Initial GPTQ model commit
Browse files- trainer_state.json +826 -0
trainer_state.json
ADDED
@@ -0,0 +1,826 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.967032967032967,
|
5 |
+
"global_step": 135,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.02,
|
12 |
+
"learning_rate": 3.3333333333333333e-06,
|
13 |
+
"loss": 1.5703,
|
14 |
+
"step": 1
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.04,
|
18 |
+
"learning_rate": 6.666666666666667e-06,
|
19 |
+
"loss": 1.5859,
|
20 |
+
"step": 2
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.07,
|
24 |
+
"learning_rate": 1e-05,
|
25 |
+
"loss": 1.5566,
|
26 |
+
"step": 3
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.09,
|
30 |
+
"learning_rate": 1.3333333333333333e-05,
|
31 |
+
"loss": 1.3877,
|
32 |
+
"step": 4
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.11,
|
36 |
+
"learning_rate": 1.6666666666666667e-05,
|
37 |
+
"loss": 1.2617,
|
38 |
+
"step": 5
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.13,
|
42 |
+
"learning_rate": 2e-05,
|
43 |
+
"loss": 1.1963,
|
44 |
+
"step": 6
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.15,
|
48 |
+
"learning_rate": 1.9997034698451396e-05,
|
49 |
+
"loss": 1.1924,
|
50 |
+
"step": 7
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.18,
|
54 |
+
"learning_rate": 1.998814055240823e-05,
|
55 |
+
"loss": 1.126,
|
56 |
+
"step": 8
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.2,
|
60 |
+
"learning_rate": 1.9973322836635517e-05,
|
61 |
+
"loss": 1.123,
|
62 |
+
"step": 9
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 0.22,
|
66 |
+
"learning_rate": 1.995259033893236e-05,
|
67 |
+
"loss": 1.0996,
|
68 |
+
"step": 10
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 0.24,
|
72 |
+
"learning_rate": 1.9925955354920265e-05,
|
73 |
+
"loss": 1.0957,
|
74 |
+
"step": 11
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 0.26,
|
78 |
+
"learning_rate": 1.9893433680751105e-05,
|
79 |
+
"loss": 1.0791,
|
80 |
+
"step": 12
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.29,
|
84 |
+
"learning_rate": 1.985504460373903e-05,
|
85 |
+
"loss": 1.0791,
|
86 |
+
"step": 13
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.31,
|
90 |
+
"learning_rate": 1.9810810890921943e-05,
|
91 |
+
"loss": 1.0762,
|
92 |
+
"step": 14
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.33,
|
96 |
+
"learning_rate": 1.9760758775559275e-05,
|
97 |
+
"loss": 1.0347,
|
98 |
+
"step": 15
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 0.35,
|
102 |
+
"learning_rate": 1.9704917941574053e-05,
|
103 |
+
"loss": 1.021,
|
104 |
+
"step": 16
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 0.37,
|
108 |
+
"learning_rate": 1.9643321505948588e-05,
|
109 |
+
"loss": 1.0259,
|
110 |
+
"step": 17
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 0.4,
|
114 |
+
"learning_rate": 1.957600599908406e-05,
|
115 |
+
"loss": 1.0317,
|
116 |
+
"step": 18
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 0.42,
|
120 |
+
"learning_rate": 1.9503011343135828e-05,
|
121 |
+
"loss": 1.0078,
|
122 |
+
"step": 19
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.44,
|
126 |
+
"learning_rate": 1.9424380828337146e-05,
|
127 |
+
"loss": 0.9849,
|
128 |
+
"step": 20
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.46,
|
132 |
+
"learning_rate": 1.9340161087325483e-05,
|
133 |
+
"loss": 1.0127,
|
134 |
+
"step": 21
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.48,
|
138 |
+
"learning_rate": 1.9250402067486523e-05,
|
139 |
+
"loss": 1.0103,
|
140 |
+
"step": 22
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 0.51,
|
144 |
+
"learning_rate": 1.9155157001332374e-05,
|
145 |
+
"loss": 1.0024,
|
146 |
+
"step": 23
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 0.53,
|
150 |
+
"learning_rate": 1.905448237493147e-05,
|
151 |
+
"loss": 0.9863,
|
152 |
+
"step": 24
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 0.55,
|
156 |
+
"learning_rate": 1.894843789440892e-05,
|
157 |
+
"loss": 1.0117,
|
158 |
+
"step": 25
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 0.57,
|
162 |
+
"learning_rate": 1.8837086450537195e-05,
|
163 |
+
"loss": 0.9775,
|
164 |
+
"step": 26
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.59,
|
168 |
+
"learning_rate": 1.872049408143808e-05,
|
169 |
+
"loss": 0.9653,
|
170 |
+
"step": 27
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.62,
|
174 |
+
"learning_rate": 1.8598729933418102e-05,
|
175 |
+
"loss": 0.9487,
|
176 |
+
"step": 28
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 0.64,
|
180 |
+
"learning_rate": 1.8471866219960604e-05,
|
181 |
+
"loss": 0.9282,
|
182 |
+
"step": 29
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 0.66,
|
186 |
+
"learning_rate": 1.833997817889878e-05,
|
187 |
+
"loss": 0.9424,
|
188 |
+
"step": 30
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 0.68,
|
192 |
+
"learning_rate": 1.820314402779511e-05,
|
193 |
+
"loss": 0.9258,
|
194 |
+
"step": 31
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 0.7,
|
198 |
+
"learning_rate": 1.806144491755363e-05,
|
199 |
+
"loss": 0.9292,
|
200 |
+
"step": 32
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 0.73,
|
204 |
+
"learning_rate": 1.7914964884292543e-05,
|
205 |
+
"loss": 0.9473,
|
206 |
+
"step": 33
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.75,
|
210 |
+
"learning_rate": 1.7763790799505746e-05,
|
211 |
+
"loss": 0.9453,
|
212 |
+
"step": 34
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.77,
|
216 |
+
"learning_rate": 1.760801231854278e-05,
|
217 |
+
"loss": 0.9355,
|
218 |
+
"step": 35
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 0.79,
|
222 |
+
"learning_rate": 1.744772182743782e-05,
|
223 |
+
"loss": 0.917,
|
224 |
+
"step": 36
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 0.81,
|
228 |
+
"learning_rate": 1.728301438811916e-05,
|
229 |
+
"loss": 0.8838,
|
230 |
+
"step": 37
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 0.84,
|
234 |
+
"learning_rate": 1.711398768203178e-05,
|
235 |
+
"loss": 0.897,
|
236 |
+
"step": 38
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 0.86,
|
240 |
+
"learning_rate": 1.6940741952206342e-05,
|
241 |
+
"loss": 0.8926,
|
242 |
+
"step": 39
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 0.88,
|
246 |
+
"learning_rate": 1.676337994380903e-05,
|
247 |
+
"loss": 0.8936,
|
248 |
+
"step": 40
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.9,
|
252 |
+
"learning_rate": 1.658200684320748e-05,
|
253 |
+
"loss": 0.894,
|
254 |
+
"step": 41
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.92,
|
258 |
+
"learning_rate": 1.6396730215588913e-05,
|
259 |
+
"loss": 0.8799,
|
260 |
+
"step": 42
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 0.95,
|
264 |
+
"learning_rate": 1.6207659941167485e-05,
|
265 |
+
"loss": 0.8945,
|
266 |
+
"step": 43
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 0.97,
|
270 |
+
"learning_rate": 1.6014908150018703e-05,
|
271 |
+
"loss": 0.8838,
|
272 |
+
"step": 44
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 0.99,
|
276 |
+
"learning_rate": 1.581858915557953e-05,
|
277 |
+
"loss": 0.8745,
|
278 |
+
"step": 45
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 1.01,
|
282 |
+
"learning_rate": 1.5618819386853607e-05,
|
283 |
+
"loss": 0.9004,
|
284 |
+
"step": 46
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 1.03,
|
288 |
+
"learning_rate": 1.541571731936185e-05,
|
289 |
+
"loss": 0.7832,
|
290 |
+
"step": 47
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 1.05,
|
294 |
+
"learning_rate": 1.5209403404879305e-05,
|
295 |
+
"loss": 0.8062,
|
296 |
+
"step": 48
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.08,
|
300 |
+
"learning_rate": 1.5000000000000002e-05,
|
301 |
+
"loss": 0.8164,
|
302 |
+
"step": 49
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 1.1,
|
306 |
+
"learning_rate": 1.4787631293572094e-05,
|
307 |
+
"loss": 0.812,
|
308 |
+
"step": 50
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 1.12,
|
312 |
+
"learning_rate": 1.4572423233046386e-05,
|
313 |
+
"loss": 0.8169,
|
314 |
+
"step": 51
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 1.14,
|
318 |
+
"learning_rate": 1.4354503449781914e-05,
|
319 |
+
"loss": 0.8101,
|
320 |
+
"step": 52
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 1.16,
|
324 |
+
"learning_rate": 1.4134001183352833e-05,
|
325 |
+
"loss": 0.8052,
|
326 |
+
"step": 53
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 1.19,
|
330 |
+
"learning_rate": 1.391104720490156e-05,
|
331 |
+
"loss": 0.7568,
|
332 |
+
"step": 54
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 1.21,
|
336 |
+
"learning_rate": 1.368577373958362e-05,
|
337 |
+
"loss": 0.7974,
|
338 |
+
"step": 55
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.23,
|
342 |
+
"learning_rate": 1.3458314388150115e-05,
|
343 |
+
"loss": 0.7812,
|
344 |
+
"step": 56
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 1.25,
|
348 |
+
"learning_rate": 1.3228804047714462e-05,
|
349 |
+
"loss": 0.7603,
|
350 |
+
"step": 57
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 1.27,
|
354 |
+
"learning_rate": 1.2997378831750242e-05,
|
355 |
+
"loss": 0.7432,
|
356 |
+
"step": 58
|
357 |
+
},
|
358 |
+
{
|
359 |
+
"epoch": 1.3,
|
360 |
+
"learning_rate": 1.2764175989367717e-05,
|
361 |
+
"loss": 0.7617,
|
362 |
+
"step": 59
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 1.32,
|
366 |
+
"learning_rate": 1.2529333823916807e-05,
|
367 |
+
"loss": 0.7676,
|
368 |
+
"step": 60
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 1.34,
|
372 |
+
"learning_rate": 1.2292991610964902e-05,
|
373 |
+
"loss": 0.751,
|
374 |
+
"step": 61
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 1.36,
|
378 |
+
"learning_rate": 1.2055289515698008e-05,
|
379 |
+
"loss": 0.7607,
|
380 |
+
"step": 62
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.38,
|
384 |
+
"learning_rate": 1.1816368509794365e-05,
|
385 |
+
"loss": 0.7427,
|
386 |
+
"step": 63
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 1.41,
|
390 |
+
"learning_rate": 1.1576370287819737e-05,
|
391 |
+
"loss": 0.772,
|
392 |
+
"step": 64
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"epoch": 1.43,
|
396 |
+
"learning_rate": 1.133543718319398e-05,
|
397 |
+
"loss": 0.7065,
|
398 |
+
"step": 65
|
399 |
+
},
|
400 |
+
{
|
401 |
+
"epoch": 1.45,
|
402 |
+
"learning_rate": 1.1093712083778748e-05,
|
403 |
+
"loss": 0.75,
|
404 |
+
"step": 66
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 1.47,
|
408 |
+
"learning_rate": 1.0851338347136358e-05,
|
409 |
+
"loss": 0.7437,
|
410 |
+
"step": 67
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 1.49,
|
414 |
+
"learning_rate": 1.060845971551014e-05,
|
415 |
+
"loss": 0.6865,
|
416 |
+
"step": 68
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 1.52,
|
420 |
+
"learning_rate": 1.0365220230576592e-05,
|
421 |
+
"loss": 0.7139,
|
422 |
+
"step": 69
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.54,
|
426 |
+
"learning_rate": 1.0121764148019977e-05,
|
427 |
+
"loss": 0.7056,
|
428 |
+
"step": 70
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 1.56,
|
432 |
+
"learning_rate": 9.878235851980027e-06,
|
433 |
+
"loss": 0.6992,
|
434 |
+
"step": 71
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 1.58,
|
438 |
+
"learning_rate": 9.634779769423412e-06,
|
439 |
+
"loss": 0.7368,
|
440 |
+
"step": 72
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 1.6,
|
444 |
+
"learning_rate": 9.391540284489862e-06,
|
445 |
+
"loss": 0.6738,
|
446 |
+
"step": 73
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 1.63,
|
450 |
+
"learning_rate": 9.148661652863644e-06,
|
451 |
+
"loss": 0.6826,
|
452 |
+
"step": 74
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 1.65,
|
456 |
+
"learning_rate": 8.906287916221259e-06,
|
457 |
+
"loss": 0.7212,
|
458 |
+
"step": 75
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 1.67,
|
462 |
+
"learning_rate": 8.664562816806022e-06,
|
463 |
+
"loss": 0.6875,
|
464 |
+
"step": 76
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.69,
|
468 |
+
"learning_rate": 8.423629712180265e-06,
|
469 |
+
"loss": 0.7065,
|
470 |
+
"step": 77
|
471 |
+
},
|
472 |
+
{
|
473 |
+
"epoch": 1.71,
|
474 |
+
"learning_rate": 8.183631490205636e-06,
|
475 |
+
"loss": 0.6846,
|
476 |
+
"step": 78
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 1.74,
|
480 |
+
"learning_rate": 7.944710484301995e-06,
|
481 |
+
"loss": 0.666,
|
482 |
+
"step": 79
|
483 |
+
},
|
484 |
+
{
|
485 |
+
"epoch": 1.76,
|
486 |
+
"learning_rate": 7.707008389035102e-06,
|
487 |
+
"loss": 0.6987,
|
488 |
+
"step": 80
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"epoch": 1.78,
|
492 |
+
"learning_rate": 7.470666176083193e-06,
|
493 |
+
"loss": 0.6719,
|
494 |
+
"step": 81
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 1.8,
|
498 |
+
"learning_rate": 7.235824010632284e-06,
|
499 |
+
"loss": 0.6743,
|
500 |
+
"step": 82
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 1.82,
|
504 |
+
"learning_rate": 7.002621168249759e-06,
|
505 |
+
"loss": 0.6567,
|
506 |
+
"step": 83
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.85,
|
510 |
+
"learning_rate": 6.771195952285541e-06,
|
511 |
+
"loss": 0.6733,
|
512 |
+
"step": 84
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 1.87,
|
516 |
+
"learning_rate": 6.5416856118498874e-06,
|
517 |
+
"loss": 0.6538,
|
518 |
+
"step": 85
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 1.89,
|
522 |
+
"learning_rate": 6.314226260416383e-06,
|
523 |
+
"loss": 0.7173,
|
524 |
+
"step": 86
|
525 |
+
},
|
526 |
+
{
|
527 |
+
"epoch": 1.91,
|
528 |
+
"learning_rate": 6.088952795098442e-06,
|
529 |
+
"loss": 0.6758,
|
530 |
+
"step": 87
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 1.93,
|
534 |
+
"learning_rate": 5.8659988166471715e-06,
|
535 |
+
"loss": 0.6123,
|
536 |
+
"step": 88
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 1.96,
|
540 |
+
"learning_rate": 5.645496550218089e-06,
|
541 |
+
"loss": 0.6304,
|
542 |
+
"step": 89
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 1.98,
|
546 |
+
"learning_rate": 5.427576766953615e-06,
|
547 |
+
"loss": 0.667,
|
548 |
+
"step": 90
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.0,
|
552 |
+
"learning_rate": 5.212368706427913e-06,
|
553 |
+
"loss": 0.6431,
|
554 |
+
"step": 91
|
555 |
+
},
|
556 |
+
{
|
557 |
+
"epoch": 2.02,
|
558 |
+
"learning_rate": 5.000000000000003e-06,
|
559 |
+
"loss": 0.583,
|
560 |
+
"step": 92
|
561 |
+
},
|
562 |
+
{
|
563 |
+
"epoch": 2.04,
|
564 |
+
"learning_rate": 4.790596595120699e-06,
|
565 |
+
"loss": 0.6089,
|
566 |
+
"step": 93
|
567 |
+
},
|
568 |
+
{
|
569 |
+
"epoch": 2.07,
|
570 |
+
"learning_rate": 4.584282680638155e-06,
|
571 |
+
"loss": 0.5725,
|
572 |
+
"step": 94
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 2.09,
|
576 |
+
"learning_rate": 4.381180613146396e-06,
|
577 |
+
"loss": 0.5759,
|
578 |
+
"step": 95
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 2.11,
|
582 |
+
"learning_rate": 4.181410844420473e-06,
|
583 |
+
"loss": 0.5928,
|
584 |
+
"step": 96
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 2.13,
|
588 |
+
"learning_rate": 3.9850918499812976e-06,
|
589 |
+
"loss": 0.6167,
|
590 |
+
"step": 97
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 2.15,
|
594 |
+
"learning_rate": 3.7923400588325156e-06,
|
595 |
+
"loss": 0.5425,
|
596 |
+
"step": 98
|
597 |
+
},
|
598 |
+
{
|
599 |
+
"epoch": 2.18,
|
600 |
+
"learning_rate": 3.6032697844110896e-06,
|
601 |
+
"loss": 0.5405,
|
602 |
+
"step": 99
|
603 |
+
},
|
604 |
+
{
|
605 |
+
"epoch": 2.2,
|
606 |
+
"learning_rate": 3.4179931567925216e-06,
|
607 |
+
"loss": 0.5547,
|
608 |
+
"step": 100
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 2.22,
|
612 |
+
"learning_rate": 3.236620056190972e-06,
|
613 |
+
"loss": 0.5686,
|
614 |
+
"step": 101
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 2.24,
|
618 |
+
"learning_rate": 3.0592580477936606e-06,
|
619 |
+
"loss": 0.5874,
|
620 |
+
"step": 102
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 2.26,
|
624 |
+
"learning_rate": 2.8860123179682244e-06,
|
625 |
+
"loss": 0.5217,
|
626 |
+
"step": 103
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 2.29,
|
630 |
+
"learning_rate": 2.7169856118808414e-06,
|
631 |
+
"loss": 0.5024,
|
632 |
+
"step": 104
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 2.31,
|
636 |
+
"learning_rate": 2.5522781725621814e-06,
|
637 |
+
"loss": 0.5518,
|
638 |
+
"step": 105
|
639 |
+
},
|
640 |
+
{
|
641 |
+
"epoch": 2.33,
|
642 |
+
"learning_rate": 2.3919876814572197e-06,
|
643 |
+
"loss": 0.5415,
|
644 |
+
"step": 106
|
645 |
+
},
|
646 |
+
{
|
647 |
+
"epoch": 2.35,
|
648 |
+
"learning_rate": 2.2362092004942583e-06,
|
649 |
+
"loss": 0.5635,
|
650 |
+
"step": 107
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 2.37,
|
654 |
+
"learning_rate": 2.08503511570746e-06,
|
655 |
+
"loss": 0.5464,
|
656 |
+
"step": 108
|
657 |
+
},
|
658 |
+
{
|
659 |
+
"epoch": 2.4,
|
660 |
+
"learning_rate": 1.9385550824463727e-06,
|
661 |
+
"loss": 0.5669,
|
662 |
+
"step": 109
|
663 |
+
},
|
664 |
+
{
|
665 |
+
"epoch": 2.42,
|
666 |
+
"learning_rate": 1.7968559722048906e-06,
|
667 |
+
"loss": 0.4912,
|
668 |
+
"step": 110
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 2.44,
|
672 |
+
"learning_rate": 1.660021821101222e-06,
|
673 |
+
"loss": 0.533,
|
674 |
+
"step": 111
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.46,
|
678 |
+
"learning_rate": 1.528133780039397e-06,
|
679 |
+
"loss": 0.5229,
|
680 |
+
"step": 112
|
681 |
+
},
|
682 |
+
{
|
683 |
+
"epoch": 2.48,
|
684 |
+
"learning_rate": 1.401270066581899e-06,
|
685 |
+
"loss": 0.5544,
|
686 |
+
"step": 113
|
687 |
+
},
|
688 |
+
{
|
689 |
+
"epoch": 2.51,
|
690 |
+
"learning_rate": 1.279505918561923e-06,
|
691 |
+
"loss": 0.54,
|
692 |
+
"step": 114
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 2.53,
|
696 |
+
"learning_rate": 1.1629135494628097e-06,
|
697 |
+
"loss": 0.5322,
|
698 |
+
"step": 115
|
699 |
+
},
|
700 |
+
{
|
701 |
+
"epoch": 2.55,
|
702 |
+
"learning_rate": 1.051562105591082e-06,
|
703 |
+
"loss": 0.5723,
|
704 |
+
"step": 116
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 2.57,
|
708 |
+
"learning_rate": 9.455176250685338e-07,
|
709 |
+
"loss": 0.5669,
|
710 |
+
"step": 117
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 2.59,
|
714 |
+
"learning_rate": 8.448429986676298e-07,
|
715 |
+
"loss": 0.5312,
|
716 |
+
"step": 118
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 2.62,
|
720 |
+
"learning_rate": 7.495979325134806e-07,
|
721 |
+
"loss": 0.563,
|
722 |
+
"step": 119
|
723 |
+
},
|
724 |
+
{
|
725 |
+
"epoch": 2.64,
|
726 |
+
"learning_rate": 6.598389126745209e-07,
|
727 |
+
"loss": 0.5959,
|
728 |
+
"step": 120
|
729 |
+
},
|
730 |
+
{
|
731 |
+
"epoch": 2.66,
|
732 |
+
"learning_rate": 5.756191716628556e-07,
|
733 |
+
"loss": 0.5435,
|
734 |
+
"step": 121
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 2.68,
|
738 |
+
"learning_rate": 4.969886568641757e-07,
|
739 |
+
"loss": 0.5308,
|
740 |
+
"step": 122
|
741 |
+
},
|
742 |
+
{
|
743 |
+
"epoch": 2.7,
|
744 |
+
"learning_rate": 4.2399400091594154e-07,
|
745 |
+
"loss": 0.5337,
|
746 |
+
"step": 123
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 2.73,
|
750 |
+
"learning_rate": 3.566784940514145e-07,
|
751 |
+
"loss": 0.5068,
|
752 |
+
"step": 124
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 2.75,
|
756 |
+
"learning_rate": 2.9508205842594727e-07,
|
757 |
+
"loss": 0.5024,
|
758 |
+
"step": 125
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 2.77,
|
762 |
+
"learning_rate": 2.392412244407294e-07,
|
763 |
+
"loss": 0.554,
|
764 |
+
"step": 126
|
765 |
+
},
|
766 |
+
{
|
767 |
+
"epoch": 2.79,
|
768 |
+
"learning_rate": 1.8918910907805733e-07,
|
769 |
+
"loss": 0.5505,
|
770 |
+
"step": 127
|
771 |
+
},
|
772 |
+
{
|
773 |
+
"epoch": 2.81,
|
774 |
+
"learning_rate": 1.4495539626097289e-07,
|
775 |
+
"loss": 0.5974,
|
776 |
+
"step": 128
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 2.84,
|
780 |
+
"learning_rate": 1.0656631924889749e-07,
|
781 |
+
"loss": 0.4895,
|
782 |
+
"step": 129
|
783 |
+
},
|
784 |
+
{
|
785 |
+
"epoch": 2.86,
|
786 |
+
"learning_rate": 7.404464507973608e-08,
|
787 |
+
"loss": 0.5193,
|
788 |
+
"step": 130
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 2.88,
|
792 |
+
"learning_rate": 4.740966106764222e-08,
|
793 |
+
"loss": 0.5391,
|
794 |
+
"step": 131
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 2.9,
|
798 |
+
"learning_rate": 2.667716336448356e-08,
|
799 |
+
"loss": 0.5173,
|
800 |
+
"step": 132
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 2.92,
|
804 |
+
"learning_rate": 1.1859447591769934e-08,
|
805 |
+
"loss": 0.5071,
|
806 |
+
"step": 133
|
807 |
+
},
|
808 |
+
{
|
809 |
+
"epoch": 2.95,
|
810 |
+
"learning_rate": 2.9653015486064143e-09,
|
811 |
+
"loss": 0.4993,
|
812 |
+
"step": 134
|
813 |
+
},
|
814 |
+
{
|
815 |
+
"epoch": 2.97,
|
816 |
+
"learning_rate": 0.0,
|
817 |
+
"loss": 0.5388,
|
818 |
+
"step": 135
|
819 |
+
}
|
820 |
+
],
|
821 |
+
"max_steps": 135,
|
822 |
+
"num_train_epochs": 3,
|
823 |
+
"total_flos": 169585743495168.0,
|
824 |
+
"trial_name": null,
|
825 |
+
"trial_params": null
|
826 |
+
}
|