MHGanainy/gpt2-xl-lora-multi-3
Browse files- README.md +2 -0
- all_results.json +10 -10
- eval_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +237 -321
README.md
CHANGED
@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
# gpt2-xl-lora-multi-3
|
16 |
|
17 |
This model is a fine-tuned version of [openai-community/gpt2-xl](https://huggingface.co/openai-community/gpt2-xl) on an unknown dataset.
|
|
|
|
|
18 |
|
19 |
## Model description
|
20 |
|
|
|
15 |
# gpt2-xl-lora-multi-3
|
16 |
|
17 |
This model is a fine-tuned version of [openai-community/gpt2-xl](https://huggingface.co/openai-community/gpt2-xl) on an unknown dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 2.4837
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 2.
|
4 |
-
"eval_runtime":
|
5 |
-
"eval_samples_per_second":
|
6 |
-
"eval_steps_per_second": 0.
|
7 |
-
"perplexity": 11.
|
8 |
-
"total_flos":
|
9 |
-
"train_loss": 2.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second":
|
12 |
-
"train_steps_per_second": 3.
|
13 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 2.4837162494659424,
|
4 |
+
"eval_runtime": 2242.6453,
|
5 |
+
"eval_samples_per_second": 8.521,
|
6 |
+
"eval_steps_per_second": 0.533,
|
7 |
+
"perplexity": 11.985723695080148,
|
8 |
+
"total_flos": 8.372955480242258e+17,
|
9 |
+
"train_loss": 2.6846868539578486,
|
10 |
+
"train_runtime": 1624.688,
|
11 |
+
"train_samples_per_second": 56.585,
|
12 |
+
"train_steps_per_second": 3.537
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 2.
|
4 |
-
"eval_runtime":
|
5 |
-
"eval_samples_per_second":
|
6 |
-
"eval_steps_per_second": 0.
|
7 |
-
"perplexity": 11.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 2.4837162494659424,
|
4 |
+
"eval_runtime": 2242.6453,
|
5 |
+
"eval_samples_per_second": 8.521,
|
6 |
+
"eval_steps_per_second": 0.533,
|
7 |
+
"perplexity": 11.985723695080148
|
8 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"total_flos":
|
4 |
-
"train_loss": 2.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 3.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"total_flos": 8.372955480242258e+17,
|
4 |
+
"train_loss": 2.6846868539578486,
|
5 |
+
"train_runtime": 1624.688,
|
6 |
+
"train_samples_per_second": 56.585,
|
7 |
+
"train_steps_per_second": 3.537
|
8 |
}
|
trainer_state.json
CHANGED
@@ -3,506 +3,422 @@
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
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": 0.
|
13 |
-
"grad_norm": 0.
|
14 |
-
"learning_rate":
|
15 |
-
"loss": 2.
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm": 0.
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 2.
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm": 0.
|
28 |
-
"learning_rate": 1.
|
29 |
-
"loss": 2.
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"grad_norm": 0.
|
35 |
-
"learning_rate": 1.
|
36 |
-
"loss": 2.
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
-
"epoch": 0.
|
41 |
-
"grad_norm": 0.
|
42 |
-
"learning_rate":
|
43 |
-
"loss": 2.
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
-
"epoch": 0.
|
48 |
-
"grad_norm": 0.
|
49 |
-
"learning_rate": 2.
|
50 |
-
"loss": 2.
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
-
"epoch": 0.
|
55 |
-
"grad_norm": 0.
|
56 |
-
"learning_rate":
|
57 |
-
"loss": 2.
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
-
"epoch": 0.
|
62 |
-
"grad_norm": 0.
|
63 |
-
"learning_rate":
|
64 |
-
"loss": 2.
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
-
"epoch": 0.
|
69 |
-
"grad_norm": 0.
|
70 |
-
"learning_rate": 3.
|
71 |
-
"loss": 2.
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
-
"epoch": 0.
|
76 |
-
"grad_norm": 0.
|
77 |
-
"learning_rate":
|
78 |
-
"loss": 2.
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
-
"epoch": 0.
|
83 |
-
"grad_norm": 0.
|
84 |
-
"learning_rate":
|
85 |
-
"loss": 2.
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
-
"epoch": 0.
|
90 |
-
"grad_norm": 0.
|
91 |
-
"learning_rate":
|
92 |
-
"loss": 2.
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
-
"epoch": 0.
|
97 |
-
"grad_norm": 0.
|
98 |
-
"learning_rate":
|
99 |
-
"loss": 2.
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
-
"epoch": 0.
|
104 |
-
"grad_norm": 0.
|
105 |
-
"learning_rate":
|
106 |
-
"loss": 2.
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
-
"epoch": 0.
|
111 |
-
"grad_norm": 0.
|
112 |
-
"learning_rate":
|
113 |
-
"loss": 2.
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm": 0.
|
119 |
-
"learning_rate":
|
120 |
-
"loss": 2.
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
-
"epoch": 0.
|
125 |
-
"grad_norm": 0.
|
126 |
-
"learning_rate":
|
127 |
-
"loss": 2.
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
-
"epoch": 0.
|
132 |
-
"grad_norm": 0.
|
133 |
-
"learning_rate":
|
134 |
-
"loss": 2.
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
-
"epoch": 0.
|
139 |
-
"grad_norm": 0.
|
140 |
-
"learning_rate":
|
141 |
-
"loss": 2.
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
-
"epoch": 0.
|
146 |
-
"grad_norm": 0.
|
147 |
-
"learning_rate":
|
148 |
-
"loss": 2.
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
-
"epoch": 0.
|
153 |
-
"grad_norm": 0.
|
154 |
-
"learning_rate":
|
155 |
-
"loss": 2.
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
-
"epoch": 0.
|
160 |
-
"grad_norm": 0.
|
161 |
-
"learning_rate":
|
162 |
-
"loss": 2.
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
-
"epoch": 0.
|
167 |
-
"grad_norm": 0.
|
168 |
-
"learning_rate":
|
169 |
-
"loss": 2.
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
-
"epoch": 0.
|
174 |
-
"grad_norm": 0.
|
175 |
-
"learning_rate":
|
176 |
-
"loss": 2.
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
-
"epoch": 0.
|
181 |
-
"grad_norm": 0.
|
182 |
-
"learning_rate":
|
183 |
-
"loss": 2.
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
-
"epoch": 0.
|
188 |
-
"grad_norm": 0.
|
189 |
-
"learning_rate":
|
190 |
-
"loss": 2.
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
-
"epoch": 0.
|
195 |
-
"grad_norm": 0.
|
196 |
-
"learning_rate":
|
197 |
-
"loss": 2.
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
-
"epoch": 0.
|
202 |
-
"grad_norm": 0.
|
203 |
-
"learning_rate": 1.
|
204 |
-
"loss": 2.
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
-
"epoch": 0.
|
209 |
-
"grad_norm": 0.
|
210 |
-
"learning_rate": 1.
|
211 |
-
"loss": 2.
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
-
"epoch": 0.
|
216 |
-
"grad_norm": 0.
|
217 |
-
"learning_rate": 1.
|
218 |
-
"loss": 2.
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
-
"epoch": 0.
|
223 |
-
"grad_norm": 0.
|
224 |
-
"learning_rate": 1.
|
225 |
-
"loss": 2.
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
-
"epoch": 0.
|
230 |
-
"grad_norm": 0.
|
231 |
-
"learning_rate": 1.
|
232 |
-
"loss": 2.
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
-
"epoch": 0.
|
237 |
-
"grad_norm": 0.
|
238 |
-
"learning_rate": 1.
|
239 |
-
"loss": 2.
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
-
"epoch": 0.
|
244 |
-
"grad_norm": 0.
|
245 |
-
"learning_rate": 1.
|
246 |
-
"loss": 2.
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
-
"epoch": 0.
|
251 |
-
"grad_norm": 0.
|
252 |
-
"learning_rate": 1.
|
253 |
-
"loss": 2.
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
-
"epoch": 0.
|
258 |
-
"grad_norm": 0.
|
259 |
-
"learning_rate": 1.
|
260 |
-
"loss": 2.
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
-
"epoch": 0.
|
265 |
-
"grad_norm": 0.
|
266 |
-
"learning_rate": 1.
|
267 |
-
"loss": 2.
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
-
"epoch": 0.
|
272 |
-
"grad_norm": 0.
|
273 |
-
"learning_rate": 1.
|
274 |
-
"loss": 2.
|
275 |
"step": 3800
|
276 |
},
|
277 |
{
|
278 |
-
"epoch": 0.
|
279 |
-
"grad_norm": 0.
|
280 |
-
"learning_rate": 1.
|
281 |
-
"loss": 2.
|
282 |
"step": 3900
|
283 |
},
|
284 |
{
|
285 |
-
"epoch": 0.
|
286 |
-
"grad_norm": 0.
|
287 |
-
"learning_rate": 1.
|
288 |
-
"loss": 2.
|
289 |
"step": 4000
|
290 |
},
|
291 |
{
|
292 |
-
"epoch": 0.
|
293 |
-
"grad_norm": 0.
|
294 |
-
"learning_rate": 1.
|
295 |
-
"loss": 2.
|
296 |
"step": 4100
|
297 |
},
|
298 |
{
|
299 |
-
"epoch": 0.
|
300 |
-
"grad_norm": 0.
|
301 |
-
"learning_rate": 1.
|
302 |
-
"loss": 2.
|
303 |
"step": 4200
|
304 |
},
|
305 |
{
|
306 |
-
"epoch": 0.
|
307 |
-
"grad_norm": 0.
|
308 |
-
"learning_rate": 1.
|
309 |
-
"loss": 2.
|
310 |
"step": 4300
|
311 |
},
|
312 |
{
|
313 |
-
"epoch": 0.
|
314 |
-
"grad_norm": 0.
|
315 |
-
"learning_rate": 1.
|
316 |
-
"loss": 2.
|
317 |
"step": 4400
|
318 |
},
|
319 |
{
|
320 |
-
"epoch": 0.
|
321 |
-
"grad_norm": 0.
|
322 |
-
"learning_rate": 1.
|
323 |
-
"loss": 2.
|
324 |
"step": 4500
|
325 |
},
|
326 |
{
|
327 |
-
"epoch": 0.
|
328 |
-
"grad_norm": 0.
|
329 |
-
"learning_rate": 1.
|
330 |
-
"loss": 2.
|
331 |
"step": 4600
|
332 |
},
|
333 |
{
|
334 |
-
"epoch": 0.
|
335 |
-
"grad_norm": 0.
|
336 |
-
"learning_rate": 1.
|
337 |
-
"loss": 2.
|
338 |
"step": 4700
|
339 |
},
|
340 |
{
|
341 |
-
"epoch": 0.
|
342 |
-
"grad_norm": 0.
|
343 |
-
"learning_rate": 1.
|
344 |
-
"loss": 2.
|
345 |
"step": 4800
|
346 |
},
|
347 |
{
|
348 |
-
"epoch": 0.
|
349 |
-
"grad_norm": 0.
|
350 |
-
"learning_rate": 1.
|
351 |
-
"loss": 2.
|
352 |
"step": 4900
|
353 |
},
|
354 |
{
|
355 |
-
"epoch": 0.
|
356 |
-
"grad_norm": 0.
|
357 |
-
"learning_rate": 1.
|
358 |
-
"loss": 2.
|
359 |
"step": 5000
|
360 |
},
|
361 |
{
|
362 |
-
"epoch": 0.
|
363 |
-
"grad_norm": 0.
|
364 |
-
"learning_rate": 1.
|
365 |
-
"loss": 2.
|
366 |
"step": 5100
|
367 |
},
|
368 |
{
|
369 |
-
"epoch": 0.
|
370 |
-
"grad_norm": 0.
|
371 |
-
"learning_rate":
|
372 |
-
"loss": 2.
|
373 |
"step": 5200
|
374 |
},
|
375 |
{
|
376 |
-
"epoch": 0.
|
377 |
-
"grad_norm": 0.
|
378 |
-
"learning_rate":
|
379 |
-
"loss": 2.
|
380 |
"step": 5300
|
381 |
},
|
382 |
{
|
383 |
-
"epoch": 0.
|
384 |
-
"grad_norm": 0.
|
385 |
-
"learning_rate":
|
386 |
-
"loss": 2.
|
387 |
"step": 5400
|
388 |
},
|
389 |
{
|
390 |
-
"epoch": 0.
|
391 |
-
"grad_norm": 0.
|
392 |
-
"learning_rate":
|
393 |
-
"loss": 2.
|
394 |
"step": 5500
|
395 |
},
|
396 |
{
|
397 |
-
"epoch": 0.
|
398 |
-
"grad_norm": 0.
|
399 |
-
"learning_rate":
|
400 |
-
"loss": 2.
|
401 |
"step": 5600
|
402 |
},
|
403 |
{
|
404 |
-
"epoch": 0.
|
405 |
-
"grad_norm": 0.
|
406 |
-
"learning_rate":
|
407 |
-
"loss": 2.
|
408 |
"step": 5700
|
409 |
},
|
410 |
-
{
|
411 |
-
"epoch": 0.8380291865337379,
|
412 |
-
"grad_norm": 0.22642740607261658,
|
413 |
-
"learning_rate": 1.8310755913154726e-05,
|
414 |
-
"loss": 2.5312,
|
415 |
-
"step": 5800
|
416 |
-
},
|
417 |
-
{
|
418 |
-
"epoch": 0.8524779656119058,
|
419 |
-
"grad_norm": 0.22286154329776764,
|
420 |
-
"learning_rate": 1.6847124401206384e-05,
|
421 |
-
"loss": 2.5354,
|
422 |
-
"step": 5900
|
423 |
-
},
|
424 |
-
{
|
425 |
-
"epoch": 0.8669267446900737,
|
426 |
-
"grad_norm": 0.22432397305965424,
|
427 |
-
"learning_rate": 1.5032704252395315e-05,
|
428 |
-
"loss": 2.5355,
|
429 |
-
"step": 6000
|
430 |
-
},
|
431 |
-
{
|
432 |
-
"epoch": 0.8813755237682416,
|
433 |
-
"grad_norm": 0.21558962762355804,
|
434 |
-
"learning_rate": 1.2960450981095643e-05,
|
435 |
-
"loss": 2.5481,
|
436 |
-
"step": 6100
|
437 |
-
},
|
438 |
-
{
|
439 |
-
"epoch": 0.8958243028464095,
|
440 |
-
"grad_norm": 0.2449900358915329,
|
441 |
-
"learning_rate": 1.0736529286191087e-05,
|
442 |
-
"loss": 2.5352,
|
443 |
-
"step": 6200
|
444 |
-
},
|
445 |
-
{
|
446 |
-
"epoch": 0.9102730819245773,
|
447 |
-
"grad_norm": 0.22407910227775574,
|
448 |
-
"learning_rate": 8.474874071465144e-06,
|
449 |
-
"loss": 2.5261,
|
450 |
-
"step": 6300
|
451 |
-
},
|
452 |
-
{
|
453 |
-
"epoch": 0.9247218610027452,
|
454 |
-
"grad_norm": 0.20298534631729126,
|
455 |
-
"learning_rate": 6.291353386531074e-06,
|
456 |
-
"loss": 2.5389,
|
457 |
-
"step": 6400
|
458 |
-
},
|
459 |
-
{
|
460 |
-
"epoch": 0.9391706400809131,
|
461 |
-
"grad_norm": 0.2112707942724228,
|
462 |
-
"learning_rate": 4.297832329749687e-06,
|
463 |
-
"loss": 2.5369,
|
464 |
-
"step": 6500
|
465 |
-
},
|
466 |
-
{
|
467 |
-
"epoch": 0.9536194191590811,
|
468 |
-
"grad_norm": 0.22108633816242218,
|
469 |
-
"learning_rate": 2.596442028451194e-06,
|
470 |
-
"loss": 2.5331,
|
471 |
-
"step": 6600
|
472 |
-
},
|
473 |
-
{
|
474 |
-
"epoch": 0.968068198237249,
|
475 |
-
"grad_norm": 0.21827644109725952,
|
476 |
-
"learning_rate": 1.2743473053542842e-06,
|
477 |
-
"loss": 2.5279,
|
478 |
-
"step": 6700
|
479 |
-
},
|
480 |
-
{
|
481 |
-
"epoch": 0.9825169773154169,
|
482 |
-
"grad_norm": 0.22307871282100677,
|
483 |
-
"learning_rate": 3.9928109162008953e-07,
|
484 |
-
"loss": 2.5346,
|
485 |
-
"step": 6800
|
486 |
-
},
|
487 |
-
{
|
488 |
-
"epoch": 0.9969657563935848,
|
489 |
-
"grad_norm": 0.22532746195793152,
|
490 |
-
"learning_rate": 1.6074365378105915e-08,
|
491 |
-
"loss": 2.5428,
|
492 |
-
"step": 6900
|
493 |
-
},
|
494 |
{
|
495 |
"epoch": 1.0,
|
496 |
-
"step":
|
497 |
-
"total_flos":
|
498 |
-
"train_loss": 2.
|
499 |
-
"train_runtime":
|
500 |
-
"train_samples_per_second":
|
501 |
-
"train_steps_per_second": 3.
|
502 |
}
|
503 |
],
|
504 |
"logging_steps": 100,
|
505 |
-
"max_steps":
|
506 |
"num_input_tokens_seen": 0,
|
507 |
"num_train_epochs": 1,
|
508 |
"save_steps": 500,
|
@@ -518,7 +434,7 @@
|
|
518 |
"attributes": {}
|
519 |
}
|
520 |
},
|
521 |
-
"total_flos":
|
522 |
"train_batch_size": 2,
|
523 |
"trial_name": null,
|
524 |
"trial_params": null
|
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 5746,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.01740341106856944,
|
13 |
+
"grad_norm": 0.046589791774749756,
|
14 |
+
"learning_rate": 4.351610095735422e-07,
|
15 |
+
"loss": 2.7999,
|
16 |
"step": 100
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.03480682213713888,
|
20 |
+
"grad_norm": 0.04616040736436844,
|
21 |
+
"learning_rate": 8.703220191470844e-07,
|
22 |
+
"loss": 2.7996,
|
23 |
"step": 200
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.05221023320570832,
|
27 |
+
"grad_norm": 0.05549981817603111,
|
28 |
+
"learning_rate": 1.305483028720627e-06,
|
29 |
+
"loss": 2.797,
|
30 |
"step": 300
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.06961364427427776,
|
34 |
+
"grad_norm": 0.063571035861969,
|
35 |
+
"learning_rate": 1.7406440382941688e-06,
|
36 |
+
"loss": 2.7909,
|
37 |
"step": 400
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.0870170553428472,
|
41 |
+
"grad_norm": 0.08422163128852844,
|
42 |
+
"learning_rate": 2.1758050478677113e-06,
|
43 |
+
"loss": 2.7951,
|
44 |
"step": 500
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 0.10442046641141664,
|
48 |
+
"grad_norm": 0.09257014095783234,
|
49 |
+
"learning_rate": 2.610966057441254e-06,
|
50 |
+
"loss": 2.7803,
|
51 |
"step": 600
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.12182387747998608,
|
55 |
+
"grad_norm": 0.11055697500705719,
|
56 |
+
"learning_rate": 3.046127067014796e-06,
|
57 |
+
"loss": 2.7681,
|
58 |
"step": 700
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 0.1392272885485555,
|
62 |
+
"grad_norm": 0.10759040713310242,
|
63 |
+
"learning_rate": 3.4812880765883376e-06,
|
64 |
+
"loss": 2.7611,
|
65 |
"step": 800
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 0.15663069961712495,
|
69 |
+
"grad_norm": 0.12318646907806396,
|
70 |
+
"learning_rate": 3.9164490861618806e-06,
|
71 |
+
"loss": 2.7402,
|
72 |
"step": 900
|
73 |
},
|
74 |
{
|
75 |
+
"epoch": 0.1740341106856944,
|
76 |
+
"grad_norm": 0.12962989509105682,
|
77 |
+
"learning_rate": 4.351610095735423e-06,
|
78 |
+
"loss": 2.7451,
|
79 |
"step": 1000
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 0.19143752175426385,
|
83 |
+
"grad_norm": 0.13981275260448456,
|
84 |
+
"learning_rate": 4.786771105308965e-06,
|
85 |
+
"loss": 2.735,
|
86 |
"step": 1100
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 0.20884093282283328,
|
90 |
+
"grad_norm": 0.14711035788059235,
|
91 |
+
"learning_rate": 5.221932114882508e-06,
|
92 |
+
"loss": 2.7469,
|
93 |
"step": 1200
|
94 |
},
|
95 |
{
|
96 |
+
"epoch": 0.22624434389140272,
|
97 |
+
"grad_norm": 0.15727241337299347,
|
98 |
+
"learning_rate": 5.657093124456049e-06,
|
99 |
+
"loss": 2.7327,
|
100 |
"step": 1300
|
101 |
},
|
102 |
{
|
103 |
+
"epoch": 0.24364775495997215,
|
104 |
+
"grad_norm": 0.15055705606937408,
|
105 |
+
"learning_rate": 6.092254134029592e-06,
|
106 |
+
"loss": 2.7234,
|
107 |
"step": 1400
|
108 |
},
|
109 |
{
|
110 |
+
"epoch": 0.2610511660285416,
|
111 |
+
"grad_norm": 0.16661331057548523,
|
112 |
+
"learning_rate": 6.527415143603134e-06,
|
113 |
+
"loss": 2.7174,
|
114 |
"step": 1500
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.278454577097111,
|
118 |
+
"grad_norm": 0.17976854741573334,
|
119 |
+
"learning_rate": 6.962576153176675e-06,
|
120 |
+
"loss": 2.719,
|
121 |
"step": 1600
|
122 |
},
|
123 |
{
|
124 |
+
"epoch": 0.2958579881656805,
|
125 |
+
"grad_norm": 0.1790621429681778,
|
126 |
+
"learning_rate": 7.397737162750218e-06,
|
127 |
+
"loss": 2.7173,
|
128 |
"step": 1700
|
129 |
},
|
130 |
{
|
131 |
+
"epoch": 0.3132613992342499,
|
132 |
+
"grad_norm": 0.19079644978046417,
|
133 |
+
"learning_rate": 7.832898172323761e-06,
|
134 |
+
"loss": 2.7131,
|
135 |
"step": 1800
|
136 |
},
|
137 |
{
|
138 |
+
"epoch": 0.33066481030281936,
|
139 |
+
"grad_norm": 0.19005636870861053,
|
140 |
+
"learning_rate": 8.268059181897302e-06,
|
141 |
+
"loss": 2.7168,
|
142 |
"step": 1900
|
143 |
},
|
144 |
{
|
145 |
+
"epoch": 0.3480682213713888,
|
146 |
+
"grad_norm": 0.19910404086112976,
|
147 |
+
"learning_rate": 8.703220191470845e-06,
|
148 |
+
"loss": 2.7061,
|
149 |
"step": 2000
|
150 |
},
|
151 |
{
|
152 |
+
"epoch": 0.3654716324399582,
|
153 |
+
"grad_norm": 0.20510949194431305,
|
154 |
+
"learning_rate": 9.138381201044387e-06,
|
155 |
+
"loss": 2.6862,
|
156 |
"step": 2100
|
157 |
},
|
158 |
{
|
159 |
+
"epoch": 0.3828750435085277,
|
160 |
+
"grad_norm": 0.20418143272399902,
|
161 |
+
"learning_rate": 9.57354221061793e-06,
|
162 |
+
"loss": 2.6802,
|
163 |
"step": 2200
|
164 |
},
|
165 |
{
|
166 |
+
"epoch": 0.4002784545770971,
|
167 |
+
"grad_norm": 0.21713656187057495,
|
168 |
+
"learning_rate": 1.000870322019147e-05,
|
169 |
+
"loss": 2.6923,
|
170 |
"step": 2300
|
171 |
},
|
172 |
{
|
173 |
+
"epoch": 0.41768186564566656,
|
174 |
+
"grad_norm": 0.2298802137374878,
|
175 |
+
"learning_rate": 1.0443864229765015e-05,
|
176 |
+
"loss": 2.6818,
|
177 |
"step": 2400
|
178 |
},
|
179 |
{
|
180 |
+
"epoch": 0.43508527671423597,
|
181 |
+
"grad_norm": 0.2294008880853653,
|
182 |
+
"learning_rate": 1.0879025239338557e-05,
|
183 |
+
"loss": 2.6896,
|
184 |
"step": 2500
|
185 |
},
|
186 |
{
|
187 |
+
"epoch": 0.45248868778280543,
|
188 |
+
"grad_norm": 0.21464629471302032,
|
189 |
+
"learning_rate": 1.1314186248912098e-05,
|
190 |
+
"loss": 2.6805,
|
191 |
"step": 2600
|
192 |
},
|
193 |
{
|
194 |
+
"epoch": 0.4698920988513749,
|
195 |
+
"grad_norm": 0.25449061393737793,
|
196 |
+
"learning_rate": 1.174934725848564e-05,
|
197 |
+
"loss": 2.6806,
|
198 |
"step": 2700
|
199 |
},
|
200 |
{
|
201 |
+
"epoch": 0.4872955099199443,
|
202 |
+
"grad_norm": 0.24079586565494537,
|
203 |
+
"learning_rate": 1.2184508268059184e-05,
|
204 |
+
"loss": 2.6844,
|
205 |
"step": 2800
|
206 |
},
|
207 |
{
|
208 |
+
"epoch": 0.5046989209885138,
|
209 |
+
"grad_norm": 0.2414436638355255,
|
210 |
+
"learning_rate": 1.2619669277632725e-05,
|
211 |
+
"loss": 2.6817,
|
212 |
"step": 2900
|
213 |
},
|
214 |
{
|
215 |
+
"epoch": 0.5221023320570832,
|
216 |
+
"grad_norm": 0.2530564069747925,
|
217 |
+
"learning_rate": 1.3054830287206268e-05,
|
218 |
+
"loss": 2.6556,
|
219 |
"step": 3000
|
220 |
},
|
221 |
{
|
222 |
+
"epoch": 0.5395057431256526,
|
223 |
+
"grad_norm": 0.26441535353660583,
|
224 |
+
"learning_rate": 1.348999129677981e-05,
|
225 |
+
"loss": 2.6749,
|
226 |
"step": 3100
|
227 |
},
|
228 |
{
|
229 |
+
"epoch": 0.556909154194222,
|
230 |
+
"grad_norm": 0.2584131062030792,
|
231 |
+
"learning_rate": 1.392515230635335e-05,
|
232 |
+
"loss": 2.6575,
|
233 |
"step": 3200
|
234 |
},
|
235 |
{
|
236 |
+
"epoch": 0.5743125652627915,
|
237 |
+
"grad_norm": 0.25025609135627747,
|
238 |
+
"learning_rate": 1.4360313315926895e-05,
|
239 |
+
"loss": 2.6658,
|
240 |
"step": 3300
|
241 |
},
|
242 |
{
|
243 |
+
"epoch": 0.591715976331361,
|
244 |
+
"grad_norm": 0.26518625020980835,
|
245 |
+
"learning_rate": 1.4795474325500436e-05,
|
246 |
+
"loss": 2.6586,
|
247 |
"step": 3400
|
248 |
},
|
249 |
{
|
250 |
+
"epoch": 0.6091193873999304,
|
251 |
+
"grad_norm": 0.26597312092781067,
|
252 |
+
"learning_rate": 1.5230635335073978e-05,
|
253 |
+
"loss": 2.6451,
|
254 |
"step": 3500
|
255 |
},
|
256 |
{
|
257 |
+
"epoch": 0.6265227984684998,
|
258 |
+
"grad_norm": 0.2725384831428528,
|
259 |
+
"learning_rate": 1.5665796344647522e-05,
|
260 |
+
"loss": 2.6521,
|
261 |
"step": 3600
|
262 |
},
|
263 |
{
|
264 |
+
"epoch": 0.6439262095370692,
|
265 |
+
"grad_norm": 0.2752222716808319,
|
266 |
+
"learning_rate": 1.6100957354221064e-05,
|
267 |
+
"loss": 2.6398,
|
268 |
"step": 3700
|
269 |
},
|
270 |
{
|
271 |
+
"epoch": 0.6613296206056387,
|
272 |
+
"grad_norm": 0.2558598518371582,
|
273 |
+
"learning_rate": 1.6536118363794605e-05,
|
274 |
+
"loss": 2.6486,
|
275 |
"step": 3800
|
276 |
},
|
277 |
{
|
278 |
+
"epoch": 0.6787330316742082,
|
279 |
+
"grad_norm": 0.26938167214393616,
|
280 |
+
"learning_rate": 1.697127937336815e-05,
|
281 |
+
"loss": 2.641,
|
282 |
"step": 3900
|
283 |
},
|
284 |
{
|
285 |
+
"epoch": 0.6961364427427776,
|
286 |
+
"grad_norm": 0.28793784976005554,
|
287 |
+
"learning_rate": 1.740644038294169e-05,
|
288 |
+
"loss": 2.6344,
|
289 |
"step": 4000
|
290 |
},
|
291 |
{
|
292 |
+
"epoch": 0.713539853811347,
|
293 |
+
"grad_norm": 0.2677360773086548,
|
294 |
+
"learning_rate": 1.7841601392515232e-05,
|
295 |
+
"loss": 2.6542,
|
296 |
"step": 4100
|
297 |
},
|
298 |
{
|
299 |
+
"epoch": 0.7309432648799165,
|
300 |
+
"grad_norm": 0.28143930435180664,
|
301 |
+
"learning_rate": 1.8276762402088773e-05,
|
302 |
+
"loss": 2.6446,
|
303 |
"step": 4200
|
304 |
},
|
305 |
{
|
306 |
+
"epoch": 0.7483466759484859,
|
307 |
+
"grad_norm": 0.28870299458503723,
|
308 |
+
"learning_rate": 1.8711923411662314e-05,
|
309 |
+
"loss": 2.6243,
|
310 |
"step": 4300
|
311 |
},
|
312 |
{
|
313 |
+
"epoch": 0.7657500870170554,
|
314 |
+
"grad_norm": 0.296633780002594,
|
315 |
+
"learning_rate": 1.914708442123586e-05,
|
316 |
+
"loss": 2.6306,
|
317 |
"step": 4400
|
318 |
},
|
319 |
{
|
320 |
+
"epoch": 0.7831534980856247,
|
321 |
+
"grad_norm": 0.2806219160556793,
|
322 |
+
"learning_rate": 1.95822454308094e-05,
|
323 |
+
"loss": 2.6356,
|
324 |
"step": 4500
|
325 |
},
|
326 |
{
|
327 |
+
"epoch": 0.8005569091541942,
|
328 |
+
"grad_norm": 0.2914940416812897,
|
329 |
+
"learning_rate": 1.999940297883134e-05,
|
330 |
+
"loss": 2.644,
|
331 |
"step": 4600
|
332 |
},
|
333 |
{
|
334 |
+
"epoch": 0.8179603202227637,
|
335 |
+
"grad_norm": 0.28510311245918274,
|
336 |
+
"learning_rate": 1.9599117132813187e-05,
|
337 |
+
"loss": 2.6357,
|
338 |
"step": 4700
|
339 |
},
|
340 |
{
|
341 |
+
"epoch": 0.8353637312913331,
|
342 |
+
"grad_norm": 0.3171123266220093,
|
343 |
+
"learning_rate": 1.8486908682093175e-05,
|
344 |
+
"loss": 2.6307,
|
345 |
"step": 4800
|
346 |
},
|
347 |
{
|
348 |
+
"epoch": 0.8527671423599026,
|
349 |
+
"grad_norm": 0.2955775558948517,
|
350 |
+
"learning_rate": 1.674526503944611e-05,
|
351 |
+
"loss": 2.6315,
|
352 |
"step": 4900
|
353 |
},
|
354 |
{
|
355 |
+
"epoch": 0.8701705534284719,
|
356 |
+
"grad_norm": 0.2767013907432556,
|
357 |
+
"learning_rate": 1.450335594635761e-05,
|
358 |
+
"loss": 2.6138,
|
359 |
"step": 5000
|
360 |
},
|
361 |
{
|
362 |
+
"epoch": 0.8875739644970414,
|
363 |
+
"grad_norm": 0.27960339188575745,
|
364 |
+
"learning_rate": 1.1927453544210397e-05,
|
365 |
+
"loss": 2.6305,
|
366 |
"step": 5100
|
367 |
},
|
368 |
{
|
369 |
+
"epoch": 0.9049773755656109,
|
370 |
+
"grad_norm": 0.31521016359329224,
|
371 |
+
"learning_rate": 9.20860073020234e-06,
|
372 |
+
"loss": 2.6249,
|
373 |
"step": 5200
|
374 |
},
|
375 |
{
|
376 |
+
"epoch": 0.9223807866341803,
|
377 |
+
"grad_norm": 0.2640378773212433,
|
378 |
+
"learning_rate": 6.548442379624425e-06,
|
379 |
+
"loss": 2.6257,
|
380 |
"step": 5300
|
381 |
},
|
382 |
{
|
383 |
+
"epoch": 0.9397841977027498,
|
384 |
+
"grad_norm": 0.28068870306015015,
|
385 |
+
"learning_rate": 4.144270267924306e-06,
|
386 |
+
"loss": 2.6261,
|
387 |
"step": 5400
|
388 |
},
|
389 |
{
|
390 |
+
"epoch": 0.9571876087713191,
|
391 |
+
"grad_norm": 0.2999429702758789,
|
392 |
+
"learning_rate": 2.1743908422712135e-06,
|
393 |
+
"loss": 2.6245,
|
394 |
"step": 5500
|
395 |
},
|
396 |
{
|
397 |
+
"epoch": 0.9745910198398886,
|
398 |
+
"grad_norm": 0.2793658971786499,
|
399 |
+
"learning_rate": 7.849010480670938e-07,
|
400 |
+
"loss": 2.6209,
|
401 |
"step": 5600
|
402 |
},
|
403 |
{
|
404 |
+
"epoch": 0.9919944309084581,
|
405 |
+
"grad_norm": 0.30049070715904236,
|
406 |
+
"learning_rate": 7.885298685522235e-08,
|
407 |
+
"loss": 2.6215,
|
408 |
"step": 5700
|
409 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
410 |
{
|
411 |
"epoch": 1.0,
|
412 |
+
"step": 5746,
|
413 |
+
"total_flos": 8.372955480242258e+17,
|
414 |
+
"train_loss": 2.6846868539578486,
|
415 |
+
"train_runtime": 1624.688,
|
416 |
+
"train_samples_per_second": 56.585,
|
417 |
+
"train_steps_per_second": 3.537
|
418 |
}
|
419 |
],
|
420 |
"logging_steps": 100,
|
421 |
+
"max_steps": 5746,
|
422 |
"num_input_tokens_seen": 0,
|
423 |
"num_train_epochs": 1,
|
424 |
"save_steps": 500,
|
|
|
434 |
"attributes": {}
|
435 |
}
|
436 |
},
|
437 |
+
"total_flos": 8.372955480242258e+17,
|
438 |
"train_batch_size": 2,
|
439 |
"trial_name": null,
|
440 |
"trial_params": null
|