End of training
Browse files- all_results.json +20 -0
- eval_results.json +9 -0
- generated_predictions.txt +0 -0
- predict_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +250 -0
all_results.json
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 12.35,
|
3 |
+
"eval_exact_match": 0.8516899766899767,
|
4 |
+
"eval_loss": 0.025585556402802467,
|
5 |
+
"eval_runtime": 3775.4217,
|
6 |
+
"eval_samples": 17160,
|
7 |
+
"eval_samples_per_second": 4.545,
|
8 |
+
"eval_steps_per_second": 1.136,
|
9 |
+
"predict_exact_match": 0.8497615057367539,
|
10 |
+
"predict_loss": 0.030499974265694618,
|
11 |
+
"predict_runtime": 8423.2902,
|
12 |
+
"predict_samples": 38785,
|
13 |
+
"predict_samples_per_second": 4.604,
|
14 |
+
"predict_steps_per_second": 1.151,
|
15 |
+
"train_loss": 0.13643742847442628,
|
16 |
+
"train_runtime": 83186.8691,
|
17 |
+
"train_samples": 124597,
|
18 |
+
"train_samples_per_second": 18.464,
|
19 |
+
"train_steps_per_second": 0.036
|
20 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 12.35,
|
3 |
+
"eval_exact_match": 0.8516899766899767,
|
4 |
+
"eval_loss": 0.025585556402802467,
|
5 |
+
"eval_runtime": 3775.4217,
|
6 |
+
"eval_samples": 17160,
|
7 |
+
"eval_samples_per_second": 4.545,
|
8 |
+
"eval_steps_per_second": 1.136
|
9 |
+
}
|
generated_predictions.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
predict_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"predict_exact_match": 0.8497615057367539,
|
3 |
+
"predict_loss": 0.030499974265694618,
|
4 |
+
"predict_runtime": 8423.2902,
|
5 |
+
"predict_samples": 38785,
|
6 |
+
"predict_samples_per_second": 4.604,
|
7 |
+
"predict_steps_per_second": 1.151
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 12.35,
|
3 |
+
"train_loss": 0.13643742847442628,
|
4 |
+
"train_runtime": 83186.8691,
|
5 |
+
"train_samples": 124597,
|
6 |
+
"train_samples_per_second": 18.464,
|
7 |
+
"train_steps_per_second": 0.036
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,250 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.025585556402802467,
|
3 |
+
"best_model_checkpoint": "/data/wheld3/mt5-base-pointer-top_v2/checkpoint-2600",
|
4 |
+
"epoch": 12.345168539325842,
|
5 |
+
"global_step": 3000,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.82,
|
12 |
+
"learning_rate": 0.0009333333333333333,
|
13 |
+
"loss": 1.4545,
|
14 |
+
"step": 200
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.82,
|
18 |
+
"eval_exact_match": 0.12942890442890442,
|
19 |
+
"eval_loss": 0.25418975949287415,
|
20 |
+
"eval_runtime": 1925.1292,
|
21 |
+
"eval_samples_per_second": 8.914,
|
22 |
+
"eval_steps_per_second": 2.228,
|
23 |
+
"step": 200
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 1.65,
|
27 |
+
"learning_rate": 0.0008666666666666667,
|
28 |
+
"loss": 0.1878,
|
29 |
+
"step": 400
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"epoch": 1.65,
|
33 |
+
"eval_exact_match": 0.21276223776223777,
|
34 |
+
"eval_loss": 0.06684188544750214,
|
35 |
+
"eval_runtime": 1936.6986,
|
36 |
+
"eval_samples_per_second": 8.86,
|
37 |
+
"eval_steps_per_second": 2.215,
|
38 |
+
"step": 400
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 2.47,
|
42 |
+
"learning_rate": 0.0008,
|
43 |
+
"loss": 0.0796,
|
44 |
+
"step": 600
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 2.47,
|
48 |
+
"eval_exact_match": 0.22756410256410256,
|
49 |
+
"eval_loss": 0.04656381905078888,
|
50 |
+
"eval_runtime": 1952.2349,
|
51 |
+
"eval_samples_per_second": 8.79,
|
52 |
+
"eval_steps_per_second": 2.197,
|
53 |
+
"step": 600
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"epoch": 3.29,
|
57 |
+
"learning_rate": 0.0007333333333333333,
|
58 |
+
"loss": 0.0536,
|
59 |
+
"step": 800
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 3.29,
|
63 |
+
"eval_exact_match": 0.23088578088578088,
|
64 |
+
"eval_loss": 0.03555314987897873,
|
65 |
+
"eval_runtime": 1939.2428,
|
66 |
+
"eval_samples_per_second": 8.849,
|
67 |
+
"eval_steps_per_second": 2.212,
|
68 |
+
"step": 800
|
69 |
+
},
|
70 |
+
{
|
71 |
+
"epoch": 4.12,
|
72 |
+
"learning_rate": 0.0006666666666666666,
|
73 |
+
"loss": 0.0424,
|
74 |
+
"step": 1000
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 4.12,
|
78 |
+
"eval_exact_match": 0.23280885780885782,
|
79 |
+
"eval_loss": 0.031658221036195755,
|
80 |
+
"eval_runtime": 1929.3944,
|
81 |
+
"eval_samples_per_second": 8.894,
|
82 |
+
"eval_steps_per_second": 2.223,
|
83 |
+
"step": 1000
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 4.94,
|
87 |
+
"learning_rate": 0.0006,
|
88 |
+
"loss": 0.0356,
|
89 |
+
"step": 1200
|
90 |
+
},
|
91 |
+
{
|
92 |
+
"epoch": 4.94,
|
93 |
+
"eval_exact_match": 0.23403263403263402,
|
94 |
+
"eval_loss": 0.02952779084444046,
|
95 |
+
"eval_runtime": 1929.8018,
|
96 |
+
"eval_samples_per_second": 8.892,
|
97 |
+
"eval_steps_per_second": 2.223,
|
98 |
+
"step": 1200
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 5.76,
|
102 |
+
"learning_rate": 0.0005333333333333334,
|
103 |
+
"loss": 0.0306,
|
104 |
+
"step": 1400
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 5.76,
|
108 |
+
"eval_exact_match": 0.23572261072261072,
|
109 |
+
"eval_loss": 0.028804348781704903,
|
110 |
+
"eval_runtime": 1939.2007,
|
111 |
+
"eval_samples_per_second": 8.849,
|
112 |
+
"eval_steps_per_second": 2.212,
|
113 |
+
"step": 1400
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 6.58,
|
117 |
+
"learning_rate": 0.00046666666666666666,
|
118 |
+
"loss": 0.0277,
|
119 |
+
"step": 1600
|
120 |
+
},
|
121 |
+
{
|
122 |
+
"epoch": 6.58,
|
123 |
+
"eval_exact_match": 0.23513986013986013,
|
124 |
+
"eval_loss": 0.027136022225022316,
|
125 |
+
"eval_runtime": 1941.0005,
|
126 |
+
"eval_samples_per_second": 8.841,
|
127 |
+
"eval_steps_per_second": 2.21,
|
128 |
+
"step": 1600
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 7.41,
|
132 |
+
"learning_rate": 0.0004,
|
133 |
+
"loss": 0.0243,
|
134 |
+
"step": 1800
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 7.41,
|
138 |
+
"eval_exact_match": 0.23513986013986013,
|
139 |
+
"eval_loss": 0.02723020501434803,
|
140 |
+
"eval_runtime": 1956.0015,
|
141 |
+
"eval_samples_per_second": 8.773,
|
142 |
+
"eval_steps_per_second": 2.193,
|
143 |
+
"step": 1800
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 8.23,
|
147 |
+
"learning_rate": 0.0003333333333333333,
|
148 |
+
"loss": 0.0225,
|
149 |
+
"step": 2000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 8.23,
|
153 |
+
"eval_exact_match": 0.2353146853146853,
|
154 |
+
"eval_loss": 0.02720719203352928,
|
155 |
+
"eval_runtime": 1939.1308,
|
156 |
+
"eval_samples_per_second": 8.849,
|
157 |
+
"eval_steps_per_second": 2.212,
|
158 |
+
"step": 2000
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 9.05,
|
162 |
+
"learning_rate": 0.0002666666666666667,
|
163 |
+
"loss": 0.0206,
|
164 |
+
"step": 2200
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 9.05,
|
168 |
+
"eval_exact_match": 0.23682983682983683,
|
169 |
+
"eval_loss": 0.026673857122659683,
|
170 |
+
"eval_runtime": 1950.8619,
|
171 |
+
"eval_samples_per_second": 8.796,
|
172 |
+
"eval_steps_per_second": 2.199,
|
173 |
+
"step": 2200
|
174 |
+
},
|
175 |
+
{
|
176 |
+
"epoch": 9.88,
|
177 |
+
"learning_rate": 0.0002,
|
178 |
+
"loss": 0.0187,
|
179 |
+
"step": 2400
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 9.88,
|
183 |
+
"eval_exact_match": 0.2367132867132867,
|
184 |
+
"eval_loss": 0.02599777653813362,
|
185 |
+
"eval_runtime": 1938.3074,
|
186 |
+
"eval_samples_per_second": 8.853,
|
187 |
+
"eval_steps_per_second": 2.213,
|
188 |
+
"step": 2400
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 10.7,
|
192 |
+
"learning_rate": 0.00013333333333333334,
|
193 |
+
"loss": 0.0173,
|
194 |
+
"step": 2600
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 10.7,
|
198 |
+
"eval_exact_match": 0.23828671328671328,
|
199 |
+
"eval_loss": 0.025585556402802467,
|
200 |
+
"eval_runtime": 1929.5665,
|
201 |
+
"eval_samples_per_second": 8.893,
|
202 |
+
"eval_steps_per_second": 2.223,
|
203 |
+
"step": 2600
|
204 |
+
},
|
205 |
+
{
|
206 |
+
"epoch": 11.52,
|
207 |
+
"learning_rate": 6.666666666666667e-05,
|
208 |
+
"loss": 0.0161,
|
209 |
+
"step": 2800
|
210 |
+
},
|
211 |
+
{
|
212 |
+
"epoch": 11.52,
|
213 |
+
"eval_exact_match": 0.23828671328671328,
|
214 |
+
"eval_loss": 0.02604704163968563,
|
215 |
+
"eval_runtime": 1944.3112,
|
216 |
+
"eval_samples_per_second": 8.826,
|
217 |
+
"eval_steps_per_second": 2.206,
|
218 |
+
"step": 2800
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 12.35,
|
222 |
+
"learning_rate": 0.0,
|
223 |
+
"loss": 0.0153,
|
224 |
+
"step": 3000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 12.35,
|
228 |
+
"eval_exact_match": 0.2377039627039627,
|
229 |
+
"eval_loss": 0.025679251179099083,
|
230 |
+
"eval_runtime": 1944.0793,
|
231 |
+
"eval_samples_per_second": 8.827,
|
232 |
+
"eval_steps_per_second": 2.207,
|
233 |
+
"step": 3000
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 12.35,
|
237 |
+
"step": 3000,
|
238 |
+
"total_flos": 7.256060777948774e+16,
|
239 |
+
"train_loss": 0.13643742847442628,
|
240 |
+
"train_runtime": 83186.8691,
|
241 |
+
"train_samples_per_second": 18.464,
|
242 |
+
"train_steps_per_second": 0.036
|
243 |
+
}
|
244 |
+
],
|
245 |
+
"max_steps": 3000,
|
246 |
+
"num_train_epochs": 13,
|
247 |
+
"total_flos": 7.256060777948774e+16,
|
248 |
+
"trial_name": null,
|
249 |
+
"trial_params": null
|
250 |
+
}
|