mikhail-panzo
commited on
Commit
•
dade094
1
Parent(s):
eb8ce83
Training in progress, step 500, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 577789320
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:eed2c86bc2c40a1659071d1ab121cc1614d73dba8e8d412c17f25aa20274aa3b
|
3 |
size 577789320
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1155772233
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:42e92a45af5353c492dbd93e2c4132cb07d4381b943c66b0c7fc1620231e6a8f
|
3 |
size 1155772233
|
last-checkpoint/rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14244
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:22229bc8272370caff8e1f2fa838a639867ecbcd2cf7b0c1722f97dc0bc4d3f7
|
3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1064
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2a1a187666ea0e44f9d015f844e1601f5b4c6844588e1b362a3c9b6a7527a74f
|
3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,714 +1,90 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "mikhail_panzo/fil_b64_le4_s8000/checkpoint-
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 2.2222222222222223,
|
13 |
-
"grad_norm":
|
14 |
"learning_rate": 2.5e-06,
|
15 |
-
"loss": 0.
|
16 |
"step": 50
|
17 |
},
|
18 |
{
|
19 |
"epoch": 4.444444444444445,
|
20 |
-
"grad_norm":
|
21 |
"learning_rate": 5e-06,
|
22 |
-
"loss": 0.
|
23 |
"step": 100
|
24 |
},
|
25 |
{
|
26 |
"epoch": 6.666666666666667,
|
27 |
-
"grad_norm": 3.
|
28 |
"learning_rate": 7.5e-06,
|
29 |
-
"loss": 0.
|
30 |
"step": 150
|
31 |
},
|
32 |
{
|
33 |
"epoch": 8.88888888888889,
|
34 |
-
"grad_norm":
|
35 |
-
"learning_rate":
|
36 |
-
"loss": 0.
|
37 |
"step": 200
|
38 |
},
|
39 |
{
|
40 |
"epoch": 11.11111111111111,
|
41 |
-
"grad_norm": 1.
|
42 |
-
"learning_rate": 1.
|
43 |
-
"loss": 0.
|
44 |
"step": 250
|
45 |
},
|
46 |
{
|
47 |
"epoch": 13.333333333333334,
|
48 |
-
"grad_norm": 1.
|
49 |
-
"learning_rate": 1.
|
50 |
-
"loss": 0.
|
51 |
"step": 300
|
52 |
},
|
53 |
{
|
54 |
"epoch": 15.555555555555555,
|
55 |
-
"grad_norm": 1.
|
56 |
-
"learning_rate": 1.
|
57 |
-
"loss": 0.
|
58 |
"step": 350
|
59 |
},
|
60 |
{
|
61 |
"epoch": 17.77777777777778,
|
62 |
-
"grad_norm":
|
63 |
-
"learning_rate":
|
64 |
-
"loss": 0.
|
65 |
"step": 400
|
66 |
},
|
67 |
{
|
68 |
"epoch": 20.0,
|
69 |
-
"grad_norm":
|
70 |
-
"learning_rate": 2.
|
71 |
-
"loss": 0.
|
72 |
"step": 450
|
73 |
},
|
74 |
{
|
75 |
"epoch": 22.22222222222222,
|
76 |
-
"grad_norm": 1.
|
77 |
-
"learning_rate": 2.
|
78 |
-
"loss": 0.
|
79 |
"step": 500
|
80 |
},
|
81 |
{
|
82 |
"epoch": 22.22222222222222,
|
83 |
-
"eval_loss": 0.
|
84 |
-
"eval_runtime":
|
85 |
-
"eval_samples_per_second":
|
86 |
-
"eval_steps_per_second": 2.
|
87 |
"step": 500
|
88 |
-
},
|
89 |
-
{
|
90 |
-
"epoch": 24.444444444444443,
|
91 |
-
"grad_norm": 1.6130620241165161,
|
92 |
-
"learning_rate": 2.7500000000000004e-05,
|
93 |
-
"loss": 0.4607,
|
94 |
-
"step": 550
|
95 |
-
},
|
96 |
-
{
|
97 |
-
"epoch": 26.666666666666668,
|
98 |
-
"grad_norm": 2.737618923187256,
|
99 |
-
"learning_rate": 3e-05,
|
100 |
-
"loss": 0.4634,
|
101 |
-
"step": 600
|
102 |
-
},
|
103 |
-
{
|
104 |
-
"epoch": 28.88888888888889,
|
105 |
-
"grad_norm": 3.0077686309814453,
|
106 |
-
"learning_rate": 3.2500000000000004e-05,
|
107 |
-
"loss": 0.4621,
|
108 |
-
"step": 650
|
109 |
-
},
|
110 |
-
{
|
111 |
-
"epoch": 31.11111111111111,
|
112 |
-
"grad_norm": 1.4074121713638306,
|
113 |
-
"learning_rate": 3.5e-05,
|
114 |
-
"loss": 0.4611,
|
115 |
-
"step": 700
|
116 |
-
},
|
117 |
-
{
|
118 |
-
"epoch": 33.333333333333336,
|
119 |
-
"grad_norm": 2.665407657623291,
|
120 |
-
"learning_rate": 3.7500000000000003e-05,
|
121 |
-
"loss": 0.4557,
|
122 |
-
"step": 750
|
123 |
-
},
|
124 |
-
{
|
125 |
-
"epoch": 35.55555555555556,
|
126 |
-
"grad_norm": 2.857210874557495,
|
127 |
-
"learning_rate": 4e-05,
|
128 |
-
"loss": 0.4493,
|
129 |
-
"step": 800
|
130 |
-
},
|
131 |
-
{
|
132 |
-
"epoch": 37.77777777777778,
|
133 |
-
"grad_norm": 1.7210990190505981,
|
134 |
-
"learning_rate": 4.25e-05,
|
135 |
-
"loss": 0.4489,
|
136 |
-
"step": 850
|
137 |
-
},
|
138 |
-
{
|
139 |
-
"epoch": 40.0,
|
140 |
-
"grad_norm": 1.4690616130828857,
|
141 |
-
"learning_rate": 4.5e-05,
|
142 |
-
"loss": 0.4494,
|
143 |
-
"step": 900
|
144 |
-
},
|
145 |
-
{
|
146 |
-
"epoch": 42.22222222222222,
|
147 |
-
"grad_norm": 1.6961876153945923,
|
148 |
-
"learning_rate": 4.75e-05,
|
149 |
-
"loss": 0.4451,
|
150 |
-
"step": 950
|
151 |
-
},
|
152 |
-
{
|
153 |
-
"epoch": 44.44444444444444,
|
154 |
-
"grad_norm": 1.8849211931228638,
|
155 |
-
"learning_rate": 5e-05,
|
156 |
-
"loss": 0.4415,
|
157 |
-
"step": 1000
|
158 |
-
},
|
159 |
-
{
|
160 |
-
"epoch": 44.44444444444444,
|
161 |
-
"eval_loss": 0.4203202724456787,
|
162 |
-
"eval_runtime": 8.7725,
|
163 |
-
"eval_samples_per_second": 18.125,
|
164 |
-
"eval_steps_per_second": 2.28,
|
165 |
-
"step": 1000
|
166 |
-
},
|
167 |
-
{
|
168 |
-
"epoch": 46.666666666666664,
|
169 |
-
"grad_norm": 4.348108768463135,
|
170 |
-
"learning_rate": 5.25e-05,
|
171 |
-
"loss": 0.4444,
|
172 |
-
"step": 1050
|
173 |
-
},
|
174 |
-
{
|
175 |
-
"epoch": 48.888888888888886,
|
176 |
-
"grad_norm": 2.2666618824005127,
|
177 |
-
"learning_rate": 5.500000000000001e-05,
|
178 |
-
"loss": 0.436,
|
179 |
-
"step": 1100
|
180 |
-
},
|
181 |
-
{
|
182 |
-
"epoch": 51.111111111111114,
|
183 |
-
"grad_norm": 2.367915630340576,
|
184 |
-
"learning_rate": 5.7499999999999995e-05,
|
185 |
-
"loss": 0.4358,
|
186 |
-
"step": 1150
|
187 |
-
},
|
188 |
-
{
|
189 |
-
"epoch": 53.333333333333336,
|
190 |
-
"grad_norm": 1.8919609785079956,
|
191 |
-
"learning_rate": 6e-05,
|
192 |
-
"loss": 0.432,
|
193 |
-
"step": 1200
|
194 |
-
},
|
195 |
-
{
|
196 |
-
"epoch": 55.55555555555556,
|
197 |
-
"grad_norm": 2.736358642578125,
|
198 |
-
"learning_rate": 6.25e-05,
|
199 |
-
"loss": 0.4373,
|
200 |
-
"step": 1250
|
201 |
-
},
|
202 |
-
{
|
203 |
-
"epoch": 57.77777777777778,
|
204 |
-
"grad_norm": 2.9132068157196045,
|
205 |
-
"learning_rate": 6.500000000000001e-05,
|
206 |
-
"loss": 0.4317,
|
207 |
-
"step": 1300
|
208 |
-
},
|
209 |
-
{
|
210 |
-
"epoch": 60.0,
|
211 |
-
"grad_norm": 1.5804557800292969,
|
212 |
-
"learning_rate": 6.750000000000001e-05,
|
213 |
-
"loss": 0.4288,
|
214 |
-
"step": 1350
|
215 |
-
},
|
216 |
-
{
|
217 |
-
"epoch": 62.22222222222222,
|
218 |
-
"grad_norm": 2.7998554706573486,
|
219 |
-
"learning_rate": 7e-05,
|
220 |
-
"loss": 0.4291,
|
221 |
-
"step": 1400
|
222 |
-
},
|
223 |
-
{
|
224 |
-
"epoch": 64.44444444444444,
|
225 |
-
"grad_norm": 1.1768637895584106,
|
226 |
-
"learning_rate": 7.25e-05,
|
227 |
-
"loss": 0.4183,
|
228 |
-
"step": 1450
|
229 |
-
},
|
230 |
-
{
|
231 |
-
"epoch": 66.66666666666667,
|
232 |
-
"grad_norm": 1.888795018196106,
|
233 |
-
"learning_rate": 7.500000000000001e-05,
|
234 |
-
"loss": 0.423,
|
235 |
-
"step": 1500
|
236 |
-
},
|
237 |
-
{
|
238 |
-
"epoch": 66.66666666666667,
|
239 |
-
"eval_loss": 0.4168904721736908,
|
240 |
-
"eval_runtime": 8.2236,
|
241 |
-
"eval_samples_per_second": 19.335,
|
242 |
-
"eval_steps_per_second": 2.432,
|
243 |
-
"step": 1500
|
244 |
-
},
|
245 |
-
{
|
246 |
-
"epoch": 68.88888888888889,
|
247 |
-
"grad_norm": 1.5093848705291748,
|
248 |
-
"learning_rate": 7.75e-05,
|
249 |
-
"loss": 0.4214,
|
250 |
-
"step": 1550
|
251 |
-
},
|
252 |
-
{
|
253 |
-
"epoch": 71.11111111111111,
|
254 |
-
"grad_norm": 3.0897819995880127,
|
255 |
-
"learning_rate": 8e-05,
|
256 |
-
"loss": 0.4217,
|
257 |
-
"step": 1600
|
258 |
-
},
|
259 |
-
{
|
260 |
-
"epoch": 73.33333333333333,
|
261 |
-
"grad_norm": 4.435822486877441,
|
262 |
-
"learning_rate": 8.25e-05,
|
263 |
-
"loss": 0.4115,
|
264 |
-
"step": 1650
|
265 |
-
},
|
266 |
-
{
|
267 |
-
"epoch": 75.55555555555556,
|
268 |
-
"grad_norm": 1.9570446014404297,
|
269 |
-
"learning_rate": 8.5e-05,
|
270 |
-
"loss": 0.414,
|
271 |
-
"step": 1700
|
272 |
-
},
|
273 |
-
{
|
274 |
-
"epoch": 77.77777777777777,
|
275 |
-
"grad_norm": 2.8186404705047607,
|
276 |
-
"learning_rate": 8.75e-05,
|
277 |
-
"loss": 0.418,
|
278 |
-
"step": 1750
|
279 |
-
},
|
280 |
-
{
|
281 |
-
"epoch": 80.0,
|
282 |
-
"grad_norm": 5.278728485107422,
|
283 |
-
"learning_rate": 9e-05,
|
284 |
-
"loss": 0.4128,
|
285 |
-
"step": 1800
|
286 |
-
},
|
287 |
-
{
|
288 |
-
"epoch": 82.22222222222223,
|
289 |
-
"grad_norm": 2.285869598388672,
|
290 |
-
"learning_rate": 9.250000000000001e-05,
|
291 |
-
"loss": 0.4144,
|
292 |
-
"step": 1850
|
293 |
-
},
|
294 |
-
{
|
295 |
-
"epoch": 84.44444444444444,
|
296 |
-
"grad_norm": 2.994182586669922,
|
297 |
-
"learning_rate": 9.5e-05,
|
298 |
-
"loss": 0.4105,
|
299 |
-
"step": 1900
|
300 |
-
},
|
301 |
-
{
|
302 |
-
"epoch": 86.66666666666667,
|
303 |
-
"grad_norm": 2.9983792304992676,
|
304 |
-
"learning_rate": 9.75e-05,
|
305 |
-
"loss": 0.414,
|
306 |
-
"step": 1950
|
307 |
-
},
|
308 |
-
{
|
309 |
-
"epoch": 88.88888888888889,
|
310 |
-
"grad_norm": 3.550610303878784,
|
311 |
-
"learning_rate": 0.0001,
|
312 |
-
"loss": 0.4108,
|
313 |
-
"step": 2000
|
314 |
-
},
|
315 |
-
{
|
316 |
-
"epoch": 88.88888888888889,
|
317 |
-
"eval_loss": 0.4182928204536438,
|
318 |
-
"eval_runtime": 7.7764,
|
319 |
-
"eval_samples_per_second": 20.446,
|
320 |
-
"eval_steps_per_second": 2.572,
|
321 |
-
"step": 2000
|
322 |
-
},
|
323 |
-
{
|
324 |
-
"epoch": 91.11111111111111,
|
325 |
-
"grad_norm": 2.4401373863220215,
|
326 |
-
"learning_rate": 9.916666666666667e-05,
|
327 |
-
"loss": 0.4159,
|
328 |
-
"step": 2050
|
329 |
-
},
|
330 |
-
{
|
331 |
-
"epoch": 93.33333333333333,
|
332 |
-
"grad_norm": 4.816738128662109,
|
333 |
-
"learning_rate": 9.835e-05,
|
334 |
-
"loss": 0.4109,
|
335 |
-
"step": 2100
|
336 |
-
},
|
337 |
-
{
|
338 |
-
"epoch": 95.55555555555556,
|
339 |
-
"grad_norm": 3.018982410430908,
|
340 |
-
"learning_rate": 9.751666666666666e-05,
|
341 |
-
"loss": 0.4072,
|
342 |
-
"step": 2150
|
343 |
-
},
|
344 |
-
{
|
345 |
-
"epoch": 97.77777777777777,
|
346 |
-
"grad_norm": 1.719527006149292,
|
347 |
-
"learning_rate": 9.668333333333334e-05,
|
348 |
-
"loss": 0.4042,
|
349 |
-
"step": 2200
|
350 |
-
},
|
351 |
-
{
|
352 |
-
"epoch": 100.0,
|
353 |
-
"grad_norm": 1.9012302160263062,
|
354 |
-
"learning_rate": 9.585000000000001e-05,
|
355 |
-
"loss": 0.4023,
|
356 |
-
"step": 2250
|
357 |
-
},
|
358 |
-
{
|
359 |
-
"epoch": 102.22222222222223,
|
360 |
-
"grad_norm": 3.062488317489624,
|
361 |
-
"learning_rate": 9.501666666666668e-05,
|
362 |
-
"loss": 0.406,
|
363 |
-
"step": 2300
|
364 |
-
},
|
365 |
-
{
|
366 |
-
"epoch": 104.44444444444444,
|
367 |
-
"grad_norm": 2.6211459636688232,
|
368 |
-
"learning_rate": 9.418333333333334e-05,
|
369 |
-
"loss": 0.405,
|
370 |
-
"step": 2350
|
371 |
-
},
|
372 |
-
{
|
373 |
-
"epoch": 106.66666666666667,
|
374 |
-
"grad_norm": 1.7554346323013306,
|
375 |
-
"learning_rate": 9.335e-05,
|
376 |
-
"loss": 0.4052,
|
377 |
-
"step": 2400
|
378 |
-
},
|
379 |
-
{
|
380 |
-
"epoch": 108.88888888888889,
|
381 |
-
"grad_norm": 1.7569692134857178,
|
382 |
-
"learning_rate": 9.251666666666667e-05,
|
383 |
-
"loss": 0.3979,
|
384 |
-
"step": 2450
|
385 |
-
},
|
386 |
-
{
|
387 |
-
"epoch": 111.11111111111111,
|
388 |
-
"grad_norm": 0.9594977498054504,
|
389 |
-
"learning_rate": 9.168333333333333e-05,
|
390 |
-
"loss": 0.396,
|
391 |
-
"step": 2500
|
392 |
-
},
|
393 |
-
{
|
394 |
-
"epoch": 111.11111111111111,
|
395 |
-
"eval_loss": 0.4142945408821106,
|
396 |
-
"eval_runtime": 8.1333,
|
397 |
-
"eval_samples_per_second": 19.549,
|
398 |
-
"eval_steps_per_second": 2.459,
|
399 |
-
"step": 2500
|
400 |
-
},
|
401 |
-
{
|
402 |
-
"epoch": 113.33333333333333,
|
403 |
-
"grad_norm": 2.79782772064209,
|
404 |
-
"learning_rate": 9.085e-05,
|
405 |
-
"loss": 0.3945,
|
406 |
-
"step": 2550
|
407 |
-
},
|
408 |
-
{
|
409 |
-
"epoch": 115.55555555555556,
|
410 |
-
"grad_norm": 2.9776482582092285,
|
411 |
-
"learning_rate": 9.001666666666667e-05,
|
412 |
-
"loss": 0.3901,
|
413 |
-
"step": 2600
|
414 |
-
},
|
415 |
-
{
|
416 |
-
"epoch": 117.77777777777777,
|
417 |
-
"grad_norm": 1.6582765579223633,
|
418 |
-
"learning_rate": 8.918333333333334e-05,
|
419 |
-
"loss": 0.3897,
|
420 |
-
"step": 2650
|
421 |
-
},
|
422 |
-
{
|
423 |
-
"epoch": 120.0,
|
424 |
-
"grad_norm": 2.3238115310668945,
|
425 |
-
"learning_rate": 8.834999999999999e-05,
|
426 |
-
"loss": 0.3946,
|
427 |
-
"step": 2700
|
428 |
-
},
|
429 |
-
{
|
430 |
-
"epoch": 122.22222222222223,
|
431 |
-
"grad_norm": 1.783127784729004,
|
432 |
-
"learning_rate": 8.751666666666668e-05,
|
433 |
-
"loss": 0.3956,
|
434 |
-
"step": 2750
|
435 |
-
},
|
436 |
-
{
|
437 |
-
"epoch": 124.44444444444444,
|
438 |
-
"grad_norm": 1.8057986497879028,
|
439 |
-
"learning_rate": 8.668333333333334e-05,
|
440 |
-
"loss": 0.3884,
|
441 |
-
"step": 2800
|
442 |
-
},
|
443 |
-
{
|
444 |
-
"epoch": 126.66666666666667,
|
445 |
-
"grad_norm": 1.6317520141601562,
|
446 |
-
"learning_rate": 8.585000000000001e-05,
|
447 |
-
"loss": 0.3895,
|
448 |
-
"step": 2850
|
449 |
-
},
|
450 |
-
{
|
451 |
-
"epoch": 128.88888888888889,
|
452 |
-
"grad_norm": 1.802009105682373,
|
453 |
-
"learning_rate": 8.501666666666667e-05,
|
454 |
-
"loss": 0.3944,
|
455 |
-
"step": 2900
|
456 |
-
},
|
457 |
-
{
|
458 |
-
"epoch": 131.11111111111111,
|
459 |
-
"grad_norm": 2.605529308319092,
|
460 |
-
"learning_rate": 8.418333333333334e-05,
|
461 |
-
"loss": 0.3951,
|
462 |
-
"step": 2950
|
463 |
-
},
|
464 |
-
{
|
465 |
-
"epoch": 133.33333333333334,
|
466 |
-
"grad_norm": 1.2800849676132202,
|
467 |
-
"learning_rate": 8.335e-05,
|
468 |
-
"loss": 0.3883,
|
469 |
-
"step": 3000
|
470 |
-
},
|
471 |
-
{
|
472 |
-
"epoch": 133.33333333333334,
|
473 |
-
"eval_loss": 0.4183387756347656,
|
474 |
-
"eval_runtime": 9.0098,
|
475 |
-
"eval_samples_per_second": 17.647,
|
476 |
-
"eval_steps_per_second": 2.22,
|
477 |
-
"step": 3000
|
478 |
-
},
|
479 |
-
{
|
480 |
-
"epoch": 135.55555555555554,
|
481 |
-
"grad_norm": 1.146140456199646,
|
482 |
-
"learning_rate": 8.251666666666668e-05,
|
483 |
-
"loss": 0.3853,
|
484 |
-
"step": 3050
|
485 |
-
},
|
486 |
-
{
|
487 |
-
"epoch": 137.77777777777777,
|
488 |
-
"grad_norm": 1.3666647672653198,
|
489 |
-
"learning_rate": 8.168333333333333e-05,
|
490 |
-
"loss": 0.3815,
|
491 |
-
"step": 3100
|
492 |
-
},
|
493 |
-
{
|
494 |
-
"epoch": 140.0,
|
495 |
-
"grad_norm": 1.4111328125,
|
496 |
-
"learning_rate": 8.085e-05,
|
497 |
-
"loss": 0.3831,
|
498 |
-
"step": 3150
|
499 |
-
},
|
500 |
-
{
|
501 |
-
"epoch": 142.22222222222223,
|
502 |
-
"grad_norm": 3.635448455810547,
|
503 |
-
"learning_rate": 8.001666666666667e-05,
|
504 |
-
"loss": 0.3838,
|
505 |
-
"step": 3200
|
506 |
-
},
|
507 |
-
{
|
508 |
-
"epoch": 144.44444444444446,
|
509 |
-
"grad_norm": 1.2508800029754639,
|
510 |
-
"learning_rate": 7.918333333333334e-05,
|
511 |
-
"loss": 0.3805,
|
512 |
-
"step": 3250
|
513 |
-
},
|
514 |
-
{
|
515 |
-
"epoch": 146.66666666666666,
|
516 |
-
"grad_norm": 1.908144235610962,
|
517 |
-
"learning_rate": 7.835000000000001e-05,
|
518 |
-
"loss": 0.3802,
|
519 |
-
"step": 3300
|
520 |
-
},
|
521 |
-
{
|
522 |
-
"epoch": 148.88888888888889,
|
523 |
-
"grad_norm": 1.8435084819793701,
|
524 |
-
"learning_rate": 7.751666666666668e-05,
|
525 |
-
"loss": 0.3764,
|
526 |
-
"step": 3350
|
527 |
-
},
|
528 |
-
{
|
529 |
-
"epoch": 151.11111111111111,
|
530 |
-
"grad_norm": 2.1144354343414307,
|
531 |
-
"learning_rate": 7.668333333333335e-05,
|
532 |
-
"loss": 0.3781,
|
533 |
-
"step": 3400
|
534 |
-
},
|
535 |
-
{
|
536 |
-
"epoch": 153.33333333333334,
|
537 |
-
"grad_norm": 2.9214673042297363,
|
538 |
-
"learning_rate": 7.585e-05,
|
539 |
-
"loss": 0.3766,
|
540 |
-
"step": 3450
|
541 |
-
},
|
542 |
-
{
|
543 |
-
"epoch": 155.55555555555554,
|
544 |
-
"grad_norm": 1.5731481313705444,
|
545 |
-
"learning_rate": 7.501666666666667e-05,
|
546 |
-
"loss": 0.3771,
|
547 |
-
"step": 3500
|
548 |
-
},
|
549 |
-
{
|
550 |
-
"epoch": 155.55555555555554,
|
551 |
-
"eval_loss": 0.42192450165748596,
|
552 |
-
"eval_runtime": 8.6123,
|
553 |
-
"eval_samples_per_second": 18.462,
|
554 |
-
"eval_steps_per_second": 2.322,
|
555 |
-
"step": 3500
|
556 |
-
},
|
557 |
-
{
|
558 |
-
"epoch": 157.77777777777777,
|
559 |
-
"grad_norm": 1.3311264514923096,
|
560 |
-
"learning_rate": 7.418333333333334e-05,
|
561 |
-
"loss": 0.3807,
|
562 |
-
"step": 3550
|
563 |
-
},
|
564 |
-
{
|
565 |
-
"epoch": 160.0,
|
566 |
-
"grad_norm": 1.186661720275879,
|
567 |
-
"learning_rate": 7.335000000000001e-05,
|
568 |
-
"loss": 0.3768,
|
569 |
-
"step": 3600
|
570 |
-
},
|
571 |
-
{
|
572 |
-
"epoch": 162.22222222222223,
|
573 |
-
"grad_norm": 1.4181159734725952,
|
574 |
-
"learning_rate": 7.251666666666666e-05,
|
575 |
-
"loss": 0.3735,
|
576 |
-
"step": 3650
|
577 |
-
},
|
578 |
-
{
|
579 |
-
"epoch": 164.44444444444446,
|
580 |
-
"grad_norm": 2.7172157764434814,
|
581 |
-
"learning_rate": 7.168333333333333e-05,
|
582 |
-
"loss": 0.3741,
|
583 |
-
"step": 3700
|
584 |
-
},
|
585 |
-
{
|
586 |
-
"epoch": 166.66666666666666,
|
587 |
-
"grad_norm": 1.6093742847442627,
|
588 |
-
"learning_rate": 7.085e-05,
|
589 |
-
"loss": 0.3721,
|
590 |
-
"step": 3750
|
591 |
-
},
|
592 |
-
{
|
593 |
-
"epoch": 168.88888888888889,
|
594 |
-
"grad_norm": 1.3624290227890015,
|
595 |
-
"learning_rate": 7.001666666666667e-05,
|
596 |
-
"loss": 0.37,
|
597 |
-
"step": 3800
|
598 |
-
},
|
599 |
-
{
|
600 |
-
"epoch": 171.11111111111111,
|
601 |
-
"grad_norm": 1.287617564201355,
|
602 |
-
"learning_rate": 6.918333333333334e-05,
|
603 |
-
"loss": 0.3655,
|
604 |
-
"step": 3850
|
605 |
-
},
|
606 |
-
{
|
607 |
-
"epoch": 173.33333333333334,
|
608 |
-
"grad_norm": 1.69022536277771,
|
609 |
-
"learning_rate": 6.835000000000001e-05,
|
610 |
-
"loss": 0.37,
|
611 |
-
"step": 3900
|
612 |
-
},
|
613 |
-
{
|
614 |
-
"epoch": 175.55555555555554,
|
615 |
-
"grad_norm": 2.1261706352233887,
|
616 |
-
"learning_rate": 6.751666666666668e-05,
|
617 |
-
"loss": 0.3736,
|
618 |
-
"step": 3950
|
619 |
-
},
|
620 |
-
{
|
621 |
-
"epoch": 177.77777777777777,
|
622 |
-
"grad_norm": 1.6689108610153198,
|
623 |
-
"learning_rate": 6.668333333333333e-05,
|
624 |
-
"loss": 0.3731,
|
625 |
-
"step": 4000
|
626 |
-
},
|
627 |
-
{
|
628 |
-
"epoch": 177.77777777777777,
|
629 |
-
"eval_loss": 0.4211507737636566,
|
630 |
-
"eval_runtime": 8.6906,
|
631 |
-
"eval_samples_per_second": 18.296,
|
632 |
-
"eval_steps_per_second": 2.301,
|
633 |
-
"step": 4000
|
634 |
-
},
|
635 |
-
{
|
636 |
-
"epoch": 180.0,
|
637 |
-
"grad_norm": 2.798970937728882,
|
638 |
-
"learning_rate": 6.585e-05,
|
639 |
-
"loss": 0.3657,
|
640 |
-
"step": 4050
|
641 |
-
},
|
642 |
-
{
|
643 |
-
"epoch": 182.22222222222223,
|
644 |
-
"grad_norm": 1.2629871368408203,
|
645 |
-
"learning_rate": 6.501666666666667e-05,
|
646 |
-
"loss": 0.3666,
|
647 |
-
"step": 4100
|
648 |
-
},
|
649 |
-
{
|
650 |
-
"epoch": 184.44444444444446,
|
651 |
-
"grad_norm": 0.970268964767456,
|
652 |
-
"learning_rate": 6.418333333333334e-05,
|
653 |
-
"loss": 0.3668,
|
654 |
-
"step": 4150
|
655 |
-
},
|
656 |
-
{
|
657 |
-
"epoch": 186.66666666666666,
|
658 |
-
"grad_norm": 1.7865740060806274,
|
659 |
-
"learning_rate": 6.335e-05,
|
660 |
-
"loss": 0.364,
|
661 |
-
"step": 4200
|
662 |
-
},
|
663 |
-
{
|
664 |
-
"epoch": 188.88888888888889,
|
665 |
-
"grad_norm": 0.8204932808876038,
|
666 |
-
"learning_rate": 6.251666666666666e-05,
|
667 |
-
"loss": 0.3647,
|
668 |
-
"step": 4250
|
669 |
-
},
|
670 |
-
{
|
671 |
-
"epoch": 191.11111111111111,
|
672 |
-
"grad_norm": 0.910372257232666,
|
673 |
-
"learning_rate": 6.168333333333333e-05,
|
674 |
-
"loss": 0.3676,
|
675 |
-
"step": 4300
|
676 |
-
},
|
677 |
-
{
|
678 |
-
"epoch": 193.33333333333334,
|
679 |
-
"grad_norm": 1.2381336688995361,
|
680 |
-
"learning_rate": 6.085000000000001e-05,
|
681 |
-
"loss": 0.3632,
|
682 |
-
"step": 4350
|
683 |
-
},
|
684 |
-
{
|
685 |
-
"epoch": 195.55555555555554,
|
686 |
-
"grad_norm": 1.588180661201477,
|
687 |
-
"learning_rate": 6.0016666666666664e-05,
|
688 |
-
"loss": 0.3621,
|
689 |
-
"step": 4400
|
690 |
-
},
|
691 |
-
{
|
692 |
-
"epoch": 197.77777777777777,
|
693 |
-
"grad_norm": 1.093023419380188,
|
694 |
-
"learning_rate": 5.918333333333333e-05,
|
695 |
-
"loss": 0.3609,
|
696 |
-
"step": 4450
|
697 |
-
},
|
698 |
-
{
|
699 |
-
"epoch": 200.0,
|
700 |
-
"grad_norm": 1.071207046508789,
|
701 |
-
"learning_rate": 5.835e-05,
|
702 |
-
"loss": 0.362,
|
703 |
-
"step": 4500
|
704 |
-
},
|
705 |
-
{
|
706 |
-
"epoch": 200.0,
|
707 |
-
"eval_loss": 0.42145419120788574,
|
708 |
-
"eval_runtime": 7.8841,
|
709 |
-
"eval_samples_per_second": 20.167,
|
710 |
-
"eval_steps_per_second": 2.537,
|
711 |
-
"step": 4500
|
712 |
}
|
713 |
],
|
714 |
"logging_steps": 50,
|
@@ -728,8 +104,8 @@
|
|
728 |
"attributes": {}
|
729 |
}
|
730 |
},
|
731 |
-
"total_flos":
|
732 |
-
"train_batch_size":
|
733 |
"trial_name": null,
|
734 |
"trial_params": null
|
735 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.4381416440010071,
|
3 |
+
"best_model_checkpoint": "mikhail_panzo/fil_b64_le4_s8000/checkpoint-500",
|
4 |
+
"epoch": 22.22222222222222,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 500,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 2.2222222222222223,
|
13 |
+
"grad_norm": 1.774868130683899,
|
14 |
"learning_rate": 2.5e-06,
|
15 |
+
"loss": 0.7871,
|
16 |
"step": 50
|
17 |
},
|
18 |
{
|
19 |
"epoch": 4.444444444444445,
|
20 |
+
"grad_norm": 4.1770920753479,
|
21 |
"learning_rate": 5e-06,
|
22 |
+
"loss": 0.7106,
|
23 |
"step": 100
|
24 |
},
|
25 |
{
|
26 |
"epoch": 6.666666666666667,
|
27 |
+
"grad_norm": 3.3354010581970215,
|
28 |
"learning_rate": 7.5e-06,
|
29 |
+
"loss": 0.6568,
|
30 |
"step": 150
|
31 |
},
|
32 |
{
|
33 |
"epoch": 8.88888888888889,
|
34 |
+
"grad_norm": 2.2018914222717285,
|
35 |
+
"learning_rate": 9.950000000000001e-06,
|
36 |
+
"loss": 0.577,
|
37 |
"step": 200
|
38 |
},
|
39 |
{
|
40 |
"epoch": 11.11111111111111,
|
41 |
+
"grad_norm": 1.4593122005462646,
|
42 |
+
"learning_rate": 1.2450000000000001e-05,
|
43 |
+
"loss": 0.5321,
|
44 |
"step": 250
|
45 |
},
|
46 |
{
|
47 |
"epoch": 13.333333333333334,
|
48 |
+
"grad_norm": 1.9025092124938965,
|
49 |
+
"learning_rate": 1.4950000000000001e-05,
|
50 |
+
"loss": 0.5224,
|
51 |
"step": 300
|
52 |
},
|
53 |
{
|
54 |
"epoch": 15.555555555555555,
|
55 |
+
"grad_norm": 1.1274640560150146,
|
56 |
+
"learning_rate": 1.745e-05,
|
57 |
+
"loss": 0.5033,
|
58 |
"step": 350
|
59 |
},
|
60 |
{
|
61 |
"epoch": 17.77777777777778,
|
62 |
+
"grad_norm": 1.615440011024475,
|
63 |
+
"learning_rate": 1.995e-05,
|
64 |
+
"loss": 0.4884,
|
65 |
"step": 400
|
66 |
},
|
67 |
{
|
68 |
"epoch": 20.0,
|
69 |
+
"grad_norm": 1.224423885345459,
|
70 |
+
"learning_rate": 2.245e-05,
|
71 |
+
"loss": 0.4891,
|
72 |
"step": 450
|
73 |
},
|
74 |
{
|
75 |
"epoch": 22.22222222222222,
|
76 |
+
"grad_norm": 1.3765957355499268,
|
77 |
+
"learning_rate": 2.495e-05,
|
78 |
+
"loss": 0.4811,
|
79 |
"step": 500
|
80 |
},
|
81 |
{
|
82 |
"epoch": 22.22222222222222,
|
83 |
+
"eval_loss": 0.4381416440010071,
|
84 |
+
"eval_runtime": 7.3061,
|
85 |
+
"eval_samples_per_second": 21.763,
|
86 |
+
"eval_steps_per_second": 2.737,
|
87 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
88 |
}
|
89 |
],
|
90 |
"logging_steps": 50,
|
|
|
104 |
"attributes": {}
|
105 |
}
|
106 |
},
|
107 |
+
"total_flos": 7257884840759520.0,
|
108 |
+
"train_batch_size": 32,
|
109 |
"trial_name": null,
|
110 |
"trial_params": null
|
111 |
}
|
last-checkpoint/training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5304
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b20fa88d77a29f6532fc31932997f0c8ed961dcd3657a7862d793575cbe78fcd
|
3 |
size 5304
|