End of training
Browse files- README.md +7 -7
- all_results.json +14 -14
- eval_results.json +10 -10
- runs/Sep24_06-17-42_data-science-wbi/events.out.tfevents.1727171357.data-science-wbi +3 -0
- train_results.json +5 -5
- trainer_state.json +211 -204
README.md
CHANGED
@@ -23,7 +23,7 @@ model-index:
|
|
23 |
metrics:
|
24 |
- name: Accuracy
|
25 |
type: accuracy
|
26 |
-
value: 0.
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -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.
|
38 |
-
- Auc Overall: 0.
|
39 |
-
- Auc Class 0: 0.
|
40 |
-
- Auc Class 1: 0.
|
41 |
-
- Auc Class 2: 0.
|
42 |
|
43 |
## Model description
|
44 |
|
|
|
23 |
metrics:
|
24 |
- name: Accuracy
|
25 |
type: accuracy
|
26 |
+
value: 0.8349146110056926
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
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: 0.4066
|
37 |
+
- Accuracy: 0.8349
|
38 |
+
- Auc Overall: 0.9379
|
39 |
+
- Auc Class 0: 0.9614
|
40 |
+
- Auc Class 1: 0.9315
|
41 |
+
- Auc Class 2: 0.9207
|
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": 6.892307692307693,
|
3 |
+
"eval_accuracy": 0.8349146110056926,
|
4 |
+
"eval_auc_class_0": 0.9613559582309583,
|
5 |
+
"eval_auc_class_1": 0.9315019107361742,
|
6 |
+
"eval_auc_class_2": 0.920726636279075,
|
7 |
+
"eval_auc_overall": 0.9378615017487358,
|
8 |
+
"eval_loss": 0.40655088424682617,
|
9 |
+
"eval_runtime": 226.5129,
|
10 |
+
"eval_samples_per_second": 9.306,
|
11 |
+
"eval_steps_per_second": 0.075,
|
12 |
+
"total_flos": 4.4565045535859466e+18,
|
13 |
+
"train_loss": 0.45718009237732204,
|
14 |
+
"train_runtime": 12452.0227,
|
15 |
+
"train_samples_per_second": 4.634,
|
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": 6.892307692307693,
|
3 |
+
"eval_accuracy": 0.8349146110056926,
|
4 |
+
"eval_auc_class_0": 0.9613559582309583,
|
5 |
+
"eval_auc_class_1": 0.9315019107361742,
|
6 |
+
"eval_auc_class_2": 0.920726636279075,
|
7 |
+
"eval_auc_overall": 0.9378615017487358,
|
8 |
+
"eval_loss": 0.40655088424682617,
|
9 |
+
"eval_runtime": 226.5129,
|
10 |
+
"eval_samples_per_second": 9.306,
|
11 |
+
"eval_steps_per_second": 0.075
|
12 |
}
|
runs/Sep24_06-17-42_data-science-wbi/events.out.tfevents.1727171357.data-science-wbi
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:047bd9e0830e3ad29e864c1d2e8102a3d401da16c1375f6a49fd3870c829092d
|
3 |
+
size 138
|
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": 6.892307692307693,
|
3 |
+
"total_flos": 4.4565045535859466e+18,
|
4 |
+
"train_loss": 0.45718009237732204,
|
5 |
+
"train_runtime": 12452.0227,
|
6 |
+
"train_samples_per_second": 4.634,
|
7 |
"train_steps_per_second": 0.009
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,277 +1,284 @@
|
|
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":
|
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.
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate": 4.
|
15 |
-
"loss": 0.
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
-
"epoch":
|
20 |
-
"step":
|
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":
|
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": 9.
|
41 |
-
"eval_steps_per_second": 0.
|
42 |
-
"step":
|
43 |
},
|
44 |
{
|
45 |
-
"epoch": 1.
|
46 |
-
"grad_norm":
|
47 |
-
"learning_rate": 4.
|
48 |
-
"loss": 0.
|
49 |
"step": 20
|
50 |
},
|
51 |
{
|
52 |
-
"epoch":
|
53 |
-
"grad_norm": 2.
|
54 |
-
"learning_rate":
|
55 |
-
"loss": 0.
|
56 |
"step": 30
|
57 |
},
|
58 |
{
|
59 |
-
"epoch":
|
60 |
-
"step":
|
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": 11.
|
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":
|
83 |
},
|
84 |
{
|
85 |
-
"epoch": 2.
|
86 |
-
"grad_norm":
|
87 |
-
"learning_rate": 3.
|
88 |
-
"loss": 0.
|
89 |
"step": 40
|
90 |
},
|
91 |
{
|
92 |
-
"epoch":
|
93 |
-
"step":
|
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":
|
113 |
-
"eval_samples_per_second": 9.
|
114 |
-
"eval_steps_per_second": 0.
|
115 |
-
"step":
|
116 |
},
|
117 |
{
|
118 |
-
"epoch": 3.
|
119 |
-
"grad_norm": 1.
|
120 |
-
"learning_rate":
|
121 |
-
"loss": 0.
|
122 |
"step": 50
|
123 |
},
|
124 |
{
|
125 |
-
"epoch":
|
126 |
-
"grad_norm":
|
127 |
-
"learning_rate": 2.
|
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":
|
160 |
-
"learning_rate":
|
161 |
-
"loss": 0.
|
162 |
"step": 70
|
163 |
},
|
164 |
{
|
165 |
-
"epoch":
|
166 |
-
"
|
167 |
-
"
|
168 |
-
"
|
169 |
-
"
|
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":
|
179 |
-
"
|
180 |
-
"
|
181 |
-
"
|
182 |
-
"
|
183 |
-
"
|
184 |
-
"
|
185 |
-
"
|
186 |
-
"
|
187 |
-
"
|
188 |
-
"
|
189 |
},
|
190 |
{
|
191 |
-
"epoch":
|
192 |
-
"
|
193 |
-
"
|
194 |
-
"
|
195 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
196 |
},
|
197 |
{
|
198 |
-
"epoch":
|
199 |
-
"grad_norm": 2.
|
200 |
-
"learning_rate":
|
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":
|
215 |
"train_steps_per_second": 0.087
|
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":
|
226 |
-
"eval_samples_per_second": 9.
|
227 |
-
"eval_steps_per_second": 0.
|
228 |
-
"step":
|
229 |
},
|
230 |
{
|
231 |
-
"epoch": 6.
|
232 |
-
"grad_norm":
|
233 |
-
"learning_rate":
|
234 |
-
"loss": 0.
|
235 |
"step": 100
|
236 |
},
|
237 |
{
|
238 |
-
"epoch":
|
239 |
-
"
|
240 |
-
"
|
241 |
-
"
|
242 |
-
"
|
243 |
-
|
244 |
-
|
245 |
-
"
|
246 |
-
"
|
247 |
-
"
|
248 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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":
|
262 |
},
|
263 |
{
|
264 |
-
"epoch":
|
265 |
-
"step":
|
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 |
],
|
273 |
"logging_steps": 10,
|
274 |
-
"max_steps":
|
275 |
"num_input_tokens_seen": 0,
|
276 |
"num_train_epochs": 7,
|
277 |
"save_steps": 500,
|
@@ -287,7 +294,7 @@
|
|
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": 0.8349146110056926,
|
3 |
+
"best_model_checkpoint": "swin-base-patch4-window7-224-in22k-MM/NMM_Classification_base_V10/checkpoint-81",
|
4 |
+
"epoch": 6.892307692307693,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 112,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.6153846153846154,
|
13 |
+
"grad_norm": 3.4730300903320312,
|
14 |
+
"learning_rate": 4.166666666666667e-05,
|
15 |
+
"loss": 0.8416,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.9846153846153847,
|
20 |
+
"step": 16,
|
21 |
+
"train_accuracy": 0.7838165716365401,
|
22 |
+
"train_auc_class_0": 0.9185630172913702,
|
23 |
+
"train_auc_class_1": 0.8867337604929844,
|
24 |
+
"train_auc_class_2": 0.8750511256900684,
|
25 |
+
"train_auc_overall": 0.8934493011581409,
|
26 |
+
"train_loss": 0.5212814211845398,
|
27 |
+
"train_runtime": 738.2478,
|
28 |
+
"train_samples_per_second": 11.166,
|
29 |
+
"train_steps_per_second": 0.088
|
30 |
},
|
31 |
{
|
32 |
+
"epoch": 0.9846153846153847,
|
33 |
+
"eval_accuracy": 0.7751423149905123,
|
34 |
+
"eval_auc_class_0": 0.9351696867321869,
|
35 |
+
"eval_auc_class_1": 0.9013257754334977,
|
36 |
+
"eval_auc_class_2": 0.896788439192459,
|
37 |
+
"eval_auc_overall": 0.9110946337860479,
|
38 |
+
"eval_loss": 0.5271921157836914,
|
39 |
+
"eval_runtime": 226.0415,
|
40 |
+
"eval_samples_per_second": 9.326,
|
41 |
+
"eval_steps_per_second": 0.075,
|
42 |
+
"step": 16
|
43 |
},
|
44 |
{
|
45 |
+
"epoch": 1.2307692307692308,
|
46 |
+
"grad_norm": 2.644275426864624,
|
47 |
+
"learning_rate": 4.600000000000001e-05,
|
48 |
+
"loss": 0.5698,
|
49 |
"step": 20
|
50 |
},
|
51 |
{
|
52 |
+
"epoch": 1.8461538461538463,
|
53 |
+
"grad_norm": 2.8130509853363037,
|
54 |
+
"learning_rate": 4.1e-05,
|
55 |
+
"loss": 0.4812,
|
56 |
"step": 30
|
57 |
},
|
58 |
{
|
59 |
+
"epoch": 1.9692307692307693,
|
60 |
+
"step": 32,
|
61 |
+
"train_accuracy": 0.8125682397185491,
|
62 |
+
"train_auc_class_0": 0.9453865517423163,
|
63 |
+
"train_auc_class_1": 0.9307891812846302,
|
64 |
+
"train_auc_class_2": 0.9068255493827921,
|
65 |
+
"train_auc_overall": 0.9276670941365796,
|
66 |
+
"train_loss": 0.45027220249176025,
|
67 |
+
"train_runtime": 742.3599,
|
68 |
+
"train_samples_per_second": 11.104,
|
69 |
+
"train_steps_per_second": 0.088
|
70 |
},
|
71 |
{
|
72 |
+
"epoch": 1.9692307692307693,
|
73 |
+
"eval_accuracy": 0.8088235294117647,
|
74 |
+
"eval_auc_class_0": 0.9525644963144962,
|
75 |
+
"eval_auc_class_1": 0.92725521408195,
|
76 |
+
"eval_auc_class_2": 0.9013002141426902,
|
77 |
+
"eval_auc_overall": 0.9270399748463788,
|
78 |
+
"eval_loss": 0.4628311097621918,
|
79 |
+
"eval_runtime": 227.1982,
|
80 |
+
"eval_samples_per_second": 9.278,
|
81 |
+
"eval_steps_per_second": 0.075,
|
82 |
+
"step": 32
|
83 |
},
|
84 |
{
|
85 |
+
"epoch": 2.4615384615384617,
|
86 |
+
"grad_norm": 4.537336349487305,
|
87 |
+
"learning_rate": 3.6e-05,
|
88 |
+
"loss": 0.4444,
|
89 |
"step": 40
|
90 |
},
|
91 |
{
|
92 |
+
"epoch": 2.953846153846154,
|
93 |
+
"step": 48,
|
94 |
+
"train_accuracy": 0.8330704840470703,
|
95 |
+
"train_auc_class_0": 0.9573463611682219,
|
96 |
+
"train_auc_class_1": 0.9445183457629789,
|
97 |
+
"train_auc_class_2": 0.9234336270362926,
|
98 |
+
"train_auc_overall": 0.9417661113224979,
|
99 |
+
"train_loss": 0.3945094048976898,
|
100 |
+
"train_runtime": 749.6504,
|
101 |
+
"train_samples_per_second": 10.996,
|
102 |
+
"train_steps_per_second": 0.087
|
103 |
},
|
104 |
{
|
105 |
+
"epoch": 2.953846153846154,
|
106 |
+
"eval_accuracy": 0.8273244781783681,
|
107 |
+
"eval_auc_class_0": 0.9561783374283375,
|
108 |
+
"eval_auc_class_1": 0.9278154008848364,
|
109 |
+
"eval_auc_class_2": 0.9137749676178002,
|
110 |
+
"eval_auc_overall": 0.932589568643658,
|
111 |
+
"eval_loss": 0.42404672503471375,
|
112 |
+
"eval_runtime": 226.6025,
|
113 |
+
"eval_samples_per_second": 9.303,
|
114 |
+
"eval_steps_per_second": 0.075,
|
115 |
+
"step": 48
|
116 |
},
|
117 |
{
|
118 |
+
"epoch": 3.076923076923077,
|
119 |
+
"grad_norm": 1.965461254119873,
|
120 |
+
"learning_rate": 3.1e-05,
|
121 |
+
"loss": 0.4265,
|
122 |
"step": 50
|
123 |
},
|
124 |
{
|
125 |
+
"epoch": 3.6923076923076925,
|
126 |
+
"grad_norm": 3.6663880348205566,
|
127 |
+
"learning_rate": 2.6000000000000002e-05,
|
128 |
+
"loss": 0.4055,
|
129 |
"step": 60
|
130 |
},
|
131 |
{
|
132 |
"epoch": 4.0,
|
133 |
+
"step": 65,
|
134 |
+
"train_accuracy": 0.8460511949532937,
|
135 |
+
"train_auc_class_0": 0.9634787762270021,
|
136 |
+
"train_auc_class_1": 0.9469802033822643,
|
137 |
+
"train_auc_class_2": 0.9308556192528558,
|
138 |
+
"train_auc_overall": 0.9471048662873741,
|
139 |
+
"train_loss": 0.37029632925987244,
|
140 |
+
"train_runtime": 746.5442,
|
141 |
+
"train_samples_per_second": 11.042,
|
142 |
+
"train_steps_per_second": 0.087
|
143 |
},
|
144 |
{
|
145 |
"epoch": 4.0,
|
146 |
+
"eval_accuracy": 0.8287476280834914,
|
147 |
+
"eval_auc_class_0": 0.9613994676494677,
|
148 |
+
"eval_auc_class_1": 0.9314889691268666,
|
149 |
+
"eval_auc_class_2": 0.9206087950151214,
|
150 |
+
"eval_auc_overall": 0.9378324105971519,
|
151 |
+
"eval_loss": 0.41696038842201233,
|
152 |
+
"eval_runtime": 226.2202,
|
153 |
+
"eval_samples_per_second": 9.318,
|
154 |
+
"eval_steps_per_second": 0.075,
|
155 |
+
"step": 65
|
156 |
},
|
157 |
{
|
158 |
+
"epoch": 4.3076923076923075,
|
159 |
+
"grad_norm": 3.821587324142456,
|
160 |
+
"learning_rate": 2.1e-05,
|
161 |
+
"loss": 0.3875,
|
162 |
"step": 70
|
163 |
},
|
164 |
{
|
165 |
+
"epoch": 4.923076923076923,
|
166 |
+
"grad_norm": 3.1884751319885254,
|
167 |
+
"learning_rate": 1.6000000000000003e-05,
|
168 |
+
"loss": 0.3914,
|
169 |
+
"step": 80
|
|
|
|
|
|
|
|
|
|
|
|
|
170 |
},
|
171 |
{
|
172 |
+
"epoch": 4.984615384615385,
|
173 |
+
"step": 81,
|
174 |
+
"train_accuracy": 0.8536940434307898,
|
175 |
+
"train_auc_class_0": 0.9671529040803416,
|
176 |
+
"train_auc_class_1": 0.9496355903643986,
|
177 |
+
"train_auc_class_2": 0.9367910485339515,
|
178 |
+
"train_auc_overall": 0.9511931809928972,
|
179 |
+
"train_loss": 0.3558380603790283,
|
180 |
+
"train_runtime": 752.3511,
|
181 |
+
"train_samples_per_second": 10.956,
|
182 |
+
"train_steps_per_second": 0.086
|
183 |
},
|
184 |
{
|
185 |
+
"epoch": 4.984615384615385,
|
186 |
+
"eval_accuracy": 0.8349146110056926,
|
187 |
+
"eval_auc_class_0": 0.9613559582309583,
|
188 |
+
"eval_auc_class_1": 0.9315019107361742,
|
189 |
+
"eval_auc_class_2": 0.920726636279075,
|
190 |
+
"eval_auc_overall": 0.9378615017487358,
|
191 |
+
"eval_loss": 0.40655088424682617,
|
192 |
+
"eval_runtime": 227.6818,
|
193 |
+
"eval_samples_per_second": 9.259,
|
194 |
+
"eval_steps_per_second": 0.075,
|
195 |
+
"step": 81
|
196 |
},
|
197 |
{
|
198 |
+
"epoch": 5.538461538461538,
|
199 |
+
"grad_norm": 2.0869638919830322,
|
200 |
+
"learning_rate": 1.1000000000000001e-05,
|
201 |
+
"loss": 0.3731,
|
202 |
"step": 90
|
203 |
},
|
204 |
{
|
205 |
+
"epoch": 5.969230769230769,
|
206 |
+
"step": 97,
|
207 |
+
"train_accuracy": 0.8507824821060294,
|
208 |
+
"train_auc_class_0": 0.9680180491855133,
|
209 |
+
"train_auc_class_1": 0.9494889730661834,
|
210 |
+
"train_auc_class_2": 0.9379508502279332,
|
211 |
+
"train_auc_overall": 0.9518192908265433,
|
212 |
+
"train_loss": 0.35146117210388184,
|
213 |
+
"train_runtime": 745.296,
|
214 |
+
"train_samples_per_second": 11.06,
|
215 |
"train_steps_per_second": 0.087
|
216 |
},
|
217 |
{
|
218 |
+
"epoch": 5.969230769230769,
|
219 |
+
"eval_accuracy": 0.8334914611005693,
|
220 |
+
"eval_auc_class_0": 0.9624193796068795,
|
221 |
+
"eval_auc_class_1": 0.9319585646645997,
|
222 |
+
"eval_auc_class_2": 0.9218326055186417,
|
223 |
+
"eval_auc_overall": 0.9387368499300403,
|
224 |
+
"eval_loss": 0.40432390570640564,
|
225 |
+
"eval_runtime": 226.3347,
|
226 |
+
"eval_samples_per_second": 9.314,
|
227 |
+
"eval_steps_per_second": 0.075,
|
228 |
+
"step": 97
|
229 |
},
|
230 |
{
|
231 |
+
"epoch": 6.153846153846154,
|
232 |
+
"grad_norm": 2.3311638832092285,
|
233 |
+
"learning_rate": 6e-06,
|
234 |
+
"loss": 0.357,
|
235 |
"step": 100
|
236 |
},
|
237 |
{
|
238 |
+
"epoch": 6.769230769230769,
|
239 |
+
"grad_norm": 2.3013312816619873,
|
240 |
+
"learning_rate": 1.0000000000000002e-06,
|
241 |
+
"loss": 0.3687,
|
242 |
+
"step": 110
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 6.892307692307693,
|
246 |
+
"step": 112,
|
247 |
+
"train_accuracy": 0.8533300982651947,
|
248 |
+
"train_auc_class_0": 0.9690346749960462,
|
249 |
+
"train_auc_class_1": 0.9513212789842702,
|
250 |
+
"train_auc_class_2": 0.9389334461695297,
|
251 |
+
"train_auc_overall": 0.9530964667166154,
|
252 |
+
"train_loss": 0.34802600741386414,
|
253 |
+
"train_runtime": 740.1343,
|
254 |
+
"train_samples_per_second": 11.137,
|
255 |
+
"train_steps_per_second": 0.088
|
256 |
},
|
257 |
{
|
258 |
+
"epoch": 6.892307692307693,
|
259 |
+
"eval_accuracy": 0.8344402277039848,
|
260 |
+
"eval_auc_class_0": 0.9626446048321048,
|
261 |
+
"eval_auc_class_1": 0.9327683396469899,
|
262 |
+
"eval_auc_class_2": 0.9223551639104368,
|
263 |
+
"eval_auc_overall": 0.9392560361298439,
|
264 |
+
"eval_loss": 0.40155309438705444,
|
265 |
+
"eval_runtime": 226.5604,
|
266 |
+
"eval_samples_per_second": 9.304,
|
267 |
+
"eval_steps_per_second": 0.075,
|
268 |
+
"step": 112
|
269 |
},
|
270 |
{
|
271 |
+
"epoch": 6.892307692307693,
|
272 |
+
"step": 112,
|
273 |
+
"total_flos": 4.4565045535859466e+18,
|
274 |
+
"train_loss": 0.45718009237732204,
|
275 |
+
"train_runtime": 12452.0227,
|
276 |
+
"train_samples_per_second": 4.634,
|
277 |
"train_steps_per_second": 0.009
|
278 |
}
|
279 |
],
|
280 |
"logging_steps": 10,
|
281 |
+
"max_steps": 112,
|
282 |
"num_input_tokens_seen": 0,
|
283 |
"num_train_epochs": 7,
|
284 |
"save_steps": 500,
|
|
|
294 |
"attributes": {}
|
295 |
}
|
296 |
},
|
297 |
+
"total_flos": 4.4565045535859466e+18,
|
298 |
"train_batch_size": 128,
|
299 |
"trial_name": null,
|
300 |
"trial_params": null
|