End of training
Browse files- README.md +4 -2
- all_results.json +9 -9
- eval_results.json +4 -4
- train_results.json +5 -5
- trainer_state.json +93 -912
README.md
CHANGED
@@ -5,6 +5,8 @@ license: apache-2.0
|
|
5 |
metrics:
|
6 |
- accuracy
|
7 |
tags:
|
|
|
|
|
8 |
- generated_from_trainer
|
9 |
model-index:
|
10 |
- name: only-lora-beans-vit-base-patch16-224-in21k
|
@@ -16,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
16 |
|
17 |
# only-lora-beans-vit-base-patch16-224-in21k
|
18 |
|
19 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on
|
20 |
It achieves the following results on the evaluation set:
|
21 |
-
- Loss:
|
22 |
- Accuracy: 0.3308
|
23 |
|
24 |
## Model description
|
|
|
5 |
metrics:
|
6 |
- accuracy
|
7 |
tags:
|
8 |
+
- image-classification
|
9 |
+
- vision
|
10 |
- generated_from_trainer
|
11 |
model-index:
|
12 |
- name: only-lora-beans-vit-base-patch16-224-in21k
|
|
|
18 |
|
19 |
# only-lora-beans-vit-base-patch16-224-in21k
|
20 |
|
21 |
+
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 beans dataset.
|
22 |
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 2.0399
|
24 |
- Accuracy: 0.3308
|
25 |
|
26 |
## Model description
|
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
"eval_accuracy": 0.3308270676691729,
|
4 |
-
"eval_loss":
|
5 |
-
"eval_runtime": 0.
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second":
|
8 |
-
"total_flos": 8.
|
9 |
-
"train_loss": 0.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second":
|
12 |
-
"train_steps_per_second": 7.
|
13 |
}
|
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
"eval_accuracy": 0.3308270676691729,
|
4 |
+
"eval_loss": 2.0399293899536133,
|
5 |
+
"eval_runtime": 0.9627,
|
6 |
+
"eval_samples_per_second": 138.153,
|
7 |
+
"eval_steps_per_second": 17.659,
|
8 |
+
"total_flos": 8.772706474360013e+17,
|
9 |
+
"train_loss": 0.9533355126014122,
|
10 |
+
"train_runtime": 174.5905,
|
11 |
+
"train_samples_per_second": 59.224,
|
12 |
+
"train_steps_per_second": 7.446
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
"eval_accuracy": 0.3308270676691729,
|
4 |
-
"eval_loss":
|
5 |
-
"eval_runtime": 0.
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second":
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
"eval_accuracy": 0.3308270676691729,
|
4 |
+
"eval_loss": 2.0399293899536133,
|
5 |
+
"eval_runtime": 0.9627,
|
6 |
+
"eval_samples_per_second": 138.153,
|
7 |
+
"eval_steps_per_second": 17.659
|
8 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
-
"total_flos": 8.
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 7.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 10.0,
|
3 |
+
"total_flos": 8.772706474360013e+17,
|
4 |
+
"train_loss": 0.9533355126014122,
|
5 |
+
"train_runtime": 174.5905,
|
6 |
+
"train_samples_per_second": 59.224,
|
7 |
+
"train_steps_per_second": 7.446
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
"best_model_checkpoint": "./beans_outputs/checkpoint-130",
|
4 |
"epoch": 10.0,
|
5 |
"eval_steps": 500,
|
@@ -8,1017 +8,198 @@
|
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
-
{
|
12 |
-
"epoch": 0.07692307692307693,
|
13 |
-
"grad_norm": 1.2493035793304443,
|
14 |
-
"learning_rate": 0.004961538461538462,
|
15 |
-
"loss": 0.8568,
|
16 |
-
"step": 10
|
17 |
-
},
|
18 |
-
{
|
19 |
-
"epoch": 0.15384615384615385,
|
20 |
-
"grad_norm": 1.4747363328933716,
|
21 |
-
"learning_rate": 0.004923076923076923,
|
22 |
-
"loss": 0.6606,
|
23 |
-
"step": 20
|
24 |
-
},
|
25 |
-
{
|
26 |
-
"epoch": 0.23076923076923078,
|
27 |
-
"grad_norm": 0.8188544511795044,
|
28 |
-
"learning_rate": 0.004884615384615385,
|
29 |
-
"loss": 0.3787,
|
30 |
-
"step": 30
|
31 |
-
},
|
32 |
-
{
|
33 |
-
"epoch": 0.3076923076923077,
|
34 |
-
"grad_norm": 1.718248724937439,
|
35 |
-
"learning_rate": 0.004846153846153846,
|
36 |
-
"loss": 0.3154,
|
37 |
-
"step": 40
|
38 |
-
},
|
39 |
-
{
|
40 |
-
"epoch": 0.38461538461538464,
|
41 |
-
"grad_norm": 2.2010860443115234,
|
42 |
-
"learning_rate": 0.004807692307692308,
|
43 |
-
"loss": 0.4932,
|
44 |
-
"step": 50
|
45 |
-
},
|
46 |
-
{
|
47 |
-
"epoch": 0.46153846153846156,
|
48 |
-
"grad_norm": 2.927755832672119,
|
49 |
-
"learning_rate": 0.0047692307692307695,
|
50 |
-
"loss": 0.561,
|
51 |
-
"step": 60
|
52 |
-
},
|
53 |
-
{
|
54 |
-
"epoch": 0.5384615384615384,
|
55 |
-
"grad_norm": 1.2203412055969238,
|
56 |
-
"learning_rate": 0.004730769230769231,
|
57 |
-
"loss": 0.3309,
|
58 |
-
"step": 70
|
59 |
-
},
|
60 |
-
{
|
61 |
-
"epoch": 0.6153846153846154,
|
62 |
-
"grad_norm": 0.08795254677534103,
|
63 |
-
"learning_rate": 0.004692307692307693,
|
64 |
-
"loss": 0.3261,
|
65 |
-
"step": 80
|
66 |
-
},
|
67 |
-
{
|
68 |
-
"epoch": 0.6923076923076923,
|
69 |
-
"grad_norm": 0.4609662890434265,
|
70 |
-
"learning_rate": 0.004653846153846154,
|
71 |
-
"loss": 0.195,
|
72 |
-
"step": 90
|
73 |
-
},
|
74 |
{
|
75 |
"epoch": 0.7692307692307693,
|
76 |
-
"grad_norm":
|
77 |
-
"learning_rate": 0.
|
78 |
-
"loss": 0.
|
79 |
"step": 100
|
80 |
},
|
81 |
-
{
|
82 |
-
"epoch": 0.8461538461538461,
|
83 |
-
"grad_norm": 2.5637238025665283,
|
84 |
-
"learning_rate": 0.0045769230769230765,
|
85 |
-
"loss": 0.5961,
|
86 |
-
"step": 110
|
87 |
-
},
|
88 |
-
{
|
89 |
-
"epoch": 0.9230769230769231,
|
90 |
-
"grad_norm": 0.32293158769607544,
|
91 |
-
"learning_rate": 0.004538461538461539,
|
92 |
-
"loss": 0.236,
|
93 |
-
"step": 120
|
94 |
-
},
|
95 |
-
{
|
96 |
-
"epoch": 1.0,
|
97 |
-
"grad_norm": 7.484982967376709,
|
98 |
-
"learning_rate": 0.0045000000000000005,
|
99 |
-
"loss": 0.6323,
|
100 |
-
"step": 130
|
101 |
-
},
|
102 |
{
|
103 |
"epoch": 1.0,
|
104 |
-
"eval_accuracy": 0.
|
105 |
-
"eval_loss": 0.
|
106 |
-
"eval_runtime": 0.
|
107 |
-
"eval_samples_per_second":
|
108 |
-
"eval_steps_per_second":
|
109 |
"step": 130
|
110 |
},
|
111 |
-
{
|
112 |
-
"epoch": 1.0769230769230769,
|
113 |
-
"grad_norm": 0.12631376087665558,
|
114 |
-
"learning_rate": 0.004461538461538462,
|
115 |
-
"loss": 0.747,
|
116 |
-
"step": 140
|
117 |
-
},
|
118 |
-
{
|
119 |
-
"epoch": 1.1538461538461537,
|
120 |
-
"grad_norm": 1.5336790084838867,
|
121 |
-
"learning_rate": 0.004423076923076923,
|
122 |
-
"loss": 0.3062,
|
123 |
-
"step": 150
|
124 |
-
},
|
125 |
-
{
|
126 |
-
"epoch": 1.2307692307692308,
|
127 |
-
"grad_norm": 0.7460920214653015,
|
128 |
-
"learning_rate": 0.004384615384615384,
|
129 |
-
"loss": 0.2088,
|
130 |
-
"step": 160
|
131 |
-
},
|
132 |
-
{
|
133 |
-
"epoch": 1.3076923076923077,
|
134 |
-
"grad_norm": 0.6954104900360107,
|
135 |
-
"learning_rate": 0.004346153846153846,
|
136 |
-
"loss": 0.2942,
|
137 |
-
"step": 170
|
138 |
-
},
|
139 |
-
{
|
140 |
-
"epoch": 1.3846153846153846,
|
141 |
-
"grad_norm": 2.2199032306671143,
|
142 |
-
"learning_rate": 0.004307692307692308,
|
143 |
-
"loss": 0.315,
|
144 |
-
"step": 180
|
145 |
-
},
|
146 |
-
{
|
147 |
-
"epoch": 1.4615384615384617,
|
148 |
-
"grad_norm": 2.654019832611084,
|
149 |
-
"learning_rate": 0.004269230769230769,
|
150 |
-
"loss": 0.4203,
|
151 |
-
"step": 190
|
152 |
-
},
|
153 |
{
|
154 |
"epoch": 1.5384615384615383,
|
155 |
-
"grad_norm": 1.
|
156 |
-
"learning_rate": 0.
|
157 |
-
"loss": 0.
|
158 |
"step": 200
|
159 |
},
|
160 |
-
{
|
161 |
-
"epoch": 1.6153846153846154,
|
162 |
-
"grad_norm": 0.9371921420097351,
|
163 |
-
"learning_rate": 0.004192307692307692,
|
164 |
-
"loss": 0.3436,
|
165 |
-
"step": 210
|
166 |
-
},
|
167 |
-
{
|
168 |
-
"epoch": 1.6923076923076923,
|
169 |
-
"grad_norm": 0.573295533657074,
|
170 |
-
"learning_rate": 0.004153846153846154,
|
171 |
-
"loss": 0.7394,
|
172 |
-
"step": 220
|
173 |
-
},
|
174 |
-
{
|
175 |
-
"epoch": 1.7692307692307692,
|
176 |
-
"grad_norm": 3.2005550861358643,
|
177 |
-
"learning_rate": 0.004115384615384615,
|
178 |
-
"loss": 0.5694,
|
179 |
-
"step": 230
|
180 |
-
},
|
181 |
-
{
|
182 |
-
"epoch": 1.8461538461538463,
|
183 |
-
"grad_norm": 1.0854452848434448,
|
184 |
-
"learning_rate": 0.004076923076923077,
|
185 |
-
"loss": 0.6043,
|
186 |
-
"step": 240
|
187 |
-
},
|
188 |
-
{
|
189 |
-
"epoch": 1.9230769230769231,
|
190 |
-
"grad_norm": 1.3743146657943726,
|
191 |
-
"learning_rate": 0.0040384615384615385,
|
192 |
-
"loss": 0.4406,
|
193 |
-
"step": 250
|
194 |
-
},
|
195 |
-
{
|
196 |
-
"epoch": 2.0,
|
197 |
-
"grad_norm": 0.11435157805681229,
|
198 |
-
"learning_rate": 0.004,
|
199 |
-
"loss": 0.4102,
|
200 |
-
"step": 260
|
201 |
-
},
|
202 |
{
|
203 |
"epoch": 2.0,
|
204 |
-
"eval_accuracy": 0.
|
205 |
-
"eval_loss": 1.
|
206 |
-
"eval_runtime": 0.
|
207 |
-
"eval_samples_per_second":
|
208 |
-
"eval_steps_per_second":
|
209 |
"step": 260
|
210 |
},
|
211 |
-
{
|
212 |
-
"epoch": 2.076923076923077,
|
213 |
-
"grad_norm": 0.9636590480804443,
|
214 |
-
"learning_rate": 0.003961538461538462,
|
215 |
-
"loss": 0.4669,
|
216 |
-
"step": 270
|
217 |
-
},
|
218 |
-
{
|
219 |
-
"epoch": 2.1538461538461537,
|
220 |
-
"grad_norm": 1.4855515956878662,
|
221 |
-
"learning_rate": 0.003923076923076923,
|
222 |
-
"loss": 0.4618,
|
223 |
-
"step": 280
|
224 |
-
},
|
225 |
-
{
|
226 |
-
"epoch": 2.230769230769231,
|
227 |
-
"grad_norm": 1.6462602615356445,
|
228 |
-
"learning_rate": 0.003884615384615385,
|
229 |
-
"loss": 0.5266,
|
230 |
-
"step": 290
|
231 |
-
},
|
232 |
{
|
233 |
"epoch": 2.3076923076923075,
|
234 |
-
"grad_norm":
|
235 |
-
"learning_rate": 0.
|
236 |
-
"loss": 0.
|
237 |
"step": 300
|
238 |
},
|
239 |
-
{
|
240 |
-
"epoch": 2.3846153846153846,
|
241 |
-
"grad_norm": 1.7253210544586182,
|
242 |
-
"learning_rate": 0.0038076923076923075,
|
243 |
-
"loss": 0.4711,
|
244 |
-
"step": 310
|
245 |
-
},
|
246 |
-
{
|
247 |
-
"epoch": 2.4615384615384617,
|
248 |
-
"grad_norm": 0.6617446541786194,
|
249 |
-
"learning_rate": 0.003769230769230769,
|
250 |
-
"loss": 0.5785,
|
251 |
-
"step": 320
|
252 |
-
},
|
253 |
-
{
|
254 |
-
"epoch": 2.5384615384615383,
|
255 |
-
"grad_norm": 0.6448426842689514,
|
256 |
-
"learning_rate": 0.003730769230769231,
|
257 |
-
"loss": 0.7907,
|
258 |
-
"step": 330
|
259 |
-
},
|
260 |
-
{
|
261 |
-
"epoch": 2.6153846153846154,
|
262 |
-
"grad_norm": 3.4137613773345947,
|
263 |
-
"learning_rate": 0.0036923076923076927,
|
264 |
-
"loss": 0.9229,
|
265 |
-
"step": 340
|
266 |
-
},
|
267 |
-
{
|
268 |
-
"epoch": 2.6923076923076925,
|
269 |
-
"grad_norm": 5.124917507171631,
|
270 |
-
"learning_rate": 0.003653846153846154,
|
271 |
-
"loss": 1.0287,
|
272 |
-
"step": 350
|
273 |
-
},
|
274 |
-
{
|
275 |
-
"epoch": 2.769230769230769,
|
276 |
-
"grad_norm": 1.2190972566604614,
|
277 |
-
"learning_rate": 0.0036153846153846154,
|
278 |
-
"loss": 0.7791,
|
279 |
-
"step": 360
|
280 |
-
},
|
281 |
-
{
|
282 |
-
"epoch": 2.8461538461538463,
|
283 |
-
"grad_norm": 0.6898893713951111,
|
284 |
-
"learning_rate": 0.003576923076923077,
|
285 |
-
"loss": 1.4277,
|
286 |
-
"step": 370
|
287 |
-
},
|
288 |
-
{
|
289 |
-
"epoch": 2.9230769230769234,
|
290 |
-
"grad_norm": 2.563831090927124,
|
291 |
-
"learning_rate": 0.003538461538461539,
|
292 |
-
"loss": 1.3955,
|
293 |
-
"step": 380
|
294 |
-
},
|
295 |
{
|
296 |
"epoch": 3.0,
|
297 |
-
"
|
298 |
-
"
|
299 |
-
"
|
300 |
-
"
|
301 |
-
|
302 |
-
{
|
303 |
-
"epoch": 3.0,
|
304 |
-
"eval_accuracy": 0.6616541353383458,
|
305 |
-
"eval_loss": 0.8284351825714111,
|
306 |
-
"eval_runtime": 0.9347,
|
307 |
-
"eval_samples_per_second": 142.291,
|
308 |
-
"eval_steps_per_second": 18.188,
|
309 |
"step": 390
|
310 |
},
|
311 |
{
|
312 |
"epoch": 3.076923076923077,
|
313 |
-
"grad_norm":
|
314 |
-
"learning_rate": 0.
|
315 |
-
"loss":
|
316 |
"step": 400
|
317 |
},
|
318 |
-
{
|
319 |
-
"epoch": 3.1538461538461537,
|
320 |
-
"grad_norm": 1.773996114730835,
|
321 |
-
"learning_rate": 0.003423076923076923,
|
322 |
-
"loss": 1.3465,
|
323 |
-
"step": 410
|
324 |
-
},
|
325 |
-
{
|
326 |
-
"epoch": 3.230769230769231,
|
327 |
-
"grad_norm": 3.1790945529937744,
|
328 |
-
"learning_rate": 0.003384615384615385,
|
329 |
-
"loss": 1.2822,
|
330 |
-
"step": 420
|
331 |
-
},
|
332 |
-
{
|
333 |
-
"epoch": 3.3076923076923075,
|
334 |
-
"grad_norm": 5.48622465133667,
|
335 |
-
"learning_rate": 0.003346153846153846,
|
336 |
-
"loss": 1.0807,
|
337 |
-
"step": 430
|
338 |
-
},
|
339 |
-
{
|
340 |
-
"epoch": 3.3846153846153846,
|
341 |
-
"grad_norm": 3.6509623527526855,
|
342 |
-
"learning_rate": 0.0033076923076923075,
|
343 |
-
"loss": 1.1806,
|
344 |
-
"step": 440
|
345 |
-
},
|
346 |
-
{
|
347 |
-
"epoch": 3.4615384615384617,
|
348 |
-
"grad_norm": 2.3183326721191406,
|
349 |
-
"learning_rate": 0.0032692307692307695,
|
350 |
-
"loss": 1.2542,
|
351 |
-
"step": 450
|
352 |
-
},
|
353 |
-
{
|
354 |
-
"epoch": 3.5384615384615383,
|
355 |
-
"grad_norm": 3.2686471939086914,
|
356 |
-
"learning_rate": 0.003230769230769231,
|
357 |
-
"loss": 1.2217,
|
358 |
-
"step": 460
|
359 |
-
},
|
360 |
-
{
|
361 |
-
"epoch": 3.6153846153846154,
|
362 |
-
"grad_norm": 0.33250316977500916,
|
363 |
-
"learning_rate": 0.003192307692307692,
|
364 |
-
"loss": 1.2107,
|
365 |
-
"step": 470
|
366 |
-
},
|
367 |
-
{
|
368 |
-
"epoch": 3.6923076923076925,
|
369 |
-
"grad_norm": 2.4644064903259277,
|
370 |
-
"learning_rate": 0.0031538461538461538,
|
371 |
-
"loss": 1.2287,
|
372 |
-
"step": 480
|
373 |
-
},
|
374 |
-
{
|
375 |
-
"epoch": 3.769230769230769,
|
376 |
-
"grad_norm": 0.9661989808082581,
|
377 |
-
"learning_rate": 0.0031153846153846153,
|
378 |
-
"loss": 1.1392,
|
379 |
-
"step": 490
|
380 |
-
},
|
381 |
{
|
382 |
"epoch": 3.8461538461538463,
|
383 |
-
"grad_norm":
|
384 |
-
"learning_rate": 0.
|
385 |
-
"loss": 1.
|
386 |
"step": 500
|
387 |
},
|
388 |
-
{
|
389 |
-
"epoch": 3.9230769230769234,
|
390 |
-
"grad_norm": 0.5606147646903992,
|
391 |
-
"learning_rate": 0.0030384615384615385,
|
392 |
-
"loss": 1.1148,
|
393 |
-
"step": 510
|
394 |
-
},
|
395 |
{
|
396 |
"epoch": 4.0,
|
397 |
-
"
|
398 |
-
"
|
399 |
-
"
|
400 |
-
"
|
401 |
-
|
402 |
-
{
|
403 |
-
"epoch": 4.0,
|
404 |
-
"eval_accuracy": 0.3383458646616541,
|
405 |
-
"eval_loss": 1.1070177555084229,
|
406 |
-
"eval_runtime": 0.9062,
|
407 |
-
"eval_samples_per_second": 146.767,
|
408 |
-
"eval_steps_per_second": 18.76,
|
409 |
"step": 520
|
410 |
},
|
411 |
-
{
|
412 |
-
"epoch": 4.076923076923077,
|
413 |
-
"grad_norm": 1.8376567363739014,
|
414 |
-
"learning_rate": 0.0029615384615384616,
|
415 |
-
"loss": 1.16,
|
416 |
-
"step": 530
|
417 |
-
},
|
418 |
-
{
|
419 |
-
"epoch": 4.153846153846154,
|
420 |
-
"grad_norm": 0.1573391705751419,
|
421 |
-
"learning_rate": 0.002923076923076923,
|
422 |
-
"loss": 1.1133,
|
423 |
-
"step": 540
|
424 |
-
},
|
425 |
-
{
|
426 |
-
"epoch": 4.230769230769231,
|
427 |
-
"grad_norm": 1.7338229417800903,
|
428 |
-
"learning_rate": 0.0028846153846153843,
|
429 |
-
"loss": 1.1604,
|
430 |
-
"step": 550
|
431 |
-
},
|
432 |
-
{
|
433 |
-
"epoch": 4.3076923076923075,
|
434 |
-
"grad_norm": 1.532076358795166,
|
435 |
-
"learning_rate": 0.002846153846153846,
|
436 |
-
"loss": 1.171,
|
437 |
-
"step": 560
|
438 |
-
},
|
439 |
-
{
|
440 |
-
"epoch": 4.384615384615385,
|
441 |
-
"grad_norm": 1.2591376304626465,
|
442 |
-
"learning_rate": 0.002807692307692308,
|
443 |
-
"loss": 1.116,
|
444 |
-
"step": 570
|
445 |
-
},
|
446 |
-
{
|
447 |
-
"epoch": 4.461538461538462,
|
448 |
-
"grad_norm": 1.7697715759277344,
|
449 |
-
"learning_rate": 0.0027692307692307695,
|
450 |
-
"loss": 1.1409,
|
451 |
-
"step": 580
|
452 |
-
},
|
453 |
-
{
|
454 |
-
"epoch": 4.538461538461538,
|
455 |
-
"grad_norm": 1.0740739107131958,
|
456 |
-
"learning_rate": 0.0027307692307692306,
|
457 |
-
"loss": 1.1458,
|
458 |
-
"step": 590
|
459 |
-
},
|
460 |
{
|
461 |
"epoch": 4.615384615384615,
|
462 |
-
"grad_norm": 0.
|
463 |
-
"learning_rate": 0.
|
464 |
-
"loss": 1.
|
465 |
"step": 600
|
466 |
},
|
467 |
-
{
|
468 |
-
"epoch": 4.6923076923076925,
|
469 |
-
"grad_norm": 1.8973933458328247,
|
470 |
-
"learning_rate": 0.0026538461538461538,
|
471 |
-
"loss": 1.1024,
|
472 |
-
"step": 610
|
473 |
-
},
|
474 |
-
{
|
475 |
-
"epoch": 4.769230769230769,
|
476 |
-
"grad_norm": 0.06765652447938919,
|
477 |
-
"learning_rate": 0.0026153846153846158,
|
478 |
-
"loss": 1.1514,
|
479 |
-
"step": 620
|
480 |
-
},
|
481 |
-
{
|
482 |
-
"epoch": 4.846153846153846,
|
483 |
-
"grad_norm": 0.8284403085708618,
|
484 |
-
"learning_rate": 0.002576923076923077,
|
485 |
-
"loss": 1.1799,
|
486 |
-
"step": 630
|
487 |
-
},
|
488 |
-
{
|
489 |
-
"epoch": 4.923076923076923,
|
490 |
-
"grad_norm": 1.611846923828125,
|
491 |
-
"learning_rate": 0.0025384615384615385,
|
492 |
-
"loss": 1.169,
|
493 |
-
"step": 640
|
494 |
-
},
|
495 |
-
{
|
496 |
-
"epoch": 5.0,
|
497 |
-
"grad_norm": 3.048553705215454,
|
498 |
-
"learning_rate": 0.0025,
|
499 |
-
"loss": 1.1069,
|
500 |
-
"step": 650
|
501 |
-
},
|
502 |
{
|
503 |
"epoch": 5.0,
|
504 |
"eval_accuracy": 0.3308270676691729,
|
505 |
-
"eval_loss": 1.
|
506 |
-
"eval_runtime": 0.
|
507 |
-
"eval_samples_per_second":
|
508 |
-
"eval_steps_per_second":
|
509 |
"step": 650
|
510 |
},
|
511 |
-
{
|
512 |
-
"epoch": 5.076923076923077,
|
513 |
-
"grad_norm": 0.6719330549240112,
|
514 |
-
"learning_rate": 0.0024615384615384616,
|
515 |
-
"loss": 1.1054,
|
516 |
-
"step": 660
|
517 |
-
},
|
518 |
-
{
|
519 |
-
"epoch": 5.153846153846154,
|
520 |
-
"grad_norm": 0.8374503254890442,
|
521 |
-
"learning_rate": 0.002423076923076923,
|
522 |
-
"loss": 1.1134,
|
523 |
-
"step": 670
|
524 |
-
},
|
525 |
-
{
|
526 |
-
"epoch": 5.230769230769231,
|
527 |
-
"grad_norm": 0.5713789463043213,
|
528 |
-
"learning_rate": 0.0023846153846153848,
|
529 |
-
"loss": 1.1534,
|
530 |
-
"step": 680
|
531 |
-
},
|
532 |
-
{
|
533 |
-
"epoch": 5.3076923076923075,
|
534 |
-
"grad_norm": 1.0185028314590454,
|
535 |
-
"learning_rate": 0.0023461538461538463,
|
536 |
-
"loss": 1.1368,
|
537 |
-
"step": 690
|
538 |
-
},
|
539 |
{
|
540 |
"epoch": 5.384615384615385,
|
541 |
-
"grad_norm":
|
542 |
-
"learning_rate": 0.
|
543 |
-
"loss": 1.
|
544 |
"step": 700
|
545 |
},
|
546 |
-
{
|
547 |
-
"epoch": 5.461538461538462,
|
548 |
-
"grad_norm": 0.9150602221488953,
|
549 |
-
"learning_rate": 0.0022692307692307695,
|
550 |
-
"loss": 1.1187,
|
551 |
-
"step": 710
|
552 |
-
},
|
553 |
-
{
|
554 |
-
"epoch": 5.538461538461538,
|
555 |
-
"grad_norm": 0.741632878780365,
|
556 |
-
"learning_rate": 0.002230769230769231,
|
557 |
-
"loss": 1.1129,
|
558 |
-
"step": 720
|
559 |
-
},
|
560 |
-
{
|
561 |
-
"epoch": 5.615384615384615,
|
562 |
-
"grad_norm": 1.5859191417694092,
|
563 |
-
"learning_rate": 0.002192307692307692,
|
564 |
-
"loss": 1.135,
|
565 |
-
"step": 730
|
566 |
-
},
|
567 |
-
{
|
568 |
-
"epoch": 5.6923076923076925,
|
569 |
-
"grad_norm": 2.2667315006256104,
|
570 |
-
"learning_rate": 0.002153846153846154,
|
571 |
-
"loss": 1.1474,
|
572 |
-
"step": 740
|
573 |
-
},
|
574 |
-
{
|
575 |
-
"epoch": 5.769230769230769,
|
576 |
-
"grad_norm": 1.3794482946395874,
|
577 |
-
"learning_rate": 0.0021153846153846153,
|
578 |
-
"loss": 1.1095,
|
579 |
-
"step": 750
|
580 |
-
},
|
581 |
-
{
|
582 |
-
"epoch": 5.846153846153846,
|
583 |
-
"grad_norm": 1.2423471212387085,
|
584 |
-
"learning_rate": 0.002076923076923077,
|
585 |
-
"loss": 1.1057,
|
586 |
-
"step": 760
|
587 |
-
},
|
588 |
-
{
|
589 |
-
"epoch": 5.923076923076923,
|
590 |
-
"grad_norm": 4.28483247756958,
|
591 |
-
"learning_rate": 0.0020384615384615385,
|
592 |
-
"loss": 1.1093,
|
593 |
-
"step": 770
|
594 |
-
},
|
595 |
-
{
|
596 |
-
"epoch": 6.0,
|
597 |
-
"grad_norm": 4.998449802398682,
|
598 |
-
"learning_rate": 0.002,
|
599 |
-
"loss": 1.1748,
|
600 |
-
"step": 780
|
601 |
-
},
|
602 |
{
|
603 |
"epoch": 6.0,
|
604 |
"eval_accuracy": 0.3308270676691729,
|
605 |
-
"eval_loss": 1.
|
606 |
-
"eval_runtime": 0.
|
607 |
-
"eval_samples_per_second":
|
608 |
-
"eval_steps_per_second": 18.
|
609 |
"step": 780
|
610 |
},
|
611 |
-
{
|
612 |
-
"epoch": 6.076923076923077,
|
613 |
-
"grad_norm": 2.81730318069458,
|
614 |
-
"learning_rate": 0.0019615384615384616,
|
615 |
-
"loss": 1.1381,
|
616 |
-
"step": 790
|
617 |
-
},
|
618 |
{
|
619 |
"epoch": 6.153846153846154,
|
620 |
-
"grad_norm":
|
621 |
-
"learning_rate": 0.
|
622 |
-
"loss": 1.
|
623 |
"step": 800
|
624 |
},
|
625 |
-
{
|
626 |
-
"epoch": 6.230769230769231,
|
627 |
-
"grad_norm": 0.8537469506263733,
|
628 |
-
"learning_rate": 0.0018846153846153845,
|
629 |
-
"loss": 1.1327,
|
630 |
-
"step": 810
|
631 |
-
},
|
632 |
-
{
|
633 |
-
"epoch": 6.3076923076923075,
|
634 |
-
"grad_norm": 0.8818097710609436,
|
635 |
-
"learning_rate": 0.0018461538461538463,
|
636 |
-
"loss": 1.1437,
|
637 |
-
"step": 820
|
638 |
-
},
|
639 |
-
{
|
640 |
-
"epoch": 6.384615384615385,
|
641 |
-
"grad_norm": 2.496903896331787,
|
642 |
-
"learning_rate": 0.0018076923076923077,
|
643 |
-
"loss": 1.1251,
|
644 |
-
"step": 830
|
645 |
-
},
|
646 |
-
{
|
647 |
-
"epoch": 6.461538461538462,
|
648 |
-
"grad_norm": 3.2999167442321777,
|
649 |
-
"learning_rate": 0.0017692307692307695,
|
650 |
-
"loss": 1.1373,
|
651 |
-
"step": 840
|
652 |
-
},
|
653 |
-
{
|
654 |
-
"epoch": 6.538461538461538,
|
655 |
-
"grad_norm": 2.9537723064422607,
|
656 |
-
"learning_rate": 0.0017307692307692308,
|
657 |
-
"loss": 1.176,
|
658 |
-
"step": 850
|
659 |
-
},
|
660 |
-
{
|
661 |
-
"epoch": 6.615384615384615,
|
662 |
-
"grad_norm": 1.5148764848709106,
|
663 |
-
"learning_rate": 0.0016923076923076924,
|
664 |
-
"loss": 1.1146,
|
665 |
-
"step": 860
|
666 |
-
},
|
667 |
-
{
|
668 |
-
"epoch": 6.6923076923076925,
|
669 |
-
"grad_norm": 1.3614485263824463,
|
670 |
-
"learning_rate": 0.0016538461538461537,
|
671 |
-
"loss": 1.0895,
|
672 |
-
"step": 870
|
673 |
-
},
|
674 |
-
{
|
675 |
-
"epoch": 6.769230769230769,
|
676 |
-
"grad_norm": 0.6686407327651978,
|
677 |
-
"learning_rate": 0.0016153846153846155,
|
678 |
-
"loss": 1.1245,
|
679 |
-
"step": 880
|
680 |
-
},
|
681 |
-
{
|
682 |
-
"epoch": 6.846153846153846,
|
683 |
-
"grad_norm": 1.6276365518569946,
|
684 |
-
"learning_rate": 0.0015769230769230769,
|
685 |
-
"loss": 1.1412,
|
686 |
-
"step": 890
|
687 |
-
},
|
688 |
{
|
689 |
"epoch": 6.923076923076923,
|
690 |
-
"grad_norm":
|
691 |
-
"learning_rate": 0.
|
692 |
-
"loss": 1.
|
693 |
"step": 900
|
694 |
},
|
695 |
-
{
|
696 |
-
"epoch": 7.0,
|
697 |
-
"grad_norm": 2.74285888671875,
|
698 |
-
"learning_rate": 0.0015,
|
699 |
-
"loss": 1.1514,
|
700 |
-
"step": 910
|
701 |
-
},
|
702 |
{
|
703 |
"epoch": 7.0,
|
704 |
"eval_accuracy": 0.3308270676691729,
|
705 |
-
"eval_loss": 1.
|
706 |
-
"eval_runtime":
|
707 |
-
"eval_samples_per_second":
|
708 |
-
"eval_steps_per_second":
|
709 |
"step": 910
|
710 |
},
|
711 |
-
{
|
712 |
-
"epoch": 7.076923076923077,
|
713 |
-
"grad_norm": 2.248867988586426,
|
714 |
-
"learning_rate": 0.0014615384615384616,
|
715 |
-
"loss": 1.0913,
|
716 |
-
"step": 920
|
717 |
-
},
|
718 |
-
{
|
719 |
-
"epoch": 7.153846153846154,
|
720 |
-
"grad_norm": 2.5912275314331055,
|
721 |
-
"learning_rate": 0.001423076923076923,
|
722 |
-
"loss": 1.1193,
|
723 |
-
"step": 930
|
724 |
-
},
|
725 |
-
{
|
726 |
-
"epoch": 7.230769230769231,
|
727 |
-
"grad_norm": 3.573967695236206,
|
728 |
-
"learning_rate": 0.0013846153846153847,
|
729 |
-
"loss": 1.1352,
|
730 |
-
"step": 940
|
731 |
-
},
|
732 |
-
{
|
733 |
-
"epoch": 7.3076923076923075,
|
734 |
-
"grad_norm": 2.077554225921631,
|
735 |
-
"learning_rate": 0.001346153846153846,
|
736 |
-
"loss": 1.0947,
|
737 |
-
"step": 950
|
738 |
-
},
|
739 |
-
{
|
740 |
-
"epoch": 7.384615384615385,
|
741 |
-
"grad_norm": 1.6364086866378784,
|
742 |
-
"learning_rate": 0.0013076923076923079,
|
743 |
-
"loss": 1.1296,
|
744 |
-
"step": 960
|
745 |
-
},
|
746 |
-
{
|
747 |
-
"epoch": 7.461538461538462,
|
748 |
-
"grad_norm": 1.1806069612503052,
|
749 |
-
"learning_rate": 0.0012692307692307692,
|
750 |
-
"loss": 1.1033,
|
751 |
-
"step": 970
|
752 |
-
},
|
753 |
-
{
|
754 |
-
"epoch": 7.538461538461538,
|
755 |
-
"grad_norm": 2.4788713455200195,
|
756 |
-
"learning_rate": 0.0012307692307692308,
|
757 |
-
"loss": 1.1148,
|
758 |
-
"step": 980
|
759 |
-
},
|
760 |
-
{
|
761 |
-
"epoch": 7.615384615384615,
|
762 |
-
"grad_norm": 0.8245519399642944,
|
763 |
-
"learning_rate": 0.0011923076923076924,
|
764 |
-
"loss": 1.1121,
|
765 |
-
"step": 990
|
766 |
-
},
|
767 |
{
|
768 |
"epoch": 7.6923076923076925,
|
769 |
-
"grad_norm":
|
770 |
-
"learning_rate": 0.
|
771 |
-
"loss": 1.
|
772 |
"step": 1000
|
773 |
},
|
774 |
-
{
|
775 |
-
"epoch": 7.769230769230769,
|
776 |
-
"grad_norm": 2.2075676918029785,
|
777 |
-
"learning_rate": 0.0011153846153846155,
|
778 |
-
"loss": 1.0956,
|
779 |
-
"step": 1010
|
780 |
-
},
|
781 |
-
{
|
782 |
-
"epoch": 7.846153846153846,
|
783 |
-
"grad_norm": 1.0866578817367554,
|
784 |
-
"learning_rate": 0.001076923076923077,
|
785 |
-
"loss": 1.1065,
|
786 |
-
"step": 1020
|
787 |
-
},
|
788 |
-
{
|
789 |
-
"epoch": 7.923076923076923,
|
790 |
-
"grad_norm": 2.4340384006500244,
|
791 |
-
"learning_rate": 0.0010384615384615384,
|
792 |
-
"loss": 1.1118,
|
793 |
-
"step": 1030
|
794 |
-
},
|
795 |
{
|
796 |
"epoch": 8.0,
|
797 |
-
"
|
798 |
-
"
|
799 |
-
"
|
800 |
-
"
|
801 |
-
|
802 |
-
{
|
803 |
-
"epoch": 8.0,
|
804 |
-
"eval_accuracy": 0.3383458646616541,
|
805 |
-
"eval_loss": 1.1004995107650757,
|
806 |
-
"eval_runtime": 0.9133,
|
807 |
-
"eval_samples_per_second": 145.621,
|
808 |
-
"eval_steps_per_second": 18.613,
|
809 |
"step": 1040
|
810 |
},
|
811 |
-
{
|
812 |
-
"epoch": 8.076923076923077,
|
813 |
-
"grad_norm": 0.689738392829895,
|
814 |
-
"learning_rate": 0.0009615384615384616,
|
815 |
-
"loss": 1.1119,
|
816 |
-
"step": 1050
|
817 |
-
},
|
818 |
-
{
|
819 |
-
"epoch": 8.153846153846153,
|
820 |
-
"grad_norm": 0.7051054835319519,
|
821 |
-
"learning_rate": 0.0009230769230769232,
|
822 |
-
"loss": 1.1076,
|
823 |
-
"step": 1060
|
824 |
-
},
|
825 |
-
{
|
826 |
-
"epoch": 8.23076923076923,
|
827 |
-
"grad_norm": 0.8233575820922852,
|
828 |
-
"learning_rate": 0.0008846153846153847,
|
829 |
-
"loss": 1.1,
|
830 |
-
"step": 1070
|
831 |
-
},
|
832 |
-
{
|
833 |
-
"epoch": 8.307692307692308,
|
834 |
-
"grad_norm": 0.4401038885116577,
|
835 |
-
"learning_rate": 0.0008461538461538462,
|
836 |
-
"loss": 1.1231,
|
837 |
-
"step": 1080
|
838 |
-
},
|
839 |
-
{
|
840 |
-
"epoch": 8.384615384615385,
|
841 |
-
"grad_norm": 2.0618515014648438,
|
842 |
-
"learning_rate": 0.0008076923076923078,
|
843 |
-
"loss": 1.0877,
|
844 |
-
"step": 1090
|
845 |
-
},
|
846 |
{
|
847 |
"epoch": 8.461538461538462,
|
848 |
-
"grad_norm":
|
849 |
-
"learning_rate": 0.
|
850 |
-
"loss": 1.
|
851 |
"step": 1100
|
852 |
},
|
853 |
-
{
|
854 |
-
"epoch": 8.538461538461538,
|
855 |
-
"grad_norm": 0.9111972451210022,
|
856 |
-
"learning_rate": 0.0007307692307692308,
|
857 |
-
"loss": 1.1033,
|
858 |
-
"step": 1110
|
859 |
-
},
|
860 |
-
{
|
861 |
-
"epoch": 8.615384615384615,
|
862 |
-
"grad_norm": 2.03430438041687,
|
863 |
-
"learning_rate": 0.0006923076923076924,
|
864 |
-
"loss": 1.0906,
|
865 |
-
"step": 1120
|
866 |
-
},
|
867 |
-
{
|
868 |
-
"epoch": 8.692307692307692,
|
869 |
-
"grad_norm": 3.2121763229370117,
|
870 |
-
"learning_rate": 0.0006538461538461539,
|
871 |
-
"loss": 1.126,
|
872 |
-
"step": 1130
|
873 |
-
},
|
874 |
-
{
|
875 |
-
"epoch": 8.76923076923077,
|
876 |
-
"grad_norm": 0.34755992889404297,
|
877 |
-
"learning_rate": 0.0006153846153846154,
|
878 |
-
"loss": 1.1051,
|
879 |
-
"step": 1140
|
880 |
-
},
|
881 |
-
{
|
882 |
-
"epoch": 8.846153846153847,
|
883 |
-
"grad_norm": 0.7230011224746704,
|
884 |
-
"learning_rate": 0.000576923076923077,
|
885 |
-
"loss": 1.0721,
|
886 |
-
"step": 1150
|
887 |
-
},
|
888 |
-
{
|
889 |
-
"epoch": 8.923076923076923,
|
890 |
-
"grad_norm": 2.8672590255737305,
|
891 |
-
"learning_rate": 0.0005384615384615385,
|
892 |
-
"loss": 1.0711,
|
893 |
-
"step": 1160
|
894 |
-
},
|
895 |
{
|
896 |
"epoch": 9.0,
|
897 |
-
"
|
898 |
-
"
|
899 |
-
"
|
900 |
-
"
|
901 |
-
|
902 |
-
{
|
903 |
-
"epoch": 9.0,
|
904 |
-
"eval_accuracy": 0.3308270676691729,
|
905 |
-
"eval_loss": 1.111855149269104,
|
906 |
-
"eval_runtime": 0.9175,
|
907 |
-
"eval_samples_per_second": 144.96,
|
908 |
-
"eval_steps_per_second": 18.529,
|
909 |
"step": 1170
|
910 |
},
|
911 |
-
{
|
912 |
-
"epoch": 9.076923076923077,
|
913 |
-
"grad_norm": 0.26905354857444763,
|
914 |
-
"learning_rate": 0.0004615384615384616,
|
915 |
-
"loss": 1.1121,
|
916 |
-
"step": 1180
|
917 |
-
},
|
918 |
-
{
|
919 |
-
"epoch": 9.153846153846153,
|
920 |
-
"grad_norm": 0.22876305878162384,
|
921 |
-
"learning_rate": 0.0004230769230769231,
|
922 |
-
"loss": 1.0967,
|
923 |
-
"step": 1190
|
924 |
-
},
|
925 |
{
|
926 |
"epoch": 9.23076923076923,
|
927 |
-
"grad_norm": 0.
|
928 |
-
"learning_rate": 0.
|
929 |
-
"loss": 1.
|
930 |
"step": 1200
|
931 |
},
|
932 |
-
{
|
933 |
-
"epoch": 9.307692307692308,
|
934 |
-
"grad_norm": 0.6317471861839294,
|
935 |
-
"learning_rate": 0.0003461538461538462,
|
936 |
-
"loss": 1.1062,
|
937 |
-
"step": 1210
|
938 |
-
},
|
939 |
-
{
|
940 |
-
"epoch": 9.384615384615385,
|
941 |
-
"grad_norm": 0.41996484994888306,
|
942 |
-
"learning_rate": 0.0003076923076923077,
|
943 |
-
"loss": 1.1162,
|
944 |
-
"step": 1220
|
945 |
-
},
|
946 |
-
{
|
947 |
-
"epoch": 9.461538461538462,
|
948 |
-
"grad_norm": 1.6321758031845093,
|
949 |
-
"learning_rate": 0.0002692307692307693,
|
950 |
-
"loss": 1.0976,
|
951 |
-
"step": 1230
|
952 |
-
},
|
953 |
-
{
|
954 |
-
"epoch": 9.538461538461538,
|
955 |
-
"grad_norm": 0.966174840927124,
|
956 |
-
"learning_rate": 0.0002307692307692308,
|
957 |
-
"loss": 1.0979,
|
958 |
-
"step": 1240
|
959 |
-
},
|
960 |
-
{
|
961 |
-
"epoch": 9.615384615384615,
|
962 |
-
"grad_norm": 0.23766568303108215,
|
963 |
-
"learning_rate": 0.00019230769230769233,
|
964 |
-
"loss": 1.1012,
|
965 |
-
"step": 1250
|
966 |
-
},
|
967 |
-
{
|
968 |
-
"epoch": 9.692307692307692,
|
969 |
-
"grad_norm": 0.930920422077179,
|
970 |
-
"learning_rate": 0.00015384615384615385,
|
971 |
-
"loss": 1.0991,
|
972 |
-
"step": 1260
|
973 |
-
},
|
974 |
-
{
|
975 |
-
"epoch": 9.76923076923077,
|
976 |
-
"grad_norm": 0.34004735946655273,
|
977 |
-
"learning_rate": 0.0001153846153846154,
|
978 |
-
"loss": 1.1009,
|
979 |
-
"step": 1270
|
980 |
-
},
|
981 |
-
{
|
982 |
-
"epoch": 9.846153846153847,
|
983 |
-
"grad_norm": 0.663216233253479,
|
984 |
-
"learning_rate": 7.692307692307693e-05,
|
985 |
-
"loss": 1.0997,
|
986 |
-
"step": 1280
|
987 |
-
},
|
988 |
-
{
|
989 |
-
"epoch": 9.923076923076923,
|
990 |
-
"grad_norm": 0.9082524180412292,
|
991 |
-
"learning_rate": 3.846153846153846e-05,
|
992 |
-
"loss": 1.0998,
|
993 |
-
"step": 1290
|
994 |
-
},
|
995 |
{
|
996 |
"epoch": 10.0,
|
997 |
-
"grad_norm":
|
998 |
-
"learning_rate": 0.
|
999 |
-
"loss": 1.
|
1000 |
"step": 1300
|
1001 |
},
|
1002 |
{
|
1003 |
"epoch": 10.0,
|
1004 |
"eval_accuracy": 0.3308270676691729,
|
1005 |
-
"eval_loss": 1.
|
1006 |
-
"eval_runtime": 0.
|
1007 |
-
"eval_samples_per_second":
|
1008 |
-
"eval_steps_per_second":
|
1009 |
"step": 1300
|
1010 |
},
|
1011 |
{
|
1012 |
"epoch": 10.0,
|
1013 |
"step": 1300,
|
1014 |
-
"total_flos": 8.
|
1015 |
-
"train_loss": 0.
|
1016 |
-
"train_runtime":
|
1017 |
-
"train_samples_per_second":
|
1018 |
-
"train_steps_per_second": 7.
|
1019 |
}
|
1020 |
],
|
1021 |
-
"logging_steps":
|
1022 |
"max_steps": 1300,
|
1023 |
"num_input_tokens_seen": 0,
|
1024 |
"num_train_epochs": 10,
|
@@ -1035,7 +216,7 @@
|
|
1035 |
"attributes": {}
|
1036 |
}
|
1037 |
},
|
1038 |
-
"total_flos": 8.
|
1039 |
"train_batch_size": 8,
|
1040 |
"trial_name": null,
|
1041 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.23085123300552368,
|
3 |
"best_model_checkpoint": "./beans_outputs/checkpoint-130",
|
4 |
"epoch": 10.0,
|
5 |
"eval_steps": 500,
|
|
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
{
|
12 |
"epoch": 0.7692307692307693,
|
13 |
+
"grad_norm": 1.3140615224838257,
|
14 |
+
"learning_rate": 0.005,
|
15 |
+
"loss": 0.4537,
|
16 |
"step": 100
|
17 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
{
|
19 |
"epoch": 1.0,
|
20 |
+
"eval_accuracy": 0.8947368421052632,
|
21 |
+
"eval_loss": 0.23085123300552368,
|
22 |
+
"eval_runtime": 0.9384,
|
23 |
+
"eval_samples_per_second": 141.736,
|
24 |
+
"eval_steps_per_second": 18.117,
|
25 |
"step": 130
|
26 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
{
|
28 |
"epoch": 1.5384615384615383,
|
29 |
+
"grad_norm": 1.5738377571105957,
|
30 |
+
"learning_rate": 0.005,
|
31 |
+
"loss": 0.5034,
|
32 |
"step": 200
|
33 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
{
|
35 |
"epoch": 2.0,
|
36 |
+
"eval_accuracy": 0.6165413533834586,
|
37 |
+
"eval_loss": 1.0887765884399414,
|
38 |
+
"eval_runtime": 0.9484,
|
39 |
+
"eval_samples_per_second": 140.237,
|
40 |
+
"eval_steps_per_second": 17.925,
|
41 |
"step": 260
|
42 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
{
|
44 |
"epoch": 2.3076923076923075,
|
45 |
+
"grad_norm": 4.012147426605225,
|
46 |
+
"learning_rate": 0.005,
|
47 |
+
"loss": 0.506,
|
48 |
"step": 300
|
49 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
{
|
51 |
"epoch": 3.0,
|
52 |
+
"eval_accuracy": 0.5714285714285714,
|
53 |
+
"eval_loss": 0.9555807709693909,
|
54 |
+
"eval_runtime": 0.9238,
|
55 |
+
"eval_samples_per_second": 143.97,
|
56 |
+
"eval_steps_per_second": 18.402,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
"step": 390
|
58 |
},
|
59 |
{
|
60 |
"epoch": 3.076923076923077,
|
61 |
+
"grad_norm": 2.720919132232666,
|
62 |
+
"learning_rate": 0.005,
|
63 |
+
"loss": 0.9151,
|
64 |
"step": 400
|
65 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
{
|
67 |
"epoch": 3.8461538461538463,
|
68 |
+
"grad_norm": 0.1517225205898285,
|
69 |
+
"learning_rate": 0.005,
|
70 |
+
"loss": 1.1528,
|
71 |
"step": 500
|
72 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
73 |
{
|
74 |
"epoch": 4.0,
|
75 |
+
"eval_accuracy": 0.3308270676691729,
|
76 |
+
"eval_loss": 1.1074310541152954,
|
77 |
+
"eval_runtime": 0.9649,
|
78 |
+
"eval_samples_per_second": 137.839,
|
79 |
+
"eval_steps_per_second": 17.618,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
80 |
"step": 520
|
81 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
{
|
83 |
"epoch": 4.615384615384615,
|
84 |
+
"grad_norm": 0.09401340782642365,
|
85 |
+
"learning_rate": 0.005,
|
86 |
+
"loss": 1.1076,
|
87 |
"step": 600
|
88 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
89 |
{
|
90 |
"epoch": 5.0,
|
91 |
"eval_accuracy": 0.3308270676691729,
|
92 |
+
"eval_loss": 1.100753664970398,
|
93 |
+
"eval_runtime": 0.982,
|
94 |
+
"eval_samples_per_second": 135.436,
|
95 |
+
"eval_steps_per_second": 17.311,
|
96 |
"step": 650
|
97 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
98 |
{
|
99 |
"epoch": 5.384615384615385,
|
100 |
+
"grad_norm": 0.5329942107200623,
|
101 |
+
"learning_rate": 0.005,
|
102 |
+
"loss": 1.1148,
|
103 |
"step": 700
|
104 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
{
|
106 |
"epoch": 6.0,
|
107 |
"eval_accuracy": 0.3308270676691729,
|
108 |
+
"eval_loss": 1.1018885374069214,
|
109 |
+
"eval_runtime": 0.9388,
|
110 |
+
"eval_samples_per_second": 141.675,
|
111 |
+
"eval_steps_per_second": 18.109,
|
112 |
"step": 780
|
113 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
{
|
115 |
"epoch": 6.153846153846154,
|
116 |
+
"grad_norm": 0.4093654453754425,
|
117 |
+
"learning_rate": 0.005,
|
118 |
+
"loss": 1.1095,
|
119 |
"step": 800
|
120 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
121 |
{
|
122 |
"epoch": 6.923076923076923,
|
123 |
+
"grad_norm": 0.4959578216075897,
|
124 |
+
"learning_rate": 0.005,
|
125 |
+
"loss": 1.1026,
|
126 |
"step": 900
|
127 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
128 |
{
|
129 |
"epoch": 7.0,
|
130 |
"eval_accuracy": 0.3308270676691729,
|
131 |
+
"eval_loss": 1.1080141067504883,
|
132 |
+
"eval_runtime": 1.0873,
|
133 |
+
"eval_samples_per_second": 122.317,
|
134 |
+
"eval_steps_per_second": 15.635,
|
135 |
"step": 910
|
136 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
137 |
{
|
138 |
"epoch": 7.6923076923076925,
|
139 |
+
"grad_norm": 0.2822917699813843,
|
140 |
+
"learning_rate": 0.005,
|
141 |
+
"loss": 1.1073,
|
142 |
"step": 1000
|
143 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
144 |
{
|
145 |
"epoch": 8.0,
|
146 |
+
"eval_accuracy": 0.3308270676691729,
|
147 |
+
"eval_loss": 1.0986037254333496,
|
148 |
+
"eval_runtime": 0.9744,
|
149 |
+
"eval_samples_per_second": 136.492,
|
150 |
+
"eval_steps_per_second": 17.446,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
151 |
"step": 1040
|
152 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
{
|
154 |
"epoch": 8.461538461538462,
|
155 |
+
"grad_norm": 0.33672869205474854,
|
156 |
+
"learning_rate": 0.005,
|
157 |
+
"loss": 1.1009,
|
158 |
"step": 1100
|
159 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
160 |
{
|
161 |
"epoch": 9.0,
|
162 |
+
"eval_accuracy": 0.3383458646616541,
|
163 |
+
"eval_loss": 1.0998892784118652,
|
164 |
+
"eval_runtime": 0.9516,
|
165 |
+
"eval_samples_per_second": 139.767,
|
166 |
+
"eval_steps_per_second": 17.865,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
167 |
"step": 1170
|
168 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
169 |
{
|
170 |
"epoch": 9.23076923076923,
|
171 |
+
"grad_norm": 0.08797406405210495,
|
172 |
+
"learning_rate": 0.005,
|
173 |
+
"loss": 1.1118,
|
174 |
"step": 1200
|
175 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
176 |
{
|
177 |
"epoch": 10.0,
|
178 |
+
"grad_norm": 0.42109444737434387,
|
179 |
+
"learning_rate": 0.005,
|
180 |
+
"loss": 1.1081,
|
181 |
"step": 1300
|
182 |
},
|
183 |
{
|
184 |
"epoch": 10.0,
|
185 |
"eval_accuracy": 0.3308270676691729,
|
186 |
+
"eval_loss": 1.098858118057251,
|
187 |
+
"eval_runtime": 0.9588,
|
188 |
+
"eval_samples_per_second": 138.708,
|
189 |
+
"eval_steps_per_second": 17.73,
|
190 |
"step": 1300
|
191 |
},
|
192 |
{
|
193 |
"epoch": 10.0,
|
194 |
"step": 1300,
|
195 |
+
"total_flos": 8.772706474360013e+17,
|
196 |
+
"train_loss": 0.9533355126014122,
|
197 |
+
"train_runtime": 174.5905,
|
198 |
+
"train_samples_per_second": 59.224,
|
199 |
+
"train_steps_per_second": 7.446
|
200 |
}
|
201 |
],
|
202 |
+
"logging_steps": 100,
|
203 |
"max_steps": 1300,
|
204 |
"num_input_tokens_seen": 0,
|
205 |
"num_train_epochs": 10,
|
|
|
216 |
"attributes": {}
|
217 |
}
|
218 |
},
|
219 |
+
"total_flos": 8.772706474360013e+17,
|
220 |
"train_batch_size": 8,
|
221 |
"trial_name": null,
|
222 |
"trial_params": null
|