Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1190 -0
trainer_state.json
ADDED
@@ -0,0 +1,1190 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 4.990881458966565,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1642,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.030395136778115502,
|
13 |
+
"grad_norm": 10.756904602050781,
|
14 |
+
"learning_rate": 2.4096385542168677e-05,
|
15 |
+
"loss": 1.6027,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.060790273556231005,
|
20 |
+
"grad_norm": 9.451399803161621,
|
21 |
+
"learning_rate": 4.8192771084337354e-05,
|
22 |
+
"loss": 0.4287,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.0911854103343465,
|
27 |
+
"grad_norm": 1.1107724905014038,
|
28 |
+
"learning_rate": 7.228915662650602e-05,
|
29 |
+
"loss": 0.0996,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.12158054711246201,
|
34 |
+
"grad_norm": 0.6546974778175354,
|
35 |
+
"learning_rate": 9.638554216867471e-05,
|
36 |
+
"loss": 0.0405,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.1519756838905775,
|
41 |
+
"grad_norm": 0.4774630665779114,
|
42 |
+
"learning_rate": 0.0001204819277108434,
|
43 |
+
"loss": 0.0249,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.182370820668693,
|
48 |
+
"grad_norm": 1.0936344861984253,
|
49 |
+
"learning_rate": 0.00014457831325301204,
|
50 |
+
"loss": 0.027,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.2127659574468085,
|
55 |
+
"grad_norm": 0.6960570216178894,
|
56 |
+
"learning_rate": 0.00016867469879518074,
|
57 |
+
"loss": 0.0187,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.24316109422492402,
|
62 |
+
"grad_norm": 0.45391130447387695,
|
63 |
+
"learning_rate": 0.00019277108433734942,
|
64 |
+
"loss": 0.0143,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.2735562310030395,
|
69 |
+
"grad_norm": 0.3450222611427307,
|
70 |
+
"learning_rate": 0.0001999900512985548,
|
71 |
+
"loss": 0.0122,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.303951367781155,
|
76 |
+
"grad_norm": 0.8266738653182983,
|
77 |
+
"learning_rate": 0.00019994132773027597,
|
78 |
+
"loss": 0.0186,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.3343465045592705,
|
83 |
+
"grad_norm": 0.4097825586795807,
|
84 |
+
"learning_rate": 0.00019985202174272956,
|
85 |
+
"loss": 0.014,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.364741641337386,
|
90 |
+
"grad_norm": 0.6124298572540283,
|
91 |
+
"learning_rate": 0.00019972216959972274,
|
92 |
+
"loss": 0.0156,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.3951367781155015,
|
97 |
+
"grad_norm": 0.5567657351493835,
|
98 |
+
"learning_rate": 0.00019955182402933334,
|
99 |
+
"loss": 0.0078,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.425531914893617,
|
104 |
+
"grad_norm": 0.6027739644050598,
|
105 |
+
"learning_rate": 0.00019934105420249908,
|
106 |
+
"loss": 0.0089,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.45592705167173253,
|
111 |
+
"grad_norm": 0.5931612849235535,
|
112 |
+
"learning_rate": 0.00019908994570492993,
|
113 |
+
"loss": 0.0143,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.48632218844984804,
|
118 |
+
"grad_norm": 0.44545474648475647,
|
119 |
+
"learning_rate": 0.00019879860050235469,
|
120 |
+
"loss": 0.0123,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.5167173252279635,
|
125 |
+
"grad_norm": 0.29651954770088196,
|
126 |
+
"learning_rate": 0.0001984671368991169,
|
127 |
+
"loss": 0.008,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.547112462006079,
|
132 |
+
"grad_norm": 0.19712242484092712,
|
133 |
+
"learning_rate": 0.0001980956894901356,
|
134 |
+
"loss": 0.006,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.5775075987841946,
|
139 |
+
"grad_norm": 0.4244577884674072,
|
140 |
+
"learning_rate": 0.00019768440910625162,
|
141 |
+
"loss": 0.0096,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.60790273556231,
|
146 |
+
"grad_norm": 0.5452653169631958,
|
147 |
+
"learning_rate": 0.00019723346275298052,
|
148 |
+
"loss": 0.0128,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.6382978723404256,
|
153 |
+
"grad_norm": 0.5842880010604858,
|
154 |
+
"learning_rate": 0.00019674303354269833,
|
155 |
+
"loss": 0.0115,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.668693009118541,
|
160 |
+
"grad_norm": 0.5381531715393066,
|
161 |
+
"learning_rate": 0.00019621332062028617,
|
162 |
+
"loss": 0.0102,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.6990881458966566,
|
167 |
+
"grad_norm": 0.7361140847206116,
|
168 |
+
"learning_rate": 0.00019564453908226515,
|
169 |
+
"loss": 0.0098,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.729483282674772,
|
174 |
+
"grad_norm": 0.45423439145088196,
|
175 |
+
"learning_rate": 0.00019503691988945367,
|
176 |
+
"loss": 0.0097,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.7598784194528876,
|
181 |
+
"grad_norm": 0.38471490144729614,
|
182 |
+
"learning_rate": 0.000194390709773183,
|
183 |
+
"loss": 0.0117,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.790273556231003,
|
188 |
+
"grad_norm": 0.36030113697052,
|
189 |
+
"learning_rate": 0.0001937061711351089,
|
190 |
+
"loss": 0.009,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.8206686930091185,
|
195 |
+
"grad_norm": 0.3028634786605835,
|
196 |
+
"learning_rate": 0.00019298358194066016,
|
197 |
+
"loss": 0.0069,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.851063829787234,
|
202 |
+
"grad_norm": 0.278566837310791,
|
203 |
+
"learning_rate": 0.0001922232356061672,
|
204 |
+
"loss": 0.011,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.8814589665653495,
|
209 |
+
"grad_norm": 0.3821546137332916,
|
210 |
+
"learning_rate": 0.00019142544087971693,
|
211 |
+
"loss": 0.0054,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.9118541033434651,
|
216 |
+
"grad_norm": 0.26465708017349243,
|
217 |
+
"learning_rate": 0.00019059052171578155,
|
218 |
+
"loss": 0.0077,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.9422492401215805,
|
223 |
+
"grad_norm": 0.4308549463748932,
|
224 |
+
"learning_rate": 0.00018971881714367295,
|
225 |
+
"loss": 0.0057,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.9726443768996961,
|
230 |
+
"grad_norm": 0.226839080452919,
|
231 |
+
"learning_rate": 0.0001888106811298755,
|
232 |
+
"loss": 0.0077,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 1.0030395136778116,
|
237 |
+
"grad_norm": 0.39627695083618164,
|
238 |
+
"learning_rate": 0.00018786648243431363,
|
239 |
+
"loss": 0.0115,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 1.033434650455927,
|
244 |
+
"grad_norm": 0.6822514533996582,
|
245 |
+
"learning_rate": 0.00018688660446061235,
|
246 |
+
"loss": 0.0103,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.0638297872340425,
|
251 |
+
"grad_norm": 0.34325116872787476,
|
252 |
+
"learning_rate": 0.00018587144510041128,
|
253 |
+
"loss": 0.0095,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.094224924012158,
|
258 |
+
"grad_norm": 1.1588419675827026,
|
259 |
+
"learning_rate": 0.00018482141657179594,
|
260 |
+
"loss": 0.0199,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 1.1246200607902737,
|
265 |
+
"grad_norm": 0.5164496898651123,
|
266 |
+
"learning_rate": 0.00018373694525191138,
|
267 |
+
"loss": 0.0113,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 1.155015197568389,
|
272 |
+
"grad_norm": 0.38176482915878296,
|
273 |
+
"learning_rate": 0.00018261847150382644,
|
274 |
+
"loss": 0.0109,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.1854103343465046,
|
279 |
+
"grad_norm": 0.43293264508247375,
|
280 |
+
"learning_rate": 0.000181466449497719,
|
281 |
+
"loss": 0.0126,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 1.21580547112462,
|
286 |
+
"grad_norm": 0.2756524980068207,
|
287 |
+
"learning_rate": 0.00018028134702645425,
|
288 |
+
"loss": 0.0057,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 1.2462006079027357,
|
293 |
+
"grad_norm": 0.3666749894618988,
|
294 |
+
"learning_rate": 0.00017906364531563185,
|
295 |
+
"loss": 0.0047,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.2765957446808511,
|
300 |
+
"grad_norm": 0.3208726942539215,
|
301 |
+
"learning_rate": 0.00017781383882817811,
|
302 |
+
"loss": 0.0153,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 1.3069908814589666,
|
307 |
+
"grad_norm": 0.36565855145454407,
|
308 |
+
"learning_rate": 0.00017653243506356332,
|
309 |
+
"loss": 0.0078,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 1.337386018237082,
|
314 |
+
"grad_norm": 0.30531516671180725,
|
315 |
+
"learning_rate": 0.00017521995435172504,
|
316 |
+
"loss": 0.0086,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 1.3677811550151975,
|
321 |
+
"grad_norm": 0.4445202648639679,
|
322 |
+
"learning_rate": 0.00017387692964178198,
|
323 |
+
"loss": 0.0046,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.3981762917933132,
|
328 |
+
"grad_norm": 0.30582693219184875,
|
329 |
+
"learning_rate": 0.00017250390628562303,
|
330 |
+
"loss": 0.0044,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.4285714285714286,
|
335 |
+
"grad_norm": 0.22891288995742798,
|
336 |
+
"learning_rate": 0.00017110144181646072,
|
337 |
+
"loss": 0.0084,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.458966565349544,
|
342 |
+
"grad_norm": 0.34802454710006714,
|
343 |
+
"learning_rate": 0.00016967010572243758,
|
344 |
+
"loss": 0.0053,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.4893617021276595,
|
349 |
+
"grad_norm": 0.4866830110549927,
|
350 |
+
"learning_rate": 0.00016821047921537858,
|
351 |
+
"loss": 0.0082,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.5197568389057752,
|
356 |
+
"grad_norm": 0.11244069784879684,
|
357 |
+
"learning_rate": 0.0001667231549947828,
|
358 |
+
"loss": 0.0043,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.5501519756838906,
|
363 |
+
"grad_norm": 0.37716448307037354,
|
364 |
+
"learning_rate": 0.00016520873700715045,
|
365 |
+
"loss": 0.0075,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.580547112462006,
|
370 |
+
"grad_norm": 0.3906088173389435,
|
371 |
+
"learning_rate": 0.00016366784020074282,
|
372 |
+
"loss": 0.0051,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.6109422492401215,
|
377 |
+
"grad_norm": 0.400751531124115,
|
378 |
+
"learning_rate": 0.00016210109027587494,
|
379 |
+
"loss": 0.0066,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.641337386018237,
|
384 |
+
"grad_norm": 0.3523465096950531,
|
385 |
+
"learning_rate": 0.00016050912343084216,
|
386 |
+
"loss": 0.0076,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.6717325227963524,
|
391 |
+
"grad_norm": 0.1771322637796402,
|
392 |
+
"learning_rate": 0.00015889258610358398,
|
393 |
+
"loss": 0.0045,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.702127659574468,
|
398 |
+
"grad_norm": 0.1138119027018547,
|
399 |
+
"learning_rate": 0.00015725213470918977,
|
400 |
+
"loss": 0.0058,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.7325227963525835,
|
405 |
+
"grad_norm": 0.32440638542175293,
|
406 |
+
"learning_rate": 0.00015558843537335338,
|
407 |
+
"loss": 0.0086,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.7629179331306992,
|
412 |
+
"grad_norm": 0.35322776436805725,
|
413 |
+
"learning_rate": 0.0001539021636618844,
|
414 |
+
"loss": 0.0063,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.7933130699088147,
|
419 |
+
"grad_norm": 0.36127084493637085,
|
420 |
+
"learning_rate": 0.0001521940043063863,
|
421 |
+
"loss": 0.0058,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.8237082066869301,
|
426 |
+
"grad_norm": 0.3828403949737549,
|
427 |
+
"learning_rate": 0.00015046465092621278,
|
428 |
+
"loss": 0.0066,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.8541033434650456,
|
433 |
+
"grad_norm": 0.2598365247249603,
|
434 |
+
"learning_rate": 0.00014871480574681477,
|
435 |
+
"loss": 0.0098,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.884498480243161,
|
440 |
+
"grad_norm": 0.42567044496536255,
|
441 |
+
"learning_rate": 0.00014694517931459317,
|
442 |
+
"loss": 0.0102,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.9148936170212765,
|
447 |
+
"grad_norm": 0.30201858282089233,
|
448 |
+
"learning_rate": 0.00014515649020837277,
|
449 |
+
"loss": 0.0059,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.9452887537993921,
|
454 |
+
"grad_norm": 0.2485564947128296,
|
455 |
+
"learning_rate": 0.00014334946474761412,
|
456 |
+
"loss": 0.0056,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.9756838905775076,
|
461 |
+
"grad_norm": 0.23254001140594482,
|
462 |
+
"learning_rate": 0.0001415248366974826,
|
463 |
+
"loss": 0.0048,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 2.0060790273556233,
|
468 |
+
"grad_norm": 0.2745867967605591,
|
469 |
+
"learning_rate": 0.00013968334697089406,
|
470 |
+
"loss": 0.0035,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 2.0364741641337387,
|
475 |
+
"grad_norm": 0.21013915538787842,
|
476 |
+
"learning_rate": 0.0001378257433276578,
|
477 |
+
"loss": 0.0027,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 2.066869300911854,
|
482 |
+
"grad_norm": 0.11735199391841888,
|
483 |
+
"learning_rate": 0.00013595278007083933,
|
484 |
+
"loss": 0.0063,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 2.0972644376899696,
|
489 |
+
"grad_norm": 0.14921429753303528,
|
490 |
+
"learning_rate": 0.00013406521774046636,
|
491 |
+
"loss": 0.0023,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 2.127659574468085,
|
496 |
+
"grad_norm": 0.12081817537546158,
|
497 |
+
"learning_rate": 0.000132163822804702,
|
498 |
+
"loss": 0.0019,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 2.1580547112462005,
|
503 |
+
"grad_norm": 0.21969059109687805,
|
504 |
+
"learning_rate": 0.00013024936734861087,
|
505 |
+
"loss": 0.0051,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 2.188449848024316,
|
510 |
+
"grad_norm": 0.15178360044956207,
|
511 |
+
"learning_rate": 0.00012832262876064427,
|
512 |
+
"loss": 0.003,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 2.2188449848024314,
|
517 |
+
"grad_norm": 0.10473588109016418,
|
518 |
+
"learning_rate": 0.00012638438941697206,
|
519 |
+
"loss": 0.0042,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 2.2492401215805473,
|
524 |
+
"grad_norm": 0.12606075406074524,
|
525 |
+
"learning_rate": 0.0001244354363637889,
|
526 |
+
"loss": 0.0035,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 2.2796352583586628,
|
531 |
+
"grad_norm": 0.17029185593128204,
|
532 |
+
"learning_rate": 0.0001224765609977246,
|
533 |
+
"loss": 0.0029,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 2.310030395136778,
|
538 |
+
"grad_norm": 0.35497593879699707,
|
539 |
+
"learning_rate": 0.00012050855874448737,
|
540 |
+
"loss": 0.0032,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 2.3404255319148937,
|
545 |
+
"grad_norm": 0.17530225217342377,
|
546 |
+
"learning_rate": 0.00011853222873587167,
|
547 |
+
"loss": 0.0039,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.370820668693009,
|
552 |
+
"grad_norm": 0.26195624470710754,
|
553 |
+
"learning_rate": 0.00011654837348526044,
|
554 |
+
"loss": 0.0063,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 2.4012158054711246,
|
559 |
+
"grad_norm": 0.19774211943149567,
|
560 |
+
"learning_rate": 0.00011455779856175488,
|
561 |
+
"loss": 0.0019,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 2.43161094224924,
|
566 |
+
"grad_norm": 0.23176009953022003,
|
567 |
+
"learning_rate": 0.00011256131226306288,
|
568 |
+
"loss": 0.0033,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 2.4620060790273555,
|
573 |
+
"grad_norm": 0.21269629895687103,
|
574 |
+
"learning_rate": 0.00011055972528727973,
|
575 |
+
"loss": 0.002,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 2.4924012158054714,
|
580 |
+
"grad_norm": 0.15158076584339142,
|
581 |
+
"learning_rate": 0.00010855385040369419,
|
582 |
+
"loss": 0.0093,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 2.522796352583587,
|
587 |
+
"grad_norm": 0.16666357219219208,
|
588 |
+
"learning_rate": 0.00010654450212275324,
|
589 |
+
"loss": 0.0045,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 2.5531914893617023,
|
594 |
+
"grad_norm": 0.20873229205608368,
|
595 |
+
"learning_rate": 0.00010453249636532007,
|
596 |
+
"loss": 0.0027,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 2.5835866261398177,
|
601 |
+
"grad_norm": 0.20907987654209137,
|
602 |
+
"learning_rate": 0.00010251865013135931,
|
603 |
+
"loss": 0.0027,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 2.613981762917933,
|
608 |
+
"grad_norm": 0.4642738997936249,
|
609 |
+
"learning_rate": 0.00010050378116818391,
|
610 |
+
"loss": 0.0047,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 2.6443768996960486,
|
615 |
+
"grad_norm": 0.16256296634674072,
|
616 |
+
"learning_rate": 9.848870763839877e-05,
|
617 |
+
"loss": 0.0032,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 2.674772036474164,
|
622 |
+
"grad_norm": 0.23262229561805725,
|
623 |
+
"learning_rate": 9.64742477876756e-05,
|
624 |
+
"loss": 0.0046,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 2.7051671732522795,
|
629 |
+
"grad_norm": 0.255894273519516,
|
630 |
+
"learning_rate": 9.44612196124941e-05,
|
631 |
+
"loss": 0.0049,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 2.735562310030395,
|
636 |
+
"grad_norm": 0.1491995006799698,
|
637 |
+
"learning_rate": 9.245044052798435e-05,
|
638 |
+
"loss": 0.0034,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 2.7659574468085104,
|
643 |
+
"grad_norm": 0.18624486029148102,
|
644 |
+
"learning_rate": 9.044272703600505e-05,
|
645 |
+
"loss": 0.0031,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 2.7963525835866263,
|
650 |
+
"grad_norm": 0.3413298428058624,
|
651 |
+
"learning_rate": 8.843889439359308e-05,
|
652 |
+
"loss": 0.0091,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 2.8267477203647418,
|
657 |
+
"grad_norm": 0.4524446725845337,
|
658 |
+
"learning_rate": 8.643975628191802e-05,
|
659 |
+
"loss": 0.0084,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 2.857142857142857,
|
664 |
+
"grad_norm": 0.2205473631620407,
|
665 |
+
"learning_rate": 8.444612447587683e-05,
|
666 |
+
"loss": 0.0038,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 2.8875379939209727,
|
671 |
+
"grad_norm": 0.3354128301143646,
|
672 |
+
"learning_rate": 8.245880851446255e-05,
|
673 |
+
"loss": 0.0039,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.917933130699088,
|
678 |
+
"grad_norm": 0.40203210711479187,
|
679 |
+
"learning_rate": 8.047861537204107e-05,
|
680 |
+
"loss": 0.0069,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.9483282674772036,
|
685 |
+
"grad_norm": 0.16645453870296478,
|
686 |
+
"learning_rate": 7.850634913066887e-05,
|
687 |
+
"loss": 0.003,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.978723404255319,
|
692 |
+
"grad_norm": 0.1760583221912384,
|
693 |
+
"learning_rate": 7.654281065358575e-05,
|
694 |
+
"loss": 0.0038,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 3.0091185410334345,
|
699 |
+
"grad_norm": 0.09641872346401215,
|
700 |
+
"learning_rate": 7.458879726001431e-05,
|
701 |
+
"loss": 0.002,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 3.0395136778115504,
|
706 |
+
"grad_norm": 0.3135606050491333,
|
707 |
+
"learning_rate": 7.264510240139836e-05,
|
708 |
+
"loss": 0.0052,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 3.069908814589666,
|
713 |
+
"grad_norm": 0.1596423238515854,
|
714 |
+
"learning_rate": 7.071251533921235e-05,
|
715 |
+
"loss": 0.0035,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 3.1003039513677813,
|
720 |
+
"grad_norm": 0.09552032500505447,
|
721 |
+
"learning_rate": 6.879182082447185e-05,
|
722 |
+
"loss": 0.0023,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 3.1306990881458967,
|
727 |
+
"grad_norm": 0.14375700056552887,
|
728 |
+
"learning_rate": 6.688379877907548e-05,
|
729 |
+
"loss": 0.0034,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 3.161094224924012,
|
734 |
+
"grad_norm": 0.18640564382076263,
|
735 |
+
"learning_rate": 6.498922397910822e-05,
|
736 |
+
"loss": 0.0021,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 3.1914893617021276,
|
741 |
+
"grad_norm": 0.15173585712909698,
|
742 |
+
"learning_rate": 6.310886574023362e-05,
|
743 |
+
"loss": 0.0019,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 3.221884498480243,
|
748 |
+
"grad_norm": 0.17078043520450592,
|
749 |
+
"learning_rate": 6.124348760530383e-05,
|
750 |
+
"loss": 0.0064,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 3.2522796352583585,
|
755 |
+
"grad_norm": 0.11285611987113953,
|
756 |
+
"learning_rate": 5.9393847034313544e-05,
|
757 |
+
"loss": 0.0022,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 3.282674772036474,
|
762 |
+
"grad_norm": 0.20150907337665558,
|
763 |
+
"learning_rate": 5.756069509682368e-05,
|
764 |
+
"loss": 0.0021,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 3.31306990881459,
|
769 |
+
"grad_norm": 0.23782743513584137,
|
770 |
+
"learning_rate": 5.574477616698061e-05,
|
771 |
+
"loss": 0.0014,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 3.3434650455927053,
|
776 |
+
"grad_norm": 0.05953775346279144,
|
777 |
+
"learning_rate": 5.3946827621253514e-05,
|
778 |
+
"loss": 0.0025,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 3.3738601823708207,
|
783 |
+
"grad_norm": 0.09737365692853928,
|
784 |
+
"learning_rate": 5.2167579539013456e-05,
|
785 |
+
"loss": 0.0016,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 3.404255319148936,
|
790 |
+
"grad_norm": 0.10946661978960037,
|
791 |
+
"learning_rate": 5.0407754406075926e-05,
|
792 |
+
"loss": 0.0018,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 3.4346504559270516,
|
797 |
+
"grad_norm": 0.18973323702812195,
|
798 |
+
"learning_rate": 4.866806682132611e-05,
|
799 |
+
"loss": 0.0025,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 3.465045592705167,
|
804 |
+
"grad_norm": 0.10450422018766403,
|
805 |
+
"learning_rate": 4.694922320654727e-05,
|
806 |
+
"loss": 0.0009,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 3.4954407294832825,
|
811 |
+
"grad_norm": 0.1521192193031311,
|
812 |
+
"learning_rate": 4.5251921519569404e-05,
|
813 |
+
"loss": 0.0026,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 3.5258358662613984,
|
818 |
+
"grad_norm": 0.11275038123130798,
|
819 |
+
"learning_rate": 4.35768509708548e-05,
|
820 |
+
"loss": 0.0014,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 3.556231003039514,
|
825 |
+
"grad_norm": 0.209175705909729,
|
826 |
+
"learning_rate": 4.1924691743635504e-05,
|
827 |
+
"loss": 0.0025,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 3.5866261398176293,
|
832 |
+
"grad_norm": 0.17129185795783997,
|
833 |
+
"learning_rate": 4.029611471771646e-05,
|
834 |
+
"loss": 0.0015,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 3.617021276595745,
|
839 |
+
"grad_norm": 0.08248832076787949,
|
840 |
+
"learning_rate": 3.869178119705648e-05,
|
841 |
+
"loss": 0.0024,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 3.6474164133738602,
|
846 |
+
"grad_norm": 0.18126089870929718,
|
847 |
+
"learning_rate": 3.711234264123747e-05,
|
848 |
+
"loss": 0.0028,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 3.6778115501519757,
|
853 |
+
"grad_norm": 0.2583954632282257,
|
854 |
+
"learning_rate": 3.555844040093129e-05,
|
855 |
+
"loss": 0.0041,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 3.708206686930091,
|
860 |
+
"grad_norm": 0.16434016823768616,
|
861 |
+
"learning_rate": 3.403070545747107e-05,
|
862 |
+
"loss": 0.0028,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 3.7386018237082066,
|
867 |
+
"grad_norm": 0.13008223474025726,
|
868 |
+
"learning_rate": 3.252975816663375e-05,
|
869 |
+
"loss": 0.0015,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 3.768996960486322,
|
874 |
+
"grad_norm": 0.0734616294503212,
|
875 |
+
"learning_rate": 3.1056208006736634e-05,
|
876 |
+
"loss": 0.0016,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 3.7993920972644375,
|
881 |
+
"grad_norm": 0.06645731627941132,
|
882 |
+
"learning_rate": 2.9610653331151216e-05,
|
883 |
+
"loss": 0.0016,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 3.829787234042553,
|
888 |
+
"grad_norm": 0.24428494274616241,
|
889 |
+
"learning_rate": 2.8193681125334393e-05,
|
890 |
+
"loss": 0.002,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 3.860182370820669,
|
895 |
+
"grad_norm": 0.22978048026561737,
|
896 |
+
"learning_rate": 2.6805866768475663e-05,
|
897 |
+
"loss": 0.0017,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 3.8905775075987843,
|
902 |
+
"grad_norm": 0.1229507103562355,
|
903 |
+
"learning_rate": 2.5447773799857244e-05,
|
904 |
+
"loss": 0.0018,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 3.9209726443768997,
|
909 |
+
"grad_norm": 0.2323068082332611,
|
910 |
+
"learning_rate": 2.4119953690022025e-05,
|
911 |
+
"loss": 0.0017,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 3.951367781155015,
|
916 |
+
"grad_norm": 0.23123906552791595,
|
917 |
+
"learning_rate": 2.2822945616841963e-05,
|
918 |
+
"loss": 0.002,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 3.9817629179331306,
|
923 |
+
"grad_norm": 0.10087133944034576,
|
924 |
+
"learning_rate": 2.1557276246578307e-05,
|
925 |
+
"loss": 0.0018,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 4.0121580547112465,
|
930 |
+
"grad_norm": 0.058749064803123474,
|
931 |
+
"learning_rate": 2.032345952002218e-05,
|
932 |
+
"loss": 0.0011,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 4.042553191489362,
|
937 |
+
"grad_norm": 0.15570074319839478,
|
938 |
+
"learning_rate": 1.9121996443802482e-05,
|
939 |
+
"loss": 0.004,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 4.072948328267477,
|
944 |
+
"grad_norm": 0.2478976845741272,
|
945 |
+
"learning_rate": 1.7953374886946006e-05,
|
946 |
+
"loss": 0.0036,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 4.103343465045593,
|
951 |
+
"grad_norm": 0.08812441676855087,
|
952 |
+
"learning_rate": 1.681806938277205e-05,
|
953 |
+
"loss": 0.0015,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 4.133738601823708,
|
958 |
+
"grad_norm": 0.0631280392408371,
|
959 |
+
"learning_rate": 1.5716540936202363e-05,
|
960 |
+
"loss": 0.0013,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 4.164133738601824,
|
965 |
+
"grad_norm": 0.061599250882864,
|
966 |
+
"learning_rate": 1.4649236836564263e-05,
|
967 |
+
"loss": 0.0078,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 4.194528875379939,
|
972 |
+
"grad_norm": 0.04393768310546875,
|
973 |
+
"learning_rate": 1.361659047596332e-05,
|
974 |
+
"loss": 0.0021,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 4.224924012158055,
|
979 |
+
"grad_norm": 0.0476597435772419,
|
980 |
+
"learning_rate": 1.2619021173299051e-05,
|
981 |
+
"loss": 0.0014,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 4.25531914893617,
|
986 |
+
"grad_norm": 0.07678095251321793,
|
987 |
+
"learning_rate": 1.1656934003995302e-05,
|
988 |
+
"loss": 0.0027,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 4.285714285714286,
|
993 |
+
"grad_norm": 0.1129755899310112,
|
994 |
+
"learning_rate": 1.0730719635514296e-05,
|
995 |
+
"loss": 0.0053,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 4.316109422492401,
|
1000 |
+
"grad_norm": 0.13822700083255768,
|
1001 |
+
"learning_rate": 9.840754168721289e-06,
|
1002 |
+
"loss": 0.002,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 4.3465045592705165,
|
1007 |
+
"grad_norm": 0.10752403736114502,
|
1008 |
+
"learning_rate": 8.987398985164108e-06,
|
1009 |
+
"loss": 0.0024,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 4.376899696048632,
|
1014 |
+
"grad_norm": 0.06812784075737,
|
1015 |
+
"learning_rate": 8.171000600329682e-06,
|
1016 |
+
"loss": 0.001,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 4.407294832826747,
|
1021 |
+
"grad_norm": 0.20128777623176575,
|
1022 |
+
"learning_rate": 7.391890522937139e-06,
|
1023 |
+
"loss": 0.0015,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 4.437689969604863,
|
1028 |
+
"grad_norm": 0.0681811273097992,
|
1029 |
+
"learning_rate": 6.6503851203245205e-06,
|
1030 |
+
"loss": 0.0073,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 4.468085106382979,
|
1035 |
+
"grad_norm": 0.04031313210725784,
|
1036 |
+
"learning_rate": 5.946785489983941e-06,
|
1037 |
+
"loss": 0.001,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 4.498480243161095,
|
1042 |
+
"grad_norm": 0.16159360110759735,
|
1043 |
+
"learning_rate": 5.2813773372971995e-06,
|
1044 |
+
"loss": 0.002,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 4.52887537993921,
|
1049 |
+
"grad_norm": 0.051200274378061295,
|
1050 |
+
"learning_rate": 4.654430859521519e-06,
|
1051 |
+
"loss": 0.0008,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 4.5592705167173255,
|
1056 |
+
"grad_norm": 0.05586516112089157,
|
1057 |
+
"learning_rate": 4.066200636072604e-06,
|
1058 |
+
"loss": 0.0024,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 4.589665653495441,
|
1063 |
+
"grad_norm": 0.22293174266815186,
|
1064 |
+
"learning_rate": 3.5169255251495283e-06,
|
1065 |
+
"loss": 0.0012,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 4.620060790273556,
|
1070 |
+
"grad_norm": 0.06470826268196106,
|
1071 |
+
"learning_rate": 3.006828566743358e-06,
|
1072 |
+
"loss": 0.0009,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 4.650455927051672,
|
1077 |
+
"grad_norm": 0.11062423884868622,
|
1078 |
+
"learning_rate": 2.536116892069007e-06,
|
1079 |
+
"loss": 0.0017,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 4.680851063829787,
|
1084 |
+
"grad_norm": 0.08456968516111374,
|
1085 |
+
"learning_rate": 2.1049816394570486e-06,
|
1086 |
+
"loss": 0.0006,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 4.711246200607903,
|
1091 |
+
"grad_norm": 0.0969931110739708,
|
1092 |
+
"learning_rate": 1.7135978767395588e-06,
|
1093 |
+
"loss": 0.0037,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 4.741641337386018,
|
1098 |
+
"grad_norm": 0.0655050203204155,
|
1099 |
+
"learning_rate": 1.3621245301617014e-06,
|
1100 |
+
"loss": 0.001,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 4.772036474164134,
|
1105 |
+
"grad_norm": 0.038389451801776886,
|
1106 |
+
"learning_rate": 1.0507043198477617e-06,
|
1107 |
+
"loss": 0.0007,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 4.802431610942249,
|
1112 |
+
"grad_norm": 0.027961203828454018,
|
1113 |
+
"learning_rate": 7.794637018477824e-07,
|
1114 |
+
"loss": 0.0013,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 4.832826747720365,
|
1119 |
+
"grad_norm": 0.24096284806728363,
|
1120 |
+
"learning_rate": 5.485128167885933e-07,
|
1121 |
+
"loss": 0.003,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 4.86322188449848,
|
1126 |
+
"grad_norm": 0.031791090965270996,
|
1127 |
+
"learning_rate": 3.579454451498099e-07,
|
1128 |
+
"loss": 0.0023,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 4.8936170212765955,
|
1133 |
+
"grad_norm": 0.028634727001190186,
|
1134 |
+
"learning_rate": 2.0783896918310508e-07,
|
1135 |
+
"loss": 0.001,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 4.924012158054711,
|
1140 |
+
"grad_norm": 0.11568836122751236,
|
1141 |
+
"learning_rate": 9.82543414901782e-08,
|
1142 |
+
"loss": 0.0038,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 4.954407294832826,
|
1147 |
+
"grad_norm": 0.13896380364894867,
|
1148 |
+
"learning_rate": 2.9236060272186395e-08,
|
1149 |
+
"loss": 0.005,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 4.984802431610943,
|
1154 |
+
"grad_norm": 0.02793799713253975,
|
1155 |
+
"learning_rate": 8.121512607317528e-10,
|
1156 |
+
"loss": 0.0016,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 4.990881458966565,
|
1161 |
+
"step": 1642,
|
1162 |
+
"total_flos": 5.715020951091302e+16,
|
1163 |
+
"train_loss": 0.018644058636483826,
|
1164 |
+
"train_runtime": 735.9075,
|
1165 |
+
"train_samples_per_second": 35.7,
|
1166 |
+
"train_steps_per_second": 2.231
|
1167 |
+
}
|
1168 |
+
],
|
1169 |
+
"logging_steps": 10,
|
1170 |
+
"max_steps": 1642,
|
1171 |
+
"num_input_tokens_seen": 0,
|
1172 |
+
"num_train_epochs": 5,
|
1173 |
+
"save_steps": 500,
|
1174 |
+
"stateful_callbacks": {
|
1175 |
+
"TrainerControl": {
|
1176 |
+
"args": {
|
1177 |
+
"should_epoch_stop": false,
|
1178 |
+
"should_evaluate": false,
|
1179 |
+
"should_log": false,
|
1180 |
+
"should_save": true,
|
1181 |
+
"should_training_stop": true
|
1182 |
+
},
|
1183 |
+
"attributes": {}
|
1184 |
+
}
|
1185 |
+
},
|
1186 |
+
"total_flos": 5.715020951091302e+16,
|
1187 |
+
"train_batch_size": 16,
|
1188 |
+
"trial_name": null,
|
1189 |
+
"trial_params": null
|
1190 |
+
}
|