chizhik
commited on
Commit
·
e9b62f8
1
Parent(s):
f8c3dc1
re-trained model after eliminating annotation errors
Browse files- README.md +0 -4
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +318 -284
- training_args.bin +1 -1
README.md
DELETED
@@ -1,4 +0,0 @@
|
|
1 |
-
---
|
2 |
-
license: afl-3.0
|
3 |
-
inference: false
|
4 |
-
---
|
|
|
|
|
|
|
|
|
|
optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 997361565
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5276bb8be599a0e1a57ad1a71f3975ae5e7e6b7cabcbcbde26fe5fd41cbd0853
|
3 |
size 997361565
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 498694189
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f5a7aee6788246a2848ebff5ac3618a37d758a581da5a7ead3c2bf124d51f6d4
|
3 |
size 498694189
|
rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14567
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:573ba4da91fda3dbf4c67c67e7334810f4f5c1c52e4466d2c4c1a4fe2dc8d0d4
|
3 |
size 14567
|
scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 623
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b71b07341c8d2987279f87286ec850204a8e1765279b2b57ded211a066f7e1b3
|
3 |
size 623
|
trainer_state.json
CHANGED
@@ -1,382 +1,416 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "./CARES/checkpoints/roberta-stratified/run-
|
4 |
-
"epoch":
|
5 |
-
"global_step":
|
6 |
"is_hyper_param_search": true,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
-
"epoch":
|
12 |
-
"eval_loss": 0.
|
13 |
-
"eval_macro_f1": 0.
|
14 |
-
"eval_macro_precision": 0.
|
15 |
-
"eval_macro_recall": 0.
|
16 |
-
"eval_micro_f1": 0.
|
17 |
-
"eval_micro_precision":
|
18 |
-
"eval_micro_recall": 0.
|
19 |
-
"eval_runtime": 2.
|
20 |
-
"eval_samples_per_second":
|
21 |
-
"eval_steps_per_second":
|
22 |
"step": 142
|
23 |
},
|
24 |
{
|
25 |
-
"epoch":
|
26 |
-
"eval_loss": 0.
|
27 |
-
"eval_macro_f1": 0.
|
28 |
-
"eval_macro_precision": 0.
|
29 |
-
"eval_macro_recall": 0.
|
30 |
-
"eval_micro_f1": 0.
|
31 |
-
"eval_micro_precision": 0.
|
32 |
-
"eval_micro_recall": 0.
|
33 |
-
"eval_runtime": 2.
|
34 |
-
"eval_samples_per_second":
|
35 |
-
"eval_steps_per_second":
|
36 |
"step": 284
|
37 |
},
|
38 |
{
|
39 |
-
"epoch":
|
40 |
-
"eval_loss": 0.
|
41 |
-
"eval_macro_f1": 0.
|
42 |
-
"eval_macro_precision": 0.
|
43 |
-
"eval_macro_recall": 0.
|
44 |
-
"eval_micro_f1": 0.
|
45 |
-
"eval_micro_precision": 0.
|
46 |
-
"eval_micro_recall": 0.
|
47 |
-
"eval_runtime": 2.
|
48 |
-
"eval_samples_per_second":
|
49 |
-
"eval_steps_per_second":
|
50 |
"step": 426
|
51 |
},
|
52 |
{
|
53 |
-
"epoch":
|
54 |
-
"learning_rate":
|
55 |
-
"loss": 0.
|
56 |
"step": 500
|
57 |
},
|
58 |
{
|
59 |
-
"epoch":
|
60 |
-
"eval_loss": 0.
|
61 |
-
"eval_macro_f1": 0.
|
62 |
-
"eval_macro_precision": 0.
|
63 |
-
"eval_macro_recall": 0.
|
64 |
-
"eval_micro_f1": 0.
|
65 |
-
"eval_micro_precision": 0.
|
66 |
-
"eval_micro_recall": 0.
|
67 |
-
"eval_runtime": 2.
|
68 |
-
"eval_samples_per_second":
|
69 |
-
"eval_steps_per_second":
|
70 |
"step": 568
|
71 |
},
|
72 |
{
|
73 |
-
"epoch":
|
74 |
-
"eval_loss": 0.
|
75 |
-
"eval_macro_f1": 0.
|
76 |
-
"eval_macro_precision": 0.
|
77 |
-
"eval_macro_recall": 0.
|
78 |
-
"eval_micro_f1": 0.
|
79 |
-
"eval_micro_precision": 0.
|
80 |
-
"eval_micro_recall": 0.
|
81 |
-
"eval_runtime": 2.
|
82 |
-
"eval_samples_per_second":
|
83 |
-
"eval_steps_per_second":
|
84 |
"step": 710
|
85 |
},
|
86 |
{
|
87 |
-
"epoch":
|
88 |
-
"eval_loss": 0.
|
89 |
-
"eval_macro_f1": 0.
|
90 |
-
"eval_macro_precision": 0.
|
91 |
-
"eval_macro_recall": 0.
|
92 |
-
"eval_micro_f1": 0.
|
93 |
-
"eval_micro_precision": 0.
|
94 |
-
"eval_micro_recall": 0.
|
95 |
-
"eval_runtime": 2.
|
96 |
-
"eval_samples_per_second":
|
97 |
-
"eval_steps_per_second":
|
98 |
"step": 852
|
99 |
},
|
100 |
{
|
101 |
-
"epoch":
|
102 |
-
"eval_loss": 0.
|
103 |
-
"eval_macro_f1": 0.
|
104 |
-
"eval_macro_precision": 0.
|
105 |
-
"eval_macro_recall": 0.
|
106 |
-
"eval_micro_f1": 0.
|
107 |
-
"eval_micro_precision": 0.
|
108 |
-
"eval_micro_recall": 0.
|
109 |
-
"eval_runtime": 2.
|
110 |
-
"eval_samples_per_second":
|
111 |
-
"eval_steps_per_second":
|
112 |
"step": 994
|
113 |
},
|
114 |
{
|
115 |
-
"epoch":
|
116 |
-
"learning_rate":
|
117 |
-
"loss": 0.
|
118 |
"step": 1000
|
119 |
},
|
120 |
{
|
121 |
-
"epoch":
|
122 |
-
"eval_loss": 0.
|
123 |
-
"eval_macro_f1": 0.
|
124 |
-
"eval_macro_precision": 0.
|
125 |
-
"eval_macro_recall": 0.
|
126 |
-
"eval_micro_f1": 0.
|
127 |
-
"eval_micro_precision": 0.
|
128 |
-
"eval_micro_recall": 0.
|
129 |
-
"eval_runtime": 2.
|
130 |
-
"eval_samples_per_second":
|
131 |
-
"eval_steps_per_second":
|
132 |
"step": 1136
|
133 |
},
|
134 |
{
|
135 |
-
"epoch":
|
136 |
-
"eval_loss": 0.
|
137 |
-
"eval_macro_f1": 0.
|
138 |
-
"eval_macro_precision": 0.
|
139 |
-
"eval_macro_recall": 0.
|
140 |
-
"eval_micro_f1": 0.
|
141 |
-
"eval_micro_precision": 0.
|
142 |
-
"eval_micro_recall": 0.
|
143 |
-
"eval_runtime": 2.
|
144 |
-
"eval_samples_per_second":
|
145 |
-
"eval_steps_per_second":
|
146 |
"step": 1278
|
147 |
},
|
148 |
{
|
149 |
-
"epoch":
|
150 |
-
"eval_loss": 0.
|
151 |
-
"eval_macro_f1": 0.
|
152 |
-
"eval_macro_precision": 0.
|
153 |
-
"eval_macro_recall": 0.
|
154 |
-
"eval_micro_f1": 0.
|
155 |
-
"eval_micro_precision": 0.
|
156 |
-
"eval_micro_recall": 0.
|
157 |
-
"eval_runtime": 2.
|
158 |
-
"eval_samples_per_second":
|
159 |
-
"eval_steps_per_second":
|
160 |
"step": 1420
|
161 |
},
|
162 |
{
|
163 |
-
"epoch":
|
164 |
-
"learning_rate": 2.
|
165 |
-
"loss": 0.
|
166 |
"step": 1500
|
167 |
},
|
168 |
{
|
169 |
-
"epoch":
|
170 |
-
"eval_loss": 0.
|
171 |
-
"eval_macro_f1": 0.
|
172 |
-
"eval_macro_precision": 0.
|
173 |
-
"eval_macro_recall": 0.
|
174 |
-
"eval_micro_f1": 0.
|
175 |
-
"eval_micro_precision": 0.
|
176 |
-
"eval_micro_recall": 0.
|
177 |
-
"eval_runtime": 2.
|
178 |
-
"eval_samples_per_second":
|
179 |
-
"eval_steps_per_second":
|
180 |
"step": 1562
|
181 |
},
|
182 |
{
|
183 |
-
"epoch":
|
184 |
-
"eval_loss": 0.
|
185 |
-
"eval_macro_f1": 0.
|
186 |
-
"eval_macro_precision": 0.
|
187 |
-
"eval_macro_recall": 0.
|
188 |
-
"eval_micro_f1": 0.
|
189 |
-
"eval_micro_precision": 0.
|
190 |
-
"eval_micro_recall": 0.
|
191 |
-
"eval_runtime": 2.
|
192 |
-
"eval_samples_per_second":
|
193 |
-
"eval_steps_per_second":
|
194 |
"step": 1704
|
195 |
},
|
196 |
{
|
197 |
-
"epoch":
|
198 |
-
"eval_loss": 0.
|
199 |
-
"eval_macro_f1": 0.
|
200 |
-
"eval_macro_precision": 0.
|
201 |
-
"eval_macro_recall": 0.
|
202 |
-
"eval_micro_f1": 0.
|
203 |
-
"eval_micro_precision": 0.
|
204 |
-
"eval_micro_recall": 0.
|
205 |
-
"eval_runtime": 2.
|
206 |
-
"eval_samples_per_second":
|
207 |
-
"eval_steps_per_second":
|
208 |
"step": 1846
|
209 |
},
|
210 |
{
|
211 |
-
"epoch":
|
212 |
-
"eval_loss": 0.
|
213 |
-
"eval_macro_f1": 0.
|
214 |
-
"eval_macro_precision": 0.
|
215 |
-
"eval_macro_recall": 0.
|
216 |
-
"eval_micro_f1": 0.
|
217 |
-
"eval_micro_precision": 0.
|
218 |
-
"eval_micro_recall": 0.
|
219 |
-
"eval_runtime": 2.
|
220 |
-
"eval_samples_per_second":
|
221 |
-
"eval_steps_per_second":
|
222 |
"step": 1988
|
223 |
},
|
224 |
{
|
225 |
-
"epoch":
|
226 |
-
"learning_rate": 2.
|
227 |
-
"loss": 0.
|
228 |
"step": 2000
|
229 |
},
|
230 |
{
|
231 |
-
"epoch":
|
232 |
-
"eval_loss": 0.
|
233 |
-
"eval_macro_f1": 0.
|
234 |
-
"eval_macro_precision": 0.
|
235 |
-
"eval_macro_recall": 0.
|
236 |
-
"eval_micro_f1": 0.
|
237 |
-
"eval_micro_precision": 0.
|
238 |
-
"eval_micro_recall": 0.
|
239 |
-
"eval_runtime": 2.
|
240 |
-
"eval_samples_per_second":
|
241 |
-
"eval_steps_per_second":
|
242 |
"step": 2130
|
243 |
},
|
244 |
{
|
245 |
-
"epoch":
|
246 |
-
"eval_loss": 0.
|
247 |
-
"eval_macro_f1": 0.
|
248 |
-
"eval_macro_precision": 0.
|
249 |
-
"eval_macro_recall": 0.
|
250 |
-
"eval_micro_f1": 0.
|
251 |
-
"eval_micro_precision": 0.
|
252 |
-
"eval_micro_recall": 0.
|
253 |
-
"eval_runtime": 2.
|
254 |
-
"eval_samples_per_second":
|
255 |
-
"eval_steps_per_second":
|
256 |
"step": 2272
|
257 |
},
|
258 |
{
|
259 |
-
"epoch":
|
260 |
-
"eval_loss": 0.
|
261 |
-
"eval_macro_f1": 0.
|
262 |
-
"eval_macro_precision": 0.
|
263 |
-
"eval_macro_recall": 0.
|
264 |
-
"eval_micro_f1": 0.
|
265 |
-
"eval_micro_precision": 0.
|
266 |
-
"eval_micro_recall": 0.
|
267 |
-
"eval_runtime": 2.
|
268 |
-
"eval_samples_per_second":
|
269 |
-
"eval_steps_per_second":
|
270 |
"step": 2414
|
271 |
},
|
272 |
{
|
273 |
-
"epoch":
|
274 |
-
"learning_rate":
|
275 |
-
"loss": 0.
|
276 |
"step": 2500
|
277 |
},
|
278 |
{
|
279 |
-
"epoch":
|
280 |
-
"eval_loss": 0.
|
281 |
-
"eval_macro_f1": 0.
|
282 |
-
"eval_macro_precision": 0.
|
283 |
-
"eval_macro_recall": 0.
|
284 |
-
"eval_micro_f1": 0.
|
285 |
-
"eval_micro_precision": 0.
|
286 |
-
"eval_micro_recall": 0.
|
287 |
-
"eval_runtime": 2.
|
288 |
-
"eval_samples_per_second":
|
289 |
-
"eval_steps_per_second":
|
290 |
"step": 2556
|
291 |
},
|
292 |
{
|
293 |
-
"epoch":
|
294 |
-
"eval_loss": 0.
|
295 |
-
"eval_macro_f1": 0.
|
296 |
-
"eval_macro_precision": 0.
|
297 |
-
"eval_macro_recall": 0.
|
298 |
-
"eval_micro_f1": 0.
|
299 |
-
"eval_micro_precision": 0.
|
300 |
-
"eval_micro_recall": 0.
|
301 |
-
"eval_runtime": 2.
|
302 |
-
"eval_samples_per_second":
|
303 |
-
"eval_steps_per_second":
|
304 |
"step": 2698
|
305 |
},
|
306 |
{
|
307 |
-
"epoch":
|
308 |
-
"eval_loss": 0.
|
309 |
-
"eval_macro_f1": 0.
|
310 |
-
"eval_macro_precision": 0.
|
311 |
-
"eval_macro_recall": 0.
|
312 |
-
"eval_micro_f1": 0.
|
313 |
-
"eval_micro_precision": 0.
|
314 |
-
"eval_micro_recall": 0.
|
315 |
-
"eval_runtime": 2.
|
316 |
-
"eval_samples_per_second":
|
317 |
-
"eval_steps_per_second":
|
318 |
"step": 2840
|
319 |
},
|
320 |
{
|
321 |
-
"epoch":
|
322 |
-
"eval_loss": 0.
|
323 |
-
"eval_macro_f1": 0.
|
324 |
-
"eval_macro_precision": 0.
|
325 |
-
"eval_macro_recall": 0.
|
326 |
-
"eval_micro_f1": 0.
|
327 |
-
"eval_micro_precision": 0.
|
328 |
-
"eval_micro_recall": 0.
|
329 |
-
"eval_runtime": 2.
|
330 |
-
"eval_samples_per_second":
|
331 |
-
"eval_steps_per_second":
|
332 |
"step": 2982
|
333 |
},
|
334 |
{
|
335 |
-
"epoch":
|
336 |
-
"learning_rate":
|
337 |
-
"loss": 0.
|
338 |
"step": 3000
|
339 |
},
|
340 |
{
|
341 |
-
"epoch":
|
342 |
-
"eval_loss": 0.
|
343 |
-
"eval_macro_f1": 0.
|
344 |
-
"eval_macro_precision": 0.
|
345 |
-
"eval_macro_recall": 0.
|
346 |
-
"eval_micro_f1": 0.
|
347 |
-
"eval_micro_precision": 0.
|
348 |
-
"eval_micro_recall": 0.
|
349 |
-
"eval_runtime": 2.
|
350 |
-
"eval_samples_per_second":
|
351 |
-
"eval_steps_per_second":
|
352 |
"step": 3124
|
353 |
},
|
354 |
{
|
355 |
-
"epoch":
|
356 |
-
"eval_loss": 0.
|
357 |
-
"eval_macro_f1": 0.
|
358 |
-
"eval_macro_precision": 0.
|
359 |
-
"eval_macro_recall": 0.
|
360 |
-
"eval_micro_f1": 0.
|
361 |
-
"eval_micro_precision": 0.
|
362 |
-
"eval_micro_recall": 0.
|
363 |
-
"eval_runtime": 2.
|
364 |
-
"eval_samples_per_second":
|
365 |
-
"eval_steps_per_second":
|
366 |
"step": 3266
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
367 |
}
|
368 |
],
|
369 |
-
"max_steps":
|
370 |
"num_train_epochs": 100,
|
371 |
-
"total_flos":
|
372 |
"trial_name": null,
|
373 |
"trial_params": {
|
374 |
-
"adam_epsilon": 6.
|
375 |
-
"learning_rate": 3.
|
376 |
-
"per_device_eval_batch_size":
|
377 |
-
"per_device_train_batch_size":
|
378 |
-
"seed":
|
379 |
-
"warmup_steps":
|
380 |
-
"weight_decay":
|
381 |
}
|
382 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.7672541719320296,
|
3 |
+
"best_model_checkpoint": "./CARES/checkpoints/roberta-stratified/run-6/checkpoint-3550",
|
4 |
+
"epoch": 50.0,
|
5 |
+
"global_step": 3550,
|
6 |
"is_hyper_param_search": true,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
+
"epoch": 2.0,
|
12 |
+
"eval_loss": 0.24808131158351898,
|
13 |
+
"eval_macro_f1": 0.05481647030175482,
|
14 |
+
"eval_macro_precision": 0.17800453514739228,
|
15 |
+
"eval_macro_recall": 0.03330847588023425,
|
16 |
+
"eval_micro_f1": 0.13594611145131658,
|
17 |
+
"eval_micro_precision": 0.9568965517241379,
|
18 |
+
"eval_micro_recall": 0.07317073170731707,
|
19 |
+
"eval_runtime": 2.8186,
|
20 |
+
"eval_samples_per_second": 342.724,
|
21 |
+
"eval_steps_per_second": 21.642,
|
22 |
"step": 142
|
23 |
},
|
24 |
{
|
25 |
+
"epoch": 4.0,
|
26 |
+
"eval_loss": 0.16814221441745758,
|
27 |
+
"eval_macro_f1": 0.37346208328620767,
|
28 |
+
"eval_macro_precision": 0.4376472360907343,
|
29 |
+
"eval_macro_recall": 0.3342883491414155,
|
30 |
+
"eval_micro_f1": 0.7248062015503876,
|
31 |
+
"eval_micro_precision": 0.8795860771401693,
|
32 |
+
"eval_micro_recall": 0.6163480553724456,
|
33 |
+
"eval_runtime": 2.8192,
|
34 |
+
"eval_samples_per_second": 342.653,
|
35 |
+
"eval_steps_per_second": 21.638,
|
36 |
"step": 284
|
37 |
},
|
38 |
{
|
39 |
+
"epoch": 6.0,
|
40 |
+
"eval_loss": 0.13150478899478912,
|
41 |
+
"eval_macro_f1": 0.4319986220599845,
|
42 |
+
"eval_macro_precision": 0.6389953058630822,
|
43 |
+
"eval_macro_recall": 0.37598442828818834,
|
44 |
+
"eval_micro_f1": 0.7641723356009071,
|
45 |
+
"eval_micro_precision": 0.895482728077945,
|
46 |
+
"eval_micro_recall": 0.6664469347396177,
|
47 |
+
"eval_runtime": 2.8191,
|
48 |
+
"eval_samples_per_second": 342.658,
|
49 |
+
"eval_steps_per_second": 21.638,
|
50 |
"step": 426
|
51 |
},
|
52 |
{
|
53 |
+
"epoch": 7.04,
|
54 |
+
"learning_rate": 2.8662355855330125e-05,
|
55 |
+
"loss": 0.2172,
|
56 |
"step": 500
|
57 |
},
|
58 |
{
|
59 |
+
"epoch": 8.0,
|
60 |
+
"eval_loss": 0.1121998205780983,
|
61 |
+
"eval_macro_f1": 0.5386615940425785,
|
62 |
+
"eval_macro_precision": 0.7631201629660762,
|
63 |
+
"eval_macro_recall": 0.46714612883136086,
|
64 |
+
"eval_micro_f1": 0.8077485380116959,
|
65 |
+
"eval_micro_precision": 0.9064807219031994,
|
66 |
+
"eval_micro_recall": 0.7284113381674358,
|
67 |
+
"eval_runtime": 2.8196,
|
68 |
+
"eval_samples_per_second": 342.598,
|
69 |
+
"eval_steps_per_second": 21.634,
|
70 |
"step": 568
|
71 |
},
|
72 |
{
|
73 |
+
"epoch": 10.0,
|
74 |
+
"eval_loss": 0.10608664900064468,
|
75 |
+
"eval_macro_f1": 0.5851323545796894,
|
76 |
+
"eval_macro_precision": 0.7572045278273729,
|
77 |
+
"eval_macro_recall": 0.5132727928856573,
|
78 |
+
"eval_micro_f1": 0.8146802325581396,
|
79 |
+
"eval_micro_precision": 0.9076923076923077,
|
80 |
+
"eval_micro_recall": 0.7389584706657878,
|
81 |
+
"eval_runtime": 2.8204,
|
82 |
+
"eval_samples_per_second": 342.507,
|
83 |
+
"eval_steps_per_second": 21.628,
|
84 |
"step": 710
|
85 |
},
|
86 |
{
|
87 |
+
"epoch": 12.0,
|
88 |
+
"eval_loss": 0.100788913667202,
|
89 |
+
"eval_macro_f1": 0.628825318779795,
|
90 |
+
"eval_macro_precision": 0.750817062947988,
|
91 |
+
"eval_macro_recall": 0.5578116563013593,
|
92 |
+
"eval_micro_f1": 0.8245363766048501,
|
93 |
+
"eval_micro_precision": 0.8982128982128982,
|
94 |
+
"eval_micro_recall": 0.7620303230059328,
|
95 |
+
"eval_runtime": 2.8211,
|
96 |
+
"eval_samples_per_second": 342.425,
|
97 |
+
"eval_steps_per_second": 21.623,
|
98 |
"step": 852
|
99 |
},
|
100 |
{
|
101 |
+
"epoch": 14.0,
|
102 |
+
"eval_loss": 0.09952697157859802,
|
103 |
+
"eval_macro_f1": 0.648367052483795,
|
104 |
+
"eval_macro_precision": 0.7408904755242313,
|
105 |
+
"eval_macro_recall": 0.5923039028415692,
|
106 |
+
"eval_micro_f1": 0.8379418970948547,
|
107 |
+
"eval_micro_precision": 0.8932835820895523,
|
108 |
+
"eval_micro_recall": 0.7890573500329597,
|
109 |
+
"eval_runtime": 2.8203,
|
110 |
+
"eval_samples_per_second": 342.519,
|
111 |
+
"eval_steps_per_second": 21.629,
|
112 |
"step": 994
|
113 |
},
|
114 |
{
|
115 |
+
"epoch": 14.08,
|
116 |
+
"learning_rate": 2.6490965260229357e-05,
|
117 |
+
"loss": 0.0388,
|
118 |
"step": 1000
|
119 |
},
|
120 |
{
|
121 |
+
"epoch": 16.0,
|
122 |
+
"eval_loss": 0.10003374516963959,
|
123 |
+
"eval_macro_f1": 0.7100297932545989,
|
124 |
+
"eval_macro_precision": 0.8731926397728601,
|
125 |
+
"eval_macro_recall": 0.6387457691830838,
|
126 |
+
"eval_micro_f1": 0.837847344354555,
|
127 |
+
"eval_micro_precision": 0.8981900452488688,
|
128 |
+
"eval_micro_recall": 0.7851021753460777,
|
129 |
+
"eval_runtime": 2.8204,
|
130 |
+
"eval_samples_per_second": 342.505,
|
131 |
+
"eval_steps_per_second": 21.628,
|
132 |
"step": 1136
|
133 |
},
|
134 |
{
|
135 |
+
"epoch": 18.0,
|
136 |
+
"eval_loss": 0.10116878896951675,
|
137 |
+
"eval_macro_f1": 0.7102542897490418,
|
138 |
+
"eval_macro_precision": 0.8708979085725551,
|
139 |
+
"eval_macro_recall": 0.6278706438272531,
|
140 |
+
"eval_micro_f1": 0.8369795342272407,
|
141 |
+
"eval_micro_precision": 0.9005315110098709,
|
142 |
+
"eval_micro_recall": 0.7818061964403428,
|
143 |
+
"eval_runtime": 2.821,
|
144 |
+
"eval_samples_per_second": 342.431,
|
145 |
+
"eval_steps_per_second": 21.624,
|
146 |
"step": 1278
|
147 |
},
|
148 |
{
|
149 |
+
"epoch": 20.0,
|
150 |
+
"eval_loss": 0.10277832299470901,
|
151 |
+
"eval_macro_f1": 0.7155479835258449,
|
152 |
+
"eval_macro_precision": 0.8815899776220457,
|
153 |
+
"eval_macro_recall": 0.635461059766739,
|
154 |
+
"eval_micro_f1": 0.8361702127659574,
|
155 |
+
"eval_micro_precision": 0.9048349961627015,
|
156 |
+
"eval_micro_recall": 0.7771918259723137,
|
157 |
+
"eval_runtime": 2.8201,
|
158 |
+
"eval_samples_per_second": 342.543,
|
159 |
+
"eval_steps_per_second": 21.631,
|
160 |
"step": 1420
|
161 |
},
|
162 |
{
|
163 |
+
"epoch": 21.13,
|
164 |
+
"learning_rate": 2.431957466512859e-05,
|
165 |
+
"loss": 0.0147,
|
166 |
"step": 1500
|
167 |
},
|
168 |
{
|
169 |
+
"epoch": 22.0,
|
170 |
+
"eval_loss": 0.10351855307817459,
|
171 |
+
"eval_macro_f1": 0.7203873361325647,
|
172 |
+
"eval_macro_precision": 0.87817690487536,
|
173 |
+
"eval_macro_recall": 0.6443209986334517,
|
174 |
+
"eval_micro_f1": 0.837389770723104,
|
175 |
+
"eval_micro_precision": 0.9006069802731411,
|
176 |
+
"eval_micro_recall": 0.7824653922214898,
|
177 |
+
"eval_runtime": 2.8203,
|
178 |
+
"eval_samples_per_second": 342.52,
|
179 |
+
"eval_steps_per_second": 21.629,
|
180 |
"step": 1562
|
181 |
},
|
182 |
{
|
183 |
+
"epoch": 24.0,
|
184 |
+
"eval_loss": 0.10557578504085541,
|
185 |
+
"eval_macro_f1": 0.7298035823714943,
|
186 |
+
"eval_macro_precision": 0.8804692132123697,
|
187 |
+
"eval_macro_recall": 0.6544575680820387,
|
188 |
+
"eval_micro_f1": 0.8399153737658673,
|
189 |
+
"eval_micro_precision": 0.9029567854435178,
|
190 |
+
"eval_micro_recall": 0.7851021753460777,
|
191 |
+
"eval_runtime": 2.8192,
|
192 |
+
"eval_samples_per_second": 342.647,
|
193 |
+
"eval_steps_per_second": 21.637,
|
194 |
"step": 1704
|
195 |
},
|
196 |
{
|
197 |
+
"epoch": 26.0,
|
198 |
+
"eval_loss": 0.10698197782039642,
|
199 |
+
"eval_macro_f1": 0.719671126038702,
|
200 |
+
"eval_macro_precision": 0.8799359996888543,
|
201 |
+
"eval_macro_recall": 0.6422398711062759,
|
202 |
+
"eval_micro_f1": 0.8375706214689265,
|
203 |
+
"eval_micro_precision": 0.9019011406844106,
|
204 |
+
"eval_micro_recall": 0.7818061964403428,
|
205 |
+
"eval_runtime": 2.8198,
|
206 |
+
"eval_samples_per_second": 342.58,
|
207 |
+
"eval_steps_per_second": 21.633,
|
208 |
"step": 1846
|
209 |
},
|
210 |
{
|
211 |
+
"epoch": 28.0,
|
212 |
+
"eval_loss": 0.10863872617483139,
|
213 |
+
"eval_macro_f1": 0.7318024399797272,
|
214 |
+
"eval_macro_precision": 0.8781101314061108,
|
215 |
+
"eval_macro_recall": 0.6548756753680312,
|
216 |
+
"eval_micro_f1": 0.8389143461402891,
|
217 |
+
"eval_micro_precision": 0.9015151515151515,
|
218 |
+
"eval_micro_recall": 0.7844429795649308,
|
219 |
+
"eval_runtime": 2.8198,
|
220 |
+
"eval_samples_per_second": 342.583,
|
221 |
+
"eval_steps_per_second": 21.633,
|
222 |
"step": 1988
|
223 |
},
|
224 |
{
|
225 |
+
"epoch": 28.17,
|
226 |
+
"learning_rate": 2.214818407002782e-05,
|
227 |
+
"loss": 0.0082,
|
228 |
"step": 2000
|
229 |
},
|
230 |
{
|
231 |
+
"epoch": 30.0,
|
232 |
+
"eval_loss": 0.1103406548500061,
|
233 |
+
"eval_macro_f1": 0.728656455732704,
|
234 |
+
"eval_macro_precision": 0.8784457053881349,
|
235 |
+
"eval_macro_recall": 0.6552278977528909,
|
236 |
+
"eval_micro_f1": 0.8400702987697715,
|
237 |
+
"eval_micro_precision": 0.8998493975903614,
|
238 |
+
"eval_micro_recall": 0.7877389584706658,
|
239 |
+
"eval_runtime": 2.8227,
|
240 |
+
"eval_samples_per_second": 342.231,
|
241 |
+
"eval_steps_per_second": 21.611,
|
242 |
"step": 2130
|
243 |
},
|
244 |
{
|
245 |
+
"epoch": 32.0,
|
246 |
+
"eval_loss": 0.11085448414087296,
|
247 |
+
"eval_macro_f1": 0.7344651790603305,
|
248 |
+
"eval_macro_precision": 0.8647882525531505,
|
249 |
+
"eval_macro_recall": 0.6579821769650391,
|
250 |
+
"eval_micro_f1": 0.8402116402116402,
|
251 |
+
"eval_micro_precision": 0.9036418816388467,
|
252 |
+
"eval_micro_recall": 0.7851021753460777,
|
253 |
+
"eval_runtime": 2.8204,
|
254 |
+
"eval_samples_per_second": 342.508,
|
255 |
+
"eval_steps_per_second": 21.628,
|
256 |
"step": 2272
|
257 |
},
|
258 |
{
|
259 |
+
"epoch": 34.0,
|
260 |
+
"eval_loss": 0.11268670856952667,
|
261 |
+
"eval_macro_f1": 0.7371526225457369,
|
262 |
+
"eval_macro_precision": 0.8672638823695995,
|
263 |
+
"eval_macro_recall": 0.6591834104386446,
|
264 |
+
"eval_micro_f1": 0.8397323001056711,
|
265 |
+
"eval_micro_precision": 0.9016641452344932,
|
266 |
+
"eval_micro_recall": 0.7857613711272248,
|
267 |
+
"eval_runtime": 2.8203,
|
268 |
+
"eval_samples_per_second": 342.515,
|
269 |
+
"eval_steps_per_second": 21.629,
|
270 |
"step": 2414
|
271 |
},
|
272 |
{
|
273 |
+
"epoch": 35.21,
|
274 |
+
"learning_rate": 1.9976793474927056e-05,
|
275 |
+
"loss": 0.0056,
|
276 |
"step": 2500
|
277 |
},
|
278 |
{
|
279 |
+
"epoch": 36.0,
|
280 |
+
"eval_loss": 0.11449939757585526,
|
281 |
+
"eval_macro_f1": 0.7435877211427984,
|
282 |
+
"eval_macro_precision": 0.8803719443456752,
|
283 |
+
"eval_macro_recall": 0.6671207777218493,
|
284 |
+
"eval_micro_f1": 0.8435852372583479,
|
285 |
+
"eval_micro_precision": 0.9036144578313253,
|
286 |
+
"eval_micro_recall": 0.7910349373764007,
|
287 |
+
"eval_runtime": 2.8219,
|
288 |
+
"eval_samples_per_second": 342.324,
|
289 |
+
"eval_steps_per_second": 21.617,
|
290 |
"step": 2556
|
291 |
},
|
292 |
{
|
293 |
+
"epoch": 38.0,
|
294 |
+
"eval_loss": 0.1153542622923851,
|
295 |
+
"eval_macro_f1": 0.7450728657264144,
|
296 |
+
"eval_macro_precision": 0.879277762192774,
|
297 |
+
"eval_macro_recall": 0.6681699293595025,
|
298 |
+
"eval_micro_f1": 0.8419570573741639,
|
299 |
+
"eval_micro_precision": 0.9033232628398792,
|
300 |
+
"eval_micro_recall": 0.7883981542518128,
|
301 |
+
"eval_runtime": 2.8191,
|
302 |
+
"eval_samples_per_second": 342.661,
|
303 |
+
"eval_steps_per_second": 21.638,
|
304 |
"step": 2698
|
305 |
},
|
306 |
{
|
307 |
+
"epoch": 40.0,
|
308 |
+
"eval_loss": 0.11756419390439987,
|
309 |
+
"eval_macro_f1": 0.744466331732378,
|
310 |
+
"eval_macro_precision": 0.8796020530253235,
|
311 |
+
"eval_macro_recall": 0.6677481100215579,
|
312 |
+
"eval_micro_f1": 0.8419570573741639,
|
313 |
+
"eval_micro_precision": 0.9033232628398792,
|
314 |
+
"eval_micro_recall": 0.7883981542518128,
|
315 |
+
"eval_runtime": 2.8181,
|
316 |
+
"eval_samples_per_second": 342.785,
|
317 |
+
"eval_steps_per_second": 21.646,
|
318 |
"step": 2840
|
319 |
},
|
320 |
{
|
321 |
+
"epoch": 42.0,
|
322 |
+
"eval_loss": 0.11846820265054703,
|
323 |
+
"eval_macro_f1": 0.7434279482403103,
|
324 |
+
"eval_macro_precision": 0.8813168271357439,
|
325 |
+
"eval_macro_recall": 0.6651456053129234,
|
326 |
+
"eval_micro_f1": 0.8427717200140696,
|
327 |
+
"eval_micro_precision": 0.9034690799396682,
|
328 |
+
"eval_micro_recall": 0.7897165458141068,
|
329 |
+
"eval_runtime": 2.8203,
|
330 |
+
"eval_samples_per_second": 342.52,
|
331 |
+
"eval_steps_per_second": 21.629,
|
332 |
"step": 2982
|
333 |
},
|
334 |
{
|
335 |
+
"epoch": 42.25,
|
336 |
+
"learning_rate": 1.7805402879826288e-05,
|
337 |
+
"loss": 0.004,
|
338 |
"step": 3000
|
339 |
},
|
340 |
{
|
341 |
+
"epoch": 44.0,
|
342 |
+
"eval_loss": 0.11996057629585266,
|
343 |
+
"eval_macro_f1": 0.7394753977857482,
|
344 |
+
"eval_macro_precision": 0.8783969916238008,
|
345 |
+
"eval_macro_recall": 0.6650662737686901,
|
346 |
+
"eval_micro_f1": 0.8427717200140696,
|
347 |
+
"eval_micro_precision": 0.9034690799396682,
|
348 |
+
"eval_micro_recall": 0.7897165458141068,
|
349 |
+
"eval_runtime": 2.82,
|
350 |
+
"eval_samples_per_second": 342.552,
|
351 |
+
"eval_steps_per_second": 21.631,
|
352 |
"step": 3124
|
353 |
},
|
354 |
{
|
355 |
+
"epoch": 46.0,
|
356 |
+
"eval_loss": 0.12046220153570175,
|
357 |
+
"eval_macro_f1": 0.7423287819151947,
|
358 |
+
"eval_macro_precision": 0.8810998402784964,
|
359 |
+
"eval_macro_recall": 0.6639465302563645,
|
360 |
+
"eval_micro_f1": 0.8432546671363156,
|
361 |
+
"eval_micro_precision": 0.905446293494705,
|
362 |
+
"eval_micro_recall": 0.7890573500329597,
|
363 |
+
"eval_runtime": 2.819,
|
364 |
+
"eval_samples_per_second": 342.677,
|
365 |
+
"eval_steps_per_second": 21.639,
|
366 |
"step": 3266
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 48.0,
|
370 |
+
"eval_loss": 0.12299305200576782,
|
371 |
+
"eval_macro_f1": 0.7584256567760465,
|
372 |
+
"eval_macro_precision": 0.9421857420436691,
|
373 |
+
"eval_macro_recall": 0.6745195368122752,
|
374 |
+
"eval_micro_f1": 0.8419570573741639,
|
375 |
+
"eval_micro_precision": 0.9033232628398792,
|
376 |
+
"eval_micro_recall": 0.7883981542518128,
|
377 |
+
"eval_runtime": 2.8184,
|
378 |
+
"eval_samples_per_second": 342.744,
|
379 |
+
"eval_steps_per_second": 21.643,
|
380 |
+
"step": 3408
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 49.3,
|
384 |
+
"learning_rate": 1.563401228472552e-05,
|
385 |
+
"loss": 0.0031,
|
386 |
+
"step": 3500
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 50.0,
|
390 |
+
"eval_loss": 0.12287621200084686,
|
391 |
+
"eval_macro_f1": 0.7672541719320296,
|
392 |
+
"eval_macro_precision": 0.9303028445349943,
|
393 |
+
"eval_macro_recall": 0.6844658294965831,
|
394 |
+
"eval_micro_f1": 0.8449122807017543,
|
395 |
+
"eval_micro_precision": 0.9032258064516129,
|
396 |
+
"eval_micro_recall": 0.7936717205009888,
|
397 |
+
"eval_runtime": 2.8187,
|
398 |
+
"eval_samples_per_second": 342.711,
|
399 |
+
"eval_steps_per_second": 21.641,
|
400 |
+
"step": 3550
|
401 |
}
|
402 |
],
|
403 |
+
"max_steps": 7100,
|
404 |
"num_train_epochs": 100,
|
405 |
+
"total_flos": 3.007255197834797e+16,
|
406 |
"trial_name": null,
|
407 |
"trial_params": {
|
408 |
+
"adam_epsilon": 6.447418463180699e-08,
|
409 |
+
"learning_rate": 3.0290898801655698e-05,
|
410 |
+
"per_device_eval_batch_size": 16,
|
411 |
+
"per_device_train_batch_size": 32,
|
412 |
+
"seed": 320,
|
413 |
+
"warmup_steps": 125,
|
414 |
+
"weight_decay": 4.5126980713116176e-08
|
415 |
}
|
416 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3439
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f8c9e36bee277bf832a3d25b8f22fe9e82c88fe3d841e639970624efd5b1bd10
|
3 |
size 3439
|