End of training
Browse files- README.md +5 -5
- all_results.json +14 -14
- eval_results.json +10 -10
- runs/Sep23_07-38-42_data-science-wbi/events.out.tfevents.1727088886.data-science-wbi +3 -0
- train_results.json +5 -5
- trainer_state.json +181 -181
README.md
CHANGED
@@ -33,12 +33,12 @@ should probably proofread and complete it, then remove this comment. -->
|
|
33 |
|
34 |
This model is a fine-tuned version of [microsoft/swin-base-patch4-window7-224-in22k](https://huggingface.co/microsoft/swin-base-patch4-window7-224-in22k) on the imagefolder dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
-
- Loss: 0.
|
37 |
- Accuracy: 0.8600
|
38 |
-
- Auc Overall: 0.
|
39 |
-
- Auc Class 0: 0.
|
40 |
-
- Auc Class 1: 0.
|
41 |
-
- Auc Class 2: 0.
|
42 |
|
43 |
## Model description
|
44 |
|
|
|
33 |
|
34 |
This model is a fine-tuned version of [microsoft/swin-base-patch4-window7-224-in22k](https://huggingface.co/microsoft/swin-base-patch4-window7-224-in22k) on the imagefolder dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
+
- Loss: 0.3584
|
37 |
- Accuracy: 0.8600
|
38 |
+
- Auc Overall: 0.9327
|
39 |
+
- Auc Class 0: 0.9624
|
40 |
+
- Auc Class 1: 0.9106
|
41 |
+
- Auc Class 2: 0.9253
|
42 |
|
43 |
## Model description
|
44 |
|
all_results.json
CHANGED
@@ -1,17 +1,17 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_auc_class_0": 0.
|
5 |
-
"eval_auc_class_1": 0.
|
6 |
-
"eval_auc_class_2": 0.
|
7 |
-
"eval_auc_overall": 0.
|
8 |
-
"eval_loss": 0.
|
9 |
-
"eval_runtime":
|
10 |
-
"eval_samples_per_second": 9.
|
11 |
-
"eval_steps_per_second": 0.
|
12 |
-
"total_flos": 4.
|
13 |
-
"train_loss": 0.
|
14 |
-
"train_runtime":
|
15 |
-
"train_samples_per_second": 4.
|
16 |
"train_steps_per_second": 0.009
|
17 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.0,
|
3 |
+
"eval_accuracy": 0.8600206611570248,
|
4 |
+
"eval_auc_class_0": 0.9623798076923077,
|
5 |
+
"eval_auc_class_1": 0.9106044580248364,
|
6 |
+
"eval_auc_class_2": 0.9252550195053197,
|
7 |
+
"eval_auc_overall": 0.932746428407488,
|
8 |
+
"eval_loss": 0.358356237411499,
|
9 |
+
"eval_runtime": 205.5157,
|
10 |
+
"eval_samples_per_second": 9.42,
|
11 |
+
"eval_steps_per_second": 0.078,
|
12 |
+
"total_flos": 4.143902092898771e+18,
|
13 |
+
"train_loss": 0.37870263712746755,
|
14 |
+
"train_runtime": 11542.069,
|
15 |
+
"train_samples_per_second": 4.583,
|
16 |
"train_steps_per_second": 0.009
|
17 |
}
|
eval_results.json
CHANGED
@@ -1,12 +1,12 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_auc_class_0": 0.
|
5 |
-
"eval_auc_class_1": 0.
|
6 |
-
"eval_auc_class_2": 0.
|
7 |
-
"eval_auc_overall": 0.
|
8 |
-
"eval_loss": 0.
|
9 |
-
"eval_runtime":
|
10 |
-
"eval_samples_per_second": 9.
|
11 |
-
"eval_steps_per_second": 0.
|
12 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.0,
|
3 |
+
"eval_accuracy": 0.8600206611570248,
|
4 |
+
"eval_auc_class_0": 0.9623798076923077,
|
5 |
+
"eval_auc_class_1": 0.9106044580248364,
|
6 |
+
"eval_auc_class_2": 0.9252550195053197,
|
7 |
+
"eval_auc_overall": 0.932746428407488,
|
8 |
+
"eval_loss": 0.358356237411499,
|
9 |
+
"eval_runtime": 205.5157,
|
10 |
+
"eval_samples_per_second": 9.42,
|
11 |
+
"eval_steps_per_second": 0.078
|
12 |
}
|
runs/Sep23_07-38-42_data-science-wbi/events.out.tfevents.1727088886.data-science-wbi
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9bd2868be35a73e31d4a9ce520d716280e3dbcb46be6d1ee9419479fc2e6b864
|
3 |
+
size 300
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"total_flos": 4.
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second": 4.
|
7 |
"train_steps_per_second": 0.009
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.0,
|
3 |
+
"total_flos": 4.143902092898771e+18,
|
4 |
+
"train_loss": 0.37870263712746755,
|
5 |
+
"train_runtime": 11542.069,
|
6 |
+
"train_samples_per_second": 4.583,
|
7 |
"train_steps_per_second": 0.009
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM/NMM_Classification_base_V10/checkpoint-
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
"global_step": 105,
|
7 |
"is_hyper_param_search": false,
|
@@ -9,264 +9,264 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0.
|
13 |
-
"grad_norm": 4.
|
14 |
"learning_rate": 4.545454545454546e-05,
|
15 |
-
"loss": 0.
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0
|
20 |
"step": 15,
|
21 |
-
"train_accuracy": 0.
|
22 |
-
"train_auc_class_0": 0.
|
23 |
-
"train_auc_class_1": 0.
|
24 |
-
"train_auc_class_2": 0.
|
25 |
-
"train_auc_overall": 0.
|
26 |
-
"train_loss": 0.
|
27 |
-
"train_runtime":
|
28 |
-
"train_samples_per_second":
|
29 |
-
"train_steps_per_second": 0.
|
30 |
},
|
31 |
{
|
32 |
-
"epoch": 0
|
33 |
-
"eval_accuracy": 0.
|
34 |
-
"eval_auc_class_0": 0.
|
35 |
-
"eval_auc_class_1": 0.
|
36 |
-
"eval_auc_class_2": 0.
|
37 |
-
"eval_auc_overall": 0.
|
38 |
-
"eval_loss": 0.
|
39 |
-
"eval_runtime":
|
40 |
-
"eval_samples_per_second":
|
41 |
-
"eval_steps_per_second": 0.
|
42 |
"step": 15
|
43 |
},
|
44 |
{
|
45 |
-
"epoch": 1.
|
46 |
-
"grad_norm":
|
47 |
"learning_rate": 4.5212765957446815e-05,
|
48 |
-
"loss": 0.
|
49 |
"step": 20
|
50 |
},
|
51 |
{
|
52 |
-
"epoch":
|
53 |
-
"grad_norm":
|
54 |
"learning_rate": 3.9893617021276594e-05,
|
55 |
-
"loss": 0.
|
56 |
"step": 30
|
57 |
},
|
58 |
{
|
59 |
-
"epoch":
|
60 |
"step": 30,
|
61 |
-
"train_accuracy": 0.
|
62 |
-
"train_auc_class_0": 0.
|
63 |
-
"train_auc_class_1": 0.
|
64 |
-
"train_auc_class_2": 0.
|
65 |
-
"train_auc_overall": 0.
|
66 |
-
"train_loss": 0.
|
67 |
-
"train_runtime":
|
68 |
-
"train_samples_per_second":
|
69 |
-
"train_steps_per_second": 0.
|
70 |
},
|
71 |
{
|
72 |
-
"epoch":
|
73 |
-
"eval_accuracy": 0.
|
74 |
-
"eval_auc_class_0": 0.
|
75 |
-
"eval_auc_class_1": 0.
|
76 |
-
"eval_auc_class_2": 0.
|
77 |
-
"eval_auc_overall": 0.
|
78 |
-
"eval_loss": 0.
|
79 |
-
"eval_runtime":
|
80 |
-
"eval_samples_per_second": 9.
|
81 |
-
"eval_steps_per_second": 0.
|
82 |
"step": 30
|
83 |
},
|
84 |
{
|
85 |
-
"epoch": 2.
|
86 |
-
"grad_norm":
|
87 |
"learning_rate": 3.4574468085106386e-05,
|
88 |
-
"loss": 0.
|
89 |
"step": 40
|
90 |
},
|
91 |
{
|
92 |
-
"epoch":
|
93 |
"step": 45,
|
94 |
-
"train_accuracy": 0.
|
95 |
-
"train_auc_class_0": 0.
|
96 |
-
"train_auc_class_1": 0.
|
97 |
-
"train_auc_class_2": 0.
|
98 |
-
"train_auc_overall": 0.
|
99 |
-
"train_loss": 0.
|
100 |
-
"train_runtime":
|
101 |
-
"train_samples_per_second":
|
102 |
-
"train_steps_per_second": 0.
|
103 |
},
|
104 |
{
|
105 |
-
"epoch":
|
106 |
-
"eval_accuracy": 0.
|
107 |
-
"eval_auc_class_0": 0.
|
108 |
-
"eval_auc_class_1": 0.
|
109 |
-
"eval_auc_class_2": 0.
|
110 |
-
"eval_auc_overall": 0.
|
111 |
-
"eval_loss": 0.
|
112 |
-
"eval_runtime": 207.
|
113 |
-
"eval_samples_per_second": 9.
|
114 |
"eval_steps_per_second": 0.077,
|
115 |
"step": 45
|
116 |
},
|
117 |
{
|
118 |
-
"epoch": 3.
|
119 |
-
"grad_norm":
|
120 |
"learning_rate": 2.925531914893617e-05,
|
121 |
-
"loss": 0.
|
122 |
"step": 50
|
123 |
},
|
124 |
{
|
125 |
-
"epoch":
|
126 |
-
"grad_norm":
|
127 |
"learning_rate": 2.393617021276596e-05,
|
128 |
-
"loss": 0.
|
129 |
"step": 60
|
130 |
},
|
131 |
{
|
132 |
"epoch": 4.0,
|
133 |
-
"step":
|
134 |
-
"train_accuracy": 0.
|
135 |
-
"train_auc_class_0": 0.
|
136 |
-
"train_auc_class_1": 0.
|
137 |
-
"train_auc_class_2": 0.
|
138 |
-
"train_auc_overall": 0.
|
139 |
-
"train_loss": 0.
|
140 |
-
"train_runtime":
|
141 |
-
"train_samples_per_second": 11.
|
142 |
-
"train_steps_per_second": 0.
|
143 |
},
|
144 |
{
|
145 |
"epoch": 4.0,
|
146 |
-
"eval_accuracy": 0.
|
147 |
-
"eval_auc_class_0": 0.
|
148 |
-
"eval_auc_class_1": 0.
|
149 |
-
"eval_auc_class_2": 0.
|
150 |
-
"eval_auc_overall": 0.
|
151 |
-
"eval_loss": 0.
|
152 |
-
"eval_runtime":
|
153 |
-
"eval_samples_per_second": 9.
|
154 |
-
"eval_steps_per_second": 0.
|
155 |
-
"step":
|
156 |
},
|
157 |
{
|
158 |
-
"epoch": 4.
|
159 |
-
"grad_norm": 2.
|
160 |
"learning_rate": 1.8617021276595745e-05,
|
161 |
-
"loss": 0.
|
162 |
"step": 70
|
163 |
},
|
164 |
{
|
165 |
-
"epoch":
|
166 |
-
"step":
|
167 |
-
"train_accuracy": 0.
|
168 |
-
"train_auc_class_0": 0.
|
169 |
-
"train_auc_class_1": 0.
|
170 |
-
"train_auc_class_2": 0.
|
171 |
-
"train_auc_overall": 0.
|
172 |
-
"train_loss": 0.
|
173 |
-
"train_runtime":
|
174 |
-
"train_samples_per_second": 10.
|
175 |
-
"train_steps_per_second": 0.
|
176 |
},
|
177 |
{
|
178 |
-
"epoch":
|
179 |
-
"eval_accuracy": 0.
|
180 |
-
"eval_auc_class_0": 0.
|
181 |
-
"eval_auc_class_1": 0.
|
182 |
-
"eval_auc_class_2": 0.
|
183 |
-
"eval_auc_overall": 0.
|
184 |
-
"eval_loss": 0.
|
185 |
-
"eval_runtime":
|
186 |
-
"eval_samples_per_second": 9.
|
187 |
-
"eval_steps_per_second": 0.
|
188 |
-
"step":
|
189 |
},
|
190 |
{
|
191 |
-
"epoch": 5.
|
192 |
-
"grad_norm": 2.
|
193 |
"learning_rate": 1.3297872340425532e-05,
|
194 |
-
"loss": 0.
|
195 |
"step": 80
|
196 |
},
|
197 |
{
|
198 |
-
"epoch":
|
199 |
-
"grad_norm":
|
200 |
"learning_rate": 7.97872340425532e-06,
|
201 |
-
"loss": 0.
|
202 |
"step": 90
|
203 |
},
|
204 |
{
|
205 |
-
"epoch":
|
206 |
-
"step":
|
207 |
-
"train_accuracy": 0.
|
208 |
-
"train_auc_class_0": 0.
|
209 |
-
"train_auc_class_1": 0.
|
210 |
-
"train_auc_class_2": 0.
|
211 |
-
"train_auc_overall": 0.
|
212 |
-
"train_loss": 0.
|
213 |
-
"train_runtime":
|
214 |
-
"train_samples_per_second": 10.
|
215 |
-
"train_steps_per_second": 0.
|
216 |
},
|
217 |
{
|
218 |
-
"epoch":
|
219 |
-
"eval_accuracy": 0.
|
220 |
-
"eval_auc_class_0": 0.
|
221 |
-
"eval_auc_class_1": 0.
|
222 |
-
"eval_auc_class_2": 0.
|
223 |
-
"eval_auc_overall": 0.
|
224 |
-
"eval_loss": 0.
|
225 |
-
"eval_runtime": 209.
|
226 |
-
"eval_samples_per_second": 9.
|
227 |
"eval_steps_per_second": 0.076,
|
228 |
-
"step":
|
229 |
},
|
230 |
{
|
231 |
-
"epoch": 6.
|
232 |
-
"grad_norm":
|
233 |
"learning_rate": 2.6595744680851065e-06,
|
234 |
-
"loss": 0.
|
235 |
"step": 100
|
236 |
},
|
237 |
{
|
238 |
-
"epoch":
|
239 |
"step": 105,
|
240 |
-
"train_accuracy": 0.
|
241 |
-
"train_auc_class_0": 0.
|
242 |
-
"train_auc_class_1": 0.
|
243 |
-
"train_auc_class_2": 0.
|
244 |
-
"train_auc_overall": 0.
|
245 |
-
"train_loss": 0.
|
246 |
-
"train_runtime":
|
247 |
-
"train_samples_per_second": 11.
|
248 |
-
"train_steps_per_second": 0.
|
249 |
},
|
250 |
{
|
251 |
-
"epoch":
|
252 |
-
"eval_accuracy": 0.
|
253 |
-
"eval_auc_class_0": 0.
|
254 |
-
"eval_auc_class_1": 0.
|
255 |
-
"eval_auc_class_2": 0.
|
256 |
-
"eval_auc_overall": 0.
|
257 |
-
"eval_loss": 0.
|
258 |
-
"eval_runtime":
|
259 |
-
"eval_samples_per_second": 9.
|
260 |
-
"eval_steps_per_second": 0.
|
261 |
"step": 105
|
262 |
},
|
263 |
{
|
264 |
-
"epoch":
|
265 |
"step": 105,
|
266 |
-
"total_flos": 4.
|
267 |
-
"train_loss": 0.
|
268 |
-
"train_runtime":
|
269 |
-
"train_samples_per_second": 4.
|
270 |
"train_steps_per_second": 0.009
|
271 |
}
|
272 |
],
|
@@ -287,7 +287,7 @@
|
|
287 |
"attributes": {}
|
288 |
}
|
289 |
},
|
290 |
-
"total_flos": 4.
|
291 |
"train_batch_size": 128,
|
292 |
"trial_name": null,
|
293 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.8600206611570248,
|
3 |
+
"best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM/NMM_Classification_base_V10/checkpoint-90",
|
4 |
+
"epoch": 7.0,
|
5 |
"eval_steps": 500,
|
6 |
"global_step": 105,
|
7 |
"is_hyper_param_search": false,
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.6666666666666666,
|
13 |
+
"grad_norm": 4.08638334274292,
|
14 |
"learning_rate": 4.545454545454546e-05,
|
15 |
+
"loss": 0.8429,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 1.0,
|
20 |
"step": 15,
|
21 |
+
"train_accuracy": 0.828215987294865,
|
22 |
+
"train_auc_class_0": 0.9164869255677786,
|
23 |
+
"train_auc_class_1": 0.770497463263411,
|
24 |
+
"train_auc_class_2": 0.8879816559016815,
|
25 |
+
"train_auc_overall": 0.858322014910957,
|
26 |
+
"train_loss": 0.4563606083393097,
|
27 |
+
"train_runtime": 687.587,
|
28 |
+
"train_samples_per_second": 10.989,
|
29 |
+
"train_steps_per_second": 0.087
|
30 |
},
|
31 |
{
|
32 |
+
"epoch": 1.0,
|
33 |
+
"eval_accuracy": 0.8274793388429752,
|
34 |
+
"eval_auc_class_0": 0.9340616414835164,
|
35 |
+
"eval_auc_class_1": 0.7507519271533845,
|
36 |
+
"eval_auc_class_2": 0.8937217290626166,
|
37 |
+
"eval_auc_overall": 0.8595117658998391,
|
38 |
+
"eval_loss": 0.49912166595458984,
|
39 |
+
"eval_runtime": 208.8157,
|
40 |
+
"eval_samples_per_second": 9.271,
|
41 |
+
"eval_steps_per_second": 0.077,
|
42 |
"step": 15
|
43 |
},
|
44 |
{
|
45 |
+
"epoch": 1.3333333333333333,
|
46 |
+
"grad_norm": 1.5998913049697876,
|
47 |
"learning_rate": 4.5212765957446815e-05,
|
48 |
+
"loss": 0.4716,
|
49 |
"step": 20
|
50 |
},
|
51 |
{
|
52 |
+
"epoch": 2.0,
|
53 |
+
"grad_norm": 2.4967057704925537,
|
54 |
"learning_rate": 3.9893617021276594e-05,
|
55 |
+
"loss": 0.3887,
|
56 |
"step": 30
|
57 |
},
|
58 |
{
|
59 |
+
"epoch": 2.0,
|
60 |
"step": 30,
|
61 |
+
"train_accuracy": 0.8553467443091582,
|
62 |
+
"train_auc_class_0": 0.9472708924769339,
|
63 |
+
"train_auc_class_1": 0.9283657851250956,
|
64 |
+
"train_auc_class_2": 0.9247901532600991,
|
65 |
+
"train_auc_overall": 0.9334756102873761,
|
66 |
+
"train_loss": 0.35952258110046387,
|
67 |
+
"train_runtime": 685.7612,
|
68 |
+
"train_samples_per_second": 11.018,
|
69 |
+
"train_steps_per_second": 0.087
|
70 |
},
|
71 |
{
|
72 |
+
"epoch": 2.0,
|
73 |
+
"eval_accuracy": 0.8502066115702479,
|
74 |
+
"eval_auc_class_0": 0.9477621336996336,
|
75 |
+
"eval_auc_class_1": 0.9035008683616039,
|
76 |
+
"eval_auc_class_2": 0.9096346486873113,
|
77 |
+
"eval_auc_overall": 0.9202992169161829,
|
78 |
+
"eval_loss": 0.3957720398902893,
|
79 |
+
"eval_runtime": 208.9311,
|
80 |
+
"eval_samples_per_second": 9.266,
|
81 |
+
"eval_steps_per_second": 0.077,
|
82 |
"step": 30
|
83 |
},
|
84 |
{
|
85 |
+
"epoch": 2.6666666666666665,
|
86 |
+
"grad_norm": 1.716306209564209,
|
87 |
"learning_rate": 3.4574468085106386e-05,
|
88 |
+
"loss": 0.3533,
|
89 |
"step": 40
|
90 |
},
|
91 |
{
|
92 |
+
"epoch": 3.0,
|
93 |
"step": 45,
|
94 |
+
"train_accuracy": 0.8750661725780836,
|
95 |
+
"train_auc_class_0": 0.9582223466702152,
|
96 |
+
"train_auc_class_1": 0.9392220276138971,
|
97 |
+
"train_auc_class_2": 0.9390688589921182,
|
98 |
+
"train_auc_overall": 0.9455044110920768,
|
99 |
+
"train_loss": 0.3131944239139557,
|
100 |
+
"train_runtime": 681.8829,
|
101 |
+
"train_samples_per_second": 11.081,
|
102 |
+
"train_steps_per_second": 0.088
|
103 |
},
|
104 |
{
|
105 |
+
"epoch": 3.0,
|
106 |
+
"eval_accuracy": 0.8589876033057852,
|
107 |
+
"eval_auc_class_0": 0.9578196543040294,
|
108 |
+
"eval_auc_class_1": 0.9092855930322143,
|
109 |
+
"eval_auc_class_2": 0.920674647298944,
|
110 |
+
"eval_auc_overall": 0.9292599648783959,
|
111 |
+
"eval_loss": 0.37213757634162903,
|
112 |
+
"eval_runtime": 207.6482,
|
113 |
+
"eval_samples_per_second": 9.323,
|
114 |
"eval_steps_per_second": 0.077,
|
115 |
"step": 45
|
116 |
},
|
117 |
{
|
118 |
+
"epoch": 3.3333333333333335,
|
119 |
+
"grad_norm": 1.727489948272705,
|
120 |
"learning_rate": 2.925531914893617e-05,
|
121 |
+
"loss": 0.3208,
|
122 |
"step": 50
|
123 |
},
|
124 |
{
|
125 |
+
"epoch": 4.0,
|
126 |
+
"grad_norm": 10.33216381072998,
|
127 |
"learning_rate": 2.393617021276596e-05,
|
128 |
+
"loss": 0.3092,
|
129 |
"step": 60
|
130 |
},
|
131 |
{
|
132 |
"epoch": 4.0,
|
133 |
+
"step": 60,
|
134 |
+
"train_accuracy": 0.8848597141344626,
|
135 |
+
"train_auc_class_0": 0.9659353652856636,
|
136 |
+
"train_auc_class_1": 0.9407955383480827,
|
137 |
+
"train_auc_class_2": 0.946874731546507,
|
138 |
+
"train_auc_overall": 0.9512018783934177,
|
139 |
+
"train_loss": 0.29102450609207153,
|
140 |
+
"train_runtime": 680.3084,
|
141 |
+
"train_samples_per_second": 11.107,
|
142 |
+
"train_steps_per_second": 0.088
|
143 |
},
|
144 |
{
|
145 |
"epoch": 4.0,
|
146 |
+
"eval_accuracy": 0.8579545454545454,
|
147 |
+
"eval_auc_class_0": 0.9606026785714286,
|
148 |
+
"eval_auc_class_1": 0.9083715281858423,
|
149 |
+
"eval_auc_class_2": 0.9231885836547031,
|
150 |
+
"eval_auc_overall": 0.9307209301373248,
|
151 |
+
"eval_loss": 0.37202903628349304,
|
152 |
+
"eval_runtime": 208.7004,
|
153 |
+
"eval_samples_per_second": 9.276,
|
154 |
+
"eval_steps_per_second": 0.077,
|
155 |
+
"step": 60
|
156 |
},
|
157 |
{
|
158 |
+
"epoch": 4.666666666666667,
|
159 |
+
"grad_norm": 2.016211748123169,
|
160 |
"learning_rate": 1.8617021276595745e-05,
|
161 |
+
"loss": 0.2891,
|
162 |
"step": 70
|
163 |
},
|
164 |
{
|
165 |
+
"epoch": 5.0,
|
166 |
+
"step": 75,
|
167 |
+
"train_accuracy": 0.8913446267866596,
|
168 |
+
"train_auc_class_0": 0.9681433160338301,
|
169 |
+
"train_auc_class_1": 0.9542337553261225,
|
170 |
+
"train_auc_class_2": 0.9506045992123575,
|
171 |
+
"train_auc_overall": 0.9576605568574367,
|
172 |
+
"train_loss": 0.2791108191013336,
|
173 |
+
"train_runtime": 690.7325,
|
174 |
+
"train_samples_per_second": 10.939,
|
175 |
+
"train_steps_per_second": 0.087
|
176 |
},
|
177 |
{
|
178 |
+
"epoch": 5.0,
|
179 |
+
"eval_accuracy": 0.8584710743801653,
|
180 |
+
"eval_auc_class_0": 0.9613524496336998,
|
181 |
+
"eval_auc_class_1": 0.9090984273731954,
|
182 |
+
"eval_auc_class_2": 0.9242633286351978,
|
183 |
+
"eval_auc_overall": 0.9315714018806975,
|
184 |
+
"eval_loss": 0.36684560775756836,
|
185 |
+
"eval_runtime": 211.7912,
|
186 |
+
"eval_samples_per_second": 9.141,
|
187 |
+
"eval_steps_per_second": 0.076,
|
188 |
+
"step": 75
|
189 |
},
|
190 |
{
|
191 |
+
"epoch": 5.333333333333333,
|
192 |
+
"grad_norm": 2.7578961849212646,
|
193 |
"learning_rate": 1.3297872340425532e-05,
|
194 |
+
"loss": 0.301,
|
195 |
"step": 80
|
196 |
},
|
197 |
{
|
198 |
+
"epoch": 6.0,
|
199 |
+
"grad_norm": 2.526148557662964,
|
200 |
"learning_rate": 7.97872340425532e-06,
|
201 |
+
"loss": 0.2852,
|
202 |
"step": 90
|
203 |
},
|
204 |
{
|
205 |
+
"epoch": 6.0,
|
206 |
+
"step": 90,
|
207 |
+
"train_accuracy": 0.8955796717840127,
|
208 |
+
"train_auc_class_0": 0.9719376552519517,
|
209 |
+
"train_auc_class_1": 0.9603400183000109,
|
210 |
+
"train_auc_class_2": 0.9560110428344394,
|
211 |
+
"train_auc_overall": 0.9627629054621339,
|
212 |
+
"train_loss": 0.2651444673538208,
|
213 |
+
"train_runtime": 689.8251,
|
214 |
+
"train_samples_per_second": 10.954,
|
215 |
+
"train_steps_per_second": 0.087
|
216 |
},
|
217 |
{
|
218 |
+
"epoch": 6.0,
|
219 |
+
"eval_accuracy": 0.8600206611570248,
|
220 |
+
"eval_auc_class_0": 0.9623798076923077,
|
221 |
+
"eval_auc_class_1": 0.9106044580248364,
|
222 |
+
"eval_auc_class_2": 0.9252550195053197,
|
223 |
+
"eval_auc_overall": 0.932746428407488,
|
224 |
+
"eval_loss": 0.358356237411499,
|
225 |
+
"eval_runtime": 209.7236,
|
226 |
+
"eval_samples_per_second": 9.231,
|
227 |
"eval_steps_per_second": 0.076,
|
228 |
+
"step": 90
|
229 |
},
|
230 |
{
|
231 |
+
"epoch": 6.666666666666667,
|
232 |
+
"grad_norm": 1.6473604440689087,
|
233 |
"learning_rate": 2.6595744680851065e-06,
|
234 |
+
"loss": 0.2804,
|
235 |
"step": 100
|
236 |
},
|
237 |
{
|
238 |
+
"epoch": 7.0,
|
239 |
"step": 105,
|
240 |
+
"train_accuracy": 0.8926680783483325,
|
241 |
+
"train_auc_class_0": 0.9692773945765316,
|
242 |
+
"train_auc_class_1": 0.9595504889107397,
|
243 |
+
"train_auc_class_2": 0.9538150093690271,
|
244 |
+
"train_auc_overall": 0.9608809642854328,
|
245 |
+
"train_loss": 0.2725497782230377,
|
246 |
+
"train_runtime": 674.8314,
|
247 |
+
"train_samples_per_second": 11.197,
|
248 |
+
"train_steps_per_second": 0.089
|
249 |
},
|
250 |
{
|
251 |
+
"epoch": 7.0,
|
252 |
+
"eval_accuracy": 0.8600206611570248,
|
253 |
+
"eval_auc_class_0": 0.9628448374542126,
|
254 |
+
"eval_auc_class_1": 0.90828447439095,
|
255 |
+
"eval_auc_class_2": 0.924919084223066,
|
256 |
+
"eval_auc_overall": 0.9320161320227428,
|
257 |
+
"eval_loss": 0.3619421422481537,
|
258 |
+
"eval_runtime": 204.9165,
|
259 |
+
"eval_samples_per_second": 9.448,
|
260 |
+
"eval_steps_per_second": 0.078,
|
261 |
"step": 105
|
262 |
},
|
263 |
{
|
264 |
+
"epoch": 7.0,
|
265 |
"step": 105,
|
266 |
+
"total_flos": 4.143902092898771e+18,
|
267 |
+
"train_loss": 0.37870263712746755,
|
268 |
+
"train_runtime": 11542.069,
|
269 |
+
"train_samples_per_second": 4.583,
|
270 |
"train_steps_per_second": 0.009
|
271 |
}
|
272 |
],
|
|
|
287 |
"attributes": {}
|
288 |
}
|
289 |
},
|
290 |
+
"total_flos": 4.143902092898771e+18,
|
291 |
"train_batch_size": 128,
|
292 |
"trial_name": null,
|
293 |
"trial_params": null
|