Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1170 -0
trainer_state.json
ADDED
@@ -0,0 +1,1170 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 0.9984496124031008,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 161,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.006201550387596899,
|
14 |
+
"grad_norm": 15.26399338835659,
|
15 |
+
"learning_rate": 0.0,
|
16 |
+
"loss": 1.9626,
|
17 |
+
"step": 1
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.012403100775193798,
|
21 |
+
"grad_norm": 14.720768346647702,
|
22 |
+
"learning_rate": 1.111111111111111e-06,
|
23 |
+
"loss": 1.9459,
|
24 |
+
"step": 2
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.018604651162790697,
|
28 |
+
"grad_norm": 15.485169481168326,
|
29 |
+
"learning_rate": 2.222222222222222e-06,
|
30 |
+
"loss": 2.3093,
|
31 |
+
"step": 3
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.024806201550387597,
|
35 |
+
"grad_norm": 15.238736100868511,
|
36 |
+
"learning_rate": 3.3333333333333333e-06,
|
37 |
+
"loss": 2.0299,
|
38 |
+
"step": 4
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.031007751937984496,
|
42 |
+
"grad_norm": 10.917974882633493,
|
43 |
+
"learning_rate": 4.444444444444444e-06,
|
44 |
+
"loss": 1.9934,
|
45 |
+
"step": 5
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.037209302325581395,
|
49 |
+
"grad_norm": 3.5887710787919214,
|
50 |
+
"learning_rate": 5.555555555555557e-06,
|
51 |
+
"loss": 1.0246,
|
52 |
+
"step": 6
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.043410852713178294,
|
56 |
+
"grad_norm": 6.2370114556413,
|
57 |
+
"learning_rate": 6.666666666666667e-06,
|
58 |
+
"loss": 1.5176,
|
59 |
+
"step": 7
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.04961240310077519,
|
63 |
+
"grad_norm": 5.919162169895783,
|
64 |
+
"learning_rate": 7.77777777777778e-06,
|
65 |
+
"loss": 1.5474,
|
66 |
+
"step": 8
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.05581395348837209,
|
70 |
+
"grad_norm": 4.325467650355213,
|
71 |
+
"learning_rate": 8.888888888888888e-06,
|
72 |
+
"loss": 1.2359,
|
73 |
+
"step": 9
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.06201550387596899,
|
77 |
+
"grad_norm": 3.619075212455386,
|
78 |
+
"learning_rate": 1e-05,
|
79 |
+
"loss": 0.9838,
|
80 |
+
"step": 10
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.06821705426356589,
|
84 |
+
"grad_norm": 3.471405941940013,
|
85 |
+
"learning_rate": 9.998932083939657e-06,
|
86 |
+
"loss": 0.9981,
|
87 |
+
"step": 11
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.07441860465116279,
|
91 |
+
"grad_norm": 2.7880671807653483,
|
92 |
+
"learning_rate": 9.995728791936505e-06,
|
93 |
+
"loss": 0.8744,
|
94 |
+
"step": 12
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.08062015503875969,
|
98 |
+
"grad_norm": 6.034968083519955,
|
99 |
+
"learning_rate": 9.990391492329341e-06,
|
100 |
+
"loss": 0.7676,
|
101 |
+
"step": 13
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.08682170542635659,
|
105 |
+
"grad_norm": 1.9803400973575944,
|
106 |
+
"learning_rate": 9.98292246503335e-06,
|
107 |
+
"loss": 0.5414,
|
108 |
+
"step": 14
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.09302325581395349,
|
112 |
+
"grad_norm": 1.234736717659378,
|
113 |
+
"learning_rate": 9.973324900566214e-06,
|
114 |
+
"loss": 0.53,
|
115 |
+
"step": 15
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.09922480620155039,
|
119 |
+
"grad_norm": 1.0213862287647824,
|
120 |
+
"learning_rate": 9.961602898685225e-06,
|
121 |
+
"loss": 0.5062,
|
122 |
+
"step": 16
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.10542635658914729,
|
126 |
+
"grad_norm": 1.1013669003760953,
|
127 |
+
"learning_rate": 9.947761466636014e-06,
|
128 |
+
"loss": 0.557,
|
129 |
+
"step": 17
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.11162790697674418,
|
133 |
+
"grad_norm": 0.7019092049936988,
|
134 |
+
"learning_rate": 9.931806517013612e-06,
|
135 |
+
"loss": 0.4931,
|
136 |
+
"step": 18
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.11782945736434108,
|
140 |
+
"grad_norm": 0.6803712338561847,
|
141 |
+
"learning_rate": 9.913744865236798e-06,
|
142 |
+
"loss": 0.493,
|
143 |
+
"step": 19
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.12403100775193798,
|
147 |
+
"grad_norm": 3.1222946795280255,
|
148 |
+
"learning_rate": 9.893584226636773e-06,
|
149 |
+
"loss": 0.5671,
|
150 |
+
"step": 20
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.13023255813953488,
|
154 |
+
"grad_norm": 0.4533299829843851,
|
155 |
+
"learning_rate": 9.871333213161438e-06,
|
156 |
+
"loss": 0.3441,
|
157 |
+
"step": 21
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.13643410852713178,
|
161 |
+
"grad_norm": 0.449537089022575,
|
162 |
+
"learning_rate": 9.847001329696653e-06,
|
163 |
+
"loss": 0.3215,
|
164 |
+
"step": 22
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.14263565891472868,
|
168 |
+
"grad_norm": 0.6774019887238769,
|
169 |
+
"learning_rate": 9.820598970006068e-06,
|
170 |
+
"loss": 0.4402,
|
171 |
+
"step": 23
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.14883720930232558,
|
175 |
+
"grad_norm": 0.45963497537736786,
|
176 |
+
"learning_rate": 9.792137412291265e-06,
|
177 |
+
"loss": 0.4022,
|
178 |
+
"step": 24
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.15503875968992248,
|
182 |
+
"grad_norm": 0.4722916828215964,
|
183 |
+
"learning_rate": 9.761628814374074e-06,
|
184 |
+
"loss": 0.4308,
|
185 |
+
"step": 25
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.16124031007751938,
|
189 |
+
"grad_norm": 0.6226919889929532,
|
190 |
+
"learning_rate": 9.729086208503174e-06,
|
191 |
+
"loss": 0.4193,
|
192 |
+
"step": 26
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.16744186046511628,
|
196 |
+
"grad_norm": 0.3944344161765665,
|
197 |
+
"learning_rate": 9.694523495787149e-06,
|
198 |
+
"loss": 0.3406,
|
199 |
+
"step": 27
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.17364341085271318,
|
203 |
+
"grad_norm": 0.6421115422344923,
|
204 |
+
"learning_rate": 9.657955440256396e-06,
|
205 |
+
"loss": 0.3788,
|
206 |
+
"step": 28
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.17984496124031008,
|
210 |
+
"grad_norm": 0.4023252512533979,
|
211 |
+
"learning_rate": 9.619397662556434e-06,
|
212 |
+
"loss": 0.3484,
|
213 |
+
"step": 29
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.18604651162790697,
|
217 |
+
"grad_norm": 0.5280821694606184,
|
218 |
+
"learning_rate": 9.578866633275289e-06,
|
219 |
+
"loss": 0.3579,
|
220 |
+
"step": 30
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.19224806201550387,
|
224 |
+
"grad_norm": 0.3639394212633886,
|
225 |
+
"learning_rate": 9.536379665907801e-06,
|
226 |
+
"loss": 0.3281,
|
227 |
+
"step": 31
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.19844961240310077,
|
231 |
+
"grad_norm": 0.3343773044134941,
|
232 |
+
"learning_rate": 9.491954909459895e-06,
|
233 |
+
"loss": 0.3163,
|
234 |
+
"step": 32
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.20465116279069767,
|
238 |
+
"grad_norm": 0.6336122974597818,
|
239 |
+
"learning_rate": 9.445611340695926e-06,
|
240 |
+
"loss": 0.3736,
|
241 |
+
"step": 33
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.21085271317829457,
|
245 |
+
"grad_norm": 0.29713001995988114,
|
246 |
+
"learning_rate": 9.397368756032445e-06,
|
247 |
+
"loss": 0.3299,
|
248 |
+
"step": 34
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.21705426356589147,
|
252 |
+
"grad_norm": 0.3516643576671143,
|
253 |
+
"learning_rate": 9.347247763081834e-06,
|
254 |
+
"loss": 0.2795,
|
255 |
+
"step": 35
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.22325581395348837,
|
259 |
+
"grad_norm": 0.46593544409154286,
|
260 |
+
"learning_rate": 9.295269771849426e-06,
|
261 |
+
"loss": 0.3805,
|
262 |
+
"step": 36
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.22945736434108527,
|
266 |
+
"grad_norm": 1.385652999055083,
|
267 |
+
"learning_rate": 9.241456985587868e-06,
|
268 |
+
"loss": 0.4296,
|
269 |
+
"step": 37
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.23565891472868217,
|
273 |
+
"grad_norm": 0.45985433788020086,
|
274 |
+
"learning_rate": 9.185832391312644e-06,
|
275 |
+
"loss": 0.3928,
|
276 |
+
"step": 38
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.24186046511627907,
|
280 |
+
"grad_norm": 0.719866169139414,
|
281 |
+
"learning_rate": 9.12841974998278e-06,
|
282 |
+
"loss": 0.3674,
|
283 |
+
"step": 39
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.24806201550387597,
|
287 |
+
"grad_norm": 0.4245555272468952,
|
288 |
+
"learning_rate": 9.069243586350976e-06,
|
289 |
+
"loss": 0.3952,
|
290 |
+
"step": 40
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.25426356589147286,
|
294 |
+
"grad_norm": 0.32715050071028695,
|
295 |
+
"learning_rate": 9.008329178487442e-06,
|
296 |
+
"loss": 0.3118,
|
297 |
+
"step": 41
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.26046511627906976,
|
301 |
+
"grad_norm": 0.323913785807112,
|
302 |
+
"learning_rate": 8.94570254698197e-06,
|
303 |
+
"loss": 0.2915,
|
304 |
+
"step": 42
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.26666666666666666,
|
308 |
+
"grad_norm": 0.44726444195044723,
|
309 |
+
"learning_rate": 8.881390443828788e-06,
|
310 |
+
"loss": 0.3854,
|
311 |
+
"step": 43
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.27286821705426356,
|
315 |
+
"grad_norm": 0.66664094212195,
|
316 |
+
"learning_rate": 8.815420340999034e-06,
|
317 |
+
"loss": 0.3049,
|
318 |
+
"step": 44
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.27906976744186046,
|
322 |
+
"grad_norm": 0.4249980060595529,
|
323 |
+
"learning_rate": 8.747820418705632e-06,
|
324 |
+
"loss": 0.3438,
|
325 |
+
"step": 45
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.28527131782945736,
|
329 |
+
"grad_norm": 0.5176821224911877,
|
330 |
+
"learning_rate": 8.67861955336566e-06,
|
331 |
+
"loss": 0.4039,
|
332 |
+
"step": 46
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.29147286821705426,
|
336 |
+
"grad_norm": 0.28158982412253863,
|
337 |
+
"learning_rate": 8.607847305265312e-06,
|
338 |
+
"loss": 0.3504,
|
339 |
+
"step": 47
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.29767441860465116,
|
343 |
+
"grad_norm": 0.5840042318469005,
|
344 |
+
"learning_rate": 8.535533905932739e-06,
|
345 |
+
"loss": 0.3256,
|
346 |
+
"step": 48
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.30387596899224806,
|
350 |
+
"grad_norm": 0.31414108766842425,
|
351 |
+
"learning_rate": 8.461710245224149e-06,
|
352 |
+
"loss": 0.3202,
|
353 |
+
"step": 49
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.31007751937984496,
|
357 |
+
"grad_norm": 1.2797772790133255,
|
358 |
+
"learning_rate": 8.386407858128707e-06,
|
359 |
+
"loss": 0.3636,
|
360 |
+
"step": 50
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.31627906976744186,
|
364 |
+
"grad_norm": 0.3482350388750869,
|
365 |
+
"learning_rate": 8.309658911297833e-06,
|
366 |
+
"loss": 0.2954,
|
367 |
+
"step": 51
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.32248062015503876,
|
371 |
+
"grad_norm": 0.42356575149686077,
|
372 |
+
"learning_rate": 8.231496189304704e-06,
|
373 |
+
"loss": 0.3755,
|
374 |
+
"step": 52
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.32868217054263565,
|
378 |
+
"grad_norm": 0.37022827136720965,
|
379 |
+
"learning_rate": 8.151953080639777e-06,
|
380 |
+
"loss": 0.3133,
|
381 |
+
"step": 53
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.33488372093023255,
|
385 |
+
"grad_norm": 0.45808972723840075,
|
386 |
+
"learning_rate": 8.071063563448341e-06,
|
387 |
+
"loss": 0.3131,
|
388 |
+
"step": 54
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.34108527131782945,
|
392 |
+
"grad_norm": 0.3131249221444263,
|
393 |
+
"learning_rate": 7.988862191016204e-06,
|
394 |
+
"loss": 0.3262,
|
395 |
+
"step": 55
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.34728682170542635,
|
399 |
+
"grad_norm": 0.5028658389613722,
|
400 |
+
"learning_rate": 7.905384077009693e-06,
|
401 |
+
"loss": 0.2962,
|
402 |
+
"step": 56
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.35348837209302325,
|
406 |
+
"grad_norm": 0.3765836806930317,
|
407 |
+
"learning_rate": 7.820664880476257e-06,
|
408 |
+
"loss": 0.3573,
|
409 |
+
"step": 57
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.35968992248062015,
|
413 |
+
"grad_norm": 1.3467161045716638,
|
414 |
+
"learning_rate": 7.734740790612137e-06,
|
415 |
+
"loss": 0.3903,
|
416 |
+
"step": 58
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.36589147286821705,
|
420 |
+
"grad_norm": 0.32717489874399847,
|
421 |
+
"learning_rate": 7.647648511303545e-06,
|
422 |
+
"loss": 0.3114,
|
423 |
+
"step": 59
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.37209302325581395,
|
427 |
+
"grad_norm": 0.3265443857577057,
|
428 |
+
"learning_rate": 7.559425245448006e-06,
|
429 |
+
"loss": 0.3298,
|
430 |
+
"step": 60
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.37829457364341085,
|
434 |
+
"grad_norm": 0.3241255915038842,
|
435 |
+
"learning_rate": 7.470108679062521e-06,
|
436 |
+
"loss": 0.236,
|
437 |
+
"step": 61
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.38449612403100775,
|
441 |
+
"grad_norm": 0.3991561591891734,
|
442 |
+
"learning_rate": 7.379736965185369e-06,
|
443 |
+
"loss": 0.3525,
|
444 |
+
"step": 62
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.39069767441860465,
|
448 |
+
"grad_norm": 0.32423623161134224,
|
449 |
+
"learning_rate": 7.288348707578409e-06,
|
450 |
+
"loss": 0.3046,
|
451 |
+
"step": 63
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.39689922480620154,
|
455 |
+
"grad_norm": 0.3171873001618541,
|
456 |
+
"learning_rate": 7.195982944236853e-06,
|
457 |
+
"loss": 0.3594,
|
458 |
+
"step": 64
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.40310077519379844,
|
462 |
+
"grad_norm": 0.28143210002099517,
|
463 |
+
"learning_rate": 7.102679130713538e-06,
|
464 |
+
"loss": 0.2633,
|
465 |
+
"step": 65
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.40930232558139534,
|
469 |
+
"grad_norm": 0.2641885107385758,
|
470 |
+
"learning_rate": 7.008477123264849e-06,
|
471 |
+
"loss": 0.274,
|
472 |
+
"step": 66
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.41550387596899224,
|
476 |
+
"grad_norm": 0.27510115989973793,
|
477 |
+
"learning_rate": 6.913417161825449e-06,
|
478 |
+
"loss": 0.2948,
|
479 |
+
"step": 67
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.42170542635658914,
|
483 |
+
"grad_norm": 0.2818463816101008,
|
484 |
+
"learning_rate": 6.817539852819149e-06,
|
485 |
+
"loss": 0.2716,
|
486 |
+
"step": 68
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.42790697674418604,
|
490 |
+
"grad_norm": 0.296666971572694,
|
491 |
+
"learning_rate": 6.720886151813194e-06,
|
492 |
+
"loss": 0.295,
|
493 |
+
"step": 69
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.43410852713178294,
|
497 |
+
"grad_norm": 1.9121660637037805,
|
498 |
+
"learning_rate": 6.6234973460234184e-06,
|
499 |
+
"loss": 0.3237,
|
500 |
+
"step": 70
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.44031007751937984,
|
504 |
+
"grad_norm": 0.38478679027888313,
|
505 |
+
"learning_rate": 6.525415036677745e-06,
|
506 |
+
"loss": 0.3505,
|
507 |
+
"step": 71
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.44651162790697674,
|
511 |
+
"grad_norm": 0.2742653943674723,
|
512 |
+
"learning_rate": 6.426681121245527e-06,
|
513 |
+
"loss": 0.3313,
|
514 |
+
"step": 72
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.45271317829457364,
|
518 |
+
"grad_norm": 0.3058386741176561,
|
519 |
+
"learning_rate": 6.327337775540362e-06,
|
520 |
+
"loss": 0.2985,
|
521 |
+
"step": 73
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.45891472868217054,
|
525 |
+
"grad_norm": 0.29164203779704884,
|
526 |
+
"learning_rate": 6.227427435703997e-06,
|
527 |
+
"loss": 0.2924,
|
528 |
+
"step": 74
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.46511627906976744,
|
532 |
+
"grad_norm": 0.31050749736435646,
|
533 |
+
"learning_rate": 6.126992780079032e-06,
|
534 |
+
"loss": 0.2931,
|
535 |
+
"step": 75
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.47131782945736433,
|
539 |
+
"grad_norm": 0.2794828298595081,
|
540 |
+
"learning_rate": 6.026076710978172e-06,
|
541 |
+
"loss": 0.2933,
|
542 |
+
"step": 76
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.47751937984496123,
|
546 |
+
"grad_norm": 0.28252753496671273,
|
547 |
+
"learning_rate": 5.924722336357793e-06,
|
548 |
+
"loss": 0.2703,
|
549 |
+
"step": 77
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.48372093023255813,
|
553 |
+
"grad_norm": 0.3272997547053633,
|
554 |
+
"learning_rate": 5.82297295140367e-06,
|
555 |
+
"loss": 0.3266,
|
556 |
+
"step": 78
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.48992248062015503,
|
560 |
+
"grad_norm": 0.26499070641459116,
|
561 |
+
"learning_rate": 5.720872020036734e-06,
|
562 |
+
"loss": 0.3111,
|
563 |
+
"step": 79
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.49612403100775193,
|
567 |
+
"grad_norm": 0.28778684221197826,
|
568 |
+
"learning_rate": 5.61846315634674e-06,
|
569 |
+
"loss": 0.2942,
|
570 |
+
"step": 80
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.5023255813953489,
|
574 |
+
"grad_norm": 0.37204267144771097,
|
575 |
+
"learning_rate": 5.515790105961785e-06,
|
576 |
+
"loss": 0.3478,
|
577 |
+
"step": 81
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.5085271317829457,
|
581 |
+
"grad_norm": 0.2847653373131156,
|
582 |
+
"learning_rate": 5.412896727361663e-06,
|
583 |
+
"loss": 0.3024,
|
584 |
+
"step": 82
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.5147286821705427,
|
588 |
+
"grad_norm": 0.3322812245008778,
|
589 |
+
"learning_rate": 5.309826973142974e-06,
|
590 |
+
"loss": 0.2983,
|
591 |
+
"step": 83
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.5209302325581395,
|
595 |
+
"grad_norm": 0.356939865277603,
|
596 |
+
"learning_rate": 5.206624871244066e-06,
|
597 |
+
"loss": 0.3077,
|
598 |
+
"step": 84
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.5271317829457365,
|
602 |
+
"grad_norm": 0.2732990191138046,
|
603 |
+
"learning_rate": 5.103334506137773e-06,
|
604 |
+
"loss": 0.2717,
|
605 |
+
"step": 85
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.5333333333333333,
|
609 |
+
"grad_norm": 0.3344747136592173,
|
610 |
+
"learning_rate": 5e-06,
|
611 |
+
"loss": 0.2807,
|
612 |
+
"step": 86
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.5395348837209303,
|
616 |
+
"grad_norm": 0.3395777579574581,
|
617 |
+
"learning_rate": 4.89666549386223e-06,
|
618 |
+
"loss": 0.3519,
|
619 |
+
"step": 87
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.5457364341085271,
|
623 |
+
"grad_norm": 0.37507314187909735,
|
624 |
+
"learning_rate": 4.793375128755934e-06,
|
625 |
+
"loss": 0.32,
|
626 |
+
"step": 88
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.5519379844961241,
|
630 |
+
"grad_norm": 0.3024187395636649,
|
631 |
+
"learning_rate": 4.690173026857028e-06,
|
632 |
+
"loss": 0.2461,
|
633 |
+
"step": 89
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.5581395348837209,
|
637 |
+
"grad_norm": 0.2754899764445255,
|
638 |
+
"learning_rate": 4.587103272638339e-06,
|
639 |
+
"loss": 0.2991,
|
640 |
+
"step": 90
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.5643410852713179,
|
644 |
+
"grad_norm": 0.34895281559601965,
|
645 |
+
"learning_rate": 4.4842098940382155e-06,
|
646 |
+
"loss": 0.3146,
|
647 |
+
"step": 91
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.5705426356589147,
|
651 |
+
"grad_norm": 0.29391978438245,
|
652 |
+
"learning_rate": 4.381536843653262e-06,
|
653 |
+
"loss": 0.3185,
|
654 |
+
"step": 92
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 0.5767441860465117,
|
658 |
+
"grad_norm": 0.27690222515112517,
|
659 |
+
"learning_rate": 4.279127979963266e-06,
|
660 |
+
"loss": 0.2876,
|
661 |
+
"step": 93
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 0.5829457364341085,
|
665 |
+
"grad_norm": 0.3201838307069794,
|
666 |
+
"learning_rate": 4.17702704859633e-06,
|
667 |
+
"loss": 0.2929,
|
668 |
+
"step": 94
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 0.5891472868217055,
|
672 |
+
"grad_norm": 1.3427423380380996,
|
673 |
+
"learning_rate": 4.075277663642208e-06,
|
674 |
+
"loss": 0.2659,
|
675 |
+
"step": 95
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 0.5953488372093023,
|
679 |
+
"grad_norm": 0.30047579019920523,
|
680 |
+
"learning_rate": 3.973923289021829e-06,
|
681 |
+
"loss": 0.2735,
|
682 |
+
"step": 96
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 0.6015503875968993,
|
686 |
+
"grad_norm": 0.2742842789808185,
|
687 |
+
"learning_rate": 3.8730072199209705e-06,
|
688 |
+
"loss": 0.2531,
|
689 |
+
"step": 97
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 0.6077519379844961,
|
693 |
+
"grad_norm": 0.2682447160880254,
|
694 |
+
"learning_rate": 3.7725725642960047e-06,
|
695 |
+
"loss": 0.2706,
|
696 |
+
"step": 98
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 0.6139534883720931,
|
700 |
+
"grad_norm": 0.2500514931866793,
|
701 |
+
"learning_rate": 3.67266222445964e-06,
|
702 |
+
"loss": 0.262,
|
703 |
+
"step": 99
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 0.6201550387596899,
|
707 |
+
"grad_norm": 0.25293752189367336,
|
708 |
+
"learning_rate": 3.573318878754475e-06,
|
709 |
+
"loss": 0.2597,
|
710 |
+
"step": 100
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 0.6263565891472869,
|
714 |
+
"grad_norm": 0.445308919718186,
|
715 |
+
"learning_rate": 3.4745849633222566e-06,
|
716 |
+
"loss": 0.2683,
|
717 |
+
"step": 101
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.6325581395348837,
|
721 |
+
"grad_norm": 0.26411175014036536,
|
722 |
+
"learning_rate": 3.3765026539765832e-06,
|
723 |
+
"loss": 0.2843,
|
724 |
+
"step": 102
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 0.6387596899224807,
|
728 |
+
"grad_norm": 0.4447506815447959,
|
729 |
+
"learning_rate": 3.2791138481868084e-06,
|
730 |
+
"loss": 0.3165,
|
731 |
+
"step": 103
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 0.6449612403100775,
|
735 |
+
"grad_norm": 0.6788210803412132,
|
736 |
+
"learning_rate": 3.1824601471808504e-06,
|
737 |
+
"loss": 0.3379,
|
738 |
+
"step": 104
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 0.6511627906976745,
|
742 |
+
"grad_norm": 0.2756366217165372,
|
743 |
+
"learning_rate": 3.0865828381745515e-06,
|
744 |
+
"loss": 0.3209,
|
745 |
+
"step": 105
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 0.6573643410852713,
|
749 |
+
"grad_norm": 0.3006265900853834,
|
750 |
+
"learning_rate": 2.991522876735154e-06,
|
751 |
+
"loss": 0.3301,
|
752 |
+
"step": 106
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 0.6635658914728683,
|
756 |
+
"grad_norm": 0.48771082441163705,
|
757 |
+
"learning_rate": 2.8973208692864623e-06,
|
758 |
+
"loss": 0.259,
|
759 |
+
"step": 107
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.6697674418604651,
|
763 |
+
"grad_norm": 0.2860117833703043,
|
764 |
+
"learning_rate": 2.804017055763149e-06,
|
765 |
+
"loss": 0.299,
|
766 |
+
"step": 108
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 0.6759689922480621,
|
770 |
+
"grad_norm": 0.32363963719457234,
|
771 |
+
"learning_rate": 2.711651292421593e-06,
|
772 |
+
"loss": 0.2555,
|
773 |
+
"step": 109
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 0.6821705426356589,
|
777 |
+
"grad_norm": 0.2850162950975944,
|
778 |
+
"learning_rate": 2.6202630348146323e-06,
|
779 |
+
"loss": 0.316,
|
780 |
+
"step": 110
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 0.6883720930232559,
|
784 |
+
"grad_norm": 0.2762800647193149,
|
785 |
+
"learning_rate": 2.529891320937481e-06,
|
786 |
+
"loss": 0.2523,
|
787 |
+
"step": 111
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 0.6945736434108527,
|
791 |
+
"grad_norm": 0.26856974090502583,
|
792 |
+
"learning_rate": 2.4405747545519966e-06,
|
793 |
+
"loss": 0.2761,
|
794 |
+
"step": 112
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 0.7007751937984497,
|
798 |
+
"grad_norm": 0.2490891936662437,
|
799 |
+
"learning_rate": 2.352351488696457e-06,
|
800 |
+
"loss": 0.2575,
|
801 |
+
"step": 113
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.7069767441860465,
|
805 |
+
"grad_norm": 0.26341989498270973,
|
806 |
+
"learning_rate": 2.265259209387867e-06,
|
807 |
+
"loss": 0.2249,
|
808 |
+
"step": 114
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 0.7131782945736435,
|
812 |
+
"grad_norm": 0.27360549708121684,
|
813 |
+
"learning_rate": 2.179335119523745e-06,
|
814 |
+
"loss": 0.2697,
|
815 |
+
"step": 115
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 0.7193798449612403,
|
819 |
+
"grad_norm": 0.25823370272340357,
|
820 |
+
"learning_rate": 2.094615922990309e-06,
|
821 |
+
"loss": 0.2856,
|
822 |
+
"step": 116
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 0.7255813953488373,
|
826 |
+
"grad_norm": 0.2861381618206921,
|
827 |
+
"learning_rate": 2.0111378089837958e-06,
|
828 |
+
"loss": 0.2735,
|
829 |
+
"step": 117
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 0.7317829457364341,
|
833 |
+
"grad_norm": 0.2644689705355804,
|
834 |
+
"learning_rate": 1.928936436551661e-06,
|
835 |
+
"loss": 0.302,
|
836 |
+
"step": 118
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 0.737984496124031,
|
840 |
+
"grad_norm": 0.2590429498790151,
|
841 |
+
"learning_rate": 1.848046919360225e-06,
|
842 |
+
"loss": 0.2951,
|
843 |
+
"step": 119
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 0.7441860465116279,
|
847 |
+
"grad_norm": 0.2706830740807015,
|
848 |
+
"learning_rate": 1.7685038106952952e-06,
|
849 |
+
"loss": 0.2236,
|
850 |
+
"step": 120
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 0.7503875968992249,
|
854 |
+
"grad_norm": 0.3323189304107019,
|
855 |
+
"learning_rate": 1.6903410887021676e-06,
|
856 |
+
"loss": 0.2847,
|
857 |
+
"step": 121
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 0.7565891472868217,
|
861 |
+
"grad_norm": 0.2747309759947305,
|
862 |
+
"learning_rate": 1.6135921418712959e-06,
|
863 |
+
"loss": 0.3403,
|
864 |
+
"step": 122
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 0.7627906976744186,
|
868 |
+
"grad_norm": 0.24338533093448128,
|
869 |
+
"learning_rate": 1.5382897547758513e-06,
|
870 |
+
"loss": 0.2742,
|
871 |
+
"step": 123
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 0.7689922480620155,
|
875 |
+
"grad_norm": 1.490134687848016,
|
876 |
+
"learning_rate": 1.4644660940672628e-06,
|
877 |
+
"loss": 0.3482,
|
878 |
+
"step": 124
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 0.7751937984496124,
|
882 |
+
"grad_norm": 0.2612434597373152,
|
883 |
+
"learning_rate": 1.3921526947346902e-06,
|
884 |
+
"loss": 0.3018,
|
885 |
+
"step": 125
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 0.7813953488372093,
|
889 |
+
"grad_norm": 0.25036249030908275,
|
890 |
+
"learning_rate": 1.321380446634342e-06,
|
891 |
+
"loss": 0.3188,
|
892 |
+
"step": 126
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 0.7875968992248062,
|
896 |
+
"grad_norm": 0.2612566749541528,
|
897 |
+
"learning_rate": 1.2521795812943704e-06,
|
898 |
+
"loss": 0.2995,
|
899 |
+
"step": 127
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 0.7937984496124031,
|
903 |
+
"grad_norm": 0.35381824882639945,
|
904 |
+
"learning_rate": 1.1845796590009684e-06,
|
905 |
+
"loss": 0.287,
|
906 |
+
"step": 128
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 0.8,
|
910 |
+
"grad_norm": 0.27735996036817234,
|
911 |
+
"learning_rate": 1.118609556171213e-06,
|
912 |
+
"loss": 0.2765,
|
913 |
+
"step": 129
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 0.8062015503875969,
|
917 |
+
"grad_norm": 0.2770800666877129,
|
918 |
+
"learning_rate": 1.0542974530180327e-06,
|
919 |
+
"loss": 0.3346,
|
920 |
+
"step": 130
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 0.8124031007751938,
|
924 |
+
"grad_norm": 0.2786390048536498,
|
925 |
+
"learning_rate": 9.916708215125586e-07,
|
926 |
+
"loss": 0.2544,
|
927 |
+
"step": 131
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 0.8186046511627907,
|
931 |
+
"grad_norm": 0.2502926836308708,
|
932 |
+
"learning_rate": 9.307564136490255e-07,
|
933 |
+
"loss": 0.2888,
|
934 |
+
"step": 132
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 0.8248062015503876,
|
938 |
+
"grad_norm": 0.29354461108159396,
|
939 |
+
"learning_rate": 8.715802500172215e-07,
|
940 |
+
"loss": 0.2817,
|
941 |
+
"step": 133
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 0.8310077519379845,
|
945 |
+
"grad_norm": 0.29537836943938983,
|
946 |
+
"learning_rate": 8.141676086873574e-07,
|
947 |
+
"loss": 0.2752,
|
948 |
+
"step": 134
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 0.8372093023255814,
|
952 |
+
"grad_norm": 0.25157759814716646,
|
953 |
+
"learning_rate": 7.585430144121319e-07,
|
954 |
+
"loss": 0.2687,
|
955 |
+
"step": 135
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 0.8434108527131783,
|
959 |
+
"grad_norm": 0.25520703041985415,
|
960 |
+
"learning_rate": 7.047302281505735e-07,
|
961 |
+
"loss": 0.2348,
|
962 |
+
"step": 136
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 0.8496124031007752,
|
966 |
+
"grad_norm": 0.2804280034486318,
|
967 |
+
"learning_rate": 6.527522369181655e-07,
|
968 |
+
"loss": 0.3057,
|
969 |
+
"step": 137
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.8558139534883721,
|
973 |
+
"grad_norm": 0.2577385480731099,
|
974 |
+
"learning_rate": 6.026312439675553e-07,
|
975 |
+
"loss": 0.2553,
|
976 |
+
"step": 138
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 0.862015503875969,
|
980 |
+
"grad_norm": 0.4111388669958174,
|
981 |
+
"learning_rate": 5.543886593040737e-07,
|
982 |
+
"loss": 0.2228,
|
983 |
+
"step": 139
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 0.8682170542635659,
|
987 |
+
"grad_norm": 0.2512038377179209,
|
988 |
+
"learning_rate": 5.080450905401057e-07,
|
989 |
+
"loss": 0.2838,
|
990 |
+
"step": 140
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 0.8744186046511628,
|
994 |
+
"grad_norm": 0.2899187916626033,
|
995 |
+
"learning_rate": 4.6362033409220077e-07,
|
996 |
+
"loss": 0.325,
|
997 |
+
"step": 141
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 0.8806201550387597,
|
1001 |
+
"grad_norm": 0.30346956388490826,
|
1002 |
+
"learning_rate": 4.211333667247125e-07,
|
1003 |
+
"loss": 0.3548,
|
1004 |
+
"step": 142
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 0.8868217054263566,
|
1008 |
+
"grad_norm": 0.25434135497158716,
|
1009 |
+
"learning_rate": 3.8060233744356634e-07,
|
1010 |
+
"loss": 0.2554,
|
1011 |
+
"step": 143
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 0.8930232558139535,
|
1015 |
+
"grad_norm": 0.2718193216598721,
|
1016 |
+
"learning_rate": 3.420445597436056e-07,
|
1017 |
+
"loss": 0.2956,
|
1018 |
+
"step": 144
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 0.8992248062015504,
|
1022 |
+
"grad_norm": 0.2581092232290721,
|
1023 |
+
"learning_rate": 3.0547650421285216e-07,
|
1024 |
+
"loss": 0.2697,
|
1025 |
+
"step": 145
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 0.9054263565891473,
|
1029 |
+
"grad_norm": 0.2522471348373793,
|
1030 |
+
"learning_rate": 2.7091379149682683e-07,
|
1031 |
+
"loss": 0.2962,
|
1032 |
+
"step": 146
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 0.9116279069767442,
|
1036 |
+
"grad_norm": 0.24916983451448027,
|
1037 |
+
"learning_rate": 2.3837118562592799e-07,
|
1038 |
+
"loss": 0.2872,
|
1039 |
+
"step": 147
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 0.9178294573643411,
|
1043 |
+
"grad_norm": 0.2599299055423855,
|
1044 |
+
"learning_rate": 2.0786258770873647e-07,
|
1045 |
+
"loss": 0.2807,
|
1046 |
+
"step": 148
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 0.924031007751938,
|
1050 |
+
"grad_norm": 0.2804131521355199,
|
1051 |
+
"learning_rate": 1.7940102999393194e-07,
|
1052 |
+
"loss": 0.2809,
|
1053 |
+
"step": 149
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 0.9302325581395349,
|
1057 |
+
"grad_norm": 0.3172859696325972,
|
1058 |
+
"learning_rate": 1.5299867030334815e-07,
|
1059 |
+
"loss": 0.3055,
|
1060 |
+
"step": 150
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 0.9364341085271318,
|
1064 |
+
"grad_norm": 0.30376879245797783,
|
1065 |
+
"learning_rate": 1.286667868385627e-07,
|
1066 |
+
"loss": 0.3647,
|
1067 |
+
"step": 151
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 0.9426356589147287,
|
1071 |
+
"grad_norm": 0.26932389563925824,
|
1072 |
+
"learning_rate": 1.0641577336322761e-07,
|
1073 |
+
"loss": 0.2771,
|
1074 |
+
"step": 152
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 0.9488372093023256,
|
1078 |
+
"grad_norm": 0.28886881022603444,
|
1079 |
+
"learning_rate": 8.625513476320291e-08,
|
1080 |
+
"loss": 0.3155,
|
1081 |
+
"step": 153
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 0.9550387596899225,
|
1085 |
+
"grad_norm": 0.2459036254604883,
|
1086 |
+
"learning_rate": 6.819348298638839e-08,
|
1087 |
+
"loss": 0.276,
|
1088 |
+
"step": 154
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 0.9612403100775194,
|
1092 |
+
"grad_norm": 0.2706567588905982,
|
1093 |
+
"learning_rate": 5.223853336398632e-08,
|
1094 |
+
"loss": 0.3116,
|
1095 |
+
"step": 155
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 0.9674418604651163,
|
1099 |
+
"grad_norm": 0.27809861760850124,
|
1100 |
+
"learning_rate": 3.839710131477492e-08,
|
1101 |
+
"loss": 0.3412,
|
1102 |
+
"step": 156
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 0.9736434108527132,
|
1106 |
+
"grad_norm": 0.2790075651739876,
|
1107 |
+
"learning_rate": 2.6675099433787212e-08,
|
1108 |
+
"loss": 0.3247,
|
1109 |
+
"step": 157
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 0.9798449612403101,
|
1113 |
+
"grad_norm": 0.2545419908224724,
|
1114 |
+
"learning_rate": 1.7077534966650767e-08,
|
1115 |
+
"loss": 0.3451,
|
1116 |
+
"step": 158
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 0.986046511627907,
|
1120 |
+
"grad_norm": 0.23665306218981744,
|
1121 |
+
"learning_rate": 9.608507670659239e-09,
|
1122 |
+
"loss": 0.2523,
|
1123 |
+
"step": 159
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 0.9922480620155039,
|
1127 |
+
"grad_norm": 0.28653358687035707,
|
1128 |
+
"learning_rate": 4.2712080634949024e-09,
|
1129 |
+
"loss": 0.2768,
|
1130 |
+
"step": 160
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 0.9984496124031008,
|
1134 |
+
"grad_norm": 0.258292218409734,
|
1135 |
+
"learning_rate": 1.0679160603449533e-09,
|
1136 |
+
"loss": 0.353,
|
1137 |
+
"step": 161
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 0.9984496124031008,
|
1141 |
+
"step": 161,
|
1142 |
+
"total_flos": 429783955537920.0,
|
1143 |
+
"train_loss": 0.41365621706343586,
|
1144 |
+
"train_runtime": 4919.9829,
|
1145 |
+
"train_samples_per_second": 2.097,
|
1146 |
+
"train_steps_per_second": 0.033
|
1147 |
+
}
|
1148 |
+
],
|
1149 |
+
"logging_steps": 1.0,
|
1150 |
+
"max_steps": 161,
|
1151 |
+
"num_input_tokens_seen": 0,
|
1152 |
+
"num_train_epochs": 1,
|
1153 |
+
"save_steps": 500,
|
1154 |
+
"stateful_callbacks": {
|
1155 |
+
"TrainerControl": {
|
1156 |
+
"args": {
|
1157 |
+
"should_epoch_stop": false,
|
1158 |
+
"should_evaluate": false,
|
1159 |
+
"should_log": false,
|
1160 |
+
"should_save": true,
|
1161 |
+
"should_training_stop": true
|
1162 |
+
},
|
1163 |
+
"attributes": {}
|
1164 |
+
}
|
1165 |
+
},
|
1166 |
+
"total_flos": 429783955537920.0,
|
1167 |
+
"train_batch_size": 1,
|
1168 |
+
"trial_name": null,
|
1169 |
+
"trial_params": null
|
1170 |
+
}
|