🍻 cheers
Browse files- README.md +7 -1
- all_results.json +12 -5
- eval_results.json +10 -0
- runs/Apr15_16-57-14_ryanserver/events.out.tfevents.1713219645.ryanserver.25404.1 +3 -0
- train_results.json +5 -5
- trainer_state.json +401 -394
README.md
CHANGED
@@ -2,6 +2,7 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: google/vit-base-patch16-224-in21k
|
4 |
tags:
|
|
|
5 |
- generated_from_trainer
|
6 |
model-index:
|
7 |
- name: ryan04152024_ALLDATA
|
@@ -13,7 +14,12 @@ should probably proofread and complete it, then remove this comment. -->
|
|
13 |
|
14 |
# ryan04152024_ALLDATA
|
15 |
|
16 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
## Model description
|
19 |
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: google/vit-base-patch16-224-in21k
|
4 |
tags:
|
5 |
+
- image-classification
|
6 |
- generated_from_trainer
|
7 |
model-index:
|
8 |
- name: ryan04152024_ALLDATA
|
|
|
14 |
|
15 |
# ryan04152024_ALLDATA
|
16 |
|
17 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the properties dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.1193
|
20 |
+
- Ordinal Mae: 0.3505
|
21 |
+
- Ordinal Accuracy: 0.7757
|
22 |
+
- Na Accuracy: 0.9411
|
23 |
|
24 |
## Model description
|
25 |
|
all_results.json
CHANGED
@@ -1,7 +1,14 @@
|
|
1 |
{
|
2 |
-
"epoch": 1.
|
3 |
-
"
|
4 |
-
"
|
5 |
-
"
|
6 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.02,
|
3 |
+
"eval_loss": 0.1193477213382721,
|
4 |
+
"eval_na_accuracy": 0.9410737752914429,
|
5 |
+
"eval_ordinal_accuracy": 0.7756819128990173,
|
6 |
+
"eval_ordinal_mae": 0.350502610206604,
|
7 |
+
"eval_runtime": 1992.5951,
|
8 |
+
"eval_samples_per_second": 25.638,
|
9 |
+
"eval_steps_per_second": 3.205,
|
10 |
+
"train_loss": 0.2225519281828883,
|
11 |
+
"train_runtime": 2924.985,
|
12 |
+
"train_samples_per_second": 17.815,
|
13 |
+
"train_steps_per_second": 1.114
|
14 |
}
|
eval_results.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.02,
|
3 |
+
"eval_loss": 0.1193477213382721,
|
4 |
+
"eval_na_accuracy": 0.9410737752914429,
|
5 |
+
"eval_ordinal_accuracy": 0.7756819128990173,
|
6 |
+
"eval_ordinal_mae": 0.350502610206604,
|
7 |
+
"eval_runtime": 1992.5951,
|
8 |
+
"eval_samples_per_second": 25.638,
|
9 |
+
"eval_steps_per_second": 3.205
|
10 |
+
}
|
runs/Apr15_16-57-14_ryanserver/events.out.tfevents.1713219645.ryanserver.25404.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:921604af27040465e89282dcd46e1b4b4a3be3d6f82fe704e4b642150afd563a
|
3 |
+
size 529
|
train_results.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
-
"epoch": 1.
|
3 |
-
"train_loss": 0.
|
4 |
-
"train_runtime":
|
5 |
-
"train_samples_per_second":
|
6 |
-
"train_steps_per_second":
|
7 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.02,
|
3 |
+
"train_loss": 0.2225519281828883,
|
4 |
+
"train_runtime": 2924.985,
|
5 |
+
"train_samples_per_second": 17.815,
|
6 |
+
"train_steps_per_second": 1.114
|
7 |
}
|
trainer_state.json
CHANGED
@@ -1,932 +1,939 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 1.
|
5 |
"eval_steps": 100,
|
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": 0.01,
|
13 |
-
"grad_norm": 0.
|
14 |
-
"learning_rate": 4.
|
15 |
-
"loss": 0.
|
16 |
"step": 25
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.02,
|
20 |
-
"grad_norm": 0.
|
21 |
-
"learning_rate": 4.
|
22 |
-
"loss": 0.
|
23 |
"step": 50
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.02,
|
27 |
-
"grad_norm": 0.
|
28 |
-
"learning_rate": 4.
|
29 |
-
"loss": 0.
|
30 |
"step": 75
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.03,
|
34 |
-
"grad_norm": 2.
|
35 |
-
"learning_rate": 4.
|
36 |
-
"loss": 0.
|
37 |
"step": 100
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.04,
|
41 |
-
"grad_norm": 0.
|
42 |
-
"learning_rate": 4.
|
43 |
-
"loss": 0.
|
44 |
"step": 125
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.05,
|
48 |
-
"grad_norm": 2.
|
49 |
-
"learning_rate": 4.
|
50 |
-
"loss": 0.
|
51 |
"step": 150
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.05,
|
55 |
-
"grad_norm":
|
56 |
-
"learning_rate": 4.
|
57 |
-
"loss": 0.
|
58 |
"step": 175
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.06,
|
62 |
-
"grad_norm": 0.
|
63 |
-
"learning_rate": 4.
|
64 |
-
"loss": 0.
|
65 |
"step": 200
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.07,
|
69 |
-
"grad_norm": 0.
|
70 |
-
"learning_rate": 4.
|
71 |
"loss": 0.3291,
|
72 |
"step": 225
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.08,
|
76 |
-
"grad_norm": 0.
|
77 |
-
"learning_rate": 4.
|
78 |
-
"loss": 0.
|
79 |
"step": 250
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.09,
|
83 |
-
"grad_norm": 0.
|
84 |
-
"learning_rate": 4.
|
85 |
-
"loss": 0.
|
86 |
"step": 275
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.09,
|
90 |
-
"grad_norm": 0.
|
91 |
-
"learning_rate": 4.
|
92 |
-
"loss": 0.
|
93 |
"step": 300
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.1,
|
97 |
-
"grad_norm": 0.
|
98 |
-
"learning_rate": 4.
|
99 |
-
"loss": 0.
|
100 |
"step": 325
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.11,
|
104 |
-
"grad_norm": 2.
|
105 |
-
"learning_rate": 4.
|
106 |
-
"loss": 0.
|
107 |
"step": 350
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.12,
|
111 |
-
"grad_norm": 0.
|
112 |
-
"learning_rate": 4.
|
113 |
-
"loss": 0.
|
114 |
"step": 375
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.13,
|
118 |
-
"grad_norm": 0.
|
119 |
-
"learning_rate": 4.
|
120 |
-
"loss": 0.
|
121 |
"step": 400
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.13,
|
125 |
-
"grad_norm": 0.
|
126 |
-
"learning_rate": 4.
|
127 |
-
"loss": 0.
|
128 |
"step": 425
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.14,
|
132 |
-
"grad_norm":
|
133 |
-
"learning_rate": 4.
|
134 |
-
"loss": 0.
|
135 |
"step": 450
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.15,
|
139 |
-
"grad_norm":
|
140 |
-
"learning_rate": 4.
|
141 |
-
"loss": 0.
|
142 |
"step": 475
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.16,
|
146 |
-
"grad_norm": 1.
|
147 |
-
"learning_rate": 4.
|
148 |
-
"loss": 0.
|
149 |
"step": 500
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.16,
|
153 |
-
"grad_norm": 3.
|
154 |
-
"learning_rate": 4.
|
155 |
-
"loss": 0.
|
156 |
"step": 525
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.17,
|
160 |
-
"grad_norm": 1.
|
161 |
-
"learning_rate": 4.
|
162 |
-
"loss": 0.
|
163 |
"step": 550
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.18,
|
167 |
-
"grad_norm": 0.
|
168 |
-
"learning_rate": 4.
|
169 |
-
"loss": 0.
|
170 |
"step": 575
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.19,
|
174 |
-
"grad_norm":
|
175 |
-
"learning_rate": 4.
|
176 |
-
"loss": 0.
|
177 |
"step": 600
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.2,
|
181 |
-
"grad_norm": 1.
|
182 |
-
"learning_rate": 4.
|
183 |
-
"loss": 0.
|
184 |
"step": 625
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.2,
|
188 |
-
"grad_norm":
|
189 |
-
"learning_rate":
|
190 |
-
"loss": 0.
|
191 |
"step": 650
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.21,
|
195 |
-
"grad_norm": 2.
|
196 |
-
"learning_rate": 3.
|
197 |
-
"loss": 0.
|
198 |
"step": 675
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.22,
|
202 |
-
"grad_norm": 1.
|
203 |
-
"learning_rate": 3.
|
204 |
-
"loss": 0.
|
205 |
"step": 700
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.23,
|
209 |
-
"grad_norm":
|
210 |
-
"learning_rate": 3.
|
211 |
-
"loss": 0.
|
212 |
"step": 725
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.23,
|
216 |
-
"grad_norm": 0.
|
217 |
-
"learning_rate": 3.
|
218 |
-
"loss": 0.
|
219 |
"step": 750
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.24,
|
223 |
-
"grad_norm": 0.
|
224 |
-
"learning_rate": 3.
|
225 |
-
"loss": 0.
|
226 |
"step": 775
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.25,
|
230 |
-
"grad_norm": 1.
|
231 |
-
"learning_rate": 3.
|
232 |
-
"loss": 0.
|
233 |
"step": 800
|
234 |
},
|
235 |
{
|
236 |
"epoch": 0.26,
|
237 |
-
"grad_norm":
|
238 |
-
"learning_rate": 3.
|
239 |
-
"loss": 0.
|
240 |
"step": 825
|
241 |
},
|
242 |
{
|
243 |
"epoch": 0.27,
|
244 |
-
"grad_norm":
|
245 |
-
"learning_rate": 3.
|
246 |
-
"loss": 0.
|
247 |
"step": 850
|
248 |
},
|
249 |
{
|
250 |
"epoch": 0.27,
|
251 |
-
"grad_norm":
|
252 |
-
"learning_rate": 3.
|
253 |
-
"loss": 0.
|
254 |
"step": 875
|
255 |
},
|
256 |
{
|
257 |
"epoch": 0.28,
|
258 |
-
"grad_norm": 1.
|
259 |
-
"learning_rate": 3.
|
260 |
-
"loss": 0.
|
261 |
"step": 900
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.29,
|
265 |
-
"grad_norm":
|
266 |
-
"learning_rate": 3.
|
267 |
-
"loss": 0.
|
268 |
"step": 925
|
269 |
},
|
270 |
{
|
271 |
"epoch": 0.3,
|
272 |
-
"grad_norm": 0.
|
273 |
-
"learning_rate": 3.
|
274 |
-
"loss": 0.
|
275 |
"step": 950
|
276 |
},
|
277 |
{
|
278 |
"epoch": 0.31,
|
279 |
-
"grad_norm": 1.
|
280 |
-
"learning_rate": 3.
|
281 |
-
"loss": 0.
|
282 |
"step": 975
|
283 |
},
|
284 |
{
|
285 |
"epoch": 0.31,
|
286 |
-
"grad_norm": 0.
|
287 |
-
"learning_rate": 3.
|
288 |
-
"loss": 0.
|
289 |
"step": 1000
|
290 |
},
|
291 |
{
|
292 |
"epoch": 0.32,
|
293 |
-
"grad_norm":
|
294 |
-
"learning_rate": 3.
|
295 |
-
"loss": 0.
|
296 |
"step": 1025
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.33,
|
300 |
-
"grad_norm": 2.
|
301 |
-
"learning_rate": 3.
|
302 |
-
"loss": 0.
|
303 |
"step": 1050
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.34,
|
307 |
-
"grad_norm": 0.
|
308 |
-
"learning_rate": 3.
|
309 |
-
"loss": 0.
|
310 |
"step": 1075
|
311 |
},
|
312 |
{
|
313 |
"epoch": 0.34,
|
314 |
-
"grad_norm":
|
315 |
-
"learning_rate": 3.
|
316 |
-
"loss": 0.
|
317 |
"step": 1100
|
318 |
},
|
319 |
{
|
320 |
"epoch": 0.35,
|
321 |
-
"grad_norm": 0.
|
322 |
-
"learning_rate": 3.
|
323 |
-
"loss": 0.
|
324 |
"step": 1125
|
325 |
},
|
326 |
{
|
327 |
"epoch": 0.36,
|
328 |
-
"grad_norm":
|
329 |
-
"learning_rate": 3.
|
330 |
-
"loss": 0.
|
331 |
"step": 1150
|
332 |
},
|
333 |
{
|
334 |
"epoch": 0.37,
|
335 |
-
"grad_norm":
|
336 |
-
"learning_rate": 3.
|
337 |
-
"loss": 0.
|
338 |
"step": 1175
|
339 |
},
|
340 |
{
|
341 |
"epoch": 0.38,
|
342 |
-
"grad_norm": 0.
|
343 |
-
"learning_rate": 3.
|
344 |
-
"loss": 0.
|
345 |
"step": 1200
|
346 |
},
|
347 |
{
|
348 |
"epoch": 0.38,
|
349 |
-
"grad_norm": 0.
|
350 |
-
"learning_rate": 3.
|
351 |
-
"loss": 0.
|
352 |
"step": 1225
|
353 |
},
|
354 |
{
|
355 |
"epoch": 0.39,
|
356 |
-
"grad_norm": 1.
|
357 |
-
"learning_rate": 3.
|
358 |
-
"loss": 0.
|
359 |
"step": 1250
|
360 |
},
|
361 |
{
|
362 |
"epoch": 0.4,
|
363 |
-
"grad_norm": 1.
|
364 |
-
"learning_rate": 3.
|
365 |
-
"loss": 0.
|
366 |
"step": 1275
|
367 |
},
|
368 |
{
|
369 |
"epoch": 0.41,
|
370 |
-
"grad_norm": 0.
|
371 |
-
"learning_rate":
|
372 |
-
"loss": 0.
|
373 |
"step": 1300
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.41,
|
377 |
-
"grad_norm": 1.
|
378 |
-
"learning_rate": 2.
|
379 |
-
"loss": 0.
|
380 |
"step": 1325
|
381 |
},
|
382 |
{
|
383 |
"epoch": 0.42,
|
384 |
-
"grad_norm": 0.
|
385 |
-
"learning_rate": 2.
|
386 |
-
"loss": 0.
|
387 |
"step": 1350
|
388 |
},
|
389 |
{
|
390 |
"epoch": 0.43,
|
391 |
-
"grad_norm": 0.
|
392 |
-
"learning_rate": 2.
|
393 |
-
"loss": 0.
|
394 |
"step": 1375
|
395 |
},
|
396 |
{
|
397 |
"epoch": 0.44,
|
398 |
-
"grad_norm": 0.
|
399 |
-
"learning_rate": 2.
|
400 |
-
"loss": 0.
|
401 |
"step": 1400
|
402 |
},
|
403 |
{
|
404 |
"epoch": 0.45,
|
405 |
-
"grad_norm":
|
406 |
-
"learning_rate": 2.
|
407 |
-
"loss": 0.
|
408 |
"step": 1425
|
409 |
},
|
410 |
{
|
411 |
"epoch": 0.45,
|
412 |
-
"grad_norm": 0.
|
413 |
-
"learning_rate": 2.
|
414 |
-
"loss": 0.
|
415 |
"step": 1450
|
416 |
},
|
417 |
{
|
418 |
"epoch": 0.46,
|
419 |
-
"grad_norm": 0.
|
420 |
-
"learning_rate": 2.
|
421 |
-
"loss": 0.
|
422 |
"step": 1475
|
423 |
},
|
424 |
{
|
425 |
"epoch": 0.47,
|
426 |
-
"grad_norm": 1.
|
427 |
-
"learning_rate": 2.
|
428 |
-
"loss": 0.
|
429 |
"step": 1500
|
430 |
},
|
431 |
{
|
432 |
"epoch": 0.48,
|
433 |
-
"grad_norm":
|
434 |
-
"learning_rate": 2.
|
435 |
-
"loss": 0.
|
436 |
"step": 1525
|
437 |
},
|
438 |
{
|
439 |
"epoch": 0.49,
|
440 |
-
"grad_norm":
|
441 |
-
"learning_rate": 2.
|
442 |
"loss": 0.2046,
|
443 |
"step": 1550
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.49,
|
447 |
-
"grad_norm":
|
448 |
-
"learning_rate": 2.
|
449 |
-
"loss": 0.
|
450 |
"step": 1575
|
451 |
},
|
452 |
{
|
453 |
"epoch": 0.5,
|
454 |
-
"grad_norm": 1.
|
455 |
-
"learning_rate": 2.
|
456 |
-
"loss": 0.
|
457 |
"step": 1600
|
458 |
},
|
459 |
{
|
460 |
"epoch": 0.51,
|
461 |
-
"grad_norm":
|
462 |
-
"learning_rate": 2.
|
463 |
-
"loss": 0.
|
464 |
"step": 1625
|
465 |
},
|
466 |
{
|
467 |
"epoch": 0.52,
|
468 |
-
"grad_norm":
|
469 |
-
"learning_rate": 2.
|
470 |
-
"loss": 0.
|
471 |
"step": 1650
|
472 |
},
|
473 |
{
|
474 |
"epoch": 0.52,
|
475 |
-
"grad_norm":
|
476 |
-
"learning_rate": 2.
|
477 |
-
"loss": 0.
|
478 |
"step": 1675
|
479 |
},
|
480 |
{
|
481 |
"epoch": 0.53,
|
482 |
-
"grad_norm": 2.
|
483 |
-
"learning_rate": 2.
|
484 |
-
"loss": 0.
|
485 |
"step": 1700
|
486 |
},
|
487 |
{
|
488 |
"epoch": 0.54,
|
489 |
-
"grad_norm": 1.
|
490 |
-
"learning_rate": 2.
|
491 |
-
"loss": 0.
|
492 |
"step": 1725
|
493 |
},
|
494 |
{
|
495 |
"epoch": 0.55,
|
496 |
-
"grad_norm": 0.
|
497 |
-
"learning_rate": 2.
|
498 |
-
"loss": 0.
|
499 |
"step": 1750
|
500 |
},
|
501 |
{
|
502 |
"epoch": 0.56,
|
503 |
-
"grad_norm":
|
504 |
-
"learning_rate": 2.
|
505 |
-
"loss": 0.
|
506 |
"step": 1775
|
507 |
},
|
508 |
{
|
509 |
"epoch": 0.56,
|
510 |
-
"grad_norm": 1.
|
511 |
-
"learning_rate": 2.
|
512 |
-
"loss": 0.
|
513 |
"step": 1800
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.57,
|
517 |
-
"grad_norm": 0.
|
518 |
-
"learning_rate": 2.
|
519 |
-
"loss": 0.
|
520 |
"step": 1825
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.58,
|
524 |
-
"grad_norm":
|
525 |
-
"learning_rate": 2.
|
526 |
-
"loss": 0.
|
527 |
"step": 1850
|
528 |
},
|
529 |
{
|
530 |
"epoch": 0.59,
|
531 |
-
"grad_norm":
|
532 |
-
"learning_rate": 2.
|
533 |
-
"loss": 0.
|
534 |
"step": 1875
|
535 |
},
|
536 |
{
|
537 |
"epoch": 0.6,
|
538 |
-
"grad_norm":
|
539 |
-
"learning_rate": 2.
|
540 |
-
"loss": 0.
|
541 |
"step": 1900
|
542 |
},
|
543 |
{
|
544 |
"epoch": 0.6,
|
545 |
-
"grad_norm": 0.
|
546 |
-
"learning_rate": 2.
|
547 |
-
"loss": 0.
|
548 |
"step": 1925
|
549 |
},
|
550 |
{
|
551 |
"epoch": 0.61,
|
552 |
-
"grad_norm": 1.
|
553 |
-
"learning_rate":
|
554 |
-
"loss": 0.
|
555 |
"step": 1950
|
556 |
},
|
557 |
{
|
558 |
"epoch": 0.62,
|
559 |
-
"grad_norm": 2.
|
560 |
-
"learning_rate": 1.
|
561 |
-
"loss": 0.
|
562 |
"step": 1975
|
563 |
},
|
564 |
{
|
565 |
"epoch": 0.63,
|
566 |
-
"grad_norm":
|
567 |
-
"learning_rate": 1.
|
568 |
-
"loss": 0.
|
569 |
"step": 2000
|
570 |
},
|
571 |
{
|
572 |
"epoch": 0.63,
|
573 |
-
"grad_norm": 0.
|
574 |
-
"learning_rate": 1.
|
575 |
-
"loss": 0.
|
576 |
"step": 2025
|
577 |
},
|
578 |
{
|
579 |
"epoch": 0.64,
|
580 |
-
"grad_norm":
|
581 |
-
"learning_rate": 1.
|
582 |
-
"loss": 0.
|
583 |
"step": 2050
|
584 |
},
|
585 |
{
|
586 |
"epoch": 0.65,
|
587 |
-
"grad_norm": 1.
|
588 |
-
"learning_rate": 1.
|
589 |
-
"loss": 0.
|
590 |
"step": 2075
|
591 |
},
|
592 |
{
|
593 |
"epoch": 0.66,
|
594 |
-
"grad_norm":
|
595 |
-
"learning_rate": 1.
|
596 |
-
"loss": 0.
|
597 |
"step": 2100
|
598 |
},
|
599 |
{
|
600 |
"epoch": 0.67,
|
601 |
-
"grad_norm": 2.
|
602 |
-
"learning_rate": 1.
|
603 |
-
"loss": 0.
|
604 |
"step": 2125
|
605 |
},
|
606 |
{
|
607 |
"epoch": 0.67,
|
608 |
-
"grad_norm": 1.
|
609 |
-
"learning_rate": 1.
|
610 |
-
"loss": 0.
|
611 |
"step": 2150
|
612 |
},
|
613 |
{
|
614 |
"epoch": 0.68,
|
615 |
-
"grad_norm": 2.
|
616 |
-
"learning_rate": 1.
|
617 |
"loss": 0.1887,
|
618 |
"step": 2175
|
619 |
},
|
620 |
{
|
621 |
"epoch": 0.69,
|
622 |
-
"grad_norm": 2.
|
623 |
-
"learning_rate": 1.
|
624 |
-
"loss": 0.
|
625 |
"step": 2200
|
626 |
},
|
627 |
{
|
628 |
"epoch": 0.7,
|
629 |
-
"grad_norm": 2.
|
630 |
-
"learning_rate": 1.
|
631 |
-
"loss": 0.
|
632 |
"step": 2225
|
633 |
},
|
634 |
{
|
635 |
"epoch": 0.7,
|
636 |
-
"grad_norm":
|
637 |
-
"learning_rate": 1.
|
638 |
-
"loss": 0.
|
639 |
"step": 2250
|
640 |
},
|
641 |
{
|
642 |
"epoch": 0.71,
|
643 |
-
"grad_norm":
|
644 |
-
"learning_rate": 1.
|
645 |
-
"loss": 0.
|
646 |
"step": 2275
|
647 |
},
|
648 |
{
|
649 |
"epoch": 0.72,
|
650 |
-
"grad_norm": 0.
|
651 |
-
"learning_rate": 1.
|
652 |
-
"loss": 0.
|
653 |
"step": 2300
|
654 |
},
|
655 |
{
|
656 |
"epoch": 0.73,
|
657 |
-
"grad_norm":
|
658 |
-
"learning_rate": 1.
|
659 |
-
"loss": 0.
|
660 |
"step": 2325
|
661 |
},
|
662 |
{
|
663 |
"epoch": 0.74,
|
664 |
-
"grad_norm": 0.
|
665 |
-
"learning_rate": 1.
|
666 |
-
"loss": 0.
|
667 |
"step": 2350
|
668 |
},
|
669 |
{
|
670 |
"epoch": 0.74,
|
671 |
-
"grad_norm":
|
672 |
-
"learning_rate": 1.
|
673 |
-
"loss": 0.
|
674 |
"step": 2375
|
675 |
},
|
676 |
{
|
677 |
"epoch": 0.75,
|
678 |
-
"grad_norm": 0.
|
679 |
-
"learning_rate": 1.
|
680 |
-
"loss": 0.
|
681 |
"step": 2400
|
682 |
},
|
683 |
{
|
684 |
"epoch": 0.76,
|
685 |
-
"grad_norm": 0.
|
686 |
-
"learning_rate": 1.
|
687 |
-
"loss": 0.
|
688 |
"step": 2425
|
689 |
},
|
690 |
{
|
691 |
"epoch": 0.77,
|
692 |
-
"grad_norm": 0.
|
693 |
-
"learning_rate": 1.
|
694 |
-
"loss": 0.
|
695 |
"step": 2450
|
696 |
},
|
697 |
{
|
698 |
"epoch": 0.78,
|
699 |
-
"grad_norm":
|
700 |
-
"learning_rate": 1.
|
701 |
-
"loss": 0.
|
702 |
"step": 2475
|
703 |
},
|
704 |
{
|
705 |
"epoch": 0.78,
|
706 |
-
"grad_norm": 1.
|
707 |
-
"learning_rate": 1.
|
708 |
-
"loss": 0.
|
709 |
"step": 2500
|
710 |
},
|
711 |
{
|
712 |
"epoch": 0.79,
|
713 |
-
"grad_norm":
|
714 |
-
"learning_rate": 1.
|
715 |
-
"loss": 0.
|
716 |
"step": 2525
|
717 |
},
|
718 |
{
|
719 |
"epoch": 0.8,
|
720 |
-
"grad_norm": 3.
|
721 |
-
"learning_rate": 1.
|
722 |
-
"loss": 0.
|
723 |
"step": 2550
|
724 |
},
|
725 |
{
|
726 |
"epoch": 0.81,
|
727 |
-
"grad_norm": 1.
|
728 |
-
"learning_rate": 1.
|
729 |
-
"loss": 0.
|
730 |
"step": 2575
|
731 |
},
|
732 |
{
|
733 |
"epoch": 0.81,
|
734 |
-
"grad_norm": 1.
|
735 |
-
"learning_rate":
|
736 |
-
"loss": 0.
|
737 |
"step": 2600
|
738 |
},
|
739 |
{
|
740 |
"epoch": 0.82,
|
741 |
-
"grad_norm": 1.
|
742 |
-
"learning_rate": 9.
|
743 |
-
"loss": 0.
|
744 |
"step": 2625
|
745 |
},
|
746 |
{
|
747 |
"epoch": 0.83,
|
748 |
-
"grad_norm":
|
749 |
-
"learning_rate":
|
750 |
-
"loss": 0.
|
751 |
"step": 2650
|
752 |
},
|
753 |
{
|
754 |
"epoch": 0.84,
|
755 |
-
"grad_norm": 2.
|
756 |
-
"learning_rate": 8.
|
757 |
-
"loss": 0.
|
758 |
"step": 2675
|
759 |
},
|
760 |
{
|
761 |
"epoch": 0.85,
|
762 |
-
"grad_norm": 0.
|
763 |
-
"learning_rate": 8.
|
764 |
-
"loss": 0.
|
765 |
"step": 2700
|
766 |
},
|
767 |
{
|
768 |
"epoch": 0.85,
|
769 |
-
"grad_norm":
|
770 |
-
"learning_rate":
|
771 |
-
"loss": 0.
|
772 |
"step": 2725
|
773 |
},
|
774 |
{
|
775 |
"epoch": 0.86,
|
776 |
-
"grad_norm": 0.
|
777 |
-
"learning_rate": 7.
|
778 |
-
"loss": 0.
|
779 |
"step": 2750
|
780 |
},
|
781 |
{
|
782 |
"epoch": 0.87,
|
783 |
-
"grad_norm": 2.
|
784 |
-
"learning_rate": 7.
|
785 |
-
"loss": 0.
|
786 |
"step": 2775
|
787 |
},
|
788 |
{
|
789 |
"epoch": 0.88,
|
790 |
-
"grad_norm":
|
791 |
-
"learning_rate":
|
792 |
-
"loss": 0.
|
793 |
"step": 2800
|
794 |
},
|
795 |
{
|
796 |
"epoch": 0.88,
|
797 |
-
"grad_norm": 2.
|
798 |
-
"learning_rate": 6.
|
799 |
-
"loss": 0.
|
800 |
"step": 2825
|
801 |
},
|
802 |
{
|
803 |
"epoch": 0.89,
|
804 |
-
"grad_norm":
|
805 |
-
"learning_rate":
|
806 |
-
"loss": 0.
|
807 |
"step": 2850
|
808 |
},
|
809 |
{
|
810 |
"epoch": 0.9,
|
811 |
-
"grad_norm": 2.
|
812 |
-
"learning_rate": 5.
|
813 |
-
"loss": 0.
|
814 |
"step": 2875
|
815 |
},
|
816 |
{
|
817 |
"epoch": 0.91,
|
818 |
-
"grad_norm": 2.
|
819 |
-
"learning_rate": 5.
|
820 |
-
"loss": 0.
|
821 |
"step": 2900
|
822 |
},
|
823 |
{
|
824 |
"epoch": 0.92,
|
825 |
-
"grad_norm": 0.
|
826 |
-
"learning_rate":
|
827 |
-
"loss": 0.
|
828 |
"step": 2925
|
829 |
},
|
830 |
{
|
831 |
"epoch": 0.92,
|
832 |
-
"grad_norm": 2.
|
833 |
-
"learning_rate": 4.
|
834 |
-
"loss": 0.
|
835 |
"step": 2950
|
836 |
},
|
837 |
{
|
838 |
"epoch": 0.93,
|
839 |
-
"grad_norm": 0.
|
840 |
-
"learning_rate":
|
841 |
-
"loss": 0.
|
842 |
"step": 2975
|
843 |
},
|
844 |
{
|
845 |
"epoch": 0.94,
|
846 |
-
"grad_norm": 0.
|
847 |
-
"learning_rate": 3.
|
848 |
-
"loss": 0.
|
849 |
"step": 3000
|
850 |
},
|
851 |
{
|
852 |
"epoch": 0.95,
|
853 |
-
"grad_norm":
|
854 |
-
"learning_rate": 3.
|
855 |
-
"loss": 0.
|
856 |
"step": 3025
|
857 |
},
|
858 |
{
|
859 |
"epoch": 0.96,
|
860 |
-
"grad_norm": 3.
|
861 |
-
"learning_rate":
|
862 |
-
"loss": 0.
|
863 |
"step": 3050
|
864 |
},
|
865 |
{
|
866 |
"epoch": 0.96,
|
867 |
-
"grad_norm": 2.
|
868 |
-
"learning_rate": 2.
|
869 |
"loss": 0.1839,
|
870 |
"step": 3075
|
871 |
},
|
872 |
{
|
873 |
"epoch": 0.97,
|
874 |
-
"grad_norm": 0.
|
875 |
-
"learning_rate":
|
876 |
-
"loss": 0.
|
877 |
"step": 3100
|
878 |
},
|
879 |
{
|
880 |
"epoch": 0.98,
|
881 |
-
"grad_norm":
|
882 |
-
"learning_rate":
|
883 |
-
"loss": 0.
|
884 |
"step": 3125
|
885 |
},
|
886 |
{
|
887 |
"epoch": 0.99,
|
888 |
-
"grad_norm": 0.
|
889 |
-
"learning_rate": 1.
|
890 |
-
"loss": 0.
|
891 |
"step": 3150
|
892 |
},
|
893 |
{
|
894 |
"epoch": 0.99,
|
895 |
-
"grad_norm": 0.
|
896 |
-
"learning_rate":
|
897 |
-
"loss": 0.
|
898 |
"step": 3175
|
899 |
},
|
900 |
{
|
901 |
"epoch": 1.0,
|
902 |
-
"grad_norm": 0.
|
903 |
-
"learning_rate":
|
904 |
-
"loss": 0.
|
905 |
"step": 3200
|
906 |
},
|
907 |
{
|
908 |
"epoch": 1.01,
|
909 |
-
"grad_norm":
|
910 |
-
"learning_rate":
|
911 |
-
"loss": 0.
|
912 |
"step": 3225
|
913 |
},
|
914 |
{
|
915 |
-
"epoch": 1.
|
916 |
-
"
|
917 |
-
"
|
918 |
-
"
|
919 |
-
"
|
920 |
-
|
921 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
922 |
}
|
923 |
],
|
924 |
"logging_steps": 25,
|
925 |
-
"max_steps":
|
926 |
"num_input_tokens_seen": 0,
|
927 |
"num_train_epochs": 2,
|
928 |
"save_steps": 100,
|
929 |
-
"total_flos":
|
930 |
"train_batch_size": 16,
|
931 |
"trial_name": null,
|
932 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.0200438459129346,
|
5 |
"eval_steps": 100,
|
6 |
+
"global_step": 3257,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.01,
|
13 |
+
"grad_norm": 0.6797587275505066,
|
14 |
+
"learning_rate": 4.961621123733498e-05,
|
15 |
+
"loss": 0.5293,
|
16 |
"step": 25
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.02,
|
20 |
+
"grad_norm": 0.559562087059021,
|
21 |
+
"learning_rate": 4.9232422474669945e-05,
|
22 |
+
"loss": 0.3908,
|
23 |
"step": 50
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.02,
|
27 |
+
"grad_norm": 0.6713371872901917,
|
28 |
+
"learning_rate": 4.884863371200491e-05,
|
29 |
+
"loss": 0.4101,
|
30 |
"step": 75
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.03,
|
34 |
+
"grad_norm": 2.220188617706299,
|
35 |
+
"learning_rate": 4.846484494933989e-05,
|
36 |
+
"loss": 0.3246,
|
37 |
"step": 100
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.04,
|
41 |
+
"grad_norm": 0.7794200778007507,
|
42 |
+
"learning_rate": 4.8081056186674856e-05,
|
43 |
+
"loss": 0.3521,
|
44 |
"step": 125
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.05,
|
48 |
+
"grad_norm": 2.3898696899414062,
|
49 |
+
"learning_rate": 4.769726742400983e-05,
|
50 |
+
"loss": 0.338,
|
51 |
"step": 150
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.05,
|
55 |
+
"grad_norm": 1.896946668624878,
|
56 |
+
"learning_rate": 4.731347866134479e-05,
|
57 |
+
"loss": 0.3142,
|
58 |
"step": 175
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.06,
|
62 |
+
"grad_norm": 0.35194501280784607,
|
63 |
+
"learning_rate": 4.692968989867977e-05,
|
64 |
+
"loss": 0.3061,
|
65 |
"step": 200
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.07,
|
69 |
+
"grad_norm": 0.6125898957252502,
|
70 |
+
"learning_rate": 4.654590113601474e-05,
|
71 |
"loss": 0.3291,
|
72 |
"step": 225
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.08,
|
76 |
+
"grad_norm": 0.36999431252479553,
|
77 |
+
"learning_rate": 4.616211237334971e-05,
|
78 |
+
"loss": 0.3392,
|
79 |
"step": 250
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.09,
|
83 |
+
"grad_norm": 0.5263044834136963,
|
84 |
+
"learning_rate": 4.5793675161191285e-05,
|
85 |
+
"loss": 0.3102,
|
86 |
"step": 275
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.09,
|
90 |
+
"grad_norm": 0.3138100802898407,
|
91 |
+
"learning_rate": 4.542523794903286e-05,
|
92 |
+
"loss": 0.3065,
|
93 |
"step": 300
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.1,
|
97 |
+
"grad_norm": 0.38833898305892944,
|
98 |
+
"learning_rate": 4.504144918636782e-05,
|
99 |
+
"loss": 0.3586,
|
100 |
"step": 325
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.11,
|
104 |
+
"grad_norm": 2.349430799484253,
|
105 |
+
"learning_rate": 4.4657660423702795e-05,
|
106 |
+
"loss": 0.3072,
|
107 |
"step": 350
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.12,
|
111 |
+
"grad_norm": 0.3345746397972107,
|
112 |
+
"learning_rate": 4.427387166103777e-05,
|
113 |
+
"loss": 0.2855,
|
114 |
"step": 375
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.13,
|
118 |
+
"grad_norm": 0.46446025371551514,
|
119 |
+
"learning_rate": 4.390543444887934e-05,
|
120 |
+
"loss": 0.2852,
|
121 |
"step": 400
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.13,
|
125 |
+
"grad_norm": 0.6488687992095947,
|
126 |
+
"learning_rate": 4.3521645686214306e-05,
|
127 |
+
"loss": 0.2731,
|
128 |
"step": 425
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.14,
|
132 |
+
"grad_norm": 0.7198542356491089,
|
133 |
+
"learning_rate": 4.313785692354928e-05,
|
134 |
+
"loss": 0.2986,
|
135 |
"step": 450
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.15,
|
139 |
+
"grad_norm": 1.6638888120651245,
|
140 |
+
"learning_rate": 4.2754068160884256e-05,
|
141 |
+
"loss": 0.2985,
|
142 |
"step": 475
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.16,
|
146 |
+
"grad_norm": 1.4277877807617188,
|
147 |
+
"learning_rate": 4.237027939821922e-05,
|
148 |
+
"loss": 0.2559,
|
149 |
"step": 500
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.16,
|
153 |
+
"grad_norm": 3.699174642562866,
|
154 |
+
"learning_rate": 4.198649063555419e-05,
|
155 |
+
"loss": 0.2731,
|
156 |
"step": 525
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.17,
|
160 |
+
"grad_norm": 1.9921191930770874,
|
161 |
+
"learning_rate": 4.160270187288917e-05,
|
162 |
+
"loss": 0.242,
|
163 |
"step": 550
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.18,
|
167 |
+
"grad_norm": 0.3546009361743927,
|
168 |
+
"learning_rate": 4.1218913110224135e-05,
|
169 |
+
"loss": 0.253,
|
170 |
"step": 575
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.19,
|
174 |
+
"grad_norm": 2.4208102226257324,
|
175 |
+
"learning_rate": 4.08351243475591e-05,
|
176 |
+
"loss": 0.2734,
|
177 |
"step": 600
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.2,
|
181 |
+
"grad_norm": 1.4212377071380615,
|
182 |
+
"learning_rate": 4.045133558489408e-05,
|
183 |
+
"loss": 0.2426,
|
184 |
"step": 625
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.2,
|
188 |
+
"grad_norm": 3.6487390995025635,
|
189 |
+
"learning_rate": 4.0067546822229046e-05,
|
190 |
+
"loss": 0.2798,
|
191 |
"step": 650
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.21,
|
195 |
+
"grad_norm": 2.001326322555542,
|
196 |
+
"learning_rate": 3.968375805956402e-05,
|
197 |
+
"loss": 0.2599,
|
198 |
"step": 675
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.22,
|
202 |
+
"grad_norm": 1.6598902940750122,
|
203 |
+
"learning_rate": 3.929996929689899e-05,
|
204 |
+
"loss": 0.2578,
|
205 |
"step": 700
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.23,
|
209 |
+
"grad_norm": 1.3611884117126465,
|
210 |
+
"learning_rate": 3.8916180534233957e-05,
|
211 |
+
"loss": 0.2766,
|
212 |
"step": 725
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.23,
|
216 |
+
"grad_norm": 0.8574317097663879,
|
217 |
+
"learning_rate": 3.853239177156893e-05,
|
218 |
+
"loss": 0.2417,
|
219 |
"step": 750
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.24,
|
223 |
+
"grad_norm": 0.8539307713508606,
|
224 |
+
"learning_rate": 3.81486030089039e-05,
|
225 |
+
"loss": 0.223,
|
226 |
"step": 775
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.25,
|
230 |
+
"grad_norm": 1.1913343667984009,
|
231 |
+
"learning_rate": 3.7764814246238874e-05,
|
232 |
+
"loss": 0.2602,
|
233 |
"step": 800
|
234 |
},
|
235 |
{
|
236 |
"epoch": 0.26,
|
237 |
+
"grad_norm": 0.5884910821914673,
|
238 |
+
"learning_rate": 3.738102548357384e-05,
|
239 |
+
"loss": 0.2747,
|
240 |
"step": 825
|
241 |
},
|
242 |
{
|
243 |
"epoch": 0.27,
|
244 |
+
"grad_norm": 0.9413720965385437,
|
245 |
+
"learning_rate": 3.699723672090881e-05,
|
246 |
+
"loss": 0.1835,
|
247 |
"step": 850
|
248 |
},
|
249 |
{
|
250 |
"epoch": 0.27,
|
251 |
+
"grad_norm": 2.626718521118164,
|
252 |
+
"learning_rate": 3.6613447958243785e-05,
|
253 |
+
"loss": 0.2883,
|
254 |
"step": 875
|
255 |
},
|
256 |
{
|
257 |
"epoch": 0.28,
|
258 |
+
"grad_norm": 1.573789119720459,
|
259 |
+
"learning_rate": 3.622965919557876e-05,
|
260 |
+
"loss": 0.252,
|
261 |
"step": 900
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.29,
|
265 |
+
"grad_norm": 0.6372594833374023,
|
266 |
+
"learning_rate": 3.584587043291373e-05,
|
267 |
+
"loss": 0.2624,
|
268 |
"step": 925
|
269 |
},
|
270 |
{
|
271 |
"epoch": 0.3,
|
272 |
+
"grad_norm": 0.4254266619682312,
|
273 |
+
"learning_rate": 3.5462081670248696e-05,
|
274 |
+
"loss": 0.2012,
|
275 |
"step": 950
|
276 |
},
|
277 |
{
|
278 |
"epoch": 0.31,
|
279 |
+
"grad_norm": 1.2674840688705444,
|
280 |
+
"learning_rate": 3.5078292907583664e-05,
|
281 |
+
"loss": 0.2576,
|
282 |
"step": 975
|
283 |
},
|
284 |
{
|
285 |
"epoch": 0.31,
|
286 |
+
"grad_norm": 0.6976171731948853,
|
287 |
+
"learning_rate": 3.469450414491864e-05,
|
288 |
+
"loss": 0.2374,
|
289 |
"step": 1000
|
290 |
},
|
291 |
{
|
292 |
"epoch": 0.32,
|
293 |
+
"grad_norm": 2.898336410522461,
|
294 |
+
"learning_rate": 3.4310715382253614e-05,
|
295 |
+
"loss": 0.2168,
|
296 |
"step": 1025
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.33,
|
300 |
+
"grad_norm": 2.6241133213043213,
|
301 |
+
"learning_rate": 3.3926926619588575e-05,
|
302 |
+
"loss": 0.2721,
|
303 |
"step": 1050
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.34,
|
307 |
+
"grad_norm": 0.452421635389328,
|
308 |
+
"learning_rate": 3.354313785692355e-05,
|
309 |
+
"loss": 0.1887,
|
310 |
"step": 1075
|
311 |
},
|
312 |
{
|
313 |
"epoch": 0.34,
|
314 |
+
"grad_norm": 4.101943492889404,
|
315 |
+
"learning_rate": 3.3159349094258525e-05,
|
316 |
+
"loss": 0.2597,
|
317 |
"step": 1100
|
318 |
},
|
319 |
{
|
320 |
"epoch": 0.35,
|
321 |
+
"grad_norm": 0.8068587779998779,
|
322 |
+
"learning_rate": 3.277556033159349e-05,
|
323 |
+
"loss": 0.1803,
|
324 |
"step": 1125
|
325 |
},
|
326 |
{
|
327 |
"epoch": 0.36,
|
328 |
+
"grad_norm": 3.24800705909729,
|
329 |
+
"learning_rate": 3.239177156892847e-05,
|
330 |
+
"loss": 0.2175,
|
331 |
"step": 1150
|
332 |
},
|
333 |
{
|
334 |
"epoch": 0.37,
|
335 |
+
"grad_norm": 3.0881152153015137,
|
336 |
+
"learning_rate": 3.2007982806263435e-05,
|
337 |
+
"loss": 0.1883,
|
338 |
"step": 1175
|
339 |
},
|
340 |
{
|
341 |
"epoch": 0.38,
|
342 |
+
"grad_norm": 0.5309311151504517,
|
343 |
+
"learning_rate": 3.1624194043598403e-05,
|
344 |
+
"loss": 0.243,
|
345 |
"step": 1200
|
346 |
},
|
347 |
{
|
348 |
"epoch": 0.38,
|
349 |
+
"grad_norm": 0.8343615531921387,
|
350 |
+
"learning_rate": 3.124040528093338e-05,
|
351 |
+
"loss": 0.198,
|
352 |
"step": 1225
|
353 |
},
|
354 |
{
|
355 |
"epoch": 0.39,
|
356 |
+
"grad_norm": 1.199294924736023,
|
357 |
+
"learning_rate": 3.0856616518268346e-05,
|
358 |
+
"loss": 0.2609,
|
359 |
"step": 1250
|
360 |
},
|
361 |
{
|
362 |
"epoch": 0.4,
|
363 |
+
"grad_norm": 1.7239773273468018,
|
364 |
+
"learning_rate": 3.0472827755603318e-05,
|
365 |
+
"loss": 0.2184,
|
366 |
"step": 1275
|
367 |
},
|
368 |
{
|
369 |
"epoch": 0.41,
|
370 |
+
"grad_norm": 0.6922329664230347,
|
371 |
+
"learning_rate": 3.008903899293829e-05,
|
372 |
+
"loss": 0.2026,
|
373 |
"step": 1300
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.41,
|
377 |
+
"grad_norm": 1.3839455842971802,
|
378 |
+
"learning_rate": 2.9705250230273257e-05,
|
379 |
+
"loss": 0.2365,
|
380 |
"step": 1325
|
381 |
},
|
382 |
{
|
383 |
"epoch": 0.42,
|
384 |
+
"grad_norm": 0.3912523686885834,
|
385 |
+
"learning_rate": 2.932146146760823e-05,
|
386 |
+
"loss": 0.2228,
|
387 |
"step": 1350
|
388 |
},
|
389 |
{
|
390 |
"epoch": 0.43,
|
391 |
+
"grad_norm": 0.7774330973625183,
|
392 |
+
"learning_rate": 2.8937672704943203e-05,
|
393 |
+
"loss": 0.1836,
|
394 |
"step": 1375
|
395 |
},
|
396 |
{
|
397 |
"epoch": 0.44,
|
398 |
+
"grad_norm": 0.7025958895683289,
|
399 |
+
"learning_rate": 2.855388394227817e-05,
|
400 |
+
"loss": 0.2092,
|
401 |
"step": 1400
|
402 |
},
|
403 |
{
|
404 |
"epoch": 0.45,
|
405 |
+
"grad_norm": 3.789003849029541,
|
406 |
+
"learning_rate": 2.8170095179613143e-05,
|
407 |
+
"loss": 0.217,
|
408 |
"step": 1425
|
409 |
},
|
410 |
{
|
411 |
"epoch": 0.45,
|
412 |
+
"grad_norm": 0.5001412034034729,
|
413 |
+
"learning_rate": 2.7786306416948114e-05,
|
414 |
+
"loss": 0.1771,
|
415 |
"step": 1450
|
416 |
},
|
417 |
{
|
418 |
"epoch": 0.46,
|
419 |
+
"grad_norm": 0.585847795009613,
|
420 |
+
"learning_rate": 2.7402517654283082e-05,
|
421 |
+
"loss": 0.1899,
|
422 |
"step": 1475
|
423 |
},
|
424 |
{
|
425 |
"epoch": 0.47,
|
426 |
+
"grad_norm": 1.6145225763320923,
|
427 |
+
"learning_rate": 2.7018728891618057e-05,
|
428 |
+
"loss": 0.214,
|
429 |
"step": 1500
|
430 |
},
|
431 |
{
|
432 |
"epoch": 0.48,
|
433 |
+
"grad_norm": 2.724287748336792,
|
434 |
+
"learning_rate": 2.6634940128953022e-05,
|
435 |
+
"loss": 0.1857,
|
436 |
"step": 1525
|
437 |
},
|
438 |
{
|
439 |
"epoch": 0.49,
|
440 |
+
"grad_norm": 1.1571438312530518,
|
441 |
+
"learning_rate": 2.6251151366287997e-05,
|
442 |
"loss": 0.2046,
|
443 |
"step": 1550
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.49,
|
447 |
+
"grad_norm": 3.0686943531036377,
|
448 |
+
"learning_rate": 2.5867362603622968e-05,
|
449 |
+
"loss": 0.1841,
|
450 |
"step": 1575
|
451 |
},
|
452 |
{
|
453 |
"epoch": 0.5,
|
454 |
+
"grad_norm": 1.6015386581420898,
|
455 |
+
"learning_rate": 2.5483573840957936e-05,
|
456 |
+
"loss": 0.2095,
|
457 |
"step": 1600
|
458 |
},
|
459 |
{
|
460 |
"epoch": 0.51,
|
461 |
+
"grad_norm": 1.5645371675491333,
|
462 |
+
"learning_rate": 2.5099785078292907e-05,
|
463 |
+
"loss": 0.193,
|
464 |
"step": 1625
|
465 |
},
|
466 |
{
|
467 |
"epoch": 0.52,
|
468 |
+
"grad_norm": 1.0582934617996216,
|
469 |
+
"learning_rate": 2.471599631562788e-05,
|
470 |
+
"loss": 0.2065,
|
471 |
"step": 1650
|
472 |
},
|
473 |
{
|
474 |
"epoch": 0.52,
|
475 |
+
"grad_norm": 4.154886722564697,
|
476 |
+
"learning_rate": 2.433220755296285e-05,
|
477 |
+
"loss": 0.1955,
|
478 |
"step": 1675
|
479 |
},
|
480 |
{
|
481 |
"epoch": 0.53,
|
482 |
+
"grad_norm": 2.047689437866211,
|
483 |
+
"learning_rate": 2.3948418790297822e-05,
|
484 |
+
"loss": 0.1999,
|
485 |
"step": 1700
|
486 |
},
|
487 |
{
|
488 |
"epoch": 0.54,
|
489 |
+
"grad_norm": 1.7541192770004272,
|
490 |
+
"learning_rate": 2.3564630027632793e-05,
|
491 |
+
"loss": 0.1906,
|
492 |
"step": 1725
|
493 |
},
|
494 |
{
|
495 |
"epoch": 0.55,
|
496 |
+
"grad_norm": 0.8929910063743591,
|
497 |
+
"learning_rate": 2.318084126496776e-05,
|
498 |
+
"loss": 0.2156,
|
499 |
"step": 1750
|
500 |
},
|
501 |
{
|
502 |
"epoch": 0.56,
|
503 |
+
"grad_norm": 2.3634512424468994,
|
504 |
+
"learning_rate": 2.2797052502302736e-05,
|
505 |
+
"loss": 0.251,
|
506 |
"step": 1775
|
507 |
},
|
508 |
{
|
509 |
"epoch": 0.56,
|
510 |
+
"grad_norm": 1.0741249322891235,
|
511 |
+
"learning_rate": 2.2413263739637704e-05,
|
512 |
+
"loss": 0.2221,
|
513 |
"step": 1800
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.57,
|
517 |
+
"grad_norm": 0.6083930134773254,
|
518 |
+
"learning_rate": 2.2029474976972675e-05,
|
519 |
+
"loss": 0.1922,
|
520 |
"step": 1825
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.58,
|
524 |
+
"grad_norm": 3.046408176422119,
|
525 |
+
"learning_rate": 2.1645686214307644e-05,
|
526 |
+
"loss": 0.1634,
|
527 |
"step": 1850
|
528 |
},
|
529 |
{
|
530 |
"epoch": 0.59,
|
531 |
+
"grad_norm": 1.0440948009490967,
|
532 |
+
"learning_rate": 2.126189745164262e-05,
|
533 |
+
"loss": 0.1798,
|
534 |
"step": 1875
|
535 |
},
|
536 |
{
|
537 |
"epoch": 0.6,
|
538 |
+
"grad_norm": 2.302553176879883,
|
539 |
+
"learning_rate": 2.0878108688977586e-05,
|
540 |
+
"loss": 0.179,
|
541 |
"step": 1900
|
542 |
},
|
543 |
{
|
544 |
"epoch": 0.6,
|
545 |
+
"grad_norm": 0.5894582867622375,
|
546 |
+
"learning_rate": 2.0494319926312558e-05,
|
547 |
+
"loss": 0.2181,
|
548 |
"step": 1925
|
549 |
},
|
550 |
{
|
551 |
"epoch": 0.61,
|
552 |
+
"grad_norm": 1.7287715673446655,
|
553 |
+
"learning_rate": 2.011053116364753e-05,
|
554 |
+
"loss": 0.1996,
|
555 |
"step": 1950
|
556 |
},
|
557 |
{
|
558 |
"epoch": 0.62,
|
559 |
+
"grad_norm": 2.5236544609069824,
|
560 |
+
"learning_rate": 1.97267424009825e-05,
|
561 |
+
"loss": 0.2366,
|
562 |
"step": 1975
|
563 |
},
|
564 |
{
|
565 |
"epoch": 0.63,
|
566 |
+
"grad_norm": 1.8819808959960938,
|
567 |
+
"learning_rate": 1.9342953638317472e-05,
|
568 |
+
"loss": 0.1906,
|
569 |
"step": 2000
|
570 |
},
|
571 |
{
|
572 |
"epoch": 0.63,
|
573 |
+
"grad_norm": 0.5952481031417847,
|
574 |
+
"learning_rate": 1.895916487565244e-05,
|
575 |
+
"loss": 0.1749,
|
576 |
"step": 2025
|
577 |
},
|
578 |
{
|
579 |
"epoch": 0.64,
|
580 |
+
"grad_norm": 0.9825776815414429,
|
581 |
+
"learning_rate": 1.8575376112987415e-05,
|
582 |
+
"loss": 0.2179,
|
583 |
"step": 2050
|
584 |
},
|
585 |
{
|
586 |
"epoch": 0.65,
|
587 |
+
"grad_norm": 1.5762274265289307,
|
588 |
+
"learning_rate": 1.8191587350322383e-05,
|
589 |
+
"loss": 0.169,
|
590 |
"step": 2075
|
591 |
},
|
592 |
{
|
593 |
"epoch": 0.66,
|
594 |
+
"grad_norm": 1.433192253112793,
|
595 |
+
"learning_rate": 1.7807798587657354e-05,
|
596 |
+
"loss": 0.1677,
|
597 |
"step": 2100
|
598 |
},
|
599 |
{
|
600 |
"epoch": 0.67,
|
601 |
+
"grad_norm": 2.5697052478790283,
|
602 |
+
"learning_rate": 1.7424009824992322e-05,
|
603 |
+
"loss": 0.2337,
|
604 |
"step": 2125
|
605 |
},
|
606 |
{
|
607 |
"epoch": 0.67,
|
608 |
+
"grad_norm": 1.336775541305542,
|
609 |
+
"learning_rate": 1.7040221062327297e-05,
|
610 |
+
"loss": 0.2168,
|
611 |
"step": 2150
|
612 |
},
|
613 |
{
|
614 |
"epoch": 0.68,
|
615 |
+
"grad_norm": 2.746216058731079,
|
616 |
+
"learning_rate": 1.6656432299662265e-05,
|
617 |
"loss": 0.1887,
|
618 |
"step": 2175
|
619 |
},
|
620 |
{
|
621 |
"epoch": 0.69,
|
622 |
+
"grad_norm": 2.6653735637664795,
|
623 |
+
"learning_rate": 1.6272643536997237e-05,
|
624 |
+
"loss": 0.2085,
|
625 |
"step": 2200
|
626 |
},
|
627 |
{
|
628 |
"epoch": 0.7,
|
629 |
+
"grad_norm": 2.748147487640381,
|
630 |
+
"learning_rate": 1.5888854774332208e-05,
|
631 |
+
"loss": 0.1927,
|
632 |
"step": 2225
|
633 |
},
|
634 |
{
|
635 |
"epoch": 0.7,
|
636 |
+
"grad_norm": 1.3852734565734863,
|
637 |
+
"learning_rate": 1.550506601166718e-05,
|
638 |
+
"loss": 0.1563,
|
639 |
"step": 2250
|
640 |
},
|
641 |
{
|
642 |
"epoch": 0.71,
|
643 |
+
"grad_norm": 3.5002951622009277,
|
644 |
+
"learning_rate": 1.512127724900215e-05,
|
645 |
+
"loss": 0.1662,
|
646 |
"step": 2275
|
647 |
},
|
648 |
{
|
649 |
"epoch": 0.72,
|
650 |
+
"grad_norm": 0.5453472137451172,
|
651 |
+
"learning_rate": 1.473748848633712e-05,
|
652 |
+
"loss": 0.1789,
|
653 |
"step": 2300
|
654 |
},
|
655 |
{
|
656 |
"epoch": 0.73,
|
657 |
+
"grad_norm": 2.8527770042419434,
|
658 |
+
"learning_rate": 1.4353699723672092e-05,
|
659 |
+
"loss": 0.2245,
|
660 |
"step": 2325
|
661 |
},
|
662 |
{
|
663 |
"epoch": 0.74,
|
664 |
+
"grad_norm": 0.5989676713943481,
|
665 |
+
"learning_rate": 1.3969910961007064e-05,
|
666 |
+
"loss": 0.1695,
|
667 |
"step": 2350
|
668 |
},
|
669 |
{
|
670 |
"epoch": 0.74,
|
671 |
+
"grad_norm": 0.6295841932296753,
|
672 |
+
"learning_rate": 1.3586122198342033e-05,
|
673 |
+
"loss": 0.1611,
|
674 |
"step": 2375
|
675 |
},
|
676 |
{
|
677 |
"epoch": 0.75,
|
678 |
+
"grad_norm": 0.4394790232181549,
|
679 |
+
"learning_rate": 1.3202333435677003e-05,
|
680 |
+
"loss": 0.2185,
|
681 |
"step": 2400
|
682 |
},
|
683 |
{
|
684 |
"epoch": 0.76,
|
685 |
+
"grad_norm": 0.9019514322280884,
|
686 |
+
"learning_rate": 1.2818544673011976e-05,
|
687 |
+
"loss": 0.1679,
|
688 |
"step": 2425
|
689 |
},
|
690 |
{
|
691 |
"epoch": 0.77,
|
692 |
+
"grad_norm": 0.7013410329818726,
|
693 |
+
"learning_rate": 1.2434755910346946e-05,
|
694 |
+
"loss": 0.1752,
|
695 |
"step": 2450
|
696 |
},
|
697 |
{
|
698 |
"epoch": 0.78,
|
699 |
+
"grad_norm": 0.7528170347213745,
|
700 |
+
"learning_rate": 1.2050967147681916e-05,
|
701 |
+
"loss": 0.1999,
|
702 |
"step": 2475
|
703 |
},
|
704 |
{
|
705 |
"epoch": 0.78,
|
706 |
+
"grad_norm": 1.6549068689346313,
|
707 |
+
"learning_rate": 1.1667178385016887e-05,
|
708 |
+
"loss": 0.1954,
|
709 |
"step": 2500
|
710 |
},
|
711 |
{
|
712 |
"epoch": 0.79,
|
713 |
+
"grad_norm": 2.296861410140991,
|
714 |
+
"learning_rate": 1.1283389622351858e-05,
|
715 |
+
"loss": 0.1999,
|
716 |
"step": 2525
|
717 |
},
|
718 |
{
|
719 |
"epoch": 0.8,
|
720 |
+
"grad_norm": 3.1202523708343506,
|
721 |
+
"learning_rate": 1.0899600859686828e-05,
|
722 |
+
"loss": 0.2417,
|
723 |
"step": 2550
|
724 |
},
|
725 |
{
|
726 |
"epoch": 0.81,
|
727 |
+
"grad_norm": 1.385256052017212,
|
728 |
+
"learning_rate": 1.05158120970218e-05,
|
729 |
+
"loss": 0.1598,
|
730 |
"step": 2575
|
731 |
},
|
732 |
{
|
733 |
"epoch": 0.81,
|
734 |
+
"grad_norm": 1.5936309099197388,
|
735 |
+
"learning_rate": 1.0132023334356771e-05,
|
736 |
+
"loss": 0.1629,
|
737 |
"step": 2600
|
738 |
},
|
739 |
{
|
740 |
"epoch": 0.82,
|
741 |
+
"grad_norm": 1.5769296884536743,
|
742 |
+
"learning_rate": 9.748234571691742e-06,
|
743 |
+
"loss": 0.1823,
|
744 |
"step": 2625
|
745 |
},
|
746 |
{
|
747 |
"epoch": 0.83,
|
748 |
+
"grad_norm": 1.9773831367492676,
|
749 |
+
"learning_rate": 9.364445809026712e-06,
|
750 |
+
"loss": 0.1851,
|
751 |
"step": 2650
|
752 |
},
|
753 |
{
|
754 |
"epoch": 0.84,
|
755 |
+
"grad_norm": 2.044830799102783,
|
756 |
+
"learning_rate": 8.980657046361684e-06,
|
757 |
+
"loss": 0.1704,
|
758 |
"step": 2675
|
759 |
},
|
760 |
{
|
761 |
"epoch": 0.85,
|
762 |
+
"grad_norm": 0.4414065480232239,
|
763 |
+
"learning_rate": 8.596868283696653e-06,
|
764 |
+
"loss": 0.1756,
|
765 |
"step": 2700
|
766 |
},
|
767 |
{
|
768 |
"epoch": 0.85,
|
769 |
+
"grad_norm": 2.6521832942962646,
|
770 |
+
"learning_rate": 8.213079521031625e-06,
|
771 |
+
"loss": 0.1898,
|
772 |
"step": 2725
|
773 |
},
|
774 |
{
|
775 |
"epoch": 0.86,
|
776 |
+
"grad_norm": 0.5525100231170654,
|
777 |
+
"learning_rate": 7.829290758366594e-06,
|
778 |
+
"loss": 0.192,
|
779 |
"step": 2750
|
780 |
},
|
781 |
{
|
782 |
"epoch": 0.87,
|
783 |
+
"grad_norm": 2.2816734313964844,
|
784 |
+
"learning_rate": 7.445501995701566e-06,
|
785 |
+
"loss": 0.1852,
|
786 |
"step": 2775
|
787 |
},
|
788 |
{
|
789 |
"epoch": 0.88,
|
790 |
+
"grad_norm": 1.4024970531463623,
|
791 |
+
"learning_rate": 7.061713233036537e-06,
|
792 |
+
"loss": 0.2279,
|
793 |
"step": 2800
|
794 |
},
|
795 |
{
|
796 |
"epoch": 0.88,
|
797 |
+
"grad_norm": 2.59187912940979,
|
798 |
+
"learning_rate": 6.677924470371508e-06,
|
799 |
+
"loss": 0.195,
|
800 |
"step": 2825
|
801 |
},
|
802 |
{
|
803 |
"epoch": 0.89,
|
804 |
+
"grad_norm": 4.647696018218994,
|
805 |
+
"learning_rate": 6.294135707706479e-06,
|
806 |
+
"loss": 0.1785,
|
807 |
"step": 2850
|
808 |
},
|
809 |
{
|
810 |
"epoch": 0.9,
|
811 |
+
"grad_norm": 2.7766268253326416,
|
812 |
+
"learning_rate": 5.910346945041449e-06,
|
813 |
+
"loss": 0.188,
|
814 |
"step": 2875
|
815 |
},
|
816 |
{
|
817 |
"epoch": 0.91,
|
818 |
+
"grad_norm": 2.672715425491333,
|
819 |
+
"learning_rate": 5.52655818237642e-06,
|
820 |
+
"loss": 0.1467,
|
821 |
"step": 2900
|
822 |
},
|
823 |
{
|
824 |
"epoch": 0.92,
|
825 |
+
"grad_norm": 0.6805928945541382,
|
826 |
+
"learning_rate": 5.142769419711391e-06,
|
827 |
+
"loss": 0.1561,
|
828 |
"step": 2925
|
829 |
},
|
830 |
{
|
831 |
"epoch": 0.92,
|
832 |
+
"grad_norm": 2.8108584880828857,
|
833 |
+
"learning_rate": 4.7589806570463625e-06,
|
834 |
+
"loss": 0.189,
|
835 |
"step": 2950
|
836 |
},
|
837 |
{
|
838 |
"epoch": 0.93,
|
839 |
+
"grad_norm": 0.4544536769390106,
|
840 |
+
"learning_rate": 4.375191894381333e-06,
|
841 |
+
"loss": 0.1464,
|
842 |
"step": 2975
|
843 |
},
|
844 |
{
|
845 |
"epoch": 0.94,
|
846 |
+
"grad_norm": 0.4506062865257263,
|
847 |
+
"learning_rate": 3.991403131716304e-06,
|
848 |
+
"loss": 0.1533,
|
849 |
"step": 3000
|
850 |
},
|
851 |
{
|
852 |
"epoch": 0.95,
|
853 |
+
"grad_norm": 1.9615803956985474,
|
854 |
+
"learning_rate": 3.607614369051274e-06,
|
855 |
+
"loss": 0.184,
|
856 |
"step": 3025
|
857 |
},
|
858 |
{
|
859 |
"epoch": 0.96,
|
860 |
+
"grad_norm": 3.426203489303589,
|
861 |
+
"learning_rate": 3.223825606386245e-06,
|
862 |
+
"loss": 0.1744,
|
863 |
"step": 3050
|
864 |
},
|
865 |
{
|
866 |
"epoch": 0.96,
|
867 |
+
"grad_norm": 2.0022315979003906,
|
868 |
+
"learning_rate": 2.8400368437212158e-06,
|
869 |
"loss": 0.1839,
|
870 |
"step": 3075
|
871 |
},
|
872 |
{
|
873 |
"epoch": 0.97,
|
874 |
+
"grad_norm": 0.8066006302833557,
|
875 |
+
"learning_rate": 2.4562480810561868e-06,
|
876 |
+
"loss": 0.199,
|
877 |
"step": 3100
|
878 |
},
|
879 |
{
|
880 |
"epoch": 0.98,
|
881 |
+
"grad_norm": 2.2845537662506104,
|
882 |
+
"learning_rate": 2.0724593183911578e-06,
|
883 |
+
"loss": 0.1361,
|
884 |
"step": 3125
|
885 |
},
|
886 |
{
|
887 |
"epoch": 0.99,
|
888 |
+
"grad_norm": 0.5564154982566833,
|
889 |
+
"learning_rate": 1.6886705557261286e-06,
|
890 |
+
"loss": 0.217,
|
891 |
"step": 3150
|
892 |
},
|
893 |
{
|
894 |
"epoch": 0.99,
|
895 |
+
"grad_norm": 0.6154067516326904,
|
896 |
+
"learning_rate": 1.3048817930610991e-06,
|
897 |
+
"loss": 0.1631,
|
898 |
"step": 3175
|
899 |
},
|
900 |
{
|
901 |
"epoch": 1.0,
|
902 |
+
"grad_norm": 0.6170292496681213,
|
903 |
+
"learning_rate": 9.2109303039607e-07,
|
904 |
+
"loss": 0.1161,
|
905 |
"step": 3200
|
906 |
},
|
907 |
{
|
908 |
"epoch": 1.01,
|
909 |
+
"grad_norm": 3.1419057846069336,
|
910 |
+
"learning_rate": 5.373042677310408e-07,
|
911 |
+
"loss": 0.106,
|
912 |
"step": 3225
|
913 |
},
|
914 |
{
|
915 |
+
"epoch": 1.02,
|
916 |
+
"grad_norm": 0.4944257438182831,
|
917 |
+
"learning_rate": 1.5351550506601167e-07,
|
918 |
+
"loss": 0.1072,
|
919 |
+
"step": 3250
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 1.02,
|
923 |
+
"step": 3257,
|
924 |
+
"total_flos": 4.038329842708009e+18,
|
925 |
+
"train_loss": 0.2225519281828883,
|
926 |
+
"train_runtime": 2924.985,
|
927 |
+
"train_samples_per_second": 17.815,
|
928 |
+
"train_steps_per_second": 1.114
|
929 |
}
|
930 |
],
|
931 |
"logging_steps": 25,
|
932 |
+
"max_steps": 3257,
|
933 |
"num_input_tokens_seen": 0,
|
934 |
"num_train_epochs": 2,
|
935 |
"save_steps": 100,
|
936 |
+
"total_flos": 4.038329842708009e+18,
|
937 |
"train_batch_size": 16,
|
938 |
"trial_name": null,
|
939 |
"trial_params": null
|