Gokulapriyan
commited on
Commit
•
6582283
1
Parent(s):
c88262b
End of training
Browse files
all_results.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.0,
|
3 |
+
"eval_accuracy": 0.9782978378816098,
|
4 |
+
"eval_loss": 0.05557282269001007,
|
5 |
+
"eval_runtime": 206.8353,
|
6 |
+
"eval_samples_per_second": 59.704,
|
7 |
+
"eval_steps_per_second": 1.866,
|
8 |
+
"total_flos": 4.5505049329926144e+18,
|
9 |
+
"train_loss": 0.22463261898293163,
|
10 |
+
"train_runtime": 4458.2012,
|
11 |
+
"train_samples_per_second": 31.39,
|
12 |
+
"train_steps_per_second": 0.245
|
13 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.0,
|
3 |
+
"eval_accuracy": 0.9782978378816098,
|
4 |
+
"eval_loss": 0.05557282269001007,
|
5 |
+
"eval_runtime": 206.8353,
|
6 |
+
"eval_samples_per_second": 59.704,
|
7 |
+
"eval_steps_per_second": 1.866
|
8 |
+
}
|
runs/Feb20_13-48-34_f8a2a2cfe247/events.out.tfevents.1676905625.f8a2a2cfe247.738.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c2d3ec69d8dec71ef8a28ebba2df0493950224d40012ff96d274364c4b34b6ce
|
3 |
+
size 363
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.0,
|
3 |
+
"total_flos": 4.5505049329926144e+18,
|
4 |
+
"train_loss": 0.22463261898293163,
|
5 |
+
"train_runtime": 4458.2012,
|
6 |
+
"train_samples_per_second": 31.39,
|
7 |
+
"train_steps_per_second": 0.245
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,697 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.9782978378816098,
|
3 |
+
"best_model_checkpoint": "swinv2-tiny-patch4-window8-256-finetuned-og-dataset-10e-finetuned-og-dataset-10e/checkpoint-1092",
|
4 |
+
"epoch": 1.9986282578875172,
|
5 |
+
"global_step": 1092,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.02,
|
12 |
+
"learning_rate": 4.5454545454545455e-06,
|
13 |
+
"loss": 0.2055,
|
14 |
+
"step": 10
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.04,
|
18 |
+
"learning_rate": 9.090909090909091e-06,
|
19 |
+
"loss": 0.2267,
|
20 |
+
"step": 20
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.05,
|
24 |
+
"learning_rate": 1.3636363636363637e-05,
|
25 |
+
"loss": 0.2416,
|
26 |
+
"step": 30
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.07,
|
30 |
+
"learning_rate": 1.8181818181818182e-05,
|
31 |
+
"loss": 0.206,
|
32 |
+
"step": 40
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.09,
|
36 |
+
"learning_rate": 2.272727272727273e-05,
|
37 |
+
"loss": 0.2383,
|
38 |
+
"step": 50
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.11,
|
42 |
+
"learning_rate": 2.7272727272727273e-05,
|
43 |
+
"loss": 0.2187,
|
44 |
+
"step": 60
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.13,
|
48 |
+
"learning_rate": 3.181818181818182e-05,
|
49 |
+
"loss": 0.2272,
|
50 |
+
"step": 70
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.15,
|
54 |
+
"learning_rate": 3.6363636363636364e-05,
|
55 |
+
"loss": 0.2259,
|
56 |
+
"step": 80
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.16,
|
60 |
+
"learning_rate": 4.0909090909090915e-05,
|
61 |
+
"loss": 0.2384,
|
62 |
+
"step": 90
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 0.18,
|
66 |
+
"learning_rate": 4.545454545454546e-05,
|
67 |
+
"loss": 0.2495,
|
68 |
+
"step": 100
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 0.2,
|
72 |
+
"learning_rate": 5e-05,
|
73 |
+
"loss": 0.2547,
|
74 |
+
"step": 110
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 0.22,
|
78 |
+
"learning_rate": 4.9490835030549896e-05,
|
79 |
+
"loss": 0.2233,
|
80 |
+
"step": 120
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.24,
|
84 |
+
"learning_rate": 4.89816700610998e-05,
|
85 |
+
"loss": 0.2539,
|
86 |
+
"step": 130
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.26,
|
90 |
+
"learning_rate": 4.84725050916497e-05,
|
91 |
+
"loss": 0.2629,
|
92 |
+
"step": 140
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.27,
|
96 |
+
"learning_rate": 4.79633401221996e-05,
|
97 |
+
"loss": 0.2284,
|
98 |
+
"step": 150
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 0.29,
|
102 |
+
"learning_rate": 4.745417515274949e-05,
|
103 |
+
"loss": 0.2647,
|
104 |
+
"step": 160
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 0.31,
|
108 |
+
"learning_rate": 4.694501018329939e-05,
|
109 |
+
"loss": 0.2756,
|
110 |
+
"step": 170
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 0.33,
|
114 |
+
"learning_rate": 4.643584521384929e-05,
|
115 |
+
"loss": 0.2573,
|
116 |
+
"step": 180
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 0.35,
|
120 |
+
"learning_rate": 4.592668024439919e-05,
|
121 |
+
"loss": 0.2583,
|
122 |
+
"step": 190
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.37,
|
126 |
+
"learning_rate": 4.541751527494909e-05,
|
127 |
+
"loss": 0.253,
|
128 |
+
"step": 200
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.38,
|
132 |
+
"learning_rate": 4.490835030549899e-05,
|
133 |
+
"loss": 0.2678,
|
134 |
+
"step": 210
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.4,
|
138 |
+
"learning_rate": 4.439918533604888e-05,
|
139 |
+
"loss": 0.2518,
|
140 |
+
"step": 220
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 0.42,
|
144 |
+
"learning_rate": 4.3890020366598776e-05,
|
145 |
+
"loss": 0.2511,
|
146 |
+
"step": 230
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 0.44,
|
150 |
+
"learning_rate": 4.338085539714868e-05,
|
151 |
+
"loss": 0.2686,
|
152 |
+
"step": 240
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 0.46,
|
156 |
+
"learning_rate": 4.287169042769857e-05,
|
157 |
+
"loss": 0.2609,
|
158 |
+
"step": 250
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 0.48,
|
162 |
+
"learning_rate": 4.236252545824848e-05,
|
163 |
+
"loss": 0.237,
|
164 |
+
"step": 260
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.49,
|
168 |
+
"learning_rate": 4.185336048879837e-05,
|
169 |
+
"loss": 0.2512,
|
170 |
+
"step": 270
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.51,
|
174 |
+
"learning_rate": 4.134419551934827e-05,
|
175 |
+
"loss": 0.2511,
|
176 |
+
"step": 280
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 0.53,
|
180 |
+
"learning_rate": 4.083503054989817e-05,
|
181 |
+
"loss": 0.2532,
|
182 |
+
"step": 290
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 0.55,
|
186 |
+
"learning_rate": 4.032586558044807e-05,
|
187 |
+
"loss": 0.262,
|
188 |
+
"step": 300
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 0.57,
|
192 |
+
"learning_rate": 3.981670061099796e-05,
|
193 |
+
"loss": 0.285,
|
194 |
+
"step": 310
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 0.59,
|
198 |
+
"learning_rate": 3.930753564154787e-05,
|
199 |
+
"loss": 0.2432,
|
200 |
+
"step": 320
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 0.6,
|
204 |
+
"learning_rate": 3.879837067209776e-05,
|
205 |
+
"loss": 0.2498,
|
206 |
+
"step": 330
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.62,
|
210 |
+
"learning_rate": 3.8289205702647656e-05,
|
211 |
+
"loss": 0.2676,
|
212 |
+
"step": 340
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.64,
|
216 |
+
"learning_rate": 3.778004073319756e-05,
|
217 |
+
"loss": 0.2773,
|
218 |
+
"step": 350
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 0.66,
|
222 |
+
"learning_rate": 3.727087576374745e-05,
|
223 |
+
"loss": 0.2349,
|
224 |
+
"step": 360
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 0.68,
|
228 |
+
"learning_rate": 3.676171079429735e-05,
|
229 |
+
"loss": 0.257,
|
230 |
+
"step": 370
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 0.7,
|
234 |
+
"learning_rate": 3.625254582484725e-05,
|
235 |
+
"loss": 0.2374,
|
236 |
+
"step": 380
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 0.71,
|
240 |
+
"learning_rate": 3.574338085539715e-05,
|
241 |
+
"loss": 0.2531,
|
242 |
+
"step": 390
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 0.73,
|
246 |
+
"learning_rate": 3.523421588594705e-05,
|
247 |
+
"loss": 0.237,
|
248 |
+
"step": 400
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.75,
|
252 |
+
"learning_rate": 3.472505091649695e-05,
|
253 |
+
"loss": 0.2559,
|
254 |
+
"step": 410
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.77,
|
258 |
+
"learning_rate": 3.421588594704684e-05,
|
259 |
+
"loss": 0.2477,
|
260 |
+
"step": 420
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 0.79,
|
264 |
+
"learning_rate": 3.370672097759674e-05,
|
265 |
+
"loss": 0.2202,
|
266 |
+
"step": 430
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 0.8,
|
270 |
+
"learning_rate": 3.319755600814664e-05,
|
271 |
+
"loss": 0.2655,
|
272 |
+
"step": 440
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 0.82,
|
276 |
+
"learning_rate": 3.268839103869654e-05,
|
277 |
+
"loss": 0.2348,
|
278 |
+
"step": 450
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 0.84,
|
282 |
+
"learning_rate": 3.217922606924644e-05,
|
283 |
+
"loss": 0.2322,
|
284 |
+
"step": 460
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 0.86,
|
288 |
+
"learning_rate": 3.167006109979633e-05,
|
289 |
+
"loss": 0.227,
|
290 |
+
"step": 470
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.88,
|
294 |
+
"learning_rate": 3.116089613034623e-05,
|
295 |
+
"loss": 0.2518,
|
296 |
+
"step": 480
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.9,
|
300 |
+
"learning_rate": 3.065173116089613e-05,
|
301 |
+
"loss": 0.2472,
|
302 |
+
"step": 490
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 0.91,
|
306 |
+
"learning_rate": 3.014256619144603e-05,
|
307 |
+
"loss": 0.2826,
|
308 |
+
"step": 500
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 0.93,
|
312 |
+
"learning_rate": 2.9633401221995927e-05,
|
313 |
+
"loss": 0.2387,
|
314 |
+
"step": 510
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 0.95,
|
318 |
+
"learning_rate": 2.9124236252545828e-05,
|
319 |
+
"loss": 0.2321,
|
320 |
+
"step": 520
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 0.97,
|
324 |
+
"learning_rate": 2.8615071283095725e-05,
|
325 |
+
"loss": 0.2018,
|
326 |
+
"step": 530
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 0.99,
|
330 |
+
"learning_rate": 2.8105906313645626e-05,
|
331 |
+
"loss": 0.2237,
|
332 |
+
"step": 540
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 1.0,
|
336 |
+
"eval_accuracy": 0.9735201230868896,
|
337 |
+
"eval_loss": 0.07291990518569946,
|
338 |
+
"eval_runtime": 204.6202,
|
339 |
+
"eval_samples_per_second": 60.351,
|
340 |
+
"eval_steps_per_second": 1.886,
|
341 |
+
"step": 546
|
342 |
+
},
|
343 |
+
{
|
344 |
+
"epoch": 1.01,
|
345 |
+
"learning_rate": 2.759674134419552e-05,
|
346 |
+
"loss": 0.2748,
|
347 |
+
"step": 550
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 1.03,
|
351 |
+
"learning_rate": 2.708757637474542e-05,
|
352 |
+
"loss": 0.2414,
|
353 |
+
"step": 560
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 1.04,
|
357 |
+
"learning_rate": 2.6578411405295317e-05,
|
358 |
+
"loss": 0.2247,
|
359 |
+
"step": 570
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.06,
|
363 |
+
"learning_rate": 2.606924643584521e-05,
|
364 |
+
"loss": 0.2446,
|
365 |
+
"step": 580
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"epoch": 1.08,
|
369 |
+
"learning_rate": 2.5560081466395115e-05,
|
370 |
+
"loss": 0.2291,
|
371 |
+
"step": 590
|
372 |
+
},
|
373 |
+
{
|
374 |
+
"epoch": 1.1,
|
375 |
+
"learning_rate": 2.505091649694501e-05,
|
376 |
+
"loss": 0.221,
|
377 |
+
"step": 600
|
378 |
+
},
|
379 |
+
{
|
380 |
+
"epoch": 1.12,
|
381 |
+
"learning_rate": 2.454175152749491e-05,
|
382 |
+
"loss": 0.2218,
|
383 |
+
"step": 610
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 1.14,
|
387 |
+
"learning_rate": 2.403258655804481e-05,
|
388 |
+
"loss": 0.2289,
|
389 |
+
"step": 620
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 1.15,
|
393 |
+
"learning_rate": 2.3523421588594704e-05,
|
394 |
+
"loss": 0.1879,
|
395 |
+
"step": 630
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 1.17,
|
399 |
+
"learning_rate": 2.30142566191446e-05,
|
400 |
+
"loss": 0.2292,
|
401 |
+
"step": 640
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.19,
|
405 |
+
"learning_rate": 2.2505091649694502e-05,
|
406 |
+
"loss": 0.244,
|
407 |
+
"step": 650
|
408 |
+
},
|
409 |
+
{
|
410 |
+
"epoch": 1.21,
|
411 |
+
"learning_rate": 2.19959266802444e-05,
|
412 |
+
"loss": 0.2235,
|
413 |
+
"step": 660
|
414 |
+
},
|
415 |
+
{
|
416 |
+
"epoch": 1.23,
|
417 |
+
"learning_rate": 2.1486761710794297e-05,
|
418 |
+
"loss": 0.2208,
|
419 |
+
"step": 670
|
420 |
+
},
|
421 |
+
{
|
422 |
+
"epoch": 1.25,
|
423 |
+
"learning_rate": 2.0977596741344197e-05,
|
424 |
+
"loss": 0.2306,
|
425 |
+
"step": 680
|
426 |
+
},
|
427 |
+
{
|
428 |
+
"epoch": 1.26,
|
429 |
+
"learning_rate": 2.0468431771894095e-05,
|
430 |
+
"loss": 0.2214,
|
431 |
+
"step": 690
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"epoch": 1.28,
|
435 |
+
"learning_rate": 1.9959266802443992e-05,
|
436 |
+
"loss": 0.1851,
|
437 |
+
"step": 700
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 1.3,
|
441 |
+
"learning_rate": 1.9450101832993893e-05,
|
442 |
+
"loss": 0.2152,
|
443 |
+
"step": 710
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.32,
|
447 |
+
"learning_rate": 1.894093686354379e-05,
|
448 |
+
"loss": 0.1935,
|
449 |
+
"step": 720
|
450 |
+
},
|
451 |
+
{
|
452 |
+
"epoch": 1.34,
|
453 |
+
"learning_rate": 1.8431771894093687e-05,
|
454 |
+
"loss": 0.2138,
|
455 |
+
"step": 730
|
456 |
+
},
|
457 |
+
{
|
458 |
+
"epoch": 1.35,
|
459 |
+
"learning_rate": 1.7922606924643584e-05,
|
460 |
+
"loss": 0.221,
|
461 |
+
"step": 740
|
462 |
+
},
|
463 |
+
{
|
464 |
+
"epoch": 1.37,
|
465 |
+
"learning_rate": 1.741344195519348e-05,
|
466 |
+
"loss": 0.2085,
|
467 |
+
"step": 750
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"epoch": 1.39,
|
471 |
+
"learning_rate": 1.6904276985743382e-05,
|
472 |
+
"loss": 0.1815,
|
473 |
+
"step": 760
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 1.41,
|
477 |
+
"learning_rate": 1.639511201629328e-05,
|
478 |
+
"loss": 0.2142,
|
479 |
+
"step": 770
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 1.43,
|
483 |
+
"learning_rate": 1.5885947046843177e-05,
|
484 |
+
"loss": 0.2027,
|
485 |
+
"step": 780
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.45,
|
489 |
+
"learning_rate": 1.5376782077393077e-05,
|
490 |
+
"loss": 0.1703,
|
491 |
+
"step": 790
|
492 |
+
},
|
493 |
+
{
|
494 |
+
"epoch": 1.46,
|
495 |
+
"learning_rate": 1.4867617107942975e-05,
|
496 |
+
"loss": 0.1879,
|
497 |
+
"step": 800
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"epoch": 1.48,
|
501 |
+
"learning_rate": 1.4358452138492872e-05,
|
502 |
+
"loss": 0.2016,
|
503 |
+
"step": 810
|
504 |
+
},
|
505 |
+
{
|
506 |
+
"epoch": 1.5,
|
507 |
+
"learning_rate": 1.3849287169042771e-05,
|
508 |
+
"loss": 0.1797,
|
509 |
+
"step": 820
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 1.52,
|
513 |
+
"learning_rate": 1.334012219959267e-05,
|
514 |
+
"loss": 0.2204,
|
515 |
+
"step": 830
|
516 |
+
},
|
517 |
+
{
|
518 |
+
"epoch": 1.54,
|
519 |
+
"learning_rate": 1.2830957230142567e-05,
|
520 |
+
"loss": 0.1896,
|
521 |
+
"step": 840
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 1.56,
|
525 |
+
"learning_rate": 1.2321792260692464e-05,
|
526 |
+
"loss": 0.1661,
|
527 |
+
"step": 850
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.57,
|
531 |
+
"learning_rate": 1.1812627291242363e-05,
|
532 |
+
"loss": 0.1973,
|
533 |
+
"step": 860
|
534 |
+
},
|
535 |
+
{
|
536 |
+
"epoch": 1.59,
|
537 |
+
"learning_rate": 1.1303462321792262e-05,
|
538 |
+
"loss": 0.2137,
|
539 |
+
"step": 870
|
540 |
+
},
|
541 |
+
{
|
542 |
+
"epoch": 1.61,
|
543 |
+
"learning_rate": 1.079429735234216e-05,
|
544 |
+
"loss": 0.2134,
|
545 |
+
"step": 880
|
546 |
+
},
|
547 |
+
{
|
548 |
+
"epoch": 1.63,
|
549 |
+
"learning_rate": 1.0285132382892057e-05,
|
550 |
+
"loss": 0.2033,
|
551 |
+
"step": 890
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 1.65,
|
555 |
+
"learning_rate": 9.775967413441956e-06,
|
556 |
+
"loss": 0.1814,
|
557 |
+
"step": 900
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 1.67,
|
561 |
+
"learning_rate": 9.266802443991853e-06,
|
562 |
+
"loss": 0.2123,
|
563 |
+
"step": 910
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 1.68,
|
567 |
+
"learning_rate": 8.757637474541752e-06,
|
568 |
+
"loss": 0.1829,
|
569 |
+
"step": 920
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.7,
|
573 |
+
"learning_rate": 8.248472505091651e-06,
|
574 |
+
"loss": 0.1747,
|
575 |
+
"step": 930
|
576 |
+
},
|
577 |
+
{
|
578 |
+
"epoch": 1.72,
|
579 |
+
"learning_rate": 7.739307535641548e-06,
|
580 |
+
"loss": 0.1861,
|
581 |
+
"step": 940
|
582 |
+
},
|
583 |
+
{
|
584 |
+
"epoch": 1.74,
|
585 |
+
"learning_rate": 7.230142566191446e-06,
|
586 |
+
"loss": 0.1823,
|
587 |
+
"step": 950
|
588 |
+
},
|
589 |
+
{
|
590 |
+
"epoch": 1.76,
|
591 |
+
"learning_rate": 6.7209775967413446e-06,
|
592 |
+
"loss": 0.2024,
|
593 |
+
"step": 960
|
594 |
+
},
|
595 |
+
{
|
596 |
+
"epoch": 1.78,
|
597 |
+
"learning_rate": 6.211812627291243e-06,
|
598 |
+
"loss": 0.2012,
|
599 |
+
"step": 970
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 1.79,
|
603 |
+
"learning_rate": 5.702647657841141e-06,
|
604 |
+
"loss": 0.1792,
|
605 |
+
"step": 980
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 1.81,
|
609 |
+
"learning_rate": 5.193482688391039e-06,
|
610 |
+
"loss": 0.2112,
|
611 |
+
"step": 990
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.83,
|
615 |
+
"learning_rate": 4.684317718940937e-06,
|
616 |
+
"loss": 0.1968,
|
617 |
+
"step": 1000
|
618 |
+
},
|
619 |
+
{
|
620 |
+
"epoch": 1.85,
|
621 |
+
"learning_rate": 4.175152749490835e-06,
|
622 |
+
"loss": 0.2072,
|
623 |
+
"step": 1010
|
624 |
+
},
|
625 |
+
{
|
626 |
+
"epoch": 1.87,
|
627 |
+
"learning_rate": 3.6659877800407332e-06,
|
628 |
+
"loss": 0.1716,
|
629 |
+
"step": 1020
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 1.89,
|
633 |
+
"learning_rate": 3.1568228105906318e-06,
|
634 |
+
"loss": 0.1922,
|
635 |
+
"step": 1030
|
636 |
+
},
|
637 |
+
{
|
638 |
+
"epoch": 1.9,
|
639 |
+
"learning_rate": 2.64765784114053e-06,
|
640 |
+
"loss": 0.1972,
|
641 |
+
"step": 1040
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 1.92,
|
645 |
+
"learning_rate": 2.1384928716904276e-06,
|
646 |
+
"loss": 0.2012,
|
647 |
+
"step": 1050
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 1.94,
|
651 |
+
"learning_rate": 1.6293279022403257e-06,
|
652 |
+
"loss": 0.2165,
|
653 |
+
"step": 1060
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 1.96,
|
657 |
+
"learning_rate": 1.120162932790224e-06,
|
658 |
+
"loss": 0.1747,
|
659 |
+
"step": 1070
|
660 |
+
},
|
661 |
+
{
|
662 |
+
"epoch": 1.98,
|
663 |
+
"learning_rate": 6.109979633401222e-07,
|
664 |
+
"loss": 0.169,
|
665 |
+
"step": 1080
|
666 |
+
},
|
667 |
+
{
|
668 |
+
"epoch": 1.99,
|
669 |
+
"learning_rate": 1.0183299389002036e-07,
|
670 |
+
"loss": 0.1672,
|
671 |
+
"step": 1090
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 2.0,
|
675 |
+
"eval_accuracy": 0.9782978378816098,
|
676 |
+
"eval_loss": 0.05557282269001007,
|
677 |
+
"eval_runtime": 213.2817,
|
678 |
+
"eval_samples_per_second": 57.9,
|
679 |
+
"eval_steps_per_second": 1.81,
|
680 |
+
"step": 1092
|
681 |
+
},
|
682 |
+
{
|
683 |
+
"epoch": 2.0,
|
684 |
+
"step": 1092,
|
685 |
+
"total_flos": 4.5505049329926144e+18,
|
686 |
+
"train_loss": 0.22463261898293163,
|
687 |
+
"train_runtime": 4458.2012,
|
688 |
+
"train_samples_per_second": 31.39,
|
689 |
+
"train_steps_per_second": 0.245
|
690 |
+
}
|
691 |
+
],
|
692 |
+
"max_steps": 1092,
|
693 |
+
"num_train_epochs": 2,
|
694 |
+
"total_flos": 4.5505049329926144e+18,
|
695 |
+
"trial_name": null,
|
696 |
+
"trial_params": null
|
697 |
+
}
|