Upload trainer_state.json
Browse files- trainer_state.json +232 -0
trainer_state.json
ADDED
@@ -0,0 +1,232 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 20.0,
|
5 |
+
"global_step": 3200,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 1.0,
|
12 |
+
"eval_bleu": 0.060808744840238495,
|
13 |
+
"eval_loss": 2.7135698795318604,
|
14 |
+
"eval_runtime": 315.8806,
|
15 |
+
"eval_samples_per_second": 3.603,
|
16 |
+
"eval_steps_per_second": 0.114,
|
17 |
+
"step": 160
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 2.0,
|
21 |
+
"eval_bleu": 0.23820890608280518,
|
22 |
+
"eval_loss": 1.7017812728881836,
|
23 |
+
"eval_runtime": 199.1113,
|
24 |
+
"eval_samples_per_second": 5.715,
|
25 |
+
"eval_steps_per_second": 0.181,
|
26 |
+
"step": 320
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 3.0,
|
30 |
+
"eval_bleu": 0.2333157930860921,
|
31 |
+
"eval_loss": 1.7553855180740356,
|
32 |
+
"eval_runtime": 183.4244,
|
33 |
+
"eval_samples_per_second": 6.204,
|
34 |
+
"eval_steps_per_second": 0.196,
|
35 |
+
"step": 480
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 3.12,
|
39 |
+
"learning_rate": 1.69125e-05,
|
40 |
+
"loss": 2.1899,
|
41 |
+
"step": 500
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 4.0,
|
45 |
+
"eval_bleu": 0.24403815580286167,
|
46 |
+
"eval_loss": 1.729722499847412,
|
47 |
+
"eval_runtime": 199.443,
|
48 |
+
"eval_samples_per_second": 5.706,
|
49 |
+
"eval_steps_per_second": 0.181,
|
50 |
+
"step": 640
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 5.0,
|
54 |
+
"eval_bleu": 0.24248370965297772,
|
55 |
+
"eval_loss": 1.7367736101150513,
|
56 |
+
"eval_runtime": 195.0886,
|
57 |
+
"eval_samples_per_second": 5.833,
|
58 |
+
"eval_steps_per_second": 0.185,
|
59 |
+
"step": 800
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 6.0,
|
63 |
+
"eval_bleu": 0.24708175633139415,
|
64 |
+
"eval_loss": 1.7663674354553223,
|
65 |
+
"eval_runtime": 190.8193,
|
66 |
+
"eval_samples_per_second": 5.964,
|
67 |
+
"eval_steps_per_second": 0.189,
|
68 |
+
"step": 960
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 6.25,
|
72 |
+
"learning_rate": 1.37875e-05,
|
73 |
+
"loss": 1.1629,
|
74 |
+
"step": 1000
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 7.0,
|
78 |
+
"eval_bleu": 0.24734088993827677,
|
79 |
+
"eval_loss": 1.8130639791488647,
|
80 |
+
"eval_runtime": 185.235,
|
81 |
+
"eval_samples_per_second": 6.144,
|
82 |
+
"eval_steps_per_second": 0.194,
|
83 |
+
"step": 1120
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 8.0,
|
87 |
+
"eval_bleu": 0.25035574867022287,
|
88 |
+
"eval_loss": 1.860228419303894,
|
89 |
+
"eval_runtime": 191.0903,
|
90 |
+
"eval_samples_per_second": 5.955,
|
91 |
+
"eval_steps_per_second": 0.188,
|
92 |
+
"step": 1280
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 9.0,
|
96 |
+
"eval_bleu": 0.2550337582147487,
|
97 |
+
"eval_loss": 1.9042994976043701,
|
98 |
+
"eval_runtime": 192.163,
|
99 |
+
"eval_samples_per_second": 5.922,
|
100 |
+
"eval_steps_per_second": 0.187,
|
101 |
+
"step": 1440
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 9.38,
|
105 |
+
"learning_rate": 1.06625e-05,
|
106 |
+
"loss": 0.807,
|
107 |
+
"step": 1500
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 10.0,
|
111 |
+
"eval_bleu": 0.255168022317063,
|
112 |
+
"eval_loss": 1.9512995481491089,
|
113 |
+
"eval_runtime": 193.6447,
|
114 |
+
"eval_samples_per_second": 5.877,
|
115 |
+
"eval_steps_per_second": 0.186,
|
116 |
+
"step": 1600
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 11.0,
|
120 |
+
"eval_bleu": 0.2583105775089707,
|
121 |
+
"eval_loss": 2.0014865398406982,
|
122 |
+
"eval_runtime": 194.2032,
|
123 |
+
"eval_samples_per_second": 5.86,
|
124 |
+
"eval_steps_per_second": 0.185,
|
125 |
+
"step": 1760
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"epoch": 12.0,
|
129 |
+
"eval_bleu": 0.2611697071205468,
|
130 |
+
"eval_loss": 2.036052942276001,
|
131 |
+
"eval_runtime": 190.4985,
|
132 |
+
"eval_samples_per_second": 5.974,
|
133 |
+
"eval_steps_per_second": 0.189,
|
134 |
+
"step": 1920
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 12.5,
|
138 |
+
"learning_rate": 7.537500000000001e-06,
|
139 |
+
"loss": 0.5977,
|
140 |
+
"step": 2000
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 13.0,
|
144 |
+
"eval_bleu": 0.2624028465673082,
|
145 |
+
"eval_loss": 2.0794923305511475,
|
146 |
+
"eval_runtime": 192.5774,
|
147 |
+
"eval_samples_per_second": 5.909,
|
148 |
+
"eval_steps_per_second": 0.187,
|
149 |
+
"step": 2080
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 14.0,
|
153 |
+
"eval_bleu": 0.26034097889106955,
|
154 |
+
"eval_loss": 2.1036157608032227,
|
155 |
+
"eval_runtime": 198.097,
|
156 |
+
"eval_samples_per_second": 5.745,
|
157 |
+
"eval_steps_per_second": 0.182,
|
158 |
+
"step": 2240
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 15.0,
|
162 |
+
"eval_bleu": 0.264538215714405,
|
163 |
+
"eval_loss": 2.1185383796691895,
|
164 |
+
"eval_runtime": 189.6413,
|
165 |
+
"eval_samples_per_second": 6.001,
|
166 |
+
"eval_steps_per_second": 0.19,
|
167 |
+
"step": 2400
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 15.62,
|
171 |
+
"learning_rate": 4.4125000000000005e-06,
|
172 |
+
"loss": 0.4697,
|
173 |
+
"step": 2500
|
174 |
+
},
|
175 |
+
{
|
176 |
+
"epoch": 16.0,
|
177 |
+
"eval_bleu": 0.2666872542669057,
|
178 |
+
"eval_loss": 2.1361563205718994,
|
179 |
+
"eval_runtime": 189.4756,
|
180 |
+
"eval_samples_per_second": 6.006,
|
181 |
+
"eval_steps_per_second": 0.19,
|
182 |
+
"step": 2560
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 17.0,
|
186 |
+
"eval_bleu": 0.2652516887552325,
|
187 |
+
"eval_loss": 2.162111520767212,
|
188 |
+
"eval_runtime": 193.0939,
|
189 |
+
"eval_samples_per_second": 5.894,
|
190 |
+
"eval_steps_per_second": 0.186,
|
191 |
+
"step": 2720
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 18.0,
|
195 |
+
"eval_bleu": 0.2673360550776601,
|
196 |
+
"eval_loss": 2.163081169128418,
|
197 |
+
"eval_runtime": 190.9327,
|
198 |
+
"eval_samples_per_second": 5.96,
|
199 |
+
"eval_steps_per_second": 0.189,
|
200 |
+
"step": 2880
|
201 |
+
},
|
202 |
+
{
|
203 |
+
"epoch": 18.75,
|
204 |
+
"learning_rate": 1.2875000000000002e-06,
|
205 |
+
"loss": 0.4032,
|
206 |
+
"step": 3000
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 19.0,
|
210 |
+
"eval_bleu": 0.2660614156233256,
|
211 |
+
"eval_loss": 2.1683239936828613,
|
212 |
+
"eval_runtime": 190.2616,
|
213 |
+
"eval_samples_per_second": 5.981,
|
214 |
+
"eval_steps_per_second": 0.189,
|
215 |
+
"step": 3040
|
216 |
+
},
|
217 |
+
{
|
218 |
+
"epoch": 20.0,
|
219 |
+
"eval_bleu": 0.2674984140410235,
|
220 |
+
"eval_loss": 2.173663377761841,
|
221 |
+
"eval_runtime": 189.5283,
|
222 |
+
"eval_samples_per_second": 6.004,
|
223 |
+
"eval_steps_per_second": 0.19,
|
224 |
+
"step": 3200
|
225 |
+
}
|
226 |
+
],
|
227 |
+
"max_steps": 3200,
|
228 |
+
"num_train_epochs": 20,
|
229 |
+
"total_flos": 5.530638338624717e+16,
|
230 |
+
"trial_name": null,
|
231 |
+
"trial_params": null
|
232 |
+
}
|