Initial GPTQ model commit
Browse files- trainer_state.json +412 -0
trainer_state.json
ADDED
@@ -0,0 +1,412 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.9171270718232045,
|
5 |
+
"global_step": 66,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.04,
|
12 |
+
"learning_rate": 6.666666666666667e-06,
|
13 |
+
"loss": 1.6057,
|
14 |
+
"step": 1
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.09,
|
18 |
+
"learning_rate": 1.3333333333333333e-05,
|
19 |
+
"loss": 1.6492,
|
20 |
+
"step": 2
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.13,
|
24 |
+
"learning_rate": 2e-05,
|
25 |
+
"loss": 1.5781,
|
26 |
+
"step": 3
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.18,
|
30 |
+
"learning_rate": 1.9987569212189224e-05,
|
31 |
+
"loss": 1.365,
|
32 |
+
"step": 4
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.22,
|
36 |
+
"learning_rate": 1.9950307753654016e-05,
|
37 |
+
"loss": 1.5415,
|
38 |
+
"step": 5
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.27,
|
42 |
+
"learning_rate": 1.9888308262251286e-05,
|
43 |
+
"loss": 1.2678,
|
44 |
+
"step": 6
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.31,
|
48 |
+
"learning_rate": 1.9801724878485438e-05,
|
49 |
+
"loss": 1.2053,
|
50 |
+
"step": 7
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.35,
|
54 |
+
"learning_rate": 1.969077286229078e-05,
|
55 |
+
"loss": 1.2021,
|
56 |
+
"step": 8
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.4,
|
60 |
+
"learning_rate": 1.955572805786141e-05,
|
61 |
+
"loss": 1.1851,
|
62 |
+
"step": 9
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 0.44,
|
66 |
+
"learning_rate": 1.9396926207859085e-05,
|
67 |
+
"loss": 1.1589,
|
68 |
+
"step": 10
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 0.49,
|
72 |
+
"learning_rate": 1.921476211870408e-05,
|
73 |
+
"loss": 1.1465,
|
74 |
+
"step": 11
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 0.53,
|
78 |
+
"learning_rate": 1.900968867902419e-05,
|
79 |
+
"loss": 1.1318,
|
80 |
+
"step": 12
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.57,
|
84 |
+
"learning_rate": 1.8782215733702286e-05,
|
85 |
+
"loss": 1.1279,
|
86 |
+
"step": 13
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.62,
|
90 |
+
"learning_rate": 1.8532908816321557e-05,
|
91 |
+
"loss": 1.1071,
|
92 |
+
"step": 14
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.66,
|
96 |
+
"learning_rate": 1.826238774315995e-05,
|
97 |
+
"loss": 1.0994,
|
98 |
+
"step": 15
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 0.71,
|
102 |
+
"learning_rate": 1.7971325072229227e-05,
|
103 |
+
"loss": 1.0902,
|
104 |
+
"step": 16
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 0.75,
|
108 |
+
"learning_rate": 1.766044443118978e-05,
|
109 |
+
"loss": 1.0745,
|
110 |
+
"step": 17
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 0.8,
|
114 |
+
"learning_rate": 1.7330518718298263e-05,
|
115 |
+
"loss": 1.0742,
|
116 |
+
"step": 18
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 0.84,
|
120 |
+
"learning_rate": 1.698236818086073e-05,
|
121 |
+
"loss": 1.0671,
|
122 |
+
"step": 19
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.88,
|
126 |
+
"learning_rate": 1.6616858375968596e-05,
|
127 |
+
"loss": 1.0704,
|
128 |
+
"step": 20
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.93,
|
132 |
+
"learning_rate": 1.6234898018587336e-05,
|
133 |
+
"loss": 1.0553,
|
134 |
+
"step": 21
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.97,
|
138 |
+
"learning_rate": 1.5837436722347902e-05,
|
139 |
+
"loss": 1.0565,
|
140 |
+
"step": 22
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 1.02,
|
144 |
+
"learning_rate": 1.5425462638657597e-05,
|
145 |
+
"loss": 1.0477,
|
146 |
+
"step": 23
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 1.06,
|
150 |
+
"learning_rate": 1.5000000000000002e-05,
|
151 |
+
"loss": 1.0238,
|
152 |
+
"step": 24
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 1.1,
|
156 |
+
"learning_rate": 1.4562106573531632e-05,
|
157 |
+
"loss": 1.0179,
|
158 |
+
"step": 25
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 1.15,
|
162 |
+
"learning_rate": 1.4112871031306118e-05,
|
163 |
+
"loss": 1.0312,
|
164 |
+
"step": 26
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 1.19,
|
168 |
+
"learning_rate": 1.3653410243663953e-05,
|
169 |
+
"loss": 1.0149,
|
170 |
+
"step": 27
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.24,
|
174 |
+
"learning_rate": 1.3184866502516846e-05,
|
175 |
+
"loss": 1.0096,
|
176 |
+
"step": 28
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 1.28,
|
180 |
+
"learning_rate": 1.2708404681430054e-05,
|
181 |
+
"loss": 0.9969,
|
182 |
+
"step": 29
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 1.33,
|
186 |
+
"learning_rate": 1.2225209339563144e-05,
|
187 |
+
"loss": 1.0026,
|
188 |
+
"step": 30
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 1.37,
|
192 |
+
"learning_rate": 1.1736481776669307e-05,
|
193 |
+
"loss": 0.9957,
|
194 |
+
"step": 31
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 1.41,
|
198 |
+
"learning_rate": 1.1243437046474854e-05,
|
199 |
+
"loss": 1.0106,
|
200 |
+
"step": 32
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 1.46,
|
204 |
+
"learning_rate": 1.0747300935864245e-05,
|
205 |
+
"loss": 0.9747,
|
206 |
+
"step": 33
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 1.5,
|
210 |
+
"learning_rate": 1.0249306917380731e-05,
|
211 |
+
"loss": 0.9871,
|
212 |
+
"step": 34
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.55,
|
216 |
+
"learning_rate": 9.750693082619274e-06,
|
217 |
+
"loss": 0.9625,
|
218 |
+
"step": 35
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 1.59,
|
222 |
+
"learning_rate": 9.252699064135759e-06,
|
223 |
+
"loss": 0.9851,
|
224 |
+
"step": 36
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 1.64,
|
228 |
+
"learning_rate": 8.756562953525151e-06,
|
229 |
+
"loss": 0.969,
|
230 |
+
"step": 37
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 1.68,
|
234 |
+
"learning_rate": 8.263518223330698e-06,
|
235 |
+
"loss": 0.9961,
|
236 |
+
"step": 38
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 1.72,
|
240 |
+
"learning_rate": 7.774790660436857e-06,
|
241 |
+
"loss": 0.9688,
|
242 |
+
"step": 39
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 1.77,
|
246 |
+
"learning_rate": 7.291595318569951e-06,
|
247 |
+
"loss": 0.9729,
|
248 |
+
"step": 40
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 1.81,
|
252 |
+
"learning_rate": 6.815133497483157e-06,
|
253 |
+
"loss": 0.9558,
|
254 |
+
"step": 41
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.86,
|
258 |
+
"learning_rate": 6.34658975633605e-06,
|
259 |
+
"loss": 0.9746,
|
260 |
+
"step": 42
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 1.9,
|
264 |
+
"learning_rate": 5.887128968693887e-06,
|
265 |
+
"loss": 0.9725,
|
266 |
+
"step": 43
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 1.94,
|
270 |
+
"learning_rate": 5.43789342646837e-06,
|
271 |
+
"loss": 0.9567,
|
272 |
+
"step": 44
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 1.99,
|
276 |
+
"learning_rate": 5.000000000000003e-06,
|
277 |
+
"loss": 0.959,
|
278 |
+
"step": 45
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 2.03,
|
282 |
+
"learning_rate": 4.5745373613424075e-06,
|
283 |
+
"loss": 0.9651,
|
284 |
+
"step": 46
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 2.08,
|
288 |
+
"learning_rate": 4.162563277652104e-06,
|
289 |
+
"loss": 0.9515,
|
290 |
+
"step": 47
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 2.12,
|
294 |
+
"learning_rate": 3.7651019814126656e-06,
|
295 |
+
"loss": 0.9565,
|
296 |
+
"step": 48
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.17,
|
300 |
+
"learning_rate": 3.3831416240314085e-06,
|
301 |
+
"loss": 0.9751,
|
302 |
+
"step": 49
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 2.21,
|
306 |
+
"learning_rate": 3.017631819139273e-06,
|
307 |
+
"loss": 0.9456,
|
308 |
+
"step": 50
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 2.25,
|
312 |
+
"learning_rate": 2.669481281701739e-06,
|
313 |
+
"loss": 0.9551,
|
314 |
+
"step": 51
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 2.3,
|
318 |
+
"learning_rate": 2.339555568810221e-06,
|
319 |
+
"loss": 0.9397,
|
320 |
+
"step": 52
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 2.34,
|
324 |
+
"learning_rate": 2.0286749277707783e-06,
|
325 |
+
"loss": 0.9437,
|
326 |
+
"step": 53
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 2.39,
|
330 |
+
"learning_rate": 1.7376122568400533e-06,
|
331 |
+
"loss": 0.938,
|
332 |
+
"step": 54
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 2.43,
|
336 |
+
"learning_rate": 1.467091183678444e-06,
|
337 |
+
"loss": 0.9395,
|
338 |
+
"step": 55
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 2.48,
|
342 |
+
"learning_rate": 1.2177842662977136e-06,
|
343 |
+
"loss": 0.944,
|
344 |
+
"step": 56
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 2.52,
|
348 |
+
"learning_rate": 9.903113209758098e-07,
|
349 |
+
"loss": 0.9169,
|
350 |
+
"step": 57
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 2.56,
|
354 |
+
"learning_rate": 7.852378812959227e-07,
|
355 |
+
"loss": 0.9301,
|
356 |
+
"step": 58
|
357 |
+
},
|
358 |
+
{
|
359 |
+
"epoch": 2.61,
|
360 |
+
"learning_rate": 6.030737921409169e-07,
|
361 |
+
"loss": 0.946,
|
362 |
+
"step": 59
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 2.65,
|
366 |
+
"learning_rate": 4.4427194213859216e-07,
|
367 |
+
"loss": 0.943,
|
368 |
+
"step": 60
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 2.7,
|
372 |
+
"learning_rate": 3.0922713770922155e-07,
|
373 |
+
"loss": 0.9523,
|
374 |
+
"step": 61
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 2.74,
|
378 |
+
"learning_rate": 1.9827512151456175e-07,
|
379 |
+
"loss": 0.916,
|
380 |
+
"step": 62
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 2.78,
|
384 |
+
"learning_rate": 1.1169173774871478e-07,
|
385 |
+
"loss": 0.9382,
|
386 |
+
"step": 63
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 2.83,
|
390 |
+
"learning_rate": 4.9692246345985905e-08,
|
391 |
+
"loss": 0.9543,
|
392 |
+
"step": 64
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"epoch": 2.87,
|
396 |
+
"learning_rate": 1.2430787810776556e-08,
|
397 |
+
"loss": 0.9474,
|
398 |
+
"step": 65
|
399 |
+
},
|
400 |
+
{
|
401 |
+
"epoch": 2.92,
|
402 |
+
"learning_rate": 0.0,
|
403 |
+
"loss": 0.9358,
|
404 |
+
"step": 66
|
405 |
+
}
|
406 |
+
],
|
407 |
+
"max_steps": 66,
|
408 |
+
"num_train_epochs": 3,
|
409 |
+
"total_flos": 85858745057280.0,
|
410 |
+
"trial_name": null,
|
411 |
+
"trial_params": null
|
412 |
+
}
|