MHGanainy/8-clusters-balanced-lex-best-v2-3
Browse files- all_results.json +9 -9
- eval_results.json +5 -5
- train_results.json +4 -4
- trainer_state.json +157 -157
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 1.
|
4 |
-
"eval_runtime": 63.
|
5 |
-
"eval_samples_per_second": 15.
|
6 |
-
"eval_steps_per_second": 1.
|
7 |
-
"perplexity": 6.
|
8 |
"total_flos": 1.5640093507584e+17,
|
9 |
-
"train_loss": 2.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second": 6.
|
12 |
-
"train_steps_per_second": 3.
|
13 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 1.9266563653945923,
|
4 |
+
"eval_runtime": 63.5112,
|
5 |
+
"eval_samples_per_second": 15.588,
|
6 |
+
"eval_steps_per_second": 1.952,
|
7 |
+
"perplexity": 6.866512705827251,
|
8 |
"total_flos": 1.5640093507584e+17,
|
9 |
+
"train_loss": 2.1807946249230077,
|
10 |
+
"train_runtime": 2485.1722,
|
11 |
+
"train_samples_per_second": 6.91,
|
12 |
+
"train_steps_per_second": 3.455
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 1.
|
4 |
-
"eval_runtime": 63.
|
5 |
-
"eval_samples_per_second": 15.
|
6 |
-
"eval_steps_per_second": 1.
|
7 |
-
"perplexity": 6.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 1.9266563653945923,
|
4 |
+
"eval_runtime": 63.5112,
|
5 |
+
"eval_samples_per_second": 15.588,
|
6 |
+
"eval_steps_per_second": 1.952,
|
7 |
+
"perplexity": 6.866512705827251
|
8 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
"total_flos": 1.5640093507584e+17,
|
4 |
-
"train_loss": 2.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second": 6.
|
7 |
-
"train_steps_per_second": 3.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
"total_flos": 1.5640093507584e+17,
|
4 |
+
"train_loss": 2.1807946249230077,
|
5 |
+
"train_runtime": 2485.1722,
|
6 |
+
"train_samples_per_second": 6.91,
|
7 |
+
"train_steps_per_second": 3.455
|
8 |
}
|
trainer_state.json
CHANGED
@@ -10,615 +10,615 @@
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.01164551065564225,
|
13 |
-
"grad_norm": 0.
|
14 |
"learning_rate": 2.331002331002331e-06,
|
15 |
"loss": 2.4953,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.0232910213112845,
|
20 |
-
"grad_norm": 0.
|
21 |
"learning_rate": 4.662004662004662e-06,
|
22 |
-
"loss": 2.
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.03493653196692675,
|
27 |
-
"grad_norm": 0.
|
28 |
"learning_rate": 6.993006993006993e-06,
|
29 |
"loss": 2.4409,
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.046582042622569,
|
34 |
-
"grad_norm": 0.
|
35 |
"learning_rate": 9.324009324009324e-06,
|
36 |
"loss": 2.4497,
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.05822755327821125,
|
41 |
-
"grad_norm": 0.
|
42 |
"learning_rate": 1.1655011655011657e-05,
|
43 |
-
"loss": 2.
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.0698730639338535,
|
48 |
-
"grad_norm": 0.
|
49 |
"learning_rate": 1.3986013986013986e-05,
|
50 |
-
"loss": 2.
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.08151857458949575,
|
55 |
-
"grad_norm": 0.
|
56 |
"learning_rate": 1.6317016317016318e-05,
|
57 |
-
"loss": 2.
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.093164085245138,
|
62 |
-
"grad_norm": 0.
|
63 |
"learning_rate": 1.8648018648018647e-05,
|
64 |
-
"loss": 2.
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.10480959590078025,
|
69 |
-
"grad_norm": 0.
|
70 |
"learning_rate": 1.999854282682324e-05,
|
71 |
-
"loss": 2.
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.1164551065564225,
|
76 |
-
"grad_norm": 0.
|
77 |
"learning_rate": 1.9983347507285766e-05,
|
78 |
-
"loss": 2.
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.12810061721206475,
|
83 |
-
"grad_norm": 0.
|
84 |
"learning_rate": 1.9951660332101616e-05,
|
85 |
-
"loss": 2.
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.139746127867707,
|
90 |
-
"grad_norm": 0.
|
91 |
"learning_rate": 1.9903533646470504e-05,
|
92 |
-
"loss": 2.
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.15139163852334925,
|
97 |
-
"grad_norm": 0.
|
98 |
"learning_rate": 1.9839046952618667e-05,
|
99 |
-
"loss": 2.
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.1630371491789915,
|
104 |
-
"grad_norm": 0.
|
105 |
"learning_rate": 1.9758306778466264e-05,
|
106 |
-
"loss": 2.
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.17468265983463374,
|
111 |
-
"grad_norm": 0.
|
112 |
"learning_rate": 1.966144650165e-05,
|
113 |
-
"loss": 2.
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.186328170490276,
|
118 |
-
"grad_norm": 0.
|
119 |
"learning_rate": 1.9548626129191778e-05,
|
120 |
-
"loss": 2.
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.19797368114591826,
|
125 |
-
"grad_norm": 0.
|
126 |
"learning_rate": 1.9420032033177225e-05,
|
127 |
-
"loss": 2.
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.2096191918015605,
|
132 |
-
"grad_norm": 0.
|
133 |
"learning_rate": 1.927587664288089e-05,
|
134 |
-
"loss": 2.
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.22126470245720276,
|
139 |
-
"grad_norm": 0.
|
140 |
"learning_rate": 1.911639809384656e-05,
|
141 |
-
"loss": 2.
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.232910213112845,
|
146 |
-
"grad_norm":
|
147 |
"learning_rate": 1.8941859834502484e-05,
|
148 |
-
"loss": 2.
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.24455572376848725,
|
153 |
-
"grad_norm": 0.
|
154 |
"learning_rate": 1.8752550190961288e-05,
|
155 |
-
"loss": 2.
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.2562012344241295,
|
160 |
-
"grad_norm": 0.
|
161 |
"learning_rate": 1.8548781890723614e-05,
|
162 |
-
"loss": 2.
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.26784674507977174,
|
167 |
-
"grad_norm": 0.
|
168 |
"learning_rate": 1.8330891546072095e-05,
|
169 |
-
"loss": 2.
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.279492255735414,
|
174 |
-
"grad_norm": 0.
|
175 |
"learning_rate": 1.809923909800931e-05,
|
176 |
"loss": 2.1932,
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.29113776639105626,
|
181 |
-
"grad_norm": 0.
|
182 |
"learning_rate": 1.7854207221658092e-05,
|
183 |
-
"loss": 2.
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.3027832770466985,
|
188 |
-
"grad_norm": 0.
|
189 |
"learning_rate": 1.7596200694106552e-05,
|
190 |
-
"loss": 2.
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.31442878770234073,
|
195 |
-
"grad_norm": 0.
|
196 |
"learning_rate": 1.7325645725742056e-05,
|
197 |
-
"loss": 2.
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.326074298357983,
|
202 |
-
"grad_norm": 0.
|
203 |
"learning_rate": 1.7042989256178744e-05,
|
204 |
-
"loss": 2.
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.33771980901362525,
|
209 |
-
"grad_norm": 0.
|
210 |
"learning_rate": 1.6748698215941704e-05,
|
211 |
-
"loss": 2.
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.3493653196692675,
|
216 |
-
"grad_norm": 0.
|
217 |
"learning_rate": 1.6443258755127393e-05,
|
218 |
-
"loss": 2.
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.36101083032490977,
|
223 |
-
"grad_norm": 0.
|
224 |
"learning_rate": 1.6127175440314596e-05,
|
225 |
"loss": 2.1204,
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.372656340980552,
|
230 |
-
"grad_norm": 0.
|
231 |
"learning_rate": 1.5800970421052487e-05,
|
232 |
-
"loss": 2.
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
"epoch": 0.38430185163619424,
|
237 |
-
"grad_norm": 1.
|
238 |
"learning_rate": 1.546518256730277e-05,
|
239 |
-
"loss": 2.
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
"epoch": 0.3959473622918365,
|
244 |
-
"grad_norm": 1.
|
245 |
"learning_rate": 1.5120366579260734e-05,
|
246 |
"loss": 2.1607,
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
"epoch": 0.40759287294747876,
|
251 |
-
"grad_norm": 0.
|
252 |
"learning_rate": 1.4767092071025792e-05,
|
253 |
-
"loss": 2.
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
"epoch": 0.419238383603121,
|
258 |
-
"grad_norm": 0.
|
259 |
"learning_rate": 1.4405942629635174e-05,
|
260 |
-
"loss": 2.
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.4308838942587632,
|
265 |
-
"grad_norm": 0.
|
266 |
"learning_rate": 1.4037514851015241e-05,
|
267 |
"loss": 2.215,
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
"epoch": 0.4425294049144055,
|
272 |
-
"grad_norm": 0.
|
273 |
"learning_rate": 1.3662417354442924e-05,
|
274 |
"loss": 2.1836,
|
275 |
"step": 3800
|
276 |
},
|
277 |
{
|
278 |
"epoch": 0.45417491557004774,
|
279 |
-
"grad_norm": 0.
|
280 |
"learning_rate": 1.3281269777145354e-05,
|
281 |
-
"loss": 2.
|
282 |
"step": 3900
|
283 |
},
|
284 |
{
|
285 |
"epoch": 0.46582042622569,
|
286 |
-
"grad_norm": 1.
|
287 |
"learning_rate": 1.2894701750698541e-05,
|
288 |
-
"loss": 2.
|
289 |
"step": 4000
|
290 |
},
|
291 |
{
|
292 |
"epoch": 0.47746593688133226,
|
293 |
-
"grad_norm": 0.
|
294 |
"learning_rate": 1.2503351860916024e-05,
|
295 |
-
"loss": 2.
|
296 |
"step": 4100
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.4891114475369745,
|
300 |
-
"grad_norm": 1.
|
301 |
"learning_rate": 1.2107866592945686e-05,
|
302 |
-
"loss": 2.
|
303 |
"step": 4200
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.5007569581926168,
|
307 |
-
"grad_norm": 0.
|
308 |
"learning_rate": 1.1708899263317381e-05,
|
309 |
-
"loss": 2.
|
310 |
"step": 4300
|
311 |
},
|
312 |
{
|
313 |
"epoch": 0.512402468848259,
|
314 |
-
"grad_norm": 0.
|
315 |
"learning_rate": 1.1307108940705536e-05,
|
316 |
"loss": 2.1605,
|
317 |
"step": 4400
|
318 |
},
|
319 |
{
|
320 |
"epoch": 0.5240479795039013,
|
321 |
-
"grad_norm": 0.
|
322 |
"learning_rate": 1.090315935718958e-05,
|
323 |
"loss": 2.1108,
|
324 |
"step": 4500
|
325 |
},
|
326 |
{
|
327 |
"epoch": 0.5356934901595435,
|
328 |
-
"grad_norm": 0.
|
329 |
"learning_rate": 1.0497717811810748e-05,
|
330 |
-
"loss": 2.
|
331 |
"step": 4600
|
332 |
},
|
333 |
{
|
334 |
"epoch": 0.5473390008151857,
|
335 |
-
"grad_norm": 0.
|
336 |
"learning_rate": 1.0091454068236455e-05,
|
337 |
"loss": 2.1498,
|
338 |
"step": 4700
|
339 |
},
|
340 |
{
|
341 |
"epoch": 0.558984511470828,
|
342 |
-
"grad_norm": 1.
|
343 |
"learning_rate": 9.685039248353284e-06,
|
344 |
-
"loss": 2.
|
345 |
"step": 4800
|
346 |
},
|
347 |
{
|
348 |
"epoch": 0.5706300221264703,
|
349 |
-
"grad_norm": 0.
|
350 |
"learning_rate": 9.279144723616279e-06,
|
351 |
-
"loss": 2.
|
352 |
"step": 4900
|
353 |
},
|
354 |
{
|
355 |
"epoch": 0.5822755327821125,
|
356 |
-
"grad_norm": 0.
|
357 |
"learning_rate": 8.874441005985965e-06,
|
358 |
-
"loss": 2.
|
359 |
"step": 5000
|
360 |
},
|
361 |
{
|
362 |
"epoch": 0.5939210434377548,
|
363 |
-
"grad_norm": 1.
|
364 |
"learning_rate": 8.47159664028521e-06,
|
365 |
"loss": 2.1631,
|
366 |
"step": 5100
|
367 |
},
|
368 |
{
|
369 |
"epoch": 0.605566554093397,
|
370 |
-
"grad_norm": 0.
|
371 |
"learning_rate": 8.07127709980564e-06,
|
372 |
-
"loss": 2.
|
373 |
"step": 5200
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.6172120647490392,
|
377 |
-
"grad_norm": 1.
|
378 |
"learning_rate": 7.674143686988085e-06,
|
379 |
"loss": 2.0871,
|
380 |
"step": 5300
|
381 |
},
|
382 |
{
|
383 |
"epoch": 0.6288575754046815,
|
384 |
-
"grad_norm": 0.
|
385 |
"learning_rate": 7.280852440992941e-06,
|
386 |
"loss": 2.1289,
|
387 |
"step": 5400
|
388 |
},
|
389 |
{
|
390 |
"epoch": 0.6405030860603237,
|
391 |
-
"grad_norm": 0.
|
392 |
"learning_rate": 6.89205305396518e-06,
|
393 |
-
"loss": 2.
|
394 |
"step": 5500
|
395 |
},
|
396 |
{
|
397 |
"epoch": 0.652148596715966,
|
398 |
-
"grad_norm": 0.
|
399 |
"learning_rate": 6.508387797784227e-06,
|
400 |
-
"loss": 2.
|
401 |
"step": 5600
|
402 |
},
|
403 |
{
|
404 |
"epoch": 0.6637941073716083,
|
405 |
-
"grad_norm": 0.
|
406 |
"learning_rate": 6.130490463071604e-06,
|
407 |
-
"loss": 2.
|
408 |
"step": 5700
|
409 |
},
|
410 |
{
|
411 |
"epoch": 0.6754396180272505,
|
412 |
-
"grad_norm": 0.
|
413 |
"learning_rate": 5.758985312209124e-06,
|
414 |
-
"loss": 2.
|
415 |
"step": 5800
|
416 |
},
|
417 |
{
|
418 |
"epoch": 0.6870851286828927,
|
419 |
-
"grad_norm": 1.
|
420 |
"learning_rate": 5.394486048097099e-06,
|
421 |
-
"loss": 2.
|
422 |
"step": 5900
|
423 |
},
|
424 |
{
|
425 |
"epoch": 0.698730639338535,
|
426 |
-
"grad_norm": 0.
|
427 |
"learning_rate": 5.037594800356142e-06,
|
428 |
-
"loss": 2.
|
429 |
"step": 6000
|
430 |
},
|
431 |
{
|
432 |
"epoch": 0.7103761499941772,
|
433 |
-
"grad_norm": 1.
|
434 |
"learning_rate": 4.688901130647314e-06,
|
435 |
"loss": 2.1396,
|
436 |
"step": 6100
|
437 |
},
|
438 |
{
|
439 |
"epoch": 0.7220216606498195,
|
440 |
-
"grad_norm": 0.
|
441 |
"learning_rate": 4.348981058753708e-06,
|
442 |
"loss": 2.0819,
|
443 |
"step": 6200
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.7336671713054618,
|
447 |
-
"grad_norm": 1.
|
448 |
"learning_rate": 4.018396111032394e-06,
|
449 |
-
"loss": 2.
|
450 |
"step": 6300
|
451 |
},
|
452 |
{
|
453 |
"epoch": 0.745312681961104,
|
454 |
-
"grad_norm": 0.
|
455 |
"learning_rate": 3.697692392808545e-06,
|
456 |
-
"loss": 2.
|
457 |
"step": 6400
|
458 |
},
|
459 |
{
|
460 |
"epoch": 0.7569581926167462,
|
461 |
-
"grad_norm": 0.
|
462 |
"learning_rate": 3.387399686244144e-06,
|
463 |
-
"loss": 2.
|
464 |
"step": 6500
|
465 |
},
|
466 |
{
|
467 |
"epoch": 0.7686037032723885,
|
468 |
-
"grad_norm": 0.
|
469 |
"learning_rate": 3.0880305751715402e-06,
|
470 |
-
"loss": 2.
|
471 |
"step": 6600
|
472 |
},
|
473 |
{
|
474 |
"epoch": 0.7802492139280307,
|
475 |
-
"grad_norm": 0.
|
476 |
"learning_rate": 2.800079598337505e-06,
|
477 |
-
"loss": 2.
|
478 |
"step": 6700
|
479 |
},
|
480 |
{
|
481 |
"epoch": 0.791894724583673,
|
482 |
-
"grad_norm": 0.
|
483 |
"learning_rate": 2.524022432456664e-06,
|
484 |
-
"loss": 2.
|
485 |
"step": 6800
|
486 |
},
|
487 |
{
|
488 |
"epoch": 0.8035402352393153,
|
489 |
-
"grad_norm": 0.
|
490 |
"learning_rate": 2.260315106423807e-06,
|
491 |
-
"loss": 2.
|
492 |
"step": 6900
|
493 |
},
|
494 |
{
|
495 |
"epoch": 0.8151857458949575,
|
496 |
-
"grad_norm": 0.
|
497 |
"learning_rate": 2.0093932479830935e-06,
|
498 |
"loss": 2.1367,
|
499 |
"step": 7000
|
500 |
},
|
501 |
{
|
502 |
"epoch": 0.8268312565505997,
|
503 |
-
"grad_norm": 1.
|
504 |
"learning_rate": 1.7716713640987526e-06,
|
505 |
-
"loss": 2.
|
506 |
"step": 7100
|
507 |
},
|
508 |
{
|
509 |
"epoch": 0.838476767206242,
|
510 |
-
"grad_norm": 0.
|
511 |
"learning_rate": 1.5475421562158854e-06,
|
512 |
"loss": 2.1459,
|
513 |
"step": 7200
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.8501222778618842,
|
517 |
-
"grad_norm": 0.
|
518 |
"learning_rate": 1.3373758715426444e-06,
|
519 |
-
"loss": 2.
|
520 |
"step": 7300
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.8617677885175264,
|
524 |
-
"grad_norm": 0.
|
525 |
"learning_rate": 1.141519691425379e-06,
|
526 |
-
"loss": 2.
|
527 |
"step": 7400
|
528 |
},
|
529 |
{
|
530 |
"epoch": 0.8734132991731688,
|
531 |
-
"grad_norm": 0.
|
532 |
"learning_rate": 9.60297157827106e-07,
|
533 |
-
"loss": 2.
|
534 |
"step": 7500
|
535 |
},
|
536 |
{
|
537 |
"epoch": 0.885058809828811,
|
538 |
-
"grad_norm": 0.
|
539 |
"learning_rate": 7.94007638856753e-07,
|
540 |
-
"loss": 2.
|
541 |
"step": 7600
|
542 |
},
|
543 |
{
|
544 |
"epoch": 0.8967043204844533,
|
545 |
-
"grad_norm": 0.
|
546 |
"learning_rate": 6.429258342320677e-07,
|
547 |
-
"loss": 2.
|
548 |
"step": 7700
|
549 |
},
|
550 |
{
|
551 |
"epoch": 0.9083498311400955,
|
552 |
-
"grad_norm": 0.
|
553 |
"learning_rate": 5.073013214931377e-07,
|
554 |
-
"loss": 2.
|
555 |
"step": 7800
|
556 |
},
|
557 |
{
|
558 |
"epoch": 0.9199953417957377,
|
559 |
-
"grad_norm": 0.
|
560 |
"learning_rate": 3.8735814371615554e-07,
|
561 |
-
"loss": 2.
|
562 |
"step": 7900
|
563 |
},
|
564 |
{
|
565 |
"epoch": 0.93164085245138,
|
566 |
-
"grad_norm": 0.
|
567 |
"learning_rate": 2.8329443940849577e-07,
|
568 |
"loss": 2.1349,
|
569 |
"step": 8000
|
570 |
},
|
571 |
{
|
572 |
"epoch": 0.9432863631070223,
|
573 |
-
"grad_norm": 0.
|
574 |
"learning_rate": 1.9528211519649341e-07,
|
575 |
"loss": 2.094,
|
576 |
"step": 8100
|
577 |
},
|
578 |
{
|
579 |
"epoch": 0.9549318737626645,
|
580 |
-
"grad_norm": 1.
|
581 |
"learning_rate": 1.234665618466202e-07,
|
582 |
-
"loss": 2.
|
583 |
"step": 8200
|
584 |
},
|
585 |
{
|
586 |
"epoch": 0.9665773844183068,
|
587 |
-
"grad_norm": 0.
|
588 |
"learning_rate": 6.796641408917671e-08,
|
589 |
"loss": 2.0692,
|
590 |
"step": 8300
|
591 |
},
|
592 |
{
|
593 |
"epoch": 0.978222895073949,
|
594 |
-
"grad_norm": 0.
|
595 |
"learning_rate": 2.8873354641258955e-08,
|
596 |
-
"loss": 2.
|
597 |
"step": 8400
|
598 |
},
|
599 |
{
|
600 |
"epoch": 0.9898684057295912,
|
601 |
-
"grad_norm": 1.
|
602 |
-
"learning_rate": 6.
|
603 |
"loss": 2.1445,
|
604 |
"step": 8500
|
605 |
},
|
606 |
{
|
607 |
"epoch": 1.0,
|
608 |
-
"eval_loss": 1.
|
609 |
-
"eval_runtime": 63.
|
610 |
-
"eval_samples_per_second": 15.
|
611 |
-
"eval_steps_per_second": 1.
|
612 |
"step": 8587
|
613 |
},
|
614 |
{
|
615 |
"epoch": 1.0,
|
616 |
"step": 8587,
|
617 |
"total_flos": 1.5640093507584e+17,
|
618 |
-
"train_loss": 2.
|
619 |
-
"train_runtime":
|
620 |
-
"train_samples_per_second": 6.
|
621 |
-
"train_steps_per_second": 3.
|
622 |
}
|
623 |
],
|
624 |
"logging_steps": 100,
|
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.01164551065564225,
|
13 |
+
"grad_norm": 0.14460527896881104,
|
14 |
"learning_rate": 2.331002331002331e-06,
|
15 |
"loss": 2.4953,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.0232910213112845,
|
20 |
+
"grad_norm": 0.24249859154224396,
|
21 |
"learning_rate": 4.662004662004662e-06,
|
22 |
+
"loss": 2.4574,
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.03493653196692675,
|
27 |
+
"grad_norm": 0.22802527248859406,
|
28 |
"learning_rate": 6.993006993006993e-06,
|
29 |
"loss": 2.4409,
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.046582042622569,
|
34 |
+
"grad_norm": 0.33553868532180786,
|
35 |
"learning_rate": 9.324009324009324e-06,
|
36 |
"loss": 2.4497,
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.05822755327821125,
|
41 |
+
"grad_norm": 0.36703944206237793,
|
42 |
"learning_rate": 1.1655011655011657e-05,
|
43 |
+
"loss": 2.3974,
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.0698730639338535,
|
48 |
+
"grad_norm": 0.42922335863113403,
|
49 |
"learning_rate": 1.3986013986013986e-05,
|
50 |
+
"loss": 2.3563,
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.08151857458949575,
|
55 |
+
"grad_norm": 0.4571892023086548,
|
56 |
"learning_rate": 1.6317016317016318e-05,
|
57 |
+
"loss": 2.3174,
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.093164085245138,
|
62 |
+
"grad_norm": 0.533064067363739,
|
63 |
"learning_rate": 1.8648018648018647e-05,
|
64 |
+
"loss": 2.3549,
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.10480959590078025,
|
69 |
+
"grad_norm": 0.5255058407783508,
|
70 |
"learning_rate": 1.999854282682324e-05,
|
71 |
+
"loss": 2.3053,
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.1164551065564225,
|
76 |
+
"grad_norm": 0.5629482865333557,
|
77 |
"learning_rate": 1.9983347507285766e-05,
|
78 |
+
"loss": 2.2866,
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.12810061721206475,
|
83 |
+
"grad_norm": 0.5937526226043701,
|
84 |
"learning_rate": 1.9951660332101616e-05,
|
85 |
+
"loss": 2.2721,
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.139746127867707,
|
90 |
+
"grad_norm": 0.703377366065979,
|
91 |
"learning_rate": 1.9903533646470504e-05,
|
92 |
+
"loss": 2.2648,
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.15139163852334925,
|
97 |
+
"grad_norm": 0.6113013625144958,
|
98 |
"learning_rate": 1.9839046952618667e-05,
|
99 |
+
"loss": 2.2509,
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.1630371491789915,
|
104 |
+
"grad_norm": 0.6333953738212585,
|
105 |
"learning_rate": 1.9758306778466264e-05,
|
106 |
+
"loss": 2.2475,
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.17468265983463374,
|
111 |
+
"grad_norm": 0.6210408806800842,
|
112 |
"learning_rate": 1.966144650165e-05,
|
113 |
+
"loss": 2.2575,
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.186328170490276,
|
118 |
+
"grad_norm": 0.6377599239349365,
|
119 |
"learning_rate": 1.9548626129191778e-05,
|
120 |
+
"loss": 2.2492,
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.19797368114591826,
|
125 |
+
"grad_norm": 0.6116755604743958,
|
126 |
"learning_rate": 1.9420032033177225e-05,
|
127 |
+
"loss": 2.2498,
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.2096191918015605,
|
132 |
+
"grad_norm": 0.9220359921455383,
|
133 |
"learning_rate": 1.927587664288089e-05,
|
134 |
+
"loss": 2.2231,
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.22126470245720276,
|
139 |
+
"grad_norm": 0.6930053234100342,
|
140 |
"learning_rate": 1.911639809384656e-05,
|
141 |
+
"loss": 2.1863,
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.232910213112845,
|
146 |
+
"grad_norm": 0.991111695766449,
|
147 |
"learning_rate": 1.8941859834502484e-05,
|
148 |
+
"loss": 2.1885,
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.24455572376848725,
|
153 |
+
"grad_norm": 0.9630569815635681,
|
154 |
"learning_rate": 1.8752550190961288e-05,
|
155 |
+
"loss": 2.208,
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.2562012344241295,
|
160 |
+
"grad_norm": 0.8858104348182678,
|
161 |
"learning_rate": 1.8548781890723614e-05,
|
162 |
+
"loss": 2.2048,
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.26784674507977174,
|
167 |
+
"grad_norm": 0.9373981952667236,
|
168 |
"learning_rate": 1.8330891546072095e-05,
|
169 |
+
"loss": 2.2026,
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.279492255735414,
|
174 |
+
"grad_norm": 0.6907739639282227,
|
175 |
"learning_rate": 1.809923909800931e-05,
|
176 |
"loss": 2.1932,
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.29113776639105626,
|
181 |
+
"grad_norm": 0.7434536814689636,
|
182 |
"learning_rate": 1.7854207221658092e-05,
|
183 |
+
"loss": 2.2032,
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.3027832770466985,
|
188 |
+
"grad_norm": 0.7883014678955078,
|
189 |
"learning_rate": 1.7596200694106552e-05,
|
190 |
+
"loss": 2.1869,
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.31442878770234073,
|
195 |
+
"grad_norm": 0.9146271347999573,
|
196 |
"learning_rate": 1.7325645725742056e-05,
|
197 |
+
"loss": 2.1941,
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.326074298357983,
|
202 |
+
"grad_norm": 0.9042865633964539,
|
203 |
"learning_rate": 1.7042989256178744e-05,
|
204 |
+
"loss": 2.1523,
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.33771980901362525,
|
209 |
+
"grad_norm": 0.9685861468315125,
|
210 |
"learning_rate": 1.6748698215941704e-05,
|
211 |
+
"loss": 2.1736,
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.3493653196692675,
|
216 |
+
"grad_norm": 0.7863497138023376,
|
217 |
"learning_rate": 1.6443258755127393e-05,
|
218 |
+
"loss": 2.1938,
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.36101083032490977,
|
223 |
+
"grad_norm": 0.631289005279541,
|
224 |
"learning_rate": 1.6127175440314596e-05,
|
225 |
"loss": 2.1204,
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.372656340980552,
|
230 |
+
"grad_norm": 0.8423033952713013,
|
231 |
"learning_rate": 1.5800970421052487e-05,
|
232 |
+
"loss": 2.1962,
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
"epoch": 0.38430185163619424,
|
237 |
+
"grad_norm": 1.026877522468567,
|
238 |
"learning_rate": 1.546518256730277e-05,
|
239 |
+
"loss": 2.1839,
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
"epoch": 0.3959473622918365,
|
244 |
+
"grad_norm": 1.0023020505905151,
|
245 |
"learning_rate": 1.5120366579260734e-05,
|
246 |
"loss": 2.1607,
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
"epoch": 0.40759287294747876,
|
251 |
+
"grad_norm": 0.8861306309700012,
|
252 |
"learning_rate": 1.4767092071025792e-05,
|
253 |
+
"loss": 2.1838,
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
"epoch": 0.419238383603121,
|
258 |
+
"grad_norm": 0.8740290403366089,
|
259 |
"learning_rate": 1.4405942629635174e-05,
|
260 |
+
"loss": 2.1205,
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
"epoch": 0.4308838942587632,
|
265 |
+
"grad_norm": 0.7423445582389832,
|
266 |
"learning_rate": 1.4037514851015241e-05,
|
267 |
"loss": 2.215,
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
"epoch": 0.4425294049144055,
|
272 |
+
"grad_norm": 0.9221410155296326,
|
273 |
"learning_rate": 1.3662417354442924e-05,
|
274 |
"loss": 2.1836,
|
275 |
"step": 3800
|
276 |
},
|
277 |
{
|
278 |
"epoch": 0.45417491557004774,
|
279 |
+
"grad_norm": 0.9722244143486023,
|
280 |
"learning_rate": 1.3281269777145354e-05,
|
281 |
+
"loss": 2.1934,
|
282 |
"step": 3900
|
283 |
},
|
284 |
{
|
285 |
"epoch": 0.46582042622569,
|
286 |
+
"grad_norm": 1.1752955913543701,
|
287 |
"learning_rate": 1.2894701750698541e-05,
|
288 |
+
"loss": 2.1382,
|
289 |
"step": 4000
|
290 |
},
|
291 |
{
|
292 |
"epoch": 0.47746593688133226,
|
293 |
+
"grad_norm": 0.8646457195281982,
|
294 |
"learning_rate": 1.2503351860916024e-05,
|
295 |
+
"loss": 2.1507,
|
296 |
"step": 4100
|
297 |
},
|
298 |
{
|
299 |
"epoch": 0.4891114475369745,
|
300 |
+
"grad_norm": 1.307706594467163,
|
301 |
"learning_rate": 1.2107866592945686e-05,
|
302 |
+
"loss": 2.1564,
|
303 |
"step": 4200
|
304 |
},
|
305 |
{
|
306 |
"epoch": 0.5007569581926168,
|
307 |
+
"grad_norm": 0.9780316948890686,
|
308 |
"learning_rate": 1.1708899263317381e-05,
|
309 |
+
"loss": 2.1395,
|
310 |
"step": 4300
|
311 |
},
|
312 |
{
|
313 |
"epoch": 0.512402468848259,
|
314 |
+
"grad_norm": 0.8423807621002197,
|
315 |
"learning_rate": 1.1307108940705536e-05,
|
316 |
"loss": 2.1605,
|
317 |
"step": 4400
|
318 |
},
|
319 |
{
|
320 |
"epoch": 0.5240479795039013,
|
321 |
+
"grad_norm": 0.9365679621696472,
|
322 |
"learning_rate": 1.090315935718958e-05,
|
323 |
"loss": 2.1108,
|
324 |
"step": 4500
|
325 |
},
|
326 |
{
|
327 |
"epoch": 0.5356934901595435,
|
328 |
+
"grad_norm": 0.8869751691818237,
|
329 |
"learning_rate": 1.0497717811810748e-05,
|
330 |
+
"loss": 2.1365,
|
331 |
"step": 4600
|
332 |
},
|
333 |
{
|
334 |
"epoch": 0.5473390008151857,
|
335 |
+
"grad_norm": 0.7278423309326172,
|
336 |
"learning_rate": 1.0091454068236455e-05,
|
337 |
"loss": 2.1498,
|
338 |
"step": 4700
|
339 |
},
|
340 |
{
|
341 |
"epoch": 0.558984511470828,
|
342 |
+
"grad_norm": 1.1241728067398071,
|
343 |
"learning_rate": 9.685039248353284e-06,
|
344 |
+
"loss": 2.1635,
|
345 |
"step": 4800
|
346 |
},
|
347 |
{
|
348 |
"epoch": 0.5706300221264703,
|
349 |
+
"grad_norm": 0.8585646748542786,
|
350 |
"learning_rate": 9.279144723616279e-06,
|
351 |
+
"loss": 2.072,
|
352 |
"step": 4900
|
353 |
},
|
354 |
{
|
355 |
"epoch": 0.5822755327821125,
|
356 |
+
"grad_norm": 0.768528938293457,
|
357 |
"learning_rate": 8.874441005985965e-06,
|
358 |
+
"loss": 2.1568,
|
359 |
"step": 5000
|
360 |
},
|
361 |
{
|
362 |
"epoch": 0.5939210434377548,
|
363 |
+
"grad_norm": 1.002091407775879,
|
364 |
"learning_rate": 8.47159664028521e-06,
|
365 |
"loss": 2.1631,
|
366 |
"step": 5100
|
367 |
},
|
368 |
{
|
369 |
"epoch": 0.605566554093397,
|
370 |
+
"grad_norm": 0.7817184329032898,
|
371 |
"learning_rate": 8.07127709980564e-06,
|
372 |
+
"loss": 2.1203,
|
373 |
"step": 5200
|
374 |
},
|
375 |
{
|
376 |
"epoch": 0.6172120647490392,
|
377 |
+
"grad_norm": 1.3114322423934937,
|
378 |
"learning_rate": 7.674143686988085e-06,
|
379 |
"loss": 2.0871,
|
380 |
"step": 5300
|
381 |
},
|
382 |
{
|
383 |
"epoch": 0.6288575754046815,
|
384 |
+
"grad_norm": 0.7261312007904053,
|
385 |
"learning_rate": 7.280852440992941e-06,
|
386 |
"loss": 2.1289,
|
387 |
"step": 5400
|
388 |
},
|
389 |
{
|
390 |
"epoch": 0.6405030860603237,
|
391 |
+
"grad_norm": 0.9209310412406921,
|
392 |
"learning_rate": 6.89205305396518e-06,
|
393 |
+
"loss": 2.1063,
|
394 |
"step": 5500
|
395 |
},
|
396 |
{
|
397 |
"epoch": 0.652148596715966,
|
398 |
+
"grad_norm": 0.9243987798690796,
|
399 |
"learning_rate": 6.508387797784227e-06,
|
400 |
+
"loss": 2.0956,
|
401 |
"step": 5600
|
402 |
},
|
403 |
{
|
404 |
"epoch": 0.6637941073716083,
|
405 |
+
"grad_norm": 0.9812811017036438,
|
406 |
"learning_rate": 6.130490463071604e-06,
|
407 |
+
"loss": 2.0918,
|
408 |
"step": 5700
|
409 |
},
|
410 |
{
|
411 |
"epoch": 0.6754396180272505,
|
412 |
+
"grad_norm": 0.8238828182220459,
|
413 |
"learning_rate": 5.758985312209124e-06,
|
414 |
+
"loss": 2.1203,
|
415 |
"step": 5800
|
416 |
},
|
417 |
{
|
418 |
"epoch": 0.6870851286828927,
|
419 |
+
"grad_norm": 1.1043486595153809,
|
420 |
"learning_rate": 5.394486048097099e-06,
|
421 |
+
"loss": 2.1159,
|
422 |
"step": 5900
|
423 |
},
|
424 |
{
|
425 |
"epoch": 0.698730639338535,
|
426 |
+
"grad_norm": 0.7944552898406982,
|
427 |
"learning_rate": 5.037594800356142e-06,
|
428 |
+
"loss": 2.1157,
|
429 |
"step": 6000
|
430 |
},
|
431 |
{
|
432 |
"epoch": 0.7103761499941772,
|
433 |
+
"grad_norm": 1.1705522537231445,
|
434 |
"learning_rate": 4.688901130647314e-06,
|
435 |
"loss": 2.1396,
|
436 |
"step": 6100
|
437 |
},
|
438 |
{
|
439 |
"epoch": 0.7220216606498195,
|
440 |
+
"grad_norm": 0.8150696754455566,
|
441 |
"learning_rate": 4.348981058753708e-06,
|
442 |
"loss": 2.0819,
|
443 |
"step": 6200
|
444 |
},
|
445 |
{
|
446 |
"epoch": 0.7336671713054618,
|
447 |
+
"grad_norm": 1.0273561477661133,
|
448 |
"learning_rate": 4.018396111032394e-06,
|
449 |
+
"loss": 2.1226,
|
450 |
"step": 6300
|
451 |
},
|
452 |
{
|
453 |
"epoch": 0.745312681961104,
|
454 |
+
"grad_norm": 0.9231303334236145,
|
455 |
"learning_rate": 3.697692392808545e-06,
|
456 |
+
"loss": 2.1035,
|
457 |
"step": 6400
|
458 |
},
|
459 |
{
|
460 |
"epoch": 0.7569581926167462,
|
461 |
+
"grad_norm": 0.7557884454727173,
|
462 |
"learning_rate": 3.387399686244144e-06,
|
463 |
+
"loss": 2.0954,
|
464 |
"step": 6500
|
465 |
},
|
466 |
{
|
467 |
"epoch": 0.7686037032723885,
|
468 |
+
"grad_norm": 0.8633888959884644,
|
469 |
"learning_rate": 3.0880305751715402e-06,
|
470 |
+
"loss": 2.1199,
|
471 |
"step": 6600
|
472 |
},
|
473 |
{
|
474 |
"epoch": 0.7802492139280307,
|
475 |
+
"grad_norm": 0.8371317982673645,
|
476 |
"learning_rate": 2.800079598337505e-06,
|
477 |
+
"loss": 2.1135,
|
478 |
"step": 6700
|
479 |
},
|
480 |
{
|
481 |
"epoch": 0.791894724583673,
|
482 |
+
"grad_norm": 0.7599378228187561,
|
483 |
"learning_rate": 2.524022432456664e-06,
|
484 |
+
"loss": 2.1385,
|
485 |
"step": 6800
|
486 |
},
|
487 |
{
|
488 |
"epoch": 0.8035402352393153,
|
489 |
+
"grad_norm": 0.8745434880256653,
|
490 |
"learning_rate": 2.260315106423807e-06,
|
491 |
+
"loss": 2.0778,
|
492 |
"step": 6900
|
493 |
},
|
494 |
{
|
495 |
"epoch": 0.8151857458949575,
|
496 |
+
"grad_norm": 0.871006429195404,
|
497 |
"learning_rate": 2.0093932479830935e-06,
|
498 |
"loss": 2.1367,
|
499 |
"step": 7000
|
500 |
},
|
501 |
{
|
502 |
"epoch": 0.8268312565505997,
|
503 |
+
"grad_norm": 1.034953236579895,
|
504 |
"learning_rate": 1.7716713640987526e-06,
|
505 |
+
"loss": 2.1163,
|
506 |
"step": 7100
|
507 |
},
|
508 |
{
|
509 |
"epoch": 0.838476767206242,
|
510 |
+
"grad_norm": 0.8476694226264954,
|
511 |
"learning_rate": 1.5475421562158854e-06,
|
512 |
"loss": 2.1459,
|
513 |
"step": 7200
|
514 |
},
|
515 |
{
|
516 |
"epoch": 0.8501222778618842,
|
517 |
+
"grad_norm": 0.9011842012405396,
|
518 |
"learning_rate": 1.3373758715426444e-06,
|
519 |
+
"loss": 2.1094,
|
520 |
"step": 7300
|
521 |
},
|
522 |
{
|
523 |
"epoch": 0.8617677885175264,
|
524 |
+
"grad_norm": 0.8178461790084839,
|
525 |
"learning_rate": 1.141519691425379e-06,
|
526 |
+
"loss": 2.0853,
|
527 |
"step": 7400
|
528 |
},
|
529 |
{
|
530 |
"epoch": 0.8734132991731688,
|
531 |
+
"grad_norm": 0.898789644241333,
|
532 |
"learning_rate": 9.60297157827106e-07,
|
533 |
+
"loss": 2.1359,
|
534 |
"step": 7500
|
535 |
},
|
536 |
{
|
537 |
"epoch": 0.885058809828811,
|
538 |
+
"grad_norm": 0.8626888990402222,
|
539 |
"learning_rate": 7.94007638856753e-07,
|
540 |
+
"loss": 2.1405,
|
541 |
"step": 7600
|
542 |
},
|
543 |
{
|
544 |
"epoch": 0.8967043204844533,
|
545 |
+
"grad_norm": 0.8167033791542053,
|
546 |
"learning_rate": 6.429258342320677e-07,
|
547 |
+
"loss": 2.151,
|
548 |
"step": 7700
|
549 |
},
|
550 |
{
|
551 |
"epoch": 0.9083498311400955,
|
552 |
+
"grad_norm": 0.855806827545166,
|
553 |
"learning_rate": 5.073013214931377e-07,
|
554 |
+
"loss": 2.1835,
|
555 |
"step": 7800
|
556 |
},
|
557 |
{
|
558 |
"epoch": 0.9199953417957377,
|
559 |
+
"grad_norm": 0.8845811486244202,
|
560 |
"learning_rate": 3.8735814371615554e-07,
|
561 |
+
"loss": 2.1314,
|
562 |
"step": 7900
|
563 |
},
|
564 |
{
|
565 |
"epoch": 0.93164085245138,
|
566 |
+
"grad_norm": 0.7994192838668823,
|
567 |
"learning_rate": 2.8329443940849577e-07,
|
568 |
"loss": 2.1349,
|
569 |
"step": 8000
|
570 |
},
|
571 |
{
|
572 |
"epoch": 0.9432863631070223,
|
573 |
+
"grad_norm": 0.792962372303009,
|
574 |
"learning_rate": 1.9528211519649341e-07,
|
575 |
"loss": 2.094,
|
576 |
"step": 8100
|
577 |
},
|
578 |
{
|
579 |
"epoch": 0.9549318737626645,
|
580 |
+
"grad_norm": 1.0241106748580933,
|
581 |
"learning_rate": 1.234665618466202e-07,
|
582 |
+
"loss": 2.1502,
|
583 |
"step": 8200
|
584 |
},
|
585 |
{
|
586 |
"epoch": 0.9665773844183068,
|
587 |
+
"grad_norm": 0.8663437366485596,
|
588 |
"learning_rate": 6.796641408917671e-08,
|
589 |
"loss": 2.0692,
|
590 |
"step": 8300
|
591 |
},
|
592 |
{
|
593 |
"epoch": 0.978222895073949,
|
594 |
+
"grad_norm": 0.7561269402503967,
|
595 |
"learning_rate": 2.8873354641258955e-08,
|
596 |
+
"loss": 2.0988,
|
597 |
"step": 8400
|
598 |
},
|
599 |
{
|
600 |
"epoch": 0.9898684057295912,
|
601 |
+
"grad_norm": 1.3960742950439453,
|
602 |
+
"learning_rate": 6.2519627527379836e-09,
|
603 |
"loss": 2.1445,
|
604 |
"step": 8500
|
605 |
},
|
606 |
{
|
607 |
"epoch": 1.0,
|
608 |
+
"eval_loss": 1.9266563653945923,
|
609 |
+
"eval_runtime": 63.7186,
|
610 |
+
"eval_samples_per_second": 15.537,
|
611 |
+
"eval_steps_per_second": 1.946,
|
612 |
"step": 8587
|
613 |
},
|
614 |
{
|
615 |
"epoch": 1.0,
|
616 |
"step": 8587,
|
617 |
"total_flos": 1.5640093507584e+17,
|
618 |
+
"train_loss": 2.1807946249230077,
|
619 |
+
"train_runtime": 2485.1722,
|
620 |
+
"train_samples_per_second": 6.91,
|
621 |
+
"train_steps_per_second": 3.455
|
622 |
}
|
623 |
],
|
624 |
"logging_steps": 100,
|