Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1001 -0
trainer_state.json
ADDED
@@ -0,0 +1,1001 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1370,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.072992700729927,
|
13 |
+
"grad_norm": 5.749124526977539,
|
14 |
+
"learning_rate": 1.4492753623188407e-05,
|
15 |
+
"loss": 0.9214,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.145985401459854,
|
20 |
+
"grad_norm": 2.569692611694336,
|
21 |
+
"learning_rate": 2.8985507246376814e-05,
|
22 |
+
"loss": 0.506,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.21897810218978103,
|
27 |
+
"grad_norm": 1.676221489906311,
|
28 |
+
"learning_rate": 4.347826086956522e-05,
|
29 |
+
"loss": 0.3337,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.291970802919708,
|
34 |
+
"grad_norm": 1.1421884298324585,
|
35 |
+
"learning_rate": 5.797101449275363e-05,
|
36 |
+
"loss": 0.2416,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.36496350364963503,
|
41 |
+
"grad_norm": 0.7032585144042969,
|
42 |
+
"learning_rate": 7.246376811594203e-05,
|
43 |
+
"loss": 0.1709,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.43795620437956206,
|
48 |
+
"grad_norm": 0.8790648579597473,
|
49 |
+
"learning_rate": 8.695652173913044e-05,
|
50 |
+
"loss": 0.1541,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.5109489051094891,
|
55 |
+
"grad_norm": 0.9456900358200073,
|
56 |
+
"learning_rate": 9.999985422436231e-05,
|
57 |
+
"loss": 0.1368,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.583941605839416,
|
62 |
+
"grad_norm": 0.8083729147911072,
|
63 |
+
"learning_rate": 9.998236217634196e-05,
|
64 |
+
"loss": 0.1253,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.656934306569343,
|
69 |
+
"grad_norm": 0.5535995960235596,
|
70 |
+
"learning_rate": 9.993572668745786e-05,
|
71 |
+
"loss": 0.1117,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.7299270072992701,
|
76 |
+
"grad_norm": 0.7210536599159241,
|
77 |
+
"learning_rate": 9.985997494967441e-05,
|
78 |
+
"loss": 0.0905,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.8029197080291971,
|
83 |
+
"grad_norm": 0.5236138105392456,
|
84 |
+
"learning_rate": 9.975515113189827e-05,
|
85 |
+
"loss": 0.0917,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.8759124087591241,
|
90 |
+
"grad_norm": 0.4034636318683624,
|
91 |
+
"learning_rate": 9.962131635422462e-05,
|
92 |
+
"loss": 0.0802,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.948905109489051,
|
97 |
+
"grad_norm": 0.6598319411277771,
|
98 |
+
"learning_rate": 9.945854865229965e-05,
|
99 |
+
"loss": 0.0744,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.0218978102189782,
|
104 |
+
"grad_norm": 0.6162580847740173,
|
105 |
+
"learning_rate": 9.926694293181986e-05,
|
106 |
+
"loss": 0.0808,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.094890510948905,
|
111 |
+
"grad_norm": 0.5971336960792542,
|
112 |
+
"learning_rate": 9.904661091319503e-05,
|
113 |
+
"loss": 0.0765,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.167883211678832,
|
118 |
+
"grad_norm": 0.7087730765342712,
|
119 |
+
"learning_rate": 9.879768106640687e-05,
|
120 |
+
"loss": 0.0713,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.2408759124087592,
|
125 |
+
"grad_norm": 0.6969349980354309,
|
126 |
+
"learning_rate": 9.852029853610148e-05,
|
127 |
+
"loss": 0.0665,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.313868613138686,
|
132 |
+
"grad_norm": 0.6811894178390503,
|
133 |
+
"learning_rate": 9.821462505695917e-05,
|
134 |
+
"loss": 0.0631,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.3868613138686132,
|
139 |
+
"grad_norm": 0.34767240285873413,
|
140 |
+
"learning_rate": 9.788083885939116e-05,
|
141 |
+
"loss": 0.064,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.4598540145985401,
|
146 |
+
"grad_norm": 0.3470080494880676,
|
147 |
+
"learning_rate": 9.75191345656179e-05,
|
148 |
+
"loss": 0.0557,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.5328467153284673,
|
153 |
+
"grad_norm": 0.39414817094802856,
|
154 |
+
"learning_rate": 9.712972307618981e-05,
|
155 |
+
"loss": 0.0627,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.6058394160583942,
|
160 |
+
"grad_norm": 0.5173900723457336,
|
161 |
+
"learning_rate": 9.671283144701663e-05,
|
162 |
+
"loss": 0.0573,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.6788321167883211,
|
167 |
+
"grad_norm": 0.34683912992477417,
|
168 |
+
"learning_rate": 9.626870275697682e-05,
|
169 |
+
"loss": 0.0603,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.7518248175182483,
|
174 |
+
"grad_norm": 0.5500655174255371,
|
175 |
+
"learning_rate": 9.579759596618454e-05,
|
176 |
+
"loss": 0.062,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.8248175182481752,
|
181 |
+
"grad_norm": 0.3647141456604004,
|
182 |
+
"learning_rate": 9.529978576499652e-05,
|
183 |
+
"loss": 0.0504,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.897810218978102,
|
188 |
+
"grad_norm": 0.3856804072856903,
|
189 |
+
"learning_rate": 9.477556241384724e-05,
|
190 |
+
"loss": 0.0543,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.9708029197080292,
|
195 |
+
"grad_norm": 0.35589075088500977,
|
196 |
+
"learning_rate": 9.422523157400533e-05,
|
197 |
+
"loss": 0.0496,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.0437956204379564,
|
202 |
+
"grad_norm": 0.37966451048851013,
|
203 |
+
"learning_rate": 9.36491141293504e-05,
|
204 |
+
"loss": 0.0475,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 2.116788321167883,
|
209 |
+
"grad_norm": 0.4046926200389862,
|
210 |
+
"learning_rate": 9.304754599927377e-05,
|
211 |
+
"loss": 0.0466,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.18978102189781,
|
216 |
+
"grad_norm": 0.5988451242446899,
|
217 |
+
"learning_rate": 9.242087794281243e-05,
|
218 |
+
"loss": 0.0529,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 2.2627737226277373,
|
223 |
+
"grad_norm": 0.3989510238170624,
|
224 |
+
"learning_rate": 9.176947535413046e-05,
|
225 |
+
"loss": 0.0509,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 2.335766423357664,
|
230 |
+
"grad_norm": 0.4888112246990204,
|
231 |
+
"learning_rate": 9.10937180494669e-05,
|
232 |
+
"loss": 0.0405,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 2.408759124087591,
|
237 |
+
"grad_norm": 0.5207938551902771,
|
238 |
+
"learning_rate": 9.039400004567469e-05,
|
239 |
+
"loss": 0.0472,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.4817518248175183,
|
244 |
+
"grad_norm": 0.41739702224731445,
|
245 |
+
"learning_rate": 8.967072933047945e-05,
|
246 |
+
"loss": 0.0523,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 2.554744525547445,
|
251 |
+
"grad_norm": 0.5469692349433899,
|
252 |
+
"learning_rate": 8.892432762459221e-05,
|
253 |
+
"loss": 0.0466,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.627737226277372,
|
258 |
+
"grad_norm": 0.44405660033226013,
|
259 |
+
"learning_rate": 8.815523013581488e-05,
|
260 |
+
"loss": 0.0475,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.7007299270072993,
|
265 |
+
"grad_norm": 0.42182251811027527,
|
266 |
+
"learning_rate": 8.736388530528162e-05,
|
267 |
+
"loss": 0.0423,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 2.7737226277372264,
|
272 |
+
"grad_norm": 0.323812335729599,
|
273 |
+
"learning_rate": 8.655075454598426e-05,
|
274 |
+
"loss": 0.0421,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 2.846715328467153,
|
279 |
+
"grad_norm": 0.44694793224334717,
|
280 |
+
"learning_rate": 8.571631197373422e-05,
|
281 |
+
"loss": 0.0447,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 2.9197080291970803,
|
286 |
+
"grad_norm": 0.2432471662759781,
|
287 |
+
"learning_rate": 8.486104413071755e-05,
|
288 |
+
"loss": 0.0438,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.9927007299270074,
|
293 |
+
"grad_norm": 0.42758315801620483,
|
294 |
+
"learning_rate": 8.398544970180469e-05,
|
295 |
+
"loss": 0.0466,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 3.065693430656934,
|
300 |
+
"grad_norm": 0.3887377679347992,
|
301 |
+
"learning_rate": 8.309003922377996e-05,
|
302 |
+
"loss": 0.04,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 3.1386861313868613,
|
307 |
+
"grad_norm": 0.33644479513168335,
|
308 |
+
"learning_rate": 8.217533478766068e-05,
|
309 |
+
"loss": 0.0455,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 3.2116788321167884,
|
314 |
+
"grad_norm": 0.29482313990592957,
|
315 |
+
"learning_rate": 8.124186973427911e-05,
|
316 |
+
"loss": 0.0398,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 3.2846715328467155,
|
321 |
+
"grad_norm": 0.46486881375312805,
|
322 |
+
"learning_rate": 8.029018834330506e-05,
|
323 |
+
"loss": 0.0432,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 3.3576642335766422,
|
328 |
+
"grad_norm": 0.4706019163131714,
|
329 |
+
"learning_rate": 7.932084551589027e-05,
|
330 |
+
"loss": 0.036,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 3.4306569343065694,
|
335 |
+
"grad_norm": 0.502895176410675,
|
336 |
+
"learning_rate": 7.833440645111975e-05,
|
337 |
+
"loss": 0.0372,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 3.5036496350364965,
|
342 |
+
"grad_norm": 0.37358635663986206,
|
343 |
+
"learning_rate": 7.733144631645852e-05,
|
344 |
+
"loss": 0.0415,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 3.576642335766423,
|
349 |
+
"grad_norm": 0.38203373551368713,
|
350 |
+
"learning_rate": 7.631254991238621e-05,
|
351 |
+
"loss": 0.0362,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 3.6496350364963503,
|
356 |
+
"grad_norm": 0.38717857003211975,
|
357 |
+
"learning_rate": 7.527831133141476e-05,
|
358 |
+
"loss": 0.0416,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 3.7226277372262775,
|
363 |
+
"grad_norm": 0.32451218366622925,
|
364 |
+
"learning_rate": 7.422933361168825e-05,
|
365 |
+
"loss": 0.0378,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 3.795620437956204,
|
370 |
+
"grad_norm": 0.3965657651424408,
|
371 |
+
"learning_rate": 7.316622838536673e-05,
|
372 |
+
"loss": 0.0374,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 3.8686131386861313,
|
377 |
+
"grad_norm": 0.5710194110870361,
|
378 |
+
"learning_rate": 7.208961552199913e-05,
|
379 |
+
"loss": 0.0372,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 3.9416058394160585,
|
384 |
+
"grad_norm": 0.3017335832118988,
|
385 |
+
"learning_rate": 7.100012276709302e-05,
|
386 |
+
"loss": 0.0411,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 4.014598540145985,
|
391 |
+
"grad_norm": 0.42834600806236267,
|
392 |
+
"learning_rate": 6.98983853760924e-05,
|
393 |
+
"loss": 0.0324,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 4.087591240875913,
|
398 |
+
"grad_norm": 0.2624529302120209,
|
399 |
+
"learning_rate": 6.878504574397626e-05,
|
400 |
+
"loss": 0.0362,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 4.160583941605839,
|
405 |
+
"grad_norm": 0.3475762605667114,
|
406 |
+
"learning_rate": 6.766075303069459e-05,
|
407 |
+
"loss": 0.0398,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 4.233576642335766,
|
412 |
+
"grad_norm": 0.5206601023674011,
|
413 |
+
"learning_rate": 6.65261627826597e-05,
|
414 |
+
"loss": 0.0367,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 4.306569343065694,
|
419 |
+
"grad_norm": 0.4349361062049866,
|
420 |
+
"learning_rate": 6.538193655051381e-05,
|
421 |
+
"loss": 0.0365,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 4.37956204379562,
|
426 |
+
"grad_norm": 0.5204529166221619,
|
427 |
+
"learning_rate": 6.422874150339579e-05,
|
428 |
+
"loss": 0.0358,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 4.452554744525547,
|
433 |
+
"grad_norm": 0.27471956610679626,
|
434 |
+
"learning_rate": 6.30672500399318e-05,
|
435 |
+
"loss": 0.0336,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 4.525547445255475,
|
440 |
+
"grad_norm": 0.3909718990325928,
|
441 |
+
"learning_rate": 6.189813939617682e-05,
|
442 |
+
"loss": 0.0363,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 4.598540145985401,
|
447 |
+
"grad_norm": 0.24090121686458588,
|
448 |
+
"learning_rate": 6.072209125073561e-05,
|
449 |
+
"loss": 0.0363,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 4.671532846715328,
|
454 |
+
"grad_norm": 0.3554544448852539,
|
455 |
+
"learning_rate": 5.95397913272932e-05,
|
456 |
+
"loss": 0.0359,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 4.744525547445256,
|
461 |
+
"grad_norm": 0.31269019842147827,
|
462 |
+
"learning_rate": 5.8351928994787006e-05,
|
463 |
+
"loss": 0.0405,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 4.817518248175182,
|
468 |
+
"grad_norm": 0.40918299555778503,
|
469 |
+
"learning_rate": 5.7159196865453294e-05,
|
470 |
+
"loss": 0.0361,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 4.89051094890511,
|
475 |
+
"grad_norm": 0.522287130355835,
|
476 |
+
"learning_rate": 5.596229039098271e-05,
|
477 |
+
"loss": 0.0336,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 4.963503649635037,
|
482 |
+
"grad_norm": 0.48848339915275574,
|
483 |
+
"learning_rate": 5.4761907457020077e-05,
|
484 |
+
"loss": 0.0317,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 5.036496350364963,
|
489 |
+
"grad_norm": 0.4906352162361145,
|
490 |
+
"learning_rate": 5.355874797624515e-05,
|
491 |
+
"loss": 0.0295,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 5.109489051094891,
|
496 |
+
"grad_norm": 0.24832546710968018,
|
497 |
+
"learning_rate": 5.235351348027129e-05,
|
498 |
+
"loss": 0.0299,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 5.182481751824818,
|
503 |
+
"grad_norm": 0.3912707567214966,
|
504 |
+
"learning_rate": 5.1146906710600306e-05,
|
505 |
+
"loss": 0.0306,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 5.255474452554744,
|
510 |
+
"grad_norm": 0.23428483307361603,
|
511 |
+
"learning_rate": 4.993963120887183e-05,
|
512 |
+
"loss": 0.0312,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 5.328467153284672,
|
517 |
+
"grad_norm": 0.39608365297317505,
|
518 |
+
"learning_rate": 4.8732390906646097e-05,
|
519 |
+
"loss": 0.0323,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 5.401459854014599,
|
524 |
+
"grad_norm": 0.24625132977962494,
|
525 |
+
"learning_rate": 4.75258897149594e-05,
|
526 |
+
"loss": 0.0346,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 5.474452554744525,
|
531 |
+
"grad_norm": 0.4109422564506531,
|
532 |
+
"learning_rate": 4.632083111389153e-05,
|
533 |
+
"loss": 0.0305,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 5.547445255474453,
|
538 |
+
"grad_norm": 0.4575289487838745,
|
539 |
+
"learning_rate": 4.5117917742384456e-05,
|
540 |
+
"loss": 0.0335,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 5.62043795620438,
|
545 |
+
"grad_norm": 0.3231428265571594,
|
546 |
+
"learning_rate": 4.391785098855156e-05,
|
547 |
+
"loss": 0.032,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 5.693430656934306,
|
552 |
+
"grad_norm": 0.22695614397525787,
|
553 |
+
"learning_rate": 4.272133058071595e-05,
|
554 |
+
"loss": 0.0341,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 5.766423357664234,
|
559 |
+
"grad_norm": 0.3635699450969696,
|
560 |
+
"learning_rate": 4.1529054179416875e-05,
|
561 |
+
"loss": 0.0303,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 5.839416058394161,
|
566 |
+
"grad_norm": 0.44931599497795105,
|
567 |
+
"learning_rate": 4.034171697062157e-05,
|
568 |
+
"loss": 0.0312,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 5.912408759124087,
|
573 |
+
"grad_norm": 0.38062986731529236,
|
574 |
+
"learning_rate": 3.916001126038008e-05,
|
575 |
+
"loss": 0.0298,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 5.985401459854015,
|
580 |
+
"grad_norm": 0.44949817657470703,
|
581 |
+
"learning_rate": 3.7984626071159224e-05,
|
582 |
+
"loss": 0.0301,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 6.0583941605839415,
|
587 |
+
"grad_norm": 0.2569969892501831,
|
588 |
+
"learning_rate": 3.681624674009121e-05,
|
589 |
+
"loss": 0.03,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 6.131386861313868,
|
594 |
+
"grad_norm": 0.3157390356063843,
|
595 |
+
"learning_rate": 3.5655554519370956e-05,
|
596 |
+
"loss": 0.0303,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 6.204379562043796,
|
601 |
+
"grad_norm": 0.27934950590133667,
|
602 |
+
"learning_rate": 3.450322617903543e-05,
|
603 |
+
"loss": 0.0281,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 6.2773722627737225,
|
608 |
+
"grad_norm": 0.2728220224380493,
|
609 |
+
"learning_rate": 3.3359933612356156e-05,
|
610 |
+
"loss": 0.0287,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 6.350364963503649,
|
615 |
+
"grad_norm": 0.27907413244247437,
|
616 |
+
"learning_rate": 3.2226343444075465e-05,
|
617 |
+
"loss": 0.0268,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 6.423357664233577,
|
622 |
+
"grad_norm": 0.2804538309574127,
|
623 |
+
"learning_rate": 3.110311664171458e-05,
|
624 |
+
"loss": 0.03,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 6.4963503649635035,
|
629 |
+
"grad_norm": 0.5569368600845337,
|
630 |
+
"learning_rate": 2.999090813018035e-05,
|
631 |
+
"loss": 0.028,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 6.569343065693431,
|
636 |
+
"grad_norm": 0.7178159952163696,
|
637 |
+
"learning_rate": 2.8890366409895148e-05,
|
638 |
+
"loss": 0.0325,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 6.642335766423358,
|
643 |
+
"grad_norm": 0.19943037629127502,
|
644 |
+
"learning_rate": 2.780213317867292e-05,
|
645 |
+
"loss": 0.025,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 6.7153284671532845,
|
650 |
+
"grad_norm": 0.1796846240758896,
|
651 |
+
"learning_rate": 2.672684295756147e-05,
|
652 |
+
"loss": 0.0257,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 6.788321167883212,
|
657 |
+
"grad_norm": 0.23872807621955872,
|
658 |
+
"learning_rate": 2.566512272086945e-05,
|
659 |
+
"loss": 0.0304,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 6.861313868613139,
|
664 |
+
"grad_norm": 0.2793905735015869,
|
665 |
+
"learning_rate": 2.4617591530593613e-05,
|
666 |
+
"loss": 0.0245,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 6.934306569343065,
|
671 |
+
"grad_norm": 0.2569703161716461,
|
672 |
+
"learning_rate": 2.3584860175459584e-05,
|
673 |
+
"loss": 0.0269,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 7.007299270072993,
|
678 |
+
"grad_norm": 0.27794763445854187,
|
679 |
+
"learning_rate": 2.2567530814786463e-05,
|
680 |
+
"loss": 0.0287,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 7.08029197080292,
|
685 |
+
"grad_norm": 0.237870454788208,
|
686 |
+
"learning_rate": 2.156619662738319e-05,
|
687 |
+
"loss": 0.0261,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 7.153284671532846,
|
692 |
+
"grad_norm": 0.3608488142490387,
|
693 |
+
"learning_rate": 2.0581441465680986e-05,
|
694 |
+
"loss": 0.0259,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 7.226277372262774,
|
699 |
+
"grad_norm": 0.3045063316822052,
|
700 |
+
"learning_rate": 1.961383951530394e-05,
|
701 |
+
"loss": 0.0286,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 7.299270072992701,
|
706 |
+
"grad_norm": 0.2131444364786148,
|
707 |
+
"learning_rate": 1.866395496027602e-05,
|
708 |
+
"loss": 0.0254,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 7.372262773722627,
|
713 |
+
"grad_norm": 0.25216737389564514,
|
714 |
+
"learning_rate": 1.7732341654059785e-05,
|
715 |
+
"loss": 0.0222,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 7.445255474452555,
|
720 |
+
"grad_norm": 0.2963661849498749,
|
721 |
+
"learning_rate": 1.6819542796618487e-05,
|
722 |
+
"loss": 0.0292,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 7.518248175182482,
|
727 |
+
"grad_norm": 0.4280974864959717,
|
728 |
+
"learning_rate": 1.592609061769004e-05,
|
729 |
+
"loss": 0.0263,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 7.591240875912408,
|
734 |
+
"grad_norm": 0.23306907713413239,
|
735 |
+
"learning_rate": 1.5052506066457461e-05,
|
736 |
+
"loss": 0.0211,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 7.664233576642336,
|
741 |
+
"grad_norm": 0.2898852825164795,
|
742 |
+
"learning_rate": 1.4199298507796698e-05,
|
743 |
+
"loss": 0.0209,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 7.737226277372263,
|
748 |
+
"grad_norm": 0.22745266556739807,
|
749 |
+
"learning_rate": 1.3366965425278899e-05,
|
750 |
+
"loss": 0.0218,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 7.81021897810219,
|
755 |
+
"grad_norm": 0.244761660695076,
|
756 |
+
"learning_rate": 1.2555992131100457e-05,
|
757 |
+
"loss": 0.0213,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 7.883211678832117,
|
762 |
+
"grad_norm": 0.3487051725387573,
|
763 |
+
"learning_rate": 1.1766851483109858e-05,
|
764 |
+
"loss": 0.0251,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 7.956204379562044,
|
769 |
+
"grad_norm": 0.38351985812187195,
|
770 |
+
"learning_rate": 1.1000003609096337e-05,
|
771 |
+
"loss": 0.0233,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 8.02919708029197,
|
776 |
+
"grad_norm": 0.2925453782081604,
|
777 |
+
"learning_rate": 1.0255895638501045e-05,
|
778 |
+
"loss": 0.0258,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 8.102189781021897,
|
783 |
+
"grad_norm": 0.2964305281639099,
|
784 |
+
"learning_rate": 9.534961441707307e-06,
|
785 |
+
"loss": 0.0193,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 8.175182481751825,
|
790 |
+
"grad_norm": 0.4270758628845215,
|
791 |
+
"learning_rate": 8.837621377061877e-06,
|
792 |
+
"loss": 0.0223,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 8.248175182481752,
|
797 |
+
"grad_norm": 0.22104066610336304,
|
798 |
+
"learning_rate": 8.16428204577468e-06,
|
799 |
+
"loss": 0.0217,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 8.321167883211679,
|
804 |
+
"grad_norm": 0.24666981399059296,
|
805 |
+
"learning_rate": 7.515336054840022e-06,
|
806 |
+
"loss": 0.0254,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 8.394160583941606,
|
811 |
+
"grad_norm": 0.16614589095115662,
|
812 |
+
"learning_rate": 6.8911617881174725e-06,
|
813 |
+
"loss": 0.02,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 8.467153284671532,
|
818 |
+
"grad_norm": 0.2146356999874115,
|
819 |
+
"learning_rate": 6.292123185705867e-06,
|
820 |
+
"loss": 0.024,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 8.540145985401459,
|
825 |
+
"grad_norm": 0.1602909117937088,
|
826 |
+
"learning_rate": 5.718569531739154e-06,
|
827 |
+
"loss": 0.0267,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 8.613138686131387,
|
832 |
+
"grad_norm": 0.4400380849838257,
|
833 |
+
"learning_rate": 5.170835250727663e-06,
|
834 |
+
"loss": 0.0231,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 8.686131386861314,
|
839 |
+
"grad_norm": 0.18121108412742615,
|
840 |
+
"learning_rate": 4.6492397125637525e-06,
|
841 |
+
"loss": 0.0192,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 8.75912408759124,
|
846 |
+
"grad_norm": 0.26767614483833313,
|
847 |
+
"learning_rate": 4.154087046305322e-06,
|
848 |
+
"loss": 0.0249,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 8.832116788321168,
|
853 |
+
"grad_norm": 0.2408941090106964,
|
854 |
+
"learning_rate": 3.6856659628459912e-06,
|
855 |
+
"loss": 0.0214,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 8.905109489051094,
|
860 |
+
"grad_norm": 0.2535434663295746,
|
861 |
+
"learning_rate": 3.244249586575038e-06,
|
862 |
+
"loss": 0.0203,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 8.978102189781023,
|
867 |
+
"grad_norm": 0.6523481011390686,
|
868 |
+
"learning_rate": 2.830095296125612e-06,
|
869 |
+
"loss": 0.0222,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 9.05109489051095,
|
874 |
+
"grad_norm": 0.16764762997627258,
|
875 |
+
"learning_rate": 2.4434445743037713e-06,
|
876 |
+
"loss": 0.0225,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 9.124087591240876,
|
881 |
+
"grad_norm": 0.35821276903152466,
|
882 |
+
"learning_rate": 2.0845228672860538e-06,
|
883 |
+
"loss": 0.0206,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 9.197080291970803,
|
888 |
+
"grad_norm": 0.38857728242874146,
|
889 |
+
"learning_rate": 1.7535394531675187e-06,
|
890 |
+
"loss": 0.0222,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 9.27007299270073,
|
895 |
+
"grad_norm": 0.2351900190114975,
|
896 |
+
"learning_rate": 1.4506873199370497e-06,
|
897 |
+
"loss": 0.021,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 9.343065693430656,
|
902 |
+
"grad_norm": 0.33225876092910767,
|
903 |
+
"learning_rate": 1.1761430529509899e-06,
|
904 |
+
"loss": 0.0258,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 9.416058394160585,
|
909 |
+
"grad_norm": 0.2640361785888672,
|
910 |
+
"learning_rate": 9.300667319706857e-07,
|
911 |
+
"loss": 0.0202,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 9.489051094890511,
|
916 |
+
"grad_norm": 0.15353083610534668,
|
917 |
+
"learning_rate": 7.126018378241062e-07,
|
918 |
+
"loss": 0.0215,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 9.562043795620438,
|
923 |
+
"grad_norm": 0.33280253410339355,
|
924 |
+
"learning_rate": 5.238751687458021e-07,
|
925 |
+
"loss": 0.0217,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 9.635036496350365,
|
930 |
+
"grad_norm": 0.18690501153469086,
|
931 |
+
"learning_rate": 3.639967664440802e-07,
|
932 |
+
"loss": 0.0194,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 9.708029197080291,
|
937 |
+
"grad_norm": 0.6559237837791443,
|
938 |
+
"learning_rate": 2.3305985193852742e-07,
|
939 |
+
"loss": 0.0242,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 9.78102189781022,
|
944 |
+
"grad_norm": 0.3035587966442108,
|
945 |
+
"learning_rate": 1.3114077120517376e-07,
|
946 |
+
"loss": 0.0271,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 9.854014598540147,
|
951 |
+
"grad_norm": 0.30467328429222107,
|
952 |
+
"learning_rate": 5.8298950661112017e-08,
|
953 |
+
"loss": 0.0187,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 9.927007299270073,
|
958 |
+
"grad_norm": 0.22451713681221008,
|
959 |
+
"learning_rate": 1.4576862514487089e-08,
|
960 |
+
"loss": 0.022,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 10.0,
|
965 |
+
"grad_norm": 0.33209139108657837,
|
966 |
+
"learning_rate": 0.0,
|
967 |
+
"loss": 0.0227,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 10.0,
|
972 |
+
"step": 1370,
|
973 |
+
"total_flos": 1.9634851190645568e+17,
|
974 |
+
"train_loss": 0.05370959477485531,
|
975 |
+
"train_runtime": 1800.9203,
|
976 |
+
"train_samples_per_second": 48.631,
|
977 |
+
"train_steps_per_second": 0.761
|
978 |
+
}
|
979 |
+
],
|
980 |
+
"logging_steps": 10,
|
981 |
+
"max_steps": 1370,
|
982 |
+
"num_input_tokens_seen": 0,
|
983 |
+
"num_train_epochs": 10,
|
984 |
+
"save_steps": 10000,
|
985 |
+
"stateful_callbacks": {
|
986 |
+
"TrainerControl": {
|
987 |
+
"args": {
|
988 |
+
"should_epoch_stop": false,
|
989 |
+
"should_evaluate": false,
|
990 |
+
"should_log": false,
|
991 |
+
"should_save": true,
|
992 |
+
"should_training_stop": true
|
993 |
+
},
|
994 |
+
"attributes": {}
|
995 |
+
}
|
996 |
+
},
|
997 |
+
"total_flos": 1.9634851190645568e+17,
|
998 |
+
"train_batch_size": 64,
|
999 |
+
"trial_name": null,
|
1000 |
+
"trial_params": null
|
1001 |
+
}
|