,
|
650 |
+
"grad_norm": 0.6950364708900452,
|
651 |
+
"learning_rate": 0.00011906976744186048,
|
652 |
+
"loss": 0.6059,
|
653 |
+
"step": 92
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 1.6771300448430493,
|
657 |
+
"grad_norm": 0.7365448474884033,
|
658 |
+
"learning_rate": 0.00011813953488372094,
|
659 |
+
"loss": 0.673,
|
660 |
+
"step": 93
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 1.695067264573991,
|
664 |
+
"grad_norm": 0.7940488457679749,
|
665 |
+
"learning_rate": 0.00011720930232558141,
|
666 |
+
"loss": 0.5726,
|
667 |
+
"step": 94
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 1.7130044843049328,
|
671 |
+
"grad_norm": 0.8307069540023804,
|
672 |
+
"learning_rate": 0.00011627906976744187,
|
673 |
+
"loss": 0.6949,
|
674 |
+
"step": 95
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 1.7309417040358746,
|
678 |
+
"grad_norm": 0.876649796962738,
|
679 |
+
"learning_rate": 0.00011534883720930234,
|
680 |
+
"loss": 0.6415,
|
681 |
+
"step": 96
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 1.7488789237668163,
|
685 |
+
"grad_norm": 0.9207577109336853,
|
686 |
+
"learning_rate": 0.00011441860465116279,
|
687 |
+
"loss": 0.674,
|
688 |
+
"step": 97
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 1.7668161434977578,
|
692 |
+
"grad_norm": 0.8050037026405334,
|
693 |
+
"learning_rate": 0.00011348837209302326,
|
694 |
+
"loss": 0.5656,
|
695 |
+
"step": 98
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 1.7847533632286996,
|
699 |
+
"grad_norm": 0.878441333770752,
|
700 |
+
"learning_rate": 0.00011255813953488372,
|
701 |
+
"loss": 0.7596,
|
702 |
+
"step": 99
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 1.8026905829596411,
|
706 |
+
"grad_norm": 0.91168612241745,
|
707 |
+
"learning_rate": 0.00011162790697674419,
|
708 |
+
"loss": 0.8177,
|
709 |
+
"step": 100
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 1.8206278026905829,
|
713 |
+
"grad_norm": 0.7757384777069092,
|
714 |
+
"learning_rate": 0.00011069767441860466,
|
715 |
+
"loss": 0.652,
|
716 |
+
"step": 101
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 1.8385650224215246,
|
720 |
+
"grad_norm": 0.8266631960868835,
|
721 |
+
"learning_rate": 0.00010976744186046512,
|
722 |
+
"loss": 0.8207,
|
723 |
+
"step": 102
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 1.8565022421524664,
|
727 |
+
"grad_norm": 0.7787818312644958,
|
728 |
+
"learning_rate": 0.0001088372093023256,
|
729 |
+
"loss": 0.6802,
|
730 |
+
"step": 103
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 1.8744394618834082,
|
734 |
+
"grad_norm": 0.8642717003822327,
|
735 |
+
"learning_rate": 0.00010790697674418607,
|
736 |
+
"loss": 0.7011,
|
737 |
+
"step": 104
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 1.89237668161435,
|
741 |
+
"grad_norm": 0.7537955641746521,
|
742 |
+
"learning_rate": 0.00010697674418604651,
|
743 |
+
"loss": 0.6297,
|
744 |
+
"step": 105
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 1.9103139013452914,
|
748 |
+
"grad_norm": 0.8498083353042603,
|
749 |
+
"learning_rate": 0.00010604651162790697,
|
750 |
+
"loss": 0.6764,
|
751 |
+
"step": 106
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 1.9282511210762332,
|
755 |
+
"grad_norm": 0.8197365403175354,
|
756 |
+
"learning_rate": 0.00010511627906976745,
|
757 |
+
"loss": 0.7126,
|
758 |
+
"step": 107
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 1.9461883408071747,
|
762 |
+
"grad_norm": 0.797406792640686,
|
763 |
+
"learning_rate": 0.0001041860465116279,
|
764 |
+
"loss": 0.7664,
|
765 |
+
"step": 108
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 1.9641255605381165,
|
769 |
+
"grad_norm": 0.8513347506523132,
|
770 |
+
"learning_rate": 0.00010325581395348838,
|
771 |
+
"loss": 0.7572,
|
772 |
+
"step": 109
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 1.9820627802690582,
|
776 |
+
"grad_norm": 0.8408546447753906,
|
777 |
+
"learning_rate": 0.00010232558139534885,
|
778 |
+
"loss": 0.5291,
|
779 |
+
"step": 110
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 2.008968609865471,
|
783 |
+
"grad_norm": 1.9233133792877197,
|
784 |
+
"learning_rate": 0.00010139534883720931,
|
785 |
+
"loss": 1.3532,
|
786 |
+
"step": 111
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 2.0269058295964126,
|
790 |
+
"grad_norm": 0.782871663570404,
|
791 |
+
"learning_rate": 0.00010046511627906978,
|
792 |
+
"loss": 0.4679,
|
793 |
+
"step": 112
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 2.0448430493273544,
|
797 |
+
"grad_norm": 0.7588692307472229,
|
798 |
+
"learning_rate": 9.953488372093024e-05,
|
799 |
+
"loss": 0.4962,
|
800 |
+
"step": 113
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 2.062780269058296,
|
804 |
+
"grad_norm": 0.8635579943656921,
|
805 |
+
"learning_rate": 9.86046511627907e-05,
|
806 |
+
"loss": 0.474,
|
807 |
+
"step": 114
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 2.0807174887892375,
|
811 |
+
"grad_norm": 0.8822935819625854,
|
812 |
+
"learning_rate": 9.767441860465116e-05,
|
813 |
+
"loss": 0.4129,
|
814 |
+
"step": 115
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 2.098654708520179,
|
818 |
+
"grad_norm": 0.9706945419311523,
|
819 |
+
"learning_rate": 9.674418604651163e-05,
|
820 |
+
"loss": 0.4961,
|
821 |
+
"step": 116
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 2.116591928251121,
|
825 |
+
"grad_norm": 1.1239269971847534,
|
826 |
+
"learning_rate": 9.58139534883721e-05,
|
827 |
+
"loss": 0.5805,
|
828 |
+
"step": 117
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 2.1345291479820627,
|
832 |
+
"grad_norm": 1.1263116598129272,
|
833 |
+
"learning_rate": 9.488372093023256e-05,
|
834 |
+
"loss": 0.4252,
|
835 |
+
"step": 118
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 2.1524663677130045,
|
839 |
+
"grad_norm": 1.2772897481918335,
|
840 |
+
"learning_rate": 9.395348837209302e-05,
|
841 |
+
"loss": 0.4534,
|
842 |
+
"step": 119
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 2.1704035874439462,
|
846 |
+
"grad_norm": 1.0017895698547363,
|
847 |
+
"learning_rate": 9.30232558139535e-05,
|
848 |
+
"loss": 0.3835,
|
849 |
+
"step": 120
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 2.188340807174888,
|
853 |
+
"grad_norm": 1.1340886354446411,
|
854 |
+
"learning_rate": 9.209302325581396e-05,
|
855 |
+
"loss": 0.4089,
|
856 |
+
"step": 121
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 2.2062780269058297,
|
860 |
+
"grad_norm": 1.011704921722412,
|
861 |
+
"learning_rate": 9.116279069767443e-05,
|
862 |
+
"loss": 0.4298,
|
863 |
+
"step": 122
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 2.2242152466367715,
|
867 |
+
"grad_norm": 0.9829872250556946,
|
868 |
+
"learning_rate": 9.023255813953489e-05,
|
869 |
+
"loss": 0.4568,
|
870 |
+
"step": 123
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 2.242152466367713,
|
874 |
+
"grad_norm": 1.0418094396591187,
|
875 |
+
"learning_rate": 8.930232558139535e-05,
|
876 |
+
"loss": 0.5086,
|
877 |
+
"step": 124
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 2.2600896860986546,
|
881 |
+
"grad_norm": 0.9814426302909851,
|
882 |
+
"learning_rate": 8.837209302325582e-05,
|
883 |
+
"loss": 0.5033,
|
884 |
+
"step": 125
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 2.2780269058295963,
|
888 |
+
"grad_norm": 1.0506559610366821,
|
889 |
+
"learning_rate": 8.744186046511629e-05,
|
890 |
+
"loss": 0.455,
|
891 |
+
"step": 126
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 2.295964125560538,
|
895 |
+
"grad_norm": 0.9949473738670349,
|
896 |
+
"learning_rate": 8.651162790697674e-05,
|
897 |
+
"loss": 0.4637,
|
898 |
+
"step": 127
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 2.31390134529148,
|
902 |
+
"grad_norm": 1.0186420679092407,
|
903 |
+
"learning_rate": 8.558139534883721e-05,
|
904 |
+
"loss": 0.4877,
|
905 |
+
"step": 128
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 2.3318385650224216,
|
909 |
+
"grad_norm": 1.0602444410324097,
|
910 |
+
"learning_rate": 8.465116279069768e-05,
|
911 |
+
"loss": 0.4941,
|
912 |
+
"step": 129
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.3497757847533634,
|
916 |
+
"grad_norm": 1.0107648372650146,
|
917 |
+
"learning_rate": 8.372093023255814e-05,
|
918 |
+
"loss": 0.4339,
|
919 |
+
"step": 130
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 2.367713004484305,
|
923 |
+
"grad_norm": 1.1475372314453125,
|
924 |
+
"learning_rate": 8.27906976744186e-05,
|
925 |
+
"loss": 0.5158,
|
926 |
+
"step": 131
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 2.3856502242152464,
|
930 |
+
"grad_norm": 1.0330064296722412,
|
931 |
+
"learning_rate": 8.186046511627907e-05,
|
932 |
+
"loss": 0.4266,
|
933 |
+
"step": 132
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 2.403587443946188,
|
937 |
+
"grad_norm": 1.3457512855529785,
|
938 |
+
"learning_rate": 8.093023255813953e-05,
|
939 |
+
"loss": 0.617,
|
940 |
+
"step": 133
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.42152466367713,
|
944 |
+
"grad_norm": 1.1562917232513428,
|
945 |
+
"learning_rate": 8e-05,
|
946 |
+
"loss": 0.4976,
|
947 |
+
"step": 134
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 2.4394618834080717,
|
951 |
+
"grad_norm": 1.087751030921936,
|
952 |
+
"learning_rate": 7.906976744186047e-05,
|
953 |
+
"loss": 0.4836,
|
954 |
+
"step": 135
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.4573991031390134,
|
958 |
+
"grad_norm": 1.0045045614242554,
|
959 |
+
"learning_rate": 7.813953488372094e-05,
|
960 |
+
"loss": 0.4294,
|
961 |
+
"step": 136
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 2.475336322869955,
|
965 |
+
"grad_norm": 1.0355446338653564,
|
966 |
+
"learning_rate": 7.72093023255814e-05,
|
967 |
+
"loss": 0.4472,
|
968 |
+
"step": 137
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 2.493273542600897,
|
972 |
+
"grad_norm": 1.163203239440918,
|
973 |
+
"learning_rate": 7.627906976744187e-05,
|
974 |
+
"loss": 0.4853,
|
975 |
+
"step": 138
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 2.5112107623318387,
|
979 |
+
"grad_norm": 1.0705980062484741,
|
980 |
+
"learning_rate": 7.534883720930233e-05,
|
981 |
+
"loss": 0.4142,
|
982 |
+
"step": 139
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 2.5291479820627805,
|
986 |
+
"grad_norm": 1.172975778579712,
|
987 |
+
"learning_rate": 7.441860465116279e-05,
|
988 |
+
"loss": 0.5348,
|
989 |
+
"step": 140
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 2.547085201793722,
|
993 |
+
"grad_norm": 0.9890033006668091,
|
994 |
+
"learning_rate": 7.348837209302326e-05,
|
995 |
+
"loss": 0.3867,
|
996 |
+
"step": 141
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 2.5650224215246635,
|
1000 |
+
"grad_norm": 1.3716145753860474,
|
1001 |
+
"learning_rate": 7.255813953488373e-05,
|
1002 |
+
"loss": 0.4829,
|
1003 |
+
"step": 142
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 2.5829596412556053,
|
1007 |
+
"grad_norm": 1.1363354921340942,
|
1008 |
+
"learning_rate": 7.162790697674418e-05,
|
1009 |
+
"loss": 0.4112,
|
1010 |
+
"step": 143
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 2.600896860986547,
|
1014 |
+
"grad_norm": 1.180514931678772,
|
1015 |
+
"learning_rate": 7.069767441860465e-05,
|
1016 |
+
"loss": 0.4212,
|
1017 |
+
"step": 144
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 2.618834080717489,
|
1021 |
+
"grad_norm": 1.1589065790176392,
|
1022 |
+
"learning_rate": 6.976744186046513e-05,
|
1023 |
+
"loss": 0.4382,
|
1024 |
+
"step": 145
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 2.6367713004484306,
|
1028 |
+
"grad_norm": 1.1208486557006836,
|
1029 |
+
"learning_rate": 6.883720930232558e-05,
|
1030 |
+
"loss": 0.4852,
|
1031 |
+
"step": 146
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 2.6547085201793723,
|
1035 |
+
"grad_norm": 1.1670925617218018,
|
1036 |
+
"learning_rate": 6.790697674418604e-05,
|
1037 |
+
"loss": 0.4801,
|
1038 |
+
"step": 147
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 2.672645739910314,
|
1042 |
+
"grad_norm": 1.1497890949249268,
|
1043 |
+
"learning_rate": 6.697674418604652e-05,
|
1044 |
+
"loss": 0.4581,
|
1045 |
+
"step": 148
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 2.6905829596412554,
|
1049 |
+
"grad_norm": 1.1380338668823242,
|
1050 |
+
"learning_rate": 6.604651162790698e-05,
|
1051 |
+
"loss": 0.4974,
|
1052 |
+
"step": 149
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 2.708520179372197,
|
1056 |
+
"grad_norm": 1.2095478773117065,
|
1057 |
+
"learning_rate": 6.511627906976745e-05,
|
1058 |
+
"loss": 0.4958,
|
1059 |
+
"step": 150
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 2.726457399103139,
|
1063 |
+
"grad_norm": 1.1369256973266602,
|
1064 |
+
"learning_rate": 6.418604651162791e-05,
|
1065 |
+
"loss": 0.4518,
|
1066 |
+
"step": 151
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 2.7443946188340806,
|
1070 |
+
"grad_norm": 1.1578013896942139,
|
1071 |
+
"learning_rate": 6.325581395348838e-05,
|
1072 |
+
"loss": 0.5655,
|
1073 |
+
"step": 152
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 2.7623318385650224,
|
1077 |
+
"grad_norm": 1.0805268287658691,
|
1078 |
+
"learning_rate": 6.232558139534884e-05,
|
1079 |
+
"loss": 0.5004,
|
1080 |
+
"step": 153
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 2.780269058295964,
|
1084 |
+
"grad_norm": 1.1408129930496216,
|
1085 |
+
"learning_rate": 6.139534883720931e-05,
|
1086 |
+
"loss": 0.4059,
|
1087 |
+
"step": 154
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 2.798206278026906,
|
1091 |
+
"grad_norm": 1.0206074714660645,
|
1092 |
+
"learning_rate": 6.0465116279069765e-05,
|
1093 |
+
"loss": 0.4119,
|
1094 |
+
"step": 155
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 2.8161434977578477,
|
1098 |
+
"grad_norm": 0.9685718417167664,
|
1099 |
+
"learning_rate": 5.953488372093024e-05,
|
1100 |
+
"loss": 0.4574,
|
1101 |
+
"step": 156
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 2.8340807174887894,
|
1105 |
+
"grad_norm": 1.0425866842269897,
|
1106 |
+
"learning_rate": 5.8604651162790704e-05,
|
1107 |
+
"loss": 0.4774,
|
1108 |
+
"step": 157
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 2.852017937219731,
|
1112 |
+
"grad_norm": 1.0325255393981934,
|
1113 |
+
"learning_rate": 5.767441860465117e-05,
|
1114 |
+
"loss": 0.4317,
|
1115 |
+
"step": 158
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 2.8699551569506725,
|
1119 |
+
"grad_norm": 1.0784574747085571,
|
1120 |
+
"learning_rate": 5.674418604651163e-05,
|
1121 |
+
"loss": 0.3921,
|
1122 |
+
"step": 159
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 2.8878923766816142,
|
1126 |
+
"grad_norm": 1.081007957458496,
|
1127 |
+
"learning_rate": 5.5813953488372095e-05,
|
1128 |
+
"loss": 0.4515,
|
1129 |
+
"step": 160
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 2.905829596412556,
|
1133 |
+
"grad_norm": 1.1916303634643555,
|
1134 |
+
"learning_rate": 5.488372093023256e-05,
|
1135 |
+
"loss": 0.4556,
|
1136 |
+
"step": 161
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 2.9237668161434978,
|
1140 |
+
"grad_norm": 1.2342188358306885,
|
1141 |
+
"learning_rate": 5.3953488372093034e-05,
|
1142 |
+
"loss": 0.4772,
|
1143 |
+
"step": 162
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 2.9417040358744395,
|
1147 |
+
"grad_norm": 1.0315567255020142,
|
1148 |
+
"learning_rate": 5.3023255813953486e-05,
|
1149 |
+
"loss": 0.3233,
|
1150 |
+
"step": 163
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 2.9596412556053813,
|
1154 |
+
"grad_norm": 1.3380693197250366,
|
1155 |
+
"learning_rate": 5.209302325581395e-05,
|
1156 |
+
"loss": 0.5098,
|
1157 |
+
"step": 164
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 2.977578475336323,
|
1161 |
+
"grad_norm": 1.2268236875534058,
|
1162 |
+
"learning_rate": 5.1162790697674425e-05,
|
1163 |
+
"loss": 0.4125,
|
1164 |
+
"step": 165
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 3.004484304932735,
|
1168 |
+
"grad_norm": 2.589136838912964,
|
1169 |
+
"learning_rate": 5.023255813953489e-05,
|
1170 |
+
"loss": 0.7637,
|
1171 |
+
"step": 166
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 3.022421524663677,
|
1175 |
+
"grad_norm": 0.9620673060417175,
|
1176 |
+
"learning_rate": 4.930232558139535e-05,
|
1177 |
+
"loss": 0.2501,
|
1178 |
+
"step": 167
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 3.0403587443946187,
|
1182 |
+
"grad_norm": 1.0851328372955322,
|
1183 |
+
"learning_rate": 4.8372093023255816e-05,
|
1184 |
+
"loss": 0.3299,
|
1185 |
+
"step": 168
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 3.0582959641255605,
|
1189 |
+
"grad_norm": 1.081047773361206,
|
1190 |
+
"learning_rate": 4.744186046511628e-05,
|
1191 |
+
"loss": 0.319,
|
1192 |
+
"step": 169
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 3.0762331838565022,
|
1196 |
+
"grad_norm": 0.9016939997673035,
|
1197 |
+
"learning_rate": 4.651162790697675e-05,
|
1198 |
+
"loss": 0.2505,
|
1199 |
+
"step": 170
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 3.094170403587444,
|
1203 |
+
"grad_norm": 1.279685616493225,
|
1204 |
+
"learning_rate": 4.5581395348837214e-05,
|
1205 |
+
"loss": 0.3574,
|
1206 |
+
"step": 171
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 3.1121076233183858,
|
1210 |
+
"grad_norm": 1.1288567781448364,
|
1211 |
+
"learning_rate": 4.465116279069767e-05,
|
1212 |
+
"loss": 0.2322,
|
1213 |
+
"step": 172
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 3.1300448430493275,
|
1217 |
+
"grad_norm": 1.0982707738876343,
|
1218 |
+
"learning_rate": 4.3720930232558146e-05,
|
1219 |
+
"loss": 0.2557,
|
1220 |
+
"step": 173
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 3.1479820627802693,
|
1224 |
+
"grad_norm": 1.2716487646102905,
|
1225 |
+
"learning_rate": 4.2790697674418605e-05,
|
1226 |
+
"loss": 0.2683,
|
1227 |
+
"step": 174
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 3.1659192825112106,
|
1231 |
+
"grad_norm": 1.277907371520996,
|
1232 |
+
"learning_rate": 4.186046511627907e-05,
|
1233 |
+
"loss": 0.3359,
|
1234 |
+
"step": 175
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 3.1838565022421523,
|
1238 |
+
"grad_norm": 1.767809510231018,
|
1239 |
+
"learning_rate": 4.093023255813954e-05,
|
1240 |
+
"loss": 0.3141,
|
1241 |
+
"step": 176
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 3.201793721973094,
|
1245 |
+
"grad_norm": 1.5723196268081665,
|
1246 |
+
"learning_rate": 4e-05,
|
1247 |
+
"loss": 0.2696,
|
1248 |
+
"step": 177
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 3.219730941704036,
|
1252 |
+
"grad_norm": 1.2438582181930542,
|
1253 |
+
"learning_rate": 3.906976744186047e-05,
|
1254 |
+
"loss": 0.2261,
|
1255 |
+
"step": 178
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 3.2376681614349776,
|
1259 |
+
"grad_norm": 1.3772393465042114,
|
1260 |
+
"learning_rate": 3.8139534883720935e-05,
|
1261 |
+
"loss": 0.2699,
|
1262 |
+
"step": 179
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 3.2556053811659194,
|
1266 |
+
"grad_norm": 1.1731289625167847,
|
1267 |
+
"learning_rate": 3.7209302325581394e-05,
|
1268 |
+
"loss": 0.2809,
|
1269 |
+
"step": 180
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 3.273542600896861,
|
1273 |
+
"grad_norm": 1.3203359842300415,
|
1274 |
+
"learning_rate": 3.627906976744187e-05,
|
1275 |
+
"loss": 0.2798,
|
1276 |
+
"step": 181
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 3.291479820627803,
|
1280 |
+
"grad_norm": 1.0982232093811035,
|
1281 |
+
"learning_rate": 3.5348837209302326e-05,
|
1282 |
+
"loss": 0.2503,
|
1283 |
+
"step": 182
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 3.3094170403587446,
|
1287 |
+
"grad_norm": 1.2753369808197021,
|
1288 |
+
"learning_rate": 3.441860465116279e-05,
|
1289 |
+
"loss": 0.2893,
|
1290 |
+
"step": 183
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 3.327354260089686,
|
1294 |
+
"grad_norm": 1.2293989658355713,
|
1295 |
+
"learning_rate": 3.348837209302326e-05,
|
1296 |
+
"loss": 0.2585,
|
1297 |
+
"step": 184
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 3.3452914798206277,
|
1301 |
+
"grad_norm": 1.3043240308761597,
|
1302 |
+
"learning_rate": 3.2558139534883724e-05,
|
1303 |
+
"loss": 0.2203,
|
1304 |
+
"step": 185
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 3.3632286995515694,
|
1308 |
+
"grad_norm": 1.1034027338027954,
|
1309 |
+
"learning_rate": 3.162790697674419e-05,
|
1310 |
+
"loss": 0.2501,
|
1311 |
+
"step": 186
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 3.381165919282511,
|
1315 |
+
"grad_norm": 0.9731037020683289,
|
1316 |
+
"learning_rate": 3.0697674418604656e-05,
|
1317 |
+
"loss": 0.1944,
|
1318 |
+
"step": 187
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 3.399103139013453,
|
1322 |
+
"grad_norm": 1.069287657737732,
|
1323 |
+
"learning_rate": 2.976744186046512e-05,
|
1324 |
+
"loss": 0.2041,
|
1325 |
+
"step": 188
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 3.4170403587443947,
|
1329 |
+
"grad_norm": 1.3233182430267334,
|
1330 |
+
"learning_rate": 2.8837209302325585e-05,
|
1331 |
+
"loss": 0.2713,
|
1332 |
+
"step": 189
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 3.4349775784753365,
|
1336 |
+
"grad_norm": 1.2428154945373535,
|
1337 |
+
"learning_rate": 2.7906976744186048e-05,
|
1338 |
+
"loss": 0.2002,
|
1339 |
+
"step": 190
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 3.452914798206278,
|
1343 |
+
"grad_norm": 1.20328688621521,
|
1344 |
+
"learning_rate": 2.6976744186046517e-05,
|
1345 |
+
"loss": 0.2646,
|
1346 |
+
"step": 191
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 3.4708520179372195,
|
1350 |
+
"grad_norm": 1.3479125499725342,
|
1351 |
+
"learning_rate": 2.6046511627906976e-05,
|
1352 |
+
"loss": 0.2911,
|
1353 |
+
"step": 192
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 3.4887892376681613,
|
1357 |
+
"grad_norm": 1.2266180515289307,
|
1358 |
+
"learning_rate": 2.5116279069767445e-05,
|
1359 |
+
"loss": 0.2811,
|
1360 |
+
"step": 193
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 3.506726457399103,
|
1364 |
+
"grad_norm": 1.2345128059387207,
|
1365 |
+
"learning_rate": 2.4186046511627908e-05,
|
1366 |
+
"loss": 0.2237,
|
1367 |
+
"step": 194
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 3.524663677130045,
|
1371 |
+
"grad_norm": 1.3437424898147583,
|
1372 |
+
"learning_rate": 2.3255813953488374e-05,
|
1373 |
+
"loss": 0.283,
|
1374 |
+
"step": 195
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 3.5426008968609866,
|
1378 |
+
"grad_norm": 1.4216517210006714,
|
1379 |
+
"learning_rate": 2.2325581395348837e-05,
|
1380 |
+
"loss": 0.2871,
|
1381 |
+
"step": 196
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 3.5605381165919283,
|
1385 |
+
"grad_norm": 1.1113003492355347,
|
1386 |
+
"learning_rate": 2.1395348837209303e-05,
|
1387 |
+
"loss": 0.219,
|
1388 |
+
"step": 197
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 3.57847533632287,
|
1392 |
+
"grad_norm": 1.592371940612793,
|
1393 |
+
"learning_rate": 2.046511627906977e-05,
|
1394 |
+
"loss": 0.3161,
|
1395 |
+
"step": 198
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 3.596412556053812,
|
1399 |
+
"grad_norm": 1.2963297367095947,
|
1400 |
+
"learning_rate": 1.9534883720930235e-05,
|
1401 |
+
"loss": 0.2242,
|
1402 |
+
"step": 199
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 3.6143497757847536,
|
1406 |
+
"grad_norm": 1.1588383913040161,
|
1407 |
+
"learning_rate": 1.8604651162790697e-05,
|
1408 |
+
"loss": 0.2079,
|
1409 |
+
"step": 200
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 3.6322869955156953,
|
1413 |
+
"grad_norm": 1.2683604955673218,
|
1414 |
+
"learning_rate": 1.7674418604651163e-05,
|
1415 |
+
"loss": 0.2501,
|
1416 |
+
"step": 201
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 3.6502242152466366,
|
1420 |
+
"grad_norm": 1.3655322790145874,
|
1421 |
+
"learning_rate": 1.674418604651163e-05,
|
1422 |
+
"loss": 0.2299,
|
1423 |
+
"step": 202
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 3.6681614349775784,
|
1427 |
+
"grad_norm": 1.3018665313720703,
|
1428 |
+
"learning_rate": 1.5813953488372095e-05,
|
1429 |
+
"loss": 0.2148,
|
1430 |
+
"step": 203
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 3.68609865470852,
|
1434 |
+
"grad_norm": 1.388330101966858,
|
1435 |
+
"learning_rate": 1.488372093023256e-05,
|
1436 |
+
"loss": 0.2512,
|
1437 |
+
"step": 204
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 3.704035874439462,
|
1441 |
+
"grad_norm": 1.535142421722412,
|
1442 |
+
"learning_rate": 1.3953488372093024e-05,
|
1443 |
+
"loss": 0.2579,
|
1444 |
+
"step": 205
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 3.7219730941704037,
|
1448 |
+
"grad_norm": 1.4287734031677246,
|
1449 |
+
"learning_rate": 1.3023255813953488e-05,
|
1450 |
+
"loss": 0.2708,
|
1451 |
+
"step": 206
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 3.7399103139013454,
|
1455 |
+
"grad_norm": 1.5674840211868286,
|
1456 |
+
"learning_rate": 1.2093023255813954e-05,
|
1457 |
+
"loss": 0.3017,
|
1458 |
+
"step": 207
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 3.7578475336322867,
|
1462 |
+
"grad_norm": 1.261733889579773,
|
1463 |
+
"learning_rate": 1.1162790697674418e-05,
|
1464 |
+
"loss": 0.2584,
|
1465 |
+
"step": 208
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 3.7757847533632285,
|
1469 |
+
"grad_norm": 1.4881441593170166,
|
1470 |
+
"learning_rate": 1.0232558139534884e-05,
|
1471 |
+
"loss": 0.3086,
|
1472 |
+
"step": 209
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 3.7937219730941703,
|
1476 |
+
"grad_norm": 1.1449949741363525,
|
1477 |
+
"learning_rate": 9.302325581395349e-06,
|
1478 |
+
"loss": 0.2277,
|
1479 |
+
"step": 210
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 3.811659192825112,
|
1483 |
+
"grad_norm": 1.3948498964309692,
|
1484 |
+
"learning_rate": 8.372093023255815e-06,
|
1485 |
+
"loss": 0.2668,
|
1486 |
+
"step": 211
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 3.8295964125560538,
|
1490 |
+
"grad_norm": 1.1462297439575195,
|
1491 |
+
"learning_rate": 7.44186046511628e-06,
|
1492 |
+
"loss": 0.2142,
|
1493 |
+
"step": 212
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 3.8475336322869955,
|
1497 |
+
"grad_norm": 1.4967782497406006,
|
1498 |
+
"learning_rate": 6.511627906976744e-06,
|
1499 |
+
"loss": 0.275,
|
1500 |
+
"step": 213
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 3.8654708520179373,
|
1504 |
+
"grad_norm": 1.3958649635314941,
|
1505 |
+
"learning_rate": 5.581395348837209e-06,
|
1506 |
+
"loss": 0.2833,
|
1507 |
+
"step": 214
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 3.883408071748879,
|
1511 |
+
"grad_norm": 1.4644280672073364,
|
1512 |
+
"learning_rate": 4.651162790697674e-06,
|
1513 |
+
"loss": 0.3475,
|
1514 |
+
"step": 215
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 3.901345291479821,
|
1518 |
+
"grad_norm": 1.3760302066802979,
|
1519 |
+
"learning_rate": 3.72093023255814e-06,
|
1520 |
+
"loss": 0.2892,
|
1521 |
+
"step": 216
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 3.9192825112107625,
|
1525 |
+
"grad_norm": 1.320532202720642,
|
1526 |
+
"learning_rate": 2.7906976744186046e-06,
|
1527 |
+
"loss": 0.2542,
|
1528 |
+
"step": 217
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 3.9372197309417043,
|
1532 |
+
"grad_norm": 1.1825841665267944,
|
1533 |
+
"learning_rate": 1.86046511627907e-06,
|
1534 |
+
"loss": 0.2306,
|
1535 |
+
"step": 218
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 3.9551569506726456,
|
1539 |
+
"grad_norm": 1.3918488025665283,
|
1540 |
+
"learning_rate": 9.30232558139535e-07,
|
1541 |
+
"loss": 0.2958,
|
1542 |
+
"step": 219
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 3.9730941704035874,
|
1546 |
+
"grad_norm": 1.3532480001449585,
|
1547 |
+
"learning_rate": 0.0,
|
1548 |
+
"loss": 0.2748,
|
1549 |
+
"step": 220
|
1550 |
+
}
|
1551 |
+
],
|
1552 |
+
"logging_steps": 1,
|
1553 |
+
"max_steps": 220,
|
1554 |
+
"num_input_tokens_seen": 0,
|
1555 |
+
"num_train_epochs": 4,
|
1556 |
+
"save_steps": 500,
|
1557 |
+
"stateful_callbacks": {
|
1558 |
+
"TrainerControl": {
|
1559 |
+
"args": {
|
1560 |
+
"should_epoch_stop": false,
|
1561 |
+
"should_evaluate": false,
|
1562 |
+
"should_log": false,
|
1563 |
+
"should_save": true,
|
1564 |
+
"should_training_stop": true
|
1565 |
+
},
|
1566 |
+
"attributes": {}
|
1567 |
+
}
|
1568 |
+
},
|
1569 |
+
"total_flos": 9925730977234944.0,
|
1570 |
+
"train_batch_size": 2,
|
1571 |
+
"trial_name": null,
|
1572 |
+
"trial_params": null
|
1573 |
+
}
|