Upload trainer_state.json
Browse files- trainer_state.json +348 -0
trainer_state.json
ADDED
@@ -0,0 +1,348 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 5.0,
|
5 |
+
"global_step": 25000,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.1,
|
12 |
+
"learning_rate": 1.9600000000000002e-05,
|
13 |
+
"loss": 1.5146,
|
14 |
+
"step": 500
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.2,
|
18 |
+
"learning_rate": 1.9200000000000003e-05,
|
19 |
+
"loss": 1.1216,
|
20 |
+
"step": 1000
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.3,
|
24 |
+
"learning_rate": 1.88e-05,
|
25 |
+
"loss": 1.0414,
|
26 |
+
"step": 1500
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.4,
|
30 |
+
"learning_rate": 1.8400000000000003e-05,
|
31 |
+
"loss": 1.0385,
|
32 |
+
"step": 2000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.5,
|
36 |
+
"learning_rate": 1.8e-05,
|
37 |
+
"loss": 1.0463,
|
38 |
+
"step": 2500
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.6,
|
42 |
+
"learning_rate": 1.76e-05,
|
43 |
+
"loss": 1.0464,
|
44 |
+
"step": 3000
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.7,
|
48 |
+
"learning_rate": 1.72e-05,
|
49 |
+
"loss": 1.0343,
|
50 |
+
"step": 3500
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.8,
|
54 |
+
"learning_rate": 1.6800000000000002e-05,
|
55 |
+
"loss": 1.0375,
|
56 |
+
"step": 4000
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.9,
|
60 |
+
"learning_rate": 1.64e-05,
|
61 |
+
"loss": 1.0389,
|
62 |
+
"step": 4500
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 1.0,
|
66 |
+
"learning_rate": 1.6000000000000003e-05,
|
67 |
+
"loss": 1.0303,
|
68 |
+
"step": 5000
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 1.0,
|
72 |
+
"eval_loss": 0.9517383575439453,
|
73 |
+
"eval_runtime": 520.9725,
|
74 |
+
"eval_samples_per_second": 19.195,
|
75 |
+
"eval_steps_per_second": 2.399,
|
76 |
+
"step": 5000
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 1.1,
|
80 |
+
"learning_rate": 1.5600000000000003e-05,
|
81 |
+
"loss": 1.0216,
|
82 |
+
"step": 5500
|
83 |
+
},
|
84 |
+
{
|
85 |
+
"epoch": 1.2,
|
86 |
+
"learning_rate": 1.5200000000000002e-05,
|
87 |
+
"loss": 1.0131,
|
88 |
+
"step": 6000
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 1.3,
|
92 |
+
"learning_rate": 1.48e-05,
|
93 |
+
"loss": 1.0193,
|
94 |
+
"step": 6500
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 1.4,
|
98 |
+
"learning_rate": 1.4400000000000001e-05,
|
99 |
+
"loss": 1.0079,
|
100 |
+
"step": 7000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.5,
|
104 |
+
"learning_rate": 1.4e-05,
|
105 |
+
"loss": 1.0117,
|
106 |
+
"step": 7500
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 1.6,
|
110 |
+
"learning_rate": 1.3600000000000002e-05,
|
111 |
+
"loss": 1.0169,
|
112 |
+
"step": 8000
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"epoch": 1.7,
|
116 |
+
"learning_rate": 1.3200000000000002e-05,
|
117 |
+
"loss": 1.0274,
|
118 |
+
"step": 8500
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 1.8,
|
122 |
+
"learning_rate": 1.2800000000000001e-05,
|
123 |
+
"loss": 1.0096,
|
124 |
+
"step": 9000
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 1.9,
|
128 |
+
"learning_rate": 1.2400000000000002e-05,
|
129 |
+
"loss": 1.009,
|
130 |
+
"step": 9500
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 2.0,
|
134 |
+
"learning_rate": 1.2e-05,
|
135 |
+
"loss": 1.01,
|
136 |
+
"step": 10000
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 2.0,
|
140 |
+
"eval_loss": 0.9439992308616638,
|
141 |
+
"eval_runtime": 529.1205,
|
142 |
+
"eval_samples_per_second": 18.899,
|
143 |
+
"eval_steps_per_second": 2.362,
|
144 |
+
"step": 10000
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 2.1,
|
148 |
+
"learning_rate": 1.16e-05,
|
149 |
+
"loss": 0.9979,
|
150 |
+
"step": 10500
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 2.2,
|
154 |
+
"learning_rate": 1.1200000000000001e-05,
|
155 |
+
"loss": 0.9961,
|
156 |
+
"step": 11000
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 2.3,
|
160 |
+
"learning_rate": 1.0800000000000002e-05,
|
161 |
+
"loss": 0.9941,
|
162 |
+
"step": 11500
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 2.4,
|
166 |
+
"learning_rate": 1.04e-05,
|
167 |
+
"loss": 0.9823,
|
168 |
+
"step": 12000
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 2.5,
|
172 |
+
"learning_rate": 1e-05,
|
173 |
+
"loss": 0.9886,
|
174 |
+
"step": 12500
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 2.6,
|
178 |
+
"learning_rate": 9.600000000000001e-06,
|
179 |
+
"loss": 0.9784,
|
180 |
+
"step": 13000
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"epoch": 2.7,
|
184 |
+
"learning_rate": 9.200000000000002e-06,
|
185 |
+
"loss": 0.9931,
|
186 |
+
"step": 13500
|
187 |
+
},
|
188 |
+
{
|
189 |
+
"epoch": 2.8,
|
190 |
+
"learning_rate": 8.8e-06,
|
191 |
+
"loss": 0.9944,
|
192 |
+
"step": 14000
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 2.9,
|
196 |
+
"learning_rate": 8.400000000000001e-06,
|
197 |
+
"loss": 0.9873,
|
198 |
+
"step": 14500
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 3.0,
|
202 |
+
"learning_rate": 8.000000000000001e-06,
|
203 |
+
"loss": 0.9898,
|
204 |
+
"step": 15000
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 3.0,
|
208 |
+
"eval_loss": 0.9317747950553894,
|
209 |
+
"eval_runtime": 529.6801,
|
210 |
+
"eval_samples_per_second": 18.879,
|
211 |
+
"eval_steps_per_second": 2.36,
|
212 |
+
"step": 15000
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 3.1,
|
216 |
+
"learning_rate": 7.600000000000001e-06,
|
217 |
+
"loss": 0.9791,
|
218 |
+
"step": 15500
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 3.2,
|
222 |
+
"learning_rate": 7.2000000000000005e-06,
|
223 |
+
"loss": 0.9821,
|
224 |
+
"step": 16000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 3.3,
|
228 |
+
"learning_rate": 6.800000000000001e-06,
|
229 |
+
"loss": 0.9809,
|
230 |
+
"step": 16500
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 3.4,
|
234 |
+
"learning_rate": 6.4000000000000006e-06,
|
235 |
+
"loss": 0.9701,
|
236 |
+
"step": 17000
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 3.5,
|
240 |
+
"learning_rate": 6e-06,
|
241 |
+
"loss": 0.9783,
|
242 |
+
"step": 17500
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 3.6,
|
246 |
+
"learning_rate": 5.600000000000001e-06,
|
247 |
+
"loss": 0.9763,
|
248 |
+
"step": 18000
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 3.7,
|
252 |
+
"learning_rate": 5.2e-06,
|
253 |
+
"loss": 0.9631,
|
254 |
+
"step": 18500
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 3.8,
|
258 |
+
"learning_rate": 4.800000000000001e-06,
|
259 |
+
"loss": 0.973,
|
260 |
+
"step": 19000
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 3.9,
|
264 |
+
"learning_rate": 4.4e-06,
|
265 |
+
"loss": 0.9688,
|
266 |
+
"step": 19500
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 4.0,
|
270 |
+
"learning_rate": 4.000000000000001e-06,
|
271 |
+
"loss": 0.9788,
|
272 |
+
"step": 20000
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 4.0,
|
276 |
+
"eval_loss": 0.9208911061286926,
|
277 |
+
"eval_runtime": 528.6017,
|
278 |
+
"eval_samples_per_second": 18.918,
|
279 |
+
"eval_steps_per_second": 2.365,
|
280 |
+
"step": 20000
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 4.1,
|
284 |
+
"learning_rate": 3.6000000000000003e-06,
|
285 |
+
"loss": 0.9695,
|
286 |
+
"step": 20500
|
287 |
+
},
|
288 |
+
{
|
289 |
+
"epoch": 4.2,
|
290 |
+
"learning_rate": 3.2000000000000003e-06,
|
291 |
+
"loss": 0.9725,
|
292 |
+
"step": 21000
|
293 |
+
},
|
294 |
+
{
|
295 |
+
"epoch": 4.3,
|
296 |
+
"learning_rate": 2.8000000000000003e-06,
|
297 |
+
"loss": 0.9606,
|
298 |
+
"step": 21500
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 4.4,
|
302 |
+
"learning_rate": 2.4000000000000003e-06,
|
303 |
+
"loss": 0.9611,
|
304 |
+
"step": 22000
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 4.5,
|
308 |
+
"learning_rate": 2.0000000000000003e-06,
|
309 |
+
"loss": 0.9511,
|
310 |
+
"step": 22500
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 4.6,
|
314 |
+
"learning_rate": 1.6000000000000001e-06,
|
315 |
+
"loss": 0.9571,
|
316 |
+
"step": 23000
|
317 |
+
},
|
318 |
+
{
|
319 |
+
"epoch": 4.7,
|
320 |
+
"learning_rate": 1.2000000000000002e-06,
|
321 |
+
"loss": 0.9661,
|
322 |
+
"step": 23500
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 4.8,
|
326 |
+
"learning_rate": 8.000000000000001e-07,
|
327 |
+
"loss": 0.9566,
|
328 |
+
"step": 24000
|
329 |
+
},
|
330 |
+
{
|
331 |
+
"epoch": 4.9,
|
332 |
+
"learning_rate": 4.0000000000000003e-07,
|
333 |
+
"loss": 0.9647,
|
334 |
+
"step": 24500
|
335 |
+
},
|
336 |
+
{
|
337 |
+
"epoch": 5.0,
|
338 |
+
"learning_rate": 0.0,
|
339 |
+
"loss": 0.9692,
|
340 |
+
"step": 25000
|
341 |
+
}
|
342 |
+
],
|
343 |
+
"max_steps": 25000,
|
344 |
+
"num_train_epochs": 5,
|
345 |
+
"total_flos": 5.26530926592e+16,
|
346 |
+
"trial_name": null,
|
347 |
+
"trial_params": null
|
348 |
+
}
|