tmnam20 commited on
Commit
b14197f
1 Parent(s): ccea73c

Training in progress, step 15000, checkpoint

Browse files
checkpoint-15000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:50cf0c1b5d758ceb5205608f8431621ad7c145001da5891ea2c1662e09ab6cfb
3
  size 932401285
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2f4d079e014223d19654d0d1e026f02d0e33842d65733aaab0ce5e7721c826b
3
  size 932401285
checkpoint-15000/pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ee91cc89d68e288e5f20afe63584f8cb19c443853b4cb1746222b6df120ff010
3
  size 466205145
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10c8a0d78eb821f7071465b5430e14b276c396c3054676ffb77409b1249f70d8
3
  size 466205145
checkpoint-15000/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fea505d3bb2cc8e103041c009ac11cec0d896dbfe4c76b96d22f081a5d99d8b8
3
  size 14575
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1a713585e10ca5328b118b85cc0cd1d602b0bee2bece307a8b1b8aac63ef26b
3
  size 14575
checkpoint-15000/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b4b33c6989ec640db67ac64ffd8c3b5b7340234a656f42a87bafeaadb08649a
3
  size 627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83f2d4595073786febf2d8027050eac23c012dc966e9818d4425465ef1d82d8c
3
  size 627
checkpoint-15000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "best_metric": 1.1149991750717163,
3
- "best_model_checkpoint": "/kaggle/output/checkpoint-9000",
4
- "epoch": 240.0,
5
  "eval_steps": 1000,
6
  "global_step": 15000,
7
  "is_hyper_param_search": false,
@@ -9,240 +9,240 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.02,
13
  "learning_rate": 2.7777777777777777e-11,
14
- "loss": 1.2395,
15
  "step": 1
16
  },
17
  {
18
- "epoch": 16.0,
19
  "learning_rate": 2.7638888888888893e-08,
20
- "loss": 1.1422,
21
  "step": 1000
22
  },
23
  {
24
- "epoch": 16.0,
25
- "eval_accuracy": 0.33073852295409184,
26
- "eval_loss": 1.1557353734970093,
27
- "eval_runtime": 54.1805,
28
- "eval_samples_per_second": 92.469,
29
- "eval_steps_per_second": 11.572,
30
  "step": 1000
31
  },
32
  {
33
- "epoch": 32.0,
34
  "learning_rate": 5.541666666666667e-08,
35
- "loss": 1.1401,
36
  "step": 2000
37
  },
38
  {
39
- "epoch": 32.0,
40
- "eval_accuracy": 0.3311377245508982,
41
- "eval_loss": 1.1507551670074463,
42
- "eval_runtime": 53.98,
43
- "eval_samples_per_second": 92.812,
44
- "eval_steps_per_second": 11.615,
45
  "step": 2000
46
  },
47
  {
48
- "epoch": 48.0,
49
  "learning_rate": 8.316666666666666e-08,
50
- "loss": 1.1287,
51
  "step": 3000
52
  },
53
  {
54
- "epoch": 48.0,
55
- "eval_accuracy": 0.3321357285429142,
56
- "eval_loss": 1.1441665887832642,
57
- "eval_runtime": 53.9873,
58
- "eval_samples_per_second": 92.8,
59
- "eval_steps_per_second": 11.614,
60
  "step": 3000
61
  },
62
  {
63
- "epoch": 64.0,
64
- "learning_rate": 1.1094444444444445e-07,
65
- "loss": 1.117,
66
  "step": 4000
67
  },
68
  {
69
- "epoch": 64.0,
70
- "eval_accuracy": 0.3287425149700599,
71
- "eval_loss": 1.134506344795227,
72
- "eval_runtime": 54.0067,
73
- "eval_samples_per_second": 92.766,
74
- "eval_steps_per_second": 11.61,
75
  "step": 4000
76
  },
77
  {
78
- "epoch": 80.0,
79
- "learning_rate": 1.3872222222222222e-07,
80
- "loss": 1.1111,
81
  "step": 5000
82
  },
83
  {
84
- "epoch": 80.0,
85
- "eval_accuracy": 0.32954091816367265,
86
- "eval_loss": 1.1283142566680908,
87
- "eval_runtime": 54.0921,
88
- "eval_samples_per_second": 92.62,
89
- "eval_steps_per_second": 11.591,
90
  "step": 5000
91
  },
92
  {
93
- "epoch": 96.0,
94
- "learning_rate": 1.6650000000000002e-07,
95
- "loss": 1.1053,
96
  "step": 6000
97
  },
98
  {
99
- "epoch": 96.0,
100
- "eval_accuracy": 0.32894211576846305,
101
- "eval_loss": 1.1218452453613281,
102
- "eval_runtime": 54.0607,
103
- "eval_samples_per_second": 92.674,
104
- "eval_steps_per_second": 11.598,
105
  "step": 6000
106
  },
107
  {
108
- "epoch": 112.0,
109
- "learning_rate": 1.9422222222222223e-07,
110
- "loss": 1.0952,
111
  "step": 7000
112
  },
113
  {
114
- "epoch": 112.0,
115
- "eval_accuracy": 0.33073852295409184,
116
- "eval_loss": 1.118302583694458,
117
- "eval_runtime": 53.9252,
118
- "eval_samples_per_second": 92.906,
119
- "eval_steps_per_second": 11.627,
120
  "step": 7000
121
  },
122
  {
123
- "epoch": 128.0,
124
- "learning_rate": 2.2197222222222225e-07,
125
- "loss": 1.0909,
126
  "step": 8000
127
  },
128
  {
129
- "epoch": 128.0,
130
- "eval_accuracy": 0.3281437125748503,
131
- "eval_loss": 1.1167163848876953,
132
- "eval_runtime": 54.095,
133
- "eval_samples_per_second": 92.615,
134
- "eval_steps_per_second": 11.591,
135
  "step": 8000
136
  },
137
  {
138
- "epoch": 144.0,
139
- "learning_rate": 2.4975e-07,
140
- "loss": 1.0838,
141
  "step": 9000
142
  },
143
  {
144
- "epoch": 144.0,
145
- "eval_accuracy": 0.3317365269461078,
146
- "eval_loss": 1.1149991750717163,
147
- "eval_runtime": 54.0536,
148
- "eval_samples_per_second": 92.686,
149
- "eval_steps_per_second": 11.6,
150
  "step": 9000
151
  },
152
  {
153
- "epoch": 160.0,
154
- "learning_rate": 2.775277777777778e-07,
155
- "loss": 1.0746,
156
  "step": 10000
157
  },
158
  {
159
- "epoch": 160.0,
160
- "eval_accuracy": 0.3347305389221557,
161
- "eval_loss": 1.1178398132324219,
162
- "eval_runtime": 54.1273,
163
- "eval_samples_per_second": 92.56,
164
- "eval_steps_per_second": 11.584,
165
  "step": 10000
166
  },
167
  {
168
- "epoch": 176.0,
169
- "learning_rate": 3.0530555555555556e-07,
170
- "loss": 1.0722,
171
  "step": 11000
172
  },
173
  {
174
- "epoch": 176.0,
175
- "eval_accuracy": 0.3373253493013972,
176
- "eval_loss": 1.1214321851730347,
177
- "eval_runtime": 54.046,
178
- "eval_samples_per_second": 92.699,
179
- "eval_steps_per_second": 11.601,
180
  "step": 11000
181
  },
182
  {
183
- "epoch": 192.0,
184
- "learning_rate": 3.3308333333333333e-07,
185
- "loss": 1.0617,
186
  "step": 12000
187
  },
188
  {
189
- "epoch": 192.0,
190
- "eval_accuracy": 0.34271457085828344,
191
- "eval_loss": 1.124221682548523,
192
- "eval_runtime": 54.0489,
193
- "eval_samples_per_second": 92.694,
194
- "eval_steps_per_second": 11.601,
195
  "step": 12000
196
  },
197
  {
198
- "epoch": 208.0,
199
- "learning_rate": 3.608611111111111e-07,
200
- "loss": 1.0509,
201
  "step": 13000
202
  },
203
  {
204
- "epoch": 208.0,
205
- "eval_accuracy": 0.3475049900199601,
206
- "eval_loss": 1.1377869844436646,
207
- "eval_runtime": 54.0137,
208
- "eval_samples_per_second": 92.754,
209
- "eval_steps_per_second": 11.608,
210
  "step": 13000
211
  },
212
  {
213
- "epoch": 224.0,
214
- "learning_rate": 3.8863888888888893e-07,
215
- "loss": 1.0381,
216
  "step": 14000
217
  },
218
  {
219
- "epoch": 224.0,
220
- "eval_accuracy": 0.34890219560878244,
221
- "eval_loss": 1.155613899230957,
222
- "eval_runtime": 53.9794,
223
- "eval_samples_per_second": 92.813,
224
- "eval_steps_per_second": 11.616,
225
  "step": 14000
226
  },
227
  {
228
- "epoch": 240.0,
229
- "learning_rate": 4.163888888888889e-07,
230
- "loss": 1.0255,
231
  "step": 15000
232
  },
233
  {
234
- "epoch": 240.0,
235
- "eval_accuracy": 0.3562874251497006,
236
- "eval_loss": 1.1687414646148682,
237
- "eval_runtime": 53.8948,
238
- "eval_samples_per_second": 92.959,
239
- "eval_steps_per_second": 11.634,
240
  "step": 15000
241
  }
242
  ],
243
  "logging_steps": 1000,
244
  "max_steps": 10000000,
245
- "num_train_epochs": 161291,
246
  "save_steps": 1000,
247
  "total_flos": 3.135532695552e+16,
248
  "trial_name": null,
 
1
  {
2
+ "best_metric": 1.0941977500915527,
3
+ "best_model_checkpoint": "/kaggle/output/checkpoint-15000",
4
+ "epoch": 0.6111473272490222,
5
  "eval_steps": 1000,
6
  "global_step": 15000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.0,
13
  "learning_rate": 2.7777777777777777e-11,
14
+ "loss": 1.029,
15
  "step": 1
16
  },
17
  {
18
+ "epoch": 0.04,
19
  "learning_rate": 2.7638888888888893e-08,
20
+ "loss": 1.181,
21
  "step": 1000
22
  },
23
  {
24
+ "epoch": 0.04,
25
+ "eval_accuracy": 0.3273453093812375,
26
+ "eval_loss": 1.1529844999313354,
27
+ "eval_runtime": 54.2837,
28
+ "eval_samples_per_second": 92.293,
29
+ "eval_steps_per_second": 11.55,
30
  "step": 1000
31
  },
32
  {
33
+ "epoch": 0.08,
34
  "learning_rate": 5.541666666666667e-08,
35
+ "loss": 1.1527,
36
  "step": 2000
37
  },
38
  {
39
+ "epoch": 0.08,
40
+ "eval_accuracy": 0.33013972055888224,
41
+ "eval_loss": 1.1351025104522705,
42
+ "eval_runtime": 54.0518,
43
+ "eval_samples_per_second": 92.689,
44
+ "eval_steps_per_second": 11.6,
45
  "step": 2000
46
  },
47
  {
48
+ "epoch": 0.12,
49
  "learning_rate": 8.316666666666666e-08,
50
+ "loss": 1.142,
51
  "step": 3000
52
  },
53
  {
54
+ "epoch": 0.12,
55
+ "eval_accuracy": 0.3317365269461078,
56
+ "eval_loss": 1.127414345741272,
57
+ "eval_runtime": 54.0871,
58
+ "eval_samples_per_second": 92.628,
59
+ "eval_steps_per_second": 11.592,
60
  "step": 3000
61
  },
62
  {
63
+ "epoch": 0.16,
64
+ "learning_rate": 1.1091666666666668e-07,
65
+ "loss": 1.1371,
66
  "step": 4000
67
  },
68
  {
69
+ "epoch": 0.16,
70
+ "eval_accuracy": 0.331936127744511,
71
+ "eval_loss": 1.1194497346878052,
72
+ "eval_runtime": 54.3907,
73
+ "eval_samples_per_second": 92.111,
74
+ "eval_steps_per_second": 11.528,
75
  "step": 4000
76
  },
77
  {
78
+ "epoch": 0.2,
79
+ "learning_rate": 1.3869444444444447e-07,
80
+ "loss": 1.1246,
81
  "step": 5000
82
  },
83
  {
84
+ "epoch": 0.2,
85
+ "eval_accuracy": 0.3345309381237525,
86
+ "eval_loss": 1.1131775379180908,
87
+ "eval_runtime": 54.3492,
88
+ "eval_samples_per_second": 92.182,
89
+ "eval_steps_per_second": 11.537,
90
  "step": 5000
91
  },
92
  {
93
+ "epoch": 0.24,
94
+ "learning_rate": 1.664722222222222e-07,
95
+ "loss": 1.1201,
96
  "step": 6000
97
  },
98
  {
99
+ "epoch": 0.24,
100
+ "eval_accuracy": 0.34151696606786425,
101
+ "eval_loss": 1.1093512773513794,
102
+ "eval_runtime": 54.3536,
103
+ "eval_samples_per_second": 92.174,
104
+ "eval_steps_per_second": 11.536,
105
  "step": 6000
106
  },
107
  {
108
+ "epoch": 0.29,
109
+ "learning_rate": 1.9425e-07,
110
+ "loss": 1.115,
111
  "step": 7000
112
  },
113
  {
114
+ "epoch": 0.29,
115
+ "eval_accuracy": 0.34510978043912177,
116
+ "eval_loss": 1.1066845655441284,
117
+ "eval_runtime": 53.9977,
118
+ "eval_samples_per_second": 92.782,
119
+ "eval_steps_per_second": 11.612,
120
  "step": 7000
121
  },
122
  {
123
+ "epoch": 0.33,
124
+ "learning_rate": 2.2202777777777778e-07,
125
+ "loss": 1.1125,
126
  "step": 8000
127
  },
128
  {
129
+ "epoch": 0.33,
130
+ "eval_accuracy": 0.3449101796407186,
131
+ "eval_loss": 1.1031057834625244,
132
+ "eval_runtime": 54.2562,
133
+ "eval_samples_per_second": 92.34,
134
+ "eval_steps_per_second": 11.556,
135
  "step": 8000
136
  },
137
  {
138
+ "epoch": 0.37,
139
+ "learning_rate": 2.498055555555556e-07,
140
+ "loss": 1.1072,
141
  "step": 9000
142
  },
143
  {
144
+ "epoch": 0.37,
145
+ "eval_accuracy": 0.34870259481037924,
146
+ "eval_loss": 1.1044224500656128,
147
+ "eval_runtime": 54.0795,
148
+ "eval_samples_per_second": 92.641,
149
+ "eval_steps_per_second": 11.594,
150
  "step": 9000
151
  },
152
  {
153
+ "epoch": 0.41,
154
+ "learning_rate": 2.7755555555555554e-07,
155
+ "loss": 1.1064,
156
  "step": 10000
157
  },
158
  {
159
+ "epoch": 0.41,
160
+ "eval_accuracy": 0.3499001996007984,
161
+ "eval_loss": 1.1001031398773193,
162
+ "eval_runtime": 54.0869,
163
+ "eval_samples_per_second": 92.629,
164
+ "eval_steps_per_second": 11.592,
165
  "step": 10000
166
  },
167
  {
168
+ "epoch": 0.45,
169
+ "learning_rate": 3.053333333333333e-07,
170
+ "loss": 1.1053,
171
  "step": 11000
172
  },
173
  {
174
+ "epoch": 0.45,
175
+ "eval_accuracy": 0.36087824351297404,
176
+ "eval_loss": 1.0987293720245361,
177
+ "eval_runtime": 53.9688,
178
+ "eval_samples_per_second": 92.831,
179
+ "eval_steps_per_second": 11.618,
180
  "step": 11000
181
  },
182
  {
183
+ "epoch": 0.49,
184
+ "learning_rate": 3.3311111111111114e-07,
185
+ "loss": 1.104,
186
  "step": 12000
187
  },
188
  {
189
+ "epoch": 0.49,
190
+ "eval_accuracy": 0.3594810379241517,
191
+ "eval_loss": 1.0968631505966187,
192
+ "eval_runtime": 54.0936,
193
+ "eval_samples_per_second": 92.617,
194
+ "eval_steps_per_second": 11.591,
195
  "step": 12000
196
  },
197
  {
198
+ "epoch": 0.53,
199
+ "learning_rate": 3.608888888888889e-07,
200
+ "loss": 1.1022,
201
  "step": 13000
202
  },
203
  {
204
+ "epoch": 0.53,
205
+ "eval_accuracy": 0.36087824351297404,
206
+ "eval_loss": 1.0962979793548584,
207
+ "eval_runtime": 54.0403,
208
+ "eval_samples_per_second": 92.709,
209
+ "eval_steps_per_second": 11.602,
210
  "step": 13000
211
  },
212
  {
213
+ "epoch": 0.57,
214
+ "learning_rate": 3.886666666666667e-07,
215
+ "loss": 1.1029,
216
  "step": 14000
217
  },
218
  {
219
+ "epoch": 0.57,
220
+ "eval_accuracy": 0.3528942115768463,
221
+ "eval_loss": 1.0987964868545532,
222
+ "eval_runtime": 54.06,
223
+ "eval_samples_per_second": 92.675,
224
+ "eval_steps_per_second": 11.598,
225
  "step": 14000
226
  },
227
  {
228
+ "epoch": 0.61,
229
+ "learning_rate": 4.163611111111111e-07,
230
+ "loss": 1.1011,
231
  "step": 15000
232
  },
233
  {
234
+ "epoch": 0.61,
235
+ "eval_accuracy": 0.36926147704590817,
236
+ "eval_loss": 1.0941977500915527,
237
+ "eval_runtime": 54.0807,
238
+ "eval_samples_per_second": 92.639,
239
+ "eval_steps_per_second": 11.594,
240
  "step": 15000
241
  }
242
  ],
243
  "logging_steps": 1000,
244
  "max_steps": 10000000,
245
+ "num_train_epochs": 408,
246
  "save_steps": 1000,
247
  "total_flos": 3.135532695552e+16,
248
  "trial_name": null,
checkpoint-15000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cb2cd1b7e648a386ccb79e47a451842e11001981c4363e993b678e0818db5cab
3
  size 4091
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5520446b18da09f4a43173b75839457981a344c0e709fd4bbcb249e3f60c859e
3
  size 4091