SIGRoboticsUIUC commited on
Commit
b34c524
·
verified ·
1 Parent(s): a738373

Upload folder using huggingface_hub

Browse files
experiment_cfg/metadata.json CHANGED
@@ -4,173 +4,173 @@
4
  "state": {
5
  "left_arm": {
6
  "max": [
7
- 2.187028646469116,
8
- -99.21630096435547,
9
- 98.77407836914062,
10
- 74.72486114501953,
11
  46.27594757080078
12
  ],
13
  "min": [
14
- -12.066365242004395,
15
- -99.45140838623047,
16
- 98.33625030517578,
17
- 73.35863494873047,
18
- -52.869354248046875
19
  ],
20
  "mean": [
21
- -6.931744575500488,
22
- -99.3890380859375,
23
- 98.62979888916016,
24
- 74.25790405273438,
25
- 1.3063249588012695
26
  ],
27
  "std": [
28
- 4.007203578948975,
29
- 0.0767870619893074,
30
- 0.12424216419458389,
31
- 0.4313095211982727,
32
- 48.00337600708008
33
  ],
34
  "q01": [
35
- -12.066365242004395,
36
- -99.45140838623047,
37
- 98.51138305664062,
38
- 73.58633422851562,
39
- -52.869354248046875
40
  ],
41
  "q99": [
42
- 0.15082956850528717,
43
- -99.21630096435547,
44
- 98.77407836914062,
45
  74.72486114501953,
46
  45.88522720336914
47
  ]
48
  },
49
  "gripper1": {
50
  "max": [
51
- 17.847938537597656
52
  ],
53
  "min": [
54
- 0.4510309398174286
55
  ],
56
  "mean": [
57
- 10.498392105102539
58
  ],
59
  "std": [
60
- 7.273200035095215
61
  ],
62
  "q01": [
63
  0.4510309398174286
64
  ],
65
  "q99": [
66
- 17.847938537597656
67
  ]
68
  },
69
  "right_arm": {
70
  "max": [
71
  31.229597091674805,
72
- 71.30260467529297,
73
- 98.0138168334961,
74
- 75.76736450195312,
75
  41.3212776184082
76
  ],
77
  "min": [
78
  -27.022125244140625,
79
- -97.51502990722656,
80
- -31.865285873413086,
81
  -96.60742950439453,
82
  -65.18462371826172
83
  ],
84
  "mean": [
85
- 6.173414707183838,
86
- 8.965652465820312,
87
- 43.04948425292969,
88
- -32.647727966308594,
89
- -44.51104736328125
90
  ],
91
  "std": [
92
- 17.221529006958008,
93
- 47.10391616821289,
94
- 32.55530548095703,
95
- 45.9875373840332,
96
- 20.5490779876709
97
  ],
98
  "q01": [
99
- -24.338048934936523,
100
- -97.35470581054688,
101
- -17.356649856567383,
102
- -96.44587707519531,
103
- -59.05551528930664
104
  ],
105
  "q99": [
106
- 29.63365936279297,
107
- 62.324649810791016,
108
- 97.75475311279297,
109
- 73.66720581054688,
110
- 29.715649795532308
111
  ]
112
  },
113
  "gripper2": {
114
  "max": [
115
- 71.65532684326172
116
  ],
117
  "min": [
118
  2.324263095855713
119
  ],
120
  "mean": [
121
- 23.636430740356445
122
  ],
123
  "std": [
124
- 15.703285217285156
125
  ],
126
  "q01": [
127
- 2.4376416206359863
128
  ],
129
  "q99": [
130
- 63.20068164825567
131
  ]
132
  }
133
  },
134
  "action": {
135
  "left_arm": {
136
  "max": [
137
- 2.2628026008605957,
138
- -100.0,
139
  100.0,
140
  76.97608184814453,
141
  46.617828369140625
142
  ],
143
  "min": [
144
- -12.107979774475098,
145
  -100.0,
146
- 99.9084243774414,
147
- 75.35468292236328,
148
- -52.625152587890625
149
  ],
150
  "mean": [
151
- -6.822534084320068,
152
- -100.0,
153
- 99.92558288574219,
154
- 76.36930084228516,
155
- 1.283119797706604
156
  ],
157
  "std": [
158
- 4.120677471160889,
159
- 0.0,
160
- 0.0450742170214653,
161
- 0.6158173680305481,
162
- 47.87895584106445
163
  ],
164
  "q01": [
165
- -12.107979774475098,
166
  -100.0,
167
- 99.9084243774414,
168
- 75.35468292236328,
169
- -52.625152587890625
170
  ],
171
  "q99": [
172
- 0.43668121099472046,
173
- -100.0,
174
  100.0,
175
  76.97608184814453,
176
  45.738704681396484
@@ -178,86 +178,86 @@
178
  },
179
  "gripper1": {
180
  "max": [
181
- 18.17367935180664
182
  ],
183
  "min": [
184
- 0.17905102670192719
185
  ],
186
  "mean": [
187
- 10.681233406066895
188
  ],
189
  "std": [
190
- 7.212050914764404
191
  ],
192
  "q01": [
193
- 0.2685765326023102
194
  ],
195
  "q99": [
196
- 17.99462890625
197
  ]
198
  },
199
  "right_arm": {
200
  "max": [
201
  31.447315216064453,
202
- 71.15780639648438,
203
  99.20035552978516,
204
- 77.47068786621094,
205
  42.69340896606445
206
  ],
207
  "min": [
208
  -27.511274337768555,
209
  -99.25834655761719,
210
- -32.56330490112305,
211
  -98.40870666503906,
212
  -66.81427764892578
213
  ],
214
  "mean": [
215
- 6.259029865264893,
216
- 7.35891580581665,
217
- 42.082435607910156,
218
- -33.01646423339844,
219
- -44.4665641784668
220
  ],
221
  "std": [
222
- 17.27552604675293,
223
- 47.02023696899414,
224
- 33.24121856689453,
225
- 46.35660171508789,
226
- 20.584171295166016
227
  ],
228
  "q01": [
229
- -24.231243133544922,
230
- -97.77503204345703,
231
- -18.96934700012207,
232
- -97.82244873046875,
233
- -59.10393142700195
234
  ],
235
  "q99": [
236
- 30.052480850219858,
237
  62.34033966064453,
238
- 98.48956298828125,
239
- 74.70687103271484,
240
- 30.398540496826172
241
  ]
242
  },
243
  "gripper2": {
244
  "max": [
245
- 71.90612030029297
246
  ],
247
  "min": [
248
  0.4267425239086151
249
  ],
250
  "mean": [
251
- 20.33241081237793
252
  ],
253
  "std": [
254
- 17.951961517333984
255
  ],
256
  "q01": [
257
  1.8492176532745361
258
  ],
259
  "q99": [
260
- 64.36272933960029
261
  ]
262
  }
263
  }
 
4
  "state": {
5
  "left_arm": {
6
  "max": [
7
+ 34.16289520263672,
8
+ 61.91222381591797,
9
+ 99.38703918457031,
10
+ 75.7874755859375,
11
  46.27594757080078
12
  ],
13
  "min": [
14
+ -27.90346908569336,
15
+ -99.60814666748047,
16
+ -33.27495574951172,
17
+ -88.31119537353516,
18
+ -66.83760833740234
19
  ],
20
  "mean": [
21
+ -1.3637042045593262,
22
+ -70.07746124267578,
23
+ 81.53594970703125,
24
+ 46.74204635620117,
25
+ -32.77935791015625
26
  ],
27
  "std": [
28
+ 10.373152732849121,
29
+ 52.60585021972656,
30
+ 34.18480682373047,
31
+ 48.48912811279297,
32
+ 37.322959899902344
33
  ],
34
  "q01": [
35
+ -22.021116256713867,
36
+ -99.52977752685547,
37
+ -18.82662010192871,
38
+ -66.29981231689453,
39
+ -58.38827896118164
40
  ],
41
  "q99": [
42
+ 26.69683265686035,
43
+ 51.0971794128418,
44
+ 98.94921112060547,
45
  74.72486114501953,
46
  45.88522720336914
47
  ]
48
  },
49
  "gripper1": {
50
  "max": [
51
+ 71.77835083007812
52
  ],
53
  "min": [
54
+ 0.0
55
  ],
56
  "mean": [
57
+ 10.566564559936523
58
  ],
59
  "std": [
60
+ 10.46413803100586
61
  ],
62
  "q01": [
63
  0.4510309398174286
64
  ],
65
  "q99": [
66
+ 56.572166442871094
67
  ]
68
  },
69
  "right_arm": {
70
  "max": [
71
  31.229597091674805,
72
+ 73.70741271972656,
73
+ 98.10017395019531,
74
+ 84.24878692626953,
75
  41.3212776184082
76
  ],
77
  "min": [
78
  -27.022125244140625,
79
+ -97.75550842285156,
80
+ -63.21243667602539,
81
  -96.60742950439453,
82
  -65.18462371826172
83
  ],
84
  "mean": [
85
+ -1.5135765075683594,
86
+ -25.33967399597168,
87
+ 61.376529693603516,
88
+ 0.33546876907348633,
89
+ -41.29887390136719
90
  ],
91
  "std": [
92
+ 12.458887100219727,
93
+ 60.93878936767578,
94
+ 35.7954216003418,
95
+ 60.31285858154297,
96
+ 23.256187438964844
97
  ],
98
  "q01": [
99
+ -22.887195587158203,
100
+ -97.51502990722656,
101
+ -18.134714126586914,
102
+ -95.5573501586914,
103
+ -57.9000244140625
104
  ],
105
  "q99": [
106
+ 28.61806297302246,
107
+ 63.446895599365234,
108
+ 97.84110260009766,
109
+ 73.2633285522461,
110
+ 29.41472053527832
111
  ]
112
  },
113
  "gripper2": {
114
  "max": [
115
+ 75.05668640136719
116
  ],
117
  "min": [
118
  2.324263095855713
119
  ],
120
  "mean": [
121
+ 23.147193908691406
122
  ],
123
  "std": [
124
+ 13.153349876403809
125
  ],
126
  "q01": [
127
+ 2.551020383834839
128
  ],
129
  "q99": [
130
+ 61.621315002441406
131
  ]
132
  }
133
  },
134
  "action": {
135
  "left_arm": {
136
  "max": [
137
+ 35.92695617675781,
138
+ 62.01413345336914,
139
  100.0,
140
  76.97608184814453,
141
  46.617828369140625
142
  ],
143
  "min": [
144
+ -28.066692352294922,
145
  -100.0,
146
+ -35.07326126098633,
147
+ -89.78516387939453,
148
+ -67.17948913574219
149
  ],
150
  "mean": [
151
+ -1.1567177772521973,
152
+ -70.6091537475586,
153
+ 82.1029281616211,
154
+ 47.767425537109375,
155
+ -32.80006408691406
156
  ],
157
  "std": [
158
+ 10.52210521697998,
159
+ 52.541717529296875,
160
+ 35.224395751953125,
161
+ 49.508544921875,
162
+ 37.26472473144531
163
  ],
164
  "q01": [
165
+ -21.873760223388672,
166
  -100.0,
167
+ -20.238094329833984,
168
+ -66.76124572753906,
169
+ -58.43711853027344
170
  ],
171
  "q99": [
172
+ 27.669710159301758,
173
+ 50.230565567016654,
174
  100.0,
175
  76.97608184814453,
176
  45.738704681396484
 
178
  },
179
  "gripper1": {
180
  "max": [
181
+ 74.2166519165039
182
  ],
183
  "min": [
184
+ 0.0
185
  ],
186
  "mean": [
187
+ 10.451210021972656
188
  ],
189
  "std": [
190
+ 10.781113624572754
191
  ],
192
  "q01": [
193
+ 0.08952551335096359
194
  ],
195
  "q99": [
196
+ 56.938228607177734
197
  ]
198
  },
199
  "right_arm": {
200
  "max": [
201
  31.447315216064453,
202
+ 72.47631072998047,
203
  99.20035552978516,
204
+ 85.42713928222656,
205
  42.69340896606445
206
  ],
207
  "min": [
208
  -27.511274337768555,
209
  -99.25834655761719,
210
+ -67.30342102050781,
211
  -98.40870666503906,
212
  -66.81427764892578
213
  ],
214
  "mean": [
215
+ -1.471267580986023,
216
+ -26.692821502685547,
217
+ 60.74913787841797,
218
+ 0.23357370495796204,
219
+ -41.19910430908203
220
  ],
221
  "std": [
222
+ 12.504396438598633,
223
+ 60.0944709777832,
224
+ 36.86882400512695,
225
+ 60.7491340637207,
226
+ 23.23846435546875
227
  ],
228
  "q01": [
229
+ -23.001230239868164,
230
+ -97.692626953125,
231
+ -19.68014144897461,
232
+ -96.31491088867188,
233
+ -57.749412536621094
234
  ],
235
  "q99": [
236
+ 28.823287963867188,
237
  62.34033966064453,
238
+ 98.57840728759766,
239
+ 74.03684997558594,
240
+ 29.721281051635742
241
  ]
242
  },
243
  "gripper2": {
244
  "max": [
245
+ 77.73826599121094
246
  ],
247
  "min": [
248
  0.4267425239086151
249
  ],
250
  "mean": [
251
+ 21.515247344970703
252
  ],
253
  "std": [
254
+ 14.212292671203613
255
  ],
256
  "q01": [
257
  1.8492176532745361
258
  ],
259
  "q99": [
260
+ 61.87766647338867
261
  ]
262
  }
263
  }
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2db511ed728aaf36eecfd28fa48f0d77139bf02c6b27d1f169399b53d84ebeba
3
  size 4999367032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:115fd57ab6985556b75222944abe6f1f152a92a06ef614f5756448fabd7c02a8
3
  size 4999367032
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2f04a59b96816a7a820663a91a327c11bb5a3b1ab34a2d059d64325e693869ea
3
  size 2586705312
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38e59620503cd8a0523f4851324145e9f86aba1a919018b013f473ad5324b370
3
  size 2586705312
trainer_state.json CHANGED
@@ -2,7 +2,7 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 5.4945054945054945,
6
  "eval_steps": 500,
7
  "global_step": 1000,
8
  "is_hyper_param_search": false,
@@ -10,710 +10,719 @@
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
- "epoch": 0.054945054945054944,
14
- "grad_norm": 3.1129469871520996,
15
- "learning_rate": 1.3846153846153847e-05,
16
- "loss": 1.2218,
17
  "step": 10
18
  },
19
  {
20
- "epoch": 0.10989010989010989,
21
- "grad_norm": 1.2952975034713745,
22
- "learning_rate": 2.9230769230769234e-05,
23
- "loss": 0.4319,
24
  "step": 20
25
  },
26
  {
27
- "epoch": 0.16483516483516483,
28
- "grad_norm": 2.0939278602600098,
29
- "learning_rate": 4.461538461538462e-05,
30
- "loss": 0.2786,
31
  "step": 30
32
  },
33
  {
34
- "epoch": 0.21978021978021978,
35
- "grad_norm": 1.0714941024780273,
36
- "learning_rate": 6e-05,
37
- "loss": 0.2013,
38
  "step": 40
39
  },
40
  {
41
- "epoch": 0.27472527472527475,
42
- "grad_norm": 1.486087441444397,
43
- "learning_rate": 7.538461538461539e-05,
44
- "loss": 0.1555,
45
  "step": 50
46
  },
47
  {
48
- "epoch": 0.32967032967032966,
49
- "grad_norm": 0.8458685874938965,
50
- "learning_rate": 9.076923076923078e-05,
51
- "loss": 0.1299,
52
  "step": 60
53
  },
54
  {
55
- "epoch": 0.38461538461538464,
56
- "grad_norm": 0.5655759572982788,
57
- "learning_rate": 9.999741165552688e-05,
58
- "loss": 0.1234,
59
  "step": 70
60
  },
61
  {
62
- "epoch": 0.43956043956043955,
63
- "grad_norm": 0.7847583293914795,
64
- "learning_rate": 9.996829585771195e-05,
65
- "loss": 0.1084,
66
  "step": 80
67
  },
68
  {
69
- "epoch": 0.4945054945054945,
70
- "grad_norm": 0.5260426998138428,
71
- "learning_rate": 9.990684773387432e-05,
72
- "loss": 0.0915,
73
  "step": 90
74
  },
75
  {
76
- "epoch": 0.5494505494505495,
77
- "grad_norm": 0.6431106925010681,
78
- "learning_rate": 9.981310704444089e-05,
79
- "loss": 0.0908,
80
  "step": 100
81
  },
82
  {
83
- "epoch": 0.6043956043956044,
84
- "grad_norm": 0.4315970838069916,
85
- "learning_rate": 9.968713444496294e-05,
86
- "loss": 0.0793,
87
  "step": 110
88
  },
89
  {
90
- "epoch": 0.6593406593406593,
91
- "grad_norm": 0.4451838731765747,
92
- "learning_rate": 9.952901144686857e-05,
93
- "loss": 0.0728,
94
  "step": 120
95
  },
96
  {
97
- "epoch": 0.7142857142857143,
98
- "grad_norm": 0.44112035632133484,
99
- "learning_rate": 9.933884036472022e-05,
100
- "loss": 0.0668,
101
  "step": 130
102
  },
103
  {
104
- "epoch": 0.7692307692307693,
105
- "grad_norm": 0.31693196296691895,
106
- "learning_rate": 9.911674425001134e-05,
107
- "loss": 0.0648,
108
  "step": 140
109
  },
110
  {
111
- "epoch": 0.8241758241758241,
112
- "grad_norm": 0.5056548118591309,
113
- "learning_rate": 9.886286681154495e-05,
114
- "loss": 0.0644,
115
  "step": 150
116
  },
117
  {
118
- "epoch": 0.8791208791208791,
119
- "grad_norm": 0.370319664478302,
120
- "learning_rate": 9.857737232244604e-05,
121
- "loss": 0.0578,
122
  "step": 160
123
  },
124
  {
125
- "epoch": 0.9340659340659341,
126
- "grad_norm": 0.6004385352134705,
127
- "learning_rate": 9.826044551386744e-05,
128
- "loss": 0.0561,
129
  "step": 170
130
  },
131
  {
132
- "epoch": 0.989010989010989,
133
- "grad_norm": 0.6260236501693726,
134
- "learning_rate": 9.791229145545831e-05,
135
- "loss": 0.0511,
136
  "step": 180
137
  },
138
  {
139
- "epoch": 1.043956043956044,
140
- "grad_norm": 0.3883954882621765,
141
- "learning_rate": 9.753313542267241e-05,
142
- "loss": 0.0528,
143
  "step": 190
144
  },
145
  {
146
- "epoch": 1.098901098901099,
147
- "grad_norm": 0.4612903594970703,
148
- "learning_rate": 9.712322275100208e-05,
149
- "loss": 0.0539,
150
  "step": 200
151
  },
152
  {
153
- "epoch": 1.1538461538461537,
154
- "grad_norm": 0.41645845770835876,
155
- "learning_rate": 9.668281867723223e-05,
156
- "loss": 0.0467,
157
  "step": 210
158
  },
159
  {
160
- "epoch": 1.2087912087912087,
161
- "grad_norm": 0.45377564430236816,
162
- "learning_rate": 9.621220816781708e-05,
163
- "loss": 0.0489,
164
  "step": 220
165
  },
166
  {
167
- "epoch": 1.2637362637362637,
168
- "grad_norm": 0.33625757694244385,
169
- "learning_rate": 9.571169573449075e-05,
170
- "loss": 0.0467,
171
  "step": 230
172
  },
173
  {
174
- "epoch": 1.3186813186813187,
175
- "grad_norm": 0.2890590727329254,
176
- "learning_rate": 9.51816052372307e-05,
177
- "loss": 0.0392,
178
  "step": 240
179
  },
180
  {
181
- "epoch": 1.3736263736263736,
182
- "grad_norm": 0.5078686475753784,
183
- "learning_rate": 9.46222796747021e-05,
184
- "loss": 0.0404,
185
  "step": 250
186
  },
187
  {
188
- "epoch": 1.4285714285714286,
189
- "grad_norm": 0.3641613721847534,
190
- "learning_rate": 9.403408096231812e-05,
191
- "loss": 0.0407,
192
  "step": 260
193
  },
194
  {
195
- "epoch": 1.4835164835164836,
196
- "grad_norm": 0.30617615580558777,
197
- "learning_rate": 9.341738969806011e-05,
198
- "loss": 0.0393,
199
  "step": 270
200
  },
201
  {
202
- "epoch": 1.5384615384615383,
203
- "grad_norm": 0.3459275960922241,
204
- "learning_rate": 9.277260491620907e-05,
205
- "loss": 0.0401,
206
  "step": 280
207
  },
208
  {
209
- "epoch": 1.5934065934065935,
210
- "grad_norm": 0.2681010663509369,
211
- "learning_rate": 9.210014382914784e-05,
212
- "loss": 0.0363,
213
  "step": 290
214
  },
215
  {
216
- "epoch": 1.6483516483516483,
217
- "grad_norm": 0.35169124603271484,
218
- "learning_rate": 9.140044155740101e-05,
219
- "loss": 0.0341,
220
  "step": 300
221
  },
222
  {
223
- "epoch": 1.7032967032967035,
224
- "grad_norm": 0.28844815492630005,
225
- "learning_rate": 9.067395084808709e-05,
226
- "loss": 0.0353,
227
  "step": 310
228
  },
229
  {
230
- "epoch": 1.7582417582417582,
231
- "grad_norm": 0.3767380714416504,
232
- "learning_rate": 8.992114178196558e-05,
233
- "loss": 0.0339,
234
  "step": 320
235
  },
236
  {
237
- "epoch": 1.8131868131868132,
238
- "grad_norm": 0.37167835235595703,
239
- "learning_rate": 8.914250146926788e-05,
240
- "loss": 0.032,
241
  "step": 330
242
  },
243
  {
244
- "epoch": 1.8681318681318682,
245
- "grad_norm": 0.2863631844520569,
246
- "learning_rate": 8.833853373450936e-05,
247
- "loss": 0.0339,
248
  "step": 340
249
  },
250
  {
251
- "epoch": 1.9230769230769231,
252
- "grad_norm": 0.25606516003608704,
253
- "learning_rate": 8.750975879048629e-05,
254
- "loss": 0.0302,
255
  "step": 350
256
  },
257
  {
258
- "epoch": 1.978021978021978,
259
- "grad_norm": 0.28373125195503235,
260
- "learning_rate": 8.665671290166853e-05,
261
- "loss": 0.031,
262
  "step": 360
263
  },
264
  {
265
- "epoch": 2.032967032967033,
266
- "grad_norm": 0.41454407572746277,
267
- "learning_rate": 8.577994803720606e-05,
268
- "loss": 0.0426,
269
  "step": 370
270
  },
271
  {
272
- "epoch": 2.087912087912088,
273
- "grad_norm": 0.4597318470478058,
274
- "learning_rate": 8.488003151377351e-05,
275
- "loss": 0.0366,
276
  "step": 380
277
  },
278
  {
279
- "epoch": 2.142857142857143,
280
- "grad_norm": 0.3471977412700653,
281
- "learning_rate": 8.395754562848408e-05,
282
- "loss": 0.0358,
283
  "step": 390
284
  },
285
  {
286
- "epoch": 2.197802197802198,
287
- "grad_norm": 0.37275147438049316,
288
- "learning_rate": 8.301308728211017e-05,
289
- "loss": 0.0321,
290
  "step": 400
291
  },
292
  {
293
- "epoch": 2.2527472527472527,
294
- "grad_norm": 0.29526907205581665,
295
- "learning_rate": 8.20472675928548e-05,
296
- "loss": 0.0318,
297
  "step": 410
298
  },
299
  {
300
- "epoch": 2.3076923076923075,
301
- "grad_norm": 0.31538689136505127,
302
- "learning_rate": 8.10607115009232e-05,
303
- "loss": 0.0312,
304
  "step": 420
305
  },
306
  {
307
- "epoch": 2.3626373626373627,
308
- "grad_norm": 0.3267456889152527,
309
- "learning_rate": 8.005405736415126e-05,
310
- "loss": 0.0292,
311
  "step": 430
312
  },
313
  {
314
- "epoch": 2.4175824175824174,
315
- "grad_norm": 0.23045669496059418,
316
- "learning_rate": 7.902795654495154e-05,
317
- "loss": 0.0268,
318
  "step": 440
319
  },
320
  {
321
- "epoch": 2.4725274725274726,
322
- "grad_norm": 0.28116142749786377,
323
- "learning_rate": 7.798307298884487e-05,
324
- "loss": 0.0268,
325
  "step": 450
326
  },
327
  {
328
- "epoch": 2.5274725274725274,
329
- "grad_norm": 0.3379570245742798,
330
- "learning_rate": 7.692008279484989e-05,
331
- "loss": 0.0271,
332
  "step": 460
333
  },
334
  {
335
- "epoch": 2.5824175824175826,
336
- "grad_norm": 0.2906380295753479,
337
- "learning_rate": 7.583967377800853e-05,
338
- "loss": 0.0259,
339
  "step": 470
340
  },
341
  {
342
- "epoch": 2.6373626373626373,
343
- "grad_norm": 0.3166893422603607,
344
- "learning_rate": 7.474254502433064e-05,
345
- "loss": 0.028,
346
  "step": 480
347
  },
348
  {
349
- "epoch": 2.6923076923076925,
350
- "grad_norm": 0.3112421929836273,
351
- "learning_rate": 7.362940643844564e-05,
352
- "loss": 0.0284,
353
  "step": 490
354
  },
355
  {
356
- "epoch": 2.7472527472527473,
357
- "grad_norm": 0.3011311888694763,
358
- "learning_rate": 7.250097828425384e-05,
359
- "loss": 0.0276,
360
  "step": 500
361
  },
362
  {
363
- "epoch": 2.802197802197802,
364
- "grad_norm": 0.21055926382541656,
365
- "learning_rate": 7.135799071887486e-05,
366
- "loss": 0.0238,
367
  "step": 510
368
  },
369
  {
370
- "epoch": 2.857142857142857,
371
- "grad_norm": 0.3269054591655731,
372
- "learning_rate": 7.020118332019438e-05,
373
- "loss": 0.0233,
374
  "step": 520
375
  },
376
  {
377
- "epoch": 2.912087912087912,
378
- "grad_norm": 0.2849093973636627,
379
- "learning_rate": 6.903130460831539e-05,
380
- "loss": 0.0249,
381
  "step": 530
382
  },
383
  {
384
- "epoch": 2.967032967032967,
385
- "grad_norm": 0.2573489844799042,
386
- "learning_rate": 6.784911156122305e-05,
387
- "loss": 0.0241,
388
  "step": 540
389
  },
390
  {
391
- "epoch": 3.021978021978022,
392
- "grad_norm": 0.3901364505290985,
393
- "learning_rate": 6.665536912497687e-05,
394
- "loss": 0.026,
395
  "step": 550
396
  },
397
  {
398
- "epoch": 3.076923076923077,
399
- "grad_norm": 0.41034600138664246,
400
- "learning_rate": 6.545084971874738e-05,
401
- "loss": 0.0287,
402
  "step": 560
403
  },
404
  {
405
- "epoch": 3.131868131868132,
406
- "grad_norm": 0.24250033497810364,
407
- "learning_rate": 6.42363327350168e-05,
408
- "loss": 0.0282,
409
  "step": 570
410
  },
411
  {
412
- "epoch": 3.186813186813187,
413
- "grad_norm": 0.3414716124534607,
414
- "learning_rate": 6.301260403526801e-05,
415
- "loss": 0.028,
416
  "step": 580
417
  },
418
  {
419
- "epoch": 3.241758241758242,
420
- "grad_norm": 0.22624026238918304,
421
- "learning_rate": 6.178045544148739e-05,
422
- "loss": 0.0244,
423
  "step": 590
424
  },
425
  {
426
- "epoch": 3.2967032967032965,
427
- "grad_norm": 0.21665652096271515,
428
- "learning_rate": 6.054068422381112e-05,
429
- "loss": 0.0242,
430
  "step": 600
431
  },
432
  {
433
- "epoch": 3.3516483516483517,
434
- "grad_norm": 0.19114629924297333,
435
- "learning_rate": 5.929409258464606e-05,
436
- "loss": 0.0237,
437
  "step": 610
438
  },
439
  {
440
- "epoch": 3.4065934065934065,
441
- "grad_norm": 0.21987110376358032,
442
- "learning_rate": 5.8041487139599216e-05,
443
- "loss": 0.025,
444
  "step": 620
445
  },
446
  {
447
- "epoch": 3.4615384615384617,
448
- "grad_norm": 0.29730021953582764,
449
- "learning_rate": 5.678367839555163e-05,
450
- "loss": 0.0226,
451
  "step": 630
452
  },
453
  {
454
- "epoch": 3.5164835164835164,
455
- "grad_norm": 0.2368212342262268,
456
- "learning_rate": 5.5521480226214405e-05,
457
- "loss": 0.0241,
458
  "step": 640
459
  },
460
  {
461
- "epoch": 3.571428571428571,
462
- "grad_norm": 0.23423047363758087,
463
- "learning_rate": 5.425570934550617e-05,
464
- "loss": 0.0207,
465
  "step": 650
466
  },
467
  {
468
- "epoch": 3.6263736263736264,
469
- "grad_norm": 0.18969936668872833,
470
- "learning_rate": 5.2987184779092715e-05,
471
- "loss": 0.0228,
472
  "step": 660
473
  },
474
  {
475
- "epoch": 3.6813186813186816,
476
- "grad_norm": 0.2447623908519745,
477
- "learning_rate": 5.171672733443098e-05,
478
- "loss": 0.019,
479
  "step": 670
480
  },
481
  {
482
- "epoch": 3.7362637362637363,
483
- "grad_norm": 0.16867715120315552,
484
- "learning_rate": 5.044515906965988e-05,
485
- "loss": 0.0193,
486
  "step": 680
487
  },
488
  {
489
- "epoch": 3.791208791208791,
490
- "grad_norm": 0.15809307992458344,
491
- "learning_rate": 4.917330276168208e-05,
492
- "loss": 0.0218,
493
  "step": 690
494
  },
495
  {
496
- "epoch": 3.8461538461538463,
497
- "grad_norm": 0.2224949151277542,
498
- "learning_rate": 4.790198137378056e-05,
499
- "loss": 0.0185,
500
  "step": 700
501
  },
502
  {
503
- "epoch": 3.901098901098901,
504
- "grad_norm": 0.23750196397304535,
505
- "learning_rate": 4.663201752311461e-05,
506
- "loss": 0.0194,
507
  "step": 710
508
  },
509
  {
510
- "epoch": 3.956043956043956,
511
- "grad_norm": 0.16117140650749207,
512
- "learning_rate": 4.536423294843978e-05,
513
- "loss": 0.0185,
514
  "step": 720
515
  },
516
  {
517
- "epoch": 4.010989010989011,
518
- "grad_norm": 0.21286968886852264,
519
- "learning_rate": 4.409944797839635e-05,
520
- "loss": 0.0186,
521
  "step": 730
522
  },
523
  {
524
- "epoch": 4.065934065934066,
525
- "grad_norm": 0.2058119922876358,
526
- "learning_rate": 4.283848100070988e-05,
527
- "loss": 0.0198,
528
  "step": 740
529
  },
530
  {
531
- "epoch": 4.1208791208791204,
532
- "grad_norm": 0.18852297961711884,
533
- "learning_rate": 4.1582147932648074e-05,
534
- "loss": 0.0236,
535
  "step": 750
536
  },
537
  {
538
- "epoch": 4.175824175824176,
539
- "grad_norm": 0.2005940079689026,
540
- "learning_rate": 4.033126169307584e-05,
541
- "loss": 0.0193,
542
  "step": 760
543
  },
544
  {
545
- "epoch": 4.230769230769231,
546
- "grad_norm": 0.26670485734939575,
547
- "learning_rate": 3.9086631676450586e-05,
548
- "loss": 0.0197,
549
  "step": 770
550
  },
551
  {
552
- "epoch": 4.285714285714286,
553
- "grad_norm": 0.14271725714206696,
554
- "learning_rate": 3.784906322909813e-05,
555
- "loss": 0.0175,
556
  "step": 780
557
  },
558
  {
559
- "epoch": 4.34065934065934,
560
- "grad_norm": 0.2008945196866989,
561
- "learning_rate": 3.661935712810779e-05,
562
- "loss": 0.0191,
563
  "step": 790
564
  },
565
  {
566
- "epoch": 4.395604395604396,
567
- "grad_norm": 0.14711880683898926,
568
- "learning_rate": 3.5398309063184146e-05,
569
- "loss": 0.018,
570
  "step": 800
571
  },
572
  {
573
- "epoch": 4.450549450549451,
574
- "grad_norm": 0.2107602059841156,
575
- "learning_rate": 3.418670912179057e-05,
576
- "loss": 0.0173,
577
  "step": 810
578
  },
579
  {
580
- "epoch": 4.5054945054945055,
581
- "grad_norm": 0.16945497691631317,
582
- "learning_rate": 3.298534127791785e-05,
583
  "loss": 0.0178,
584
  "step": 820
585
  },
586
  {
587
- "epoch": 4.56043956043956,
588
- "grad_norm": 0.14888402819633484,
589
- "learning_rate": 3.179498288480834e-05,
590
- "loss": 0.0181,
591
  "step": 830
592
  },
593
  {
594
- "epoch": 4.615384615384615,
595
- "grad_norm": 0.1862138956785202,
596
- "learning_rate": 3.061640417196433e-05,
597
- "loss": 0.0171,
598
  "step": 840
599
  },
600
  {
601
- "epoch": 4.670329670329671,
602
- "grad_norm": 0.17378169298171997,
603
- "learning_rate": 2.945036774676587e-05,
604
- "loss": 0.0156,
605
  "step": 850
606
  },
607
  {
608
- "epoch": 4.725274725274725,
609
- "grad_norm": 0.1787486970424652,
610
- "learning_rate": 2.8297628101020322e-05,
611
- "loss": 0.0205,
612
  "step": 860
613
  },
614
  {
615
- "epoch": 4.78021978021978,
616
- "grad_norm": 0.1538439691066742,
617
- "learning_rate": 2.71589311227634e-05,
618
- "loss": 0.0165,
619
  "step": 870
620
  },
621
  {
622
- "epoch": 4.835164835164835,
623
- "grad_norm": 0.16950446367263794,
624
- "learning_rate": 2.6035013613627224e-05,
625
- "loss": 0.0168,
626
  "step": 880
627
  },
628
  {
629
- "epoch": 4.8901098901098905,
630
- "grad_norm": 0.1429736614227295,
631
- "learning_rate": 2.492660281208779e-05,
632
- "loss": 0.0137,
633
  "step": 890
634
  },
635
  {
636
- "epoch": 4.945054945054945,
637
- "grad_norm": 0.15313905477523804,
638
- "learning_rate": 2.3834415922900415e-05,
639
- "loss": 0.0156,
640
  "step": 900
641
  },
642
  {
643
- "epoch": 5.0,
644
- "grad_norm": 0.5302070379257202,
645
- "learning_rate": 2.2759159653027374e-05,
646
- "loss": 0.0135,
647
  "step": 910
648
  },
649
  {
650
- "epoch": 5.054945054945055,
651
- "grad_norm": 0.18755768239498138,
652
- "learning_rate": 2.170152975435859e-05,
653
- "loss": 0.016,
654
  "step": 920
655
  },
656
  {
657
- "epoch": 5.1098901098901095,
658
- "grad_norm": 0.24148383736610413,
659
- "learning_rate": 2.066221057352036e-05,
660
- "loss": 0.015,
661
  "step": 930
662
  },
663
  {
664
- "epoch": 5.164835164835165,
665
- "grad_norm": 0.20809908211231232,
666
- "learning_rate": 1.9641874609064443e-05,
667
- "loss": 0.0174,
668
  "step": 940
669
  },
670
  {
671
- "epoch": 5.21978021978022,
672
- "grad_norm": 0.21820078790187836,
673
- "learning_rate": 1.8641182076323148e-05,
674
- "loss": 0.0157,
675
  "step": 950
676
  },
677
  {
678
- "epoch": 5.274725274725275,
679
- "grad_norm": 0.17557694017887115,
680
- "learning_rate": 1.76607804802126e-05,
681
- "loss": 0.0158,
682
  "step": 960
683
  },
684
  {
685
- "epoch": 5.329670329670329,
686
- "grad_norm": 0.2052188515663147,
687
- "learning_rate": 1.6701304196260166e-05,
688
- "loss": 0.0145,
689
  "step": 970
690
  },
691
  {
692
- "epoch": 5.384615384615385,
693
- "grad_norm": 0.19574584066867828,
694
- "learning_rate": 1.5763374060127624e-05,
695
- "loss": 0.0146,
696
  "step": 980
697
  },
698
  {
699
- "epoch": 5.43956043956044,
700
- "grad_norm": 0.15790286660194397,
701
- "learning_rate": 1.4847596965895011e-05,
702
- "loss": 0.0162,
703
  "step": 990
704
  },
705
  {
706
- "epoch": 5.4945054945054945,
707
- "grad_norm": 0.16527500748634338,
708
- "learning_rate": 1.3954565473365844e-05,
709
- "loss": 0.017,
710
  "step": 1000
 
 
 
 
 
 
 
 
 
711
  }
712
  ],
713
  "logging_steps": 10,
714
- "max_steps": 1300,
715
  "num_input_tokens_seen": 0,
716
- "num_train_epochs": 8,
717
  "save_steps": 1000,
718
  "stateful_callbacks": {
719
  "TrainerControl": {
@@ -722,13 +731,13 @@
722
  "should_evaluate": false,
723
  "should_log": false,
724
  "should_save": true,
725
- "should_training_stop": false
726
  },
727
  "attributes": {}
728
  }
729
  },
730
  "total_flos": 0.0,
731
- "train_batch_size": 100,
732
  "trial_name": null,
733
  "trial_params": null
734
  }
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 0.9025270758122743,
6
  "eval_steps": 500,
7
  "global_step": 1000,
8
  "is_hyper_param_search": false,
 
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
+ "epoch": 0.009025270758122744,
14
+ "grad_norm": 1.977829933166504,
15
+ "learning_rate": 1.8e-05,
16
+ "loss": 0.9407,
17
  "step": 10
18
  },
19
  {
20
+ "epoch": 0.018050541516245487,
21
+ "grad_norm": 0.9880571365356445,
22
+ "learning_rate": 3.8e-05,
23
+ "loss": 0.2558,
24
  "step": 20
25
  },
26
  {
27
+ "epoch": 0.02707581227436823,
28
+ "grad_norm": 1.016868233680725,
29
+ "learning_rate": 5.8e-05,
30
+ "loss": 0.151,
31
  "step": 30
32
  },
33
  {
34
+ "epoch": 0.036101083032490974,
35
+ "grad_norm": 1.0683897733688354,
36
+ "learning_rate": 7.800000000000001e-05,
37
+ "loss": 0.1332,
38
  "step": 40
39
  },
40
  {
41
+ "epoch": 0.04512635379061372,
42
+ "grad_norm": 0.8018341660499573,
43
+ "learning_rate": 9.8e-05,
44
+ "loss": 0.0983,
45
  "step": 50
46
  },
47
  {
48
+ "epoch": 0.05415162454873646,
49
+ "grad_norm": 0.6524431109428406,
50
+ "learning_rate": 9.997785653888835e-05,
51
+ "loss": 0.0808,
52
  "step": 60
53
  },
54
  {
55
+ "epoch": 0.0631768953068592,
56
+ "grad_norm": 0.6845595240592957,
57
+ "learning_rate": 9.990133642141359e-05,
58
+ "loss": 0.0778,
59
  "step": 70
60
  },
61
  {
62
+ "epoch": 0.07220216606498195,
63
+ "grad_norm": 0.6983832120895386,
64
+ "learning_rate": 9.977024992520602e-05,
65
+ "loss": 0.0711,
66
  "step": 80
67
  },
68
  {
69
+ "epoch": 0.0812274368231047,
70
+ "grad_norm": 0.6112372279167175,
71
+ "learning_rate": 9.95847403914247e-05,
72
+ "loss": 0.0679,
73
  "step": 90
74
  },
75
  {
76
+ "epoch": 0.09025270758122744,
77
+ "grad_norm": 0.5820932388305664,
78
+ "learning_rate": 9.934501067202117e-05,
79
+ "loss": 0.0641,
80
  "step": 100
81
  },
82
  {
83
+ "epoch": 0.09927797833935018,
84
+ "grad_norm": 0.5818979144096375,
85
+ "learning_rate": 9.905132290792394e-05,
86
+ "loss": 0.0618,
87
  "step": 110
88
  },
89
  {
90
+ "epoch": 0.10830324909747292,
91
+ "grad_norm": 0.6202858686447144,
92
+ "learning_rate": 9.870399824239117e-05,
93
+ "loss": 0.0554,
94
  "step": 120
95
  },
96
  {
97
+ "epoch": 0.11732851985559567,
98
+ "grad_norm": 0.5585798621177673,
99
+ "learning_rate": 9.830341646984521e-05,
100
+ "loss": 0.0559,
101
  "step": 130
102
  },
103
  {
104
+ "epoch": 0.1263537906137184,
105
+ "grad_norm": 0.6366000771522522,
106
+ "learning_rate": 9.785001562057309e-05,
107
+ "loss": 0.0553,
108
  "step": 140
109
  },
110
  {
111
+ "epoch": 0.13537906137184116,
112
+ "grad_norm": 0.4224783778190613,
113
+ "learning_rate": 9.734429148174675e-05,
114
+ "loss": 0.05,
115
  "step": 150
116
  },
117
  {
118
+ "epoch": 0.1444043321299639,
119
+ "grad_norm": 0.39184054732322693,
120
+ "learning_rate": 9.6786797055287e-05,
121
+ "loss": 0.0455,
122
  "step": 160
123
  },
124
  {
125
+ "epoch": 0.15342960288808663,
126
+ "grad_norm": 0.4573412835597992,
127
+ "learning_rate": 9.617814195316411e-05,
128
+ "loss": 0.0408,
129
  "step": 170
130
  },
131
  {
132
+ "epoch": 0.1624548736462094,
133
+ "grad_norm": 0.4405965805053711,
134
+ "learning_rate": 9.551899173079607e-05,
135
+ "loss": 0.0455,
136
  "step": 180
137
  },
138
  {
139
+ "epoch": 0.17148014440433212,
140
+ "grad_norm": 0.46750208735466003,
141
+ "learning_rate": 9.481006715927351e-05,
142
+ "loss": 0.0448,
143
  "step": 190
144
  },
145
  {
146
+ "epoch": 0.18050541516245489,
147
+ "grad_norm": 0.37655723094940186,
148
+ "learning_rate": 9.405214343720707e-05,
149
+ "loss": 0.0439,
150
  "step": 200
151
  },
152
  {
153
+ "epoch": 0.18953068592057762,
154
+ "grad_norm": 0.5338054895401001,
155
+ "learning_rate": 9.32460493430591e-05,
156
+ "loss": 0.047,
157
  "step": 210
158
  },
159
  {
160
+ "epoch": 0.19855595667870035,
161
+ "grad_norm": 0.4981560707092285,
162
+ "learning_rate": 9.239266632888659e-05,
163
+ "loss": 0.0418,
164
  "step": 220
165
  },
166
  {
167
+ "epoch": 0.2075812274368231,
168
+ "grad_norm": 0.4395357668399811,
169
+ "learning_rate": 9.14929275564863e-05,
170
+ "loss": 0.0408,
171
  "step": 230
172
  },
173
  {
174
+ "epoch": 0.21660649819494585,
175
+ "grad_norm": 0.44028720259666443,
176
+ "learning_rate": 9.0547816876996e-05,
177
+ "loss": 0.0409,
178
  "step": 240
179
  },
180
  {
181
+ "epoch": 0.22563176895306858,
182
+ "grad_norm": 0.37459373474121094,
183
+ "learning_rate": 8.955836775506776e-05,
184
+ "loss": 0.0426,
185
  "step": 250
186
  },
187
  {
188
+ "epoch": 0.23465703971119134,
189
+ "grad_norm": 0.4326346516609192,
190
+ "learning_rate": 8.852566213878947e-05,
191
+ "loss": 0.0429,
192
  "step": 260
193
  },
194
  {
195
+ "epoch": 0.24368231046931407,
196
+ "grad_norm": 0.45173099637031555,
197
+ "learning_rate": 8.745082927659047e-05,
198
+ "loss": 0.0384,
199
  "step": 270
200
  },
201
  {
202
+ "epoch": 0.2527075812274368,
203
+ "grad_norm": 0.33228954672813416,
204
+ "learning_rate": 8.633504448242505e-05,
205
+ "loss": 0.0422,
206
  "step": 280
207
  },
208
  {
209
+ "epoch": 0.26173285198555957,
210
+ "grad_norm": 0.364307165145874,
211
+ "learning_rate": 8.517952785058385e-05,
212
+ "loss": 0.0369,
213
  "step": 290
214
  },
215
  {
216
+ "epoch": 0.27075812274368233,
217
+ "grad_norm": 0.34387287497520447,
218
+ "learning_rate": 8.398554292153866e-05,
219
+ "loss": 0.0353,
220
  "step": 300
221
  },
222
  {
223
+ "epoch": 0.27978339350180503,
224
+ "grad_norm": 0.3851270079612732,
225
+ "learning_rate": 8.275439530027948e-05,
226
+ "loss": 0.0417,
227
  "step": 310
228
  },
229
  {
230
+ "epoch": 0.2888086642599278,
231
+ "grad_norm": 0.28114214539527893,
232
+ "learning_rate": 8.148743122865463e-05,
233
+ "loss": 0.0345,
234
  "step": 320
235
  },
236
  {
237
+ "epoch": 0.29783393501805056,
238
+ "grad_norm": 0.33744707703590393,
239
+ "learning_rate": 8.018603611327504e-05,
240
+ "loss": 0.0342,
241
  "step": 330
242
  },
243
  {
244
+ "epoch": 0.30685920577617326,
245
+ "grad_norm": 0.4103146195411682,
246
+ "learning_rate": 7.88516330105925e-05,
247
+ "loss": 0.035,
248
  "step": 340
249
  },
250
  {
251
+ "epoch": 0.315884476534296,
252
+ "grad_norm": 0.328669011592865,
253
+ "learning_rate": 7.748568107080832e-05,
254
+ "loss": 0.0327,
255
  "step": 350
256
  },
257
  {
258
+ "epoch": 0.3249097472924188,
259
+ "grad_norm": 0.4371018409729004,
260
+ "learning_rate": 7.608967394231387e-05,
261
+ "loss": 0.0359,
262
  "step": 360
263
  },
264
  {
265
+ "epoch": 0.33393501805054154,
266
+ "grad_norm": 0.4694584012031555,
267
+ "learning_rate": 7.466513813840825e-05,
268
+ "loss": 0.0339,
269
  "step": 370
270
  },
271
  {
272
+ "epoch": 0.34296028880866425,
273
+ "grad_norm": 0.38598158955574036,
274
+ "learning_rate": 7.32136313680782e-05,
275
+ "loss": 0.0289,
276
  "step": 380
277
  },
278
  {
279
+ "epoch": 0.351985559566787,
280
+ "grad_norm": 0.36169150471687317,
281
+ "learning_rate": 7.173674083266624e-05,
282
+ "loss": 0.032,
283
  "step": 390
284
  },
285
  {
286
+ "epoch": 0.36101083032490977,
287
+ "grad_norm": 0.3175145983695984,
288
+ "learning_rate": 7.023608149028937e-05,
289
+ "loss": 0.0311,
290
  "step": 400
291
  },
292
  {
293
+ "epoch": 0.3700361010830325,
294
+ "grad_norm": 0.4808519184589386,
295
+ "learning_rate": 6.871329428990602e-05,
296
+ "loss": 0.0351,
297
  "step": 410
298
  },
299
  {
300
+ "epoch": 0.37906137184115524,
301
+ "grad_norm": 0.3936750590801239,
302
+ "learning_rate": 6.71700443769625e-05,
303
+ "loss": 0.0297,
304
  "step": 420
305
  },
306
  {
307
+ "epoch": 0.388086642599278,
308
+ "grad_norm": 0.31872764229774475,
309
+ "learning_rate": 6.56080192725808e-05,
310
+ "loss": 0.029,
311
  "step": 430
312
  },
313
  {
314
+ "epoch": 0.3971119133574007,
315
+ "grad_norm": 0.3199341297149658,
316
+ "learning_rate": 6.402892702827916e-05,
317
+ "loss": 0.0305,
318
  "step": 440
319
  },
320
  {
321
+ "epoch": 0.40613718411552346,
322
+ "grad_norm": 0.29824522137641907,
323
+ "learning_rate": 6.243449435824276e-05,
324
+ "loss": 0.0313,
325
  "step": 450
326
  },
327
  {
328
+ "epoch": 0.4151624548736462,
329
+ "grad_norm": 0.3850104808807373,
330
+ "learning_rate": 6.0826464751186994e-05,
331
+ "loss": 0.0307,
332
  "step": 460
333
  },
334
  {
335
+ "epoch": 0.42418772563176893,
336
+ "grad_norm": 0.28914761543273926,
337
+ "learning_rate": 5.9206596563878357e-05,
338
+ "loss": 0.0335,
339
  "step": 470
340
  },
341
  {
342
+ "epoch": 0.4332129963898917,
343
+ "grad_norm": 0.29466763138771057,
344
+ "learning_rate": 5.757666109839702e-05,
345
+ "loss": 0.0282,
346
  "step": 480
347
  },
348
  {
349
+ "epoch": 0.44223826714801445,
350
+ "grad_norm": 0.3107747435569763,
351
+ "learning_rate": 5.5938440665244006e-05,
352
+ "loss": 0.027,
353
  "step": 490
354
  },
355
  {
356
+ "epoch": 0.45126353790613716,
357
+ "grad_norm": 0.38809922337532043,
358
+ "learning_rate": 5.4293726634410855e-05,
359
+ "loss": 0.0281,
360
  "step": 500
361
  },
362
  {
363
+ "epoch": 0.4602888086642599,
364
+ "grad_norm": 0.27953779697418213,
365
+ "learning_rate": 5.264431747654284e-05,
366
+ "loss": 0.0251,
367
  "step": 510
368
  },
369
  {
370
+ "epoch": 0.4693140794223827,
371
+ "grad_norm": 0.24429909884929657,
372
+ "learning_rate": 5.0992016796337686e-05,
373
+ "loss": 0.0258,
374
  "step": 520
375
  },
376
  {
377
+ "epoch": 0.47833935018050544,
378
+ "grad_norm": 0.21034276485443115,
379
+ "learning_rate": 4.93386313603304e-05,
380
+ "loss": 0.0271,
381
  "step": 530
382
  },
383
  {
384
+ "epoch": 0.48736462093862815,
385
+ "grad_norm": 0.29745611548423767,
386
+ "learning_rate": 4.7685969121220456e-05,
387
+ "loss": 0.0268,
388
  "step": 540
389
  },
390
  {
391
+ "epoch": 0.4963898916967509,
392
+ "grad_norm": 0.2871965765953064,
393
+ "learning_rate": 4.60358372409022e-05,
394
+ "loss": 0.0245,
395
  "step": 550
396
  },
397
  {
398
+ "epoch": 0.5054151624548736,
399
+ "grad_norm": 0.30364564061164856,
400
+ "learning_rate": 4.439004011435979e-05,
401
+ "loss": 0.0237,
402
  "step": 560
403
  },
404
  {
405
+ "epoch": 0.5144404332129964,
406
+ "grad_norm": 0.28791144490242004,
407
+ "learning_rate": 4.275037739658771e-05,
408
+ "loss": 0.0264,
409
  "step": 570
410
  },
411
  {
412
+ "epoch": 0.5234657039711191,
413
+ "grad_norm": 0.3472572863101959,
414
+ "learning_rate": 4.111864203469457e-05,
415
+ "loss": 0.0268,
416
  "step": 580
417
  },
418
  {
419
+ "epoch": 0.5324909747292419,
420
+ "grad_norm": 0.3806765377521515,
421
+ "learning_rate": 3.949661830734172e-05,
422
+ "loss": 0.0224,
423
  "step": 590
424
  },
425
  {
426
+ "epoch": 0.5415162454873647,
427
+ "grad_norm": 0.27796250581741333,
428
+ "learning_rate": 3.788607987366069e-05,
429
+ "loss": 0.0253,
430
  "step": 600
431
  },
432
  {
433
+ "epoch": 0.5505415162454874,
434
+ "grad_norm": 0.24273499846458435,
435
+ "learning_rate": 3.628878783378302e-05,
436
+ "loss": 0.0208,
437
  "step": 610
438
  },
439
  {
440
+ "epoch": 0.5595667870036101,
441
+ "grad_norm": 0.3810800611972809,
442
+ "learning_rate": 3.470648880310313e-05,
443
+ "loss": 0.0219,
444
  "step": 620
445
  },
446
  {
447
+ "epoch": 0.5685920577617328,
448
+ "grad_norm": 0.31840479373931885,
449
+ "learning_rate": 3.3140913002379995e-05,
450
+ "loss": 0.0222,
451
  "step": 630
452
  },
453
  {
454
+ "epoch": 0.5776173285198556,
455
+ "grad_norm": 0.2828656733036041,
456
+ "learning_rate": 3.1593772365766105e-05,
457
+ "loss": 0.0212,
458
  "step": 640
459
  },
460
  {
461
+ "epoch": 0.5866425992779783,
462
+ "grad_norm": 0.22635580599308014,
463
+ "learning_rate": 3.006675866883275e-05,
464
+ "loss": 0.0213,
465
  "step": 650
466
  },
467
  {
468
+ "epoch": 0.5956678700361011,
469
+ "grad_norm": 0.20679627358913422,
470
+ "learning_rate": 2.8561541678638142e-05,
471
+ "loss": 0.0239,
472
  "step": 660
473
  },
474
  {
475
+ "epoch": 0.6046931407942239,
476
+ "grad_norm": 0.20941723883152008,
477
+ "learning_rate": 2.707976732786166e-05,
478
+ "loss": 0.0198,
479
  "step": 670
480
  },
481
  {
482
+ "epoch": 0.6137184115523465,
483
+ "grad_norm": 0.23454327881336212,
484
+ "learning_rate": 2.562305591500069e-05,
485
+ "loss": 0.0199,
486
  "step": 680
487
  },
488
  {
489
+ "epoch": 0.6227436823104693,
490
+ "grad_norm": 0.2705540060997009,
491
+ "learning_rate": 2.419300033259798e-05,
492
+ "loss": 0.0207,
493
  "step": 690
494
  },
495
  {
496
+ "epoch": 0.631768953068592,
497
+ "grad_norm": 0.2031233012676239,
498
+ "learning_rate": 2.279116432543705e-05,
499
+ "loss": 0.0195,
500
  "step": 700
501
  },
502
  {
503
+ "epoch": 0.6407942238267148,
504
+ "grad_norm": 0.3306087553501129,
505
+ "learning_rate": 2.1419080780610123e-05,
506
+ "loss": 0.0237,
507
  "step": 710
508
  },
509
  {
510
+ "epoch": 0.6498194945848376,
511
+ "grad_norm": 0.2521066963672638,
512
+ "learning_rate": 2.0078250051328784e-05,
513
+ "loss": 0.0201,
514
  "step": 720
515
  },
516
  {
517
+ "epoch": 0.6588447653429603,
518
+ "grad_norm": 0.17102816700935364,
519
+ "learning_rate": 1.877013831630961e-05,
520
+ "loss": 0.0157,
521
  "step": 730
522
  },
523
  {
524
+ "epoch": 0.6678700361010831,
525
+ "grad_norm": 0.2178596407175064,
526
+ "learning_rate": 1.749617597652934e-05,
527
+ "loss": 0.0212,
528
  "step": 740
529
  },
530
  {
531
+ "epoch": 0.6768953068592057,
532
+ "grad_norm": 0.23762241005897522,
533
+ "learning_rate": 1.62577560911024e-05,
534
+ "loss": 0.0198,
535
  "step": 750
536
  },
537
  {
538
+ "epoch": 0.6859205776173285,
539
+ "grad_norm": 0.24799777567386627,
540
+ "learning_rate": 1.5056232853991209e-05,
541
+ "loss": 0.0233,
542
  "step": 760
543
  },
544
  {
545
+ "epoch": 0.6949458483754513,
546
+ "grad_norm": 0.1982533186674118,
547
+ "learning_rate": 1.389292011321498e-05,
548
+ "loss": 0.0177,
549
  "step": 770
550
  },
551
  {
552
+ "epoch": 0.703971119133574,
553
+ "grad_norm": 0.27399930357933044,
554
+ "learning_rate": 1.2769089934176126e-05,
555
+ "loss": 0.0189,
556
  "step": 780
557
  },
558
  {
559
+ "epoch": 0.7129963898916968,
560
+ "grad_norm": 0.17227351665496826,
561
+ "learning_rate": 1.1685971208675539e-05,
562
+ "loss": 0.0184,
563
  "step": 790
564
  },
565
  {
566
+ "epoch": 0.7220216606498195,
567
+ "grad_norm": 0.1276431679725647,
568
+ "learning_rate": 1.0644748311137376e-05,
569
+ "loss": 0.0183,
570
  "step": 800
571
  },
572
  {
573
+ "epoch": 0.7310469314079422,
574
+ "grad_norm": 0.20766472816467285,
575
+ "learning_rate": 9.646559803512994e-06,
576
+ "loss": 0.019,
577
  "step": 810
578
  },
579
  {
580
+ "epoch": 0.740072202166065,
581
+ "grad_norm": 0.17726139724254608,
582
+ "learning_rate": 8.692497190280224e-06,
583
  "loss": 0.0178,
584
  "step": 820
585
  },
586
  {
587
+ "epoch": 0.7490974729241877,
588
+ "grad_norm": 0.2659638226032257,
589
+ "learning_rate": 7.783603724899257e-06,
590
+ "loss": 0.0186,
591
  "step": 830
592
  },
593
  {
594
+ "epoch": 0.7581227436823105,
595
+ "grad_norm": 0.1689656376838684,
596
+ "learning_rate": 6.92087326903022e-06,
597
+ "loss": 0.0179,
598
  "step": 840
599
  },
600
  {
601
+ "epoch": 0.7671480144404332,
602
+ "grad_norm": 0.2008303552865982,
603
+ "learning_rate": 6.1052492057601275e-06,
604
+ "loss": 0.0166,
605
  "step": 850
606
  },
607
  {
608
+ "epoch": 0.776173285198556,
609
+ "grad_norm": 0.2688104212284088,
610
+ "learning_rate": 5.337623408027293e-06,
611
+ "loss": 0.0201,
612
  "step": 860
613
  },
614
  {
615
+ "epoch": 0.7851985559566786,
616
+ "grad_norm": 0.16218852996826172,
617
+ "learning_rate": 4.618835263371396e-06,
618
+ "loss": 0.0182,
619
  "step": 870
620
  },
621
  {
622
+ "epoch": 0.7942238267148014,
623
+ "grad_norm": 0.19471345841884613,
624
+ "learning_rate": 3.949670756075447e-06,
625
+ "loss": 0.0178,
626
  "step": 880
627
  },
628
  {
629
+ "epoch": 0.8032490974729242,
630
+ "grad_norm": 0.19546721875667572,
631
+ "learning_rate": 3.3308616077036115e-06,
632
+ "loss": 0.0178,
633
  "step": 890
634
  },
635
  {
636
+ "epoch": 0.8122743682310469,
637
+ "grad_norm": 0.17014919221401215,
638
+ "learning_rate": 2.7630844769743757e-06,
639
+ "loss": 0.0152,
640
  "step": 900
641
  },
642
  {
643
+ "epoch": 0.8212996389891697,
644
+ "grad_norm": 0.17462334036827087,
645
+ "learning_rate": 2.2469602198441573e-06,
646
+ "loss": 0.0159,
647
  "step": 910
648
  },
649
  {
650
+ "epoch": 0.8303249097472925,
651
+ "grad_norm": 0.1929808259010315,
652
+ "learning_rate": 1.7830532106104747e-06,
653
+ "loss": 0.0152,
654
  "step": 920
655
  },
656
  {
657
+ "epoch": 0.8393501805054152,
658
+ "grad_norm": 0.16477197408676147,
659
+ "learning_rate": 1.3718707247769135e-06,
660
+ "loss": 0.0159,
661
  "step": 930
662
  },
663
  {
664
+ "epoch": 0.8483754512635379,
665
+ "grad_norm": 0.18839409947395325,
666
+ "learning_rate": 1.0138623843548078e-06,
667
+ "loss": 0.0143,
668
  "step": 940
669
  },
670
  {
671
+ "epoch": 0.8574007220216606,
672
+ "grad_norm": 0.19429394602775574,
673
+ "learning_rate": 7.094196662081831e-07,
674
+ "loss": 0.0173,
675
  "step": 950
676
  },
677
  {
678
+ "epoch": 0.8664259927797834,
679
+ "grad_norm": 0.15573933720588684,
680
+ "learning_rate": 4.5887547397955864e-07,
681
+ "loss": 0.0152,
682
  "step": 960
683
  },
684
  {
685
+ "epoch": 0.8754512635379061,
686
+ "grad_norm": 0.16420625150203705,
687
+ "learning_rate": 2.625037740646763e-07,
688
+ "loss": 0.0181,
689
  "step": 970
690
  },
691
  {
692
+ "epoch": 0.8844765342960289,
693
+ "grad_norm": 0.14045004546642303,
694
+ "learning_rate": 1.2051929603428825e-07,
695
+ "loss": 0.0163,
696
  "step": 980
697
  },
698
  {
699
+ "epoch": 0.8935018050541517,
700
+ "grad_norm": 0.20185790956020355,
701
+ "learning_rate": 3.3077297830541584e-08,
702
+ "loss": 0.0189,
703
  "step": 990
704
  },
705
  {
706
+ "epoch": 0.9025270758122743,
707
+ "grad_norm": 0.2897006571292877,
708
+ "learning_rate": 2.7339599464326627e-10,
709
+ "loss": 0.0205,
710
  "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.9025270758122743,
714
+ "step": 1000,
715
+ "total_flos": 0.0,
716
+ "train_loss": 0.04500286555290222,
717
+ "train_runtime": 685.5246,
718
+ "train_samples_per_second": 70.019,
719
+ "train_steps_per_second": 1.459
720
  }
721
  ],
722
  "logging_steps": 10,
723
+ "max_steps": 1000,
724
  "num_input_tokens_seen": 0,
725
+ "num_train_epochs": 1,
726
  "save_steps": 1000,
727
  "stateful_callbacks": {
728
  "TrainerControl": {
 
731
  "should_evaluate": false,
732
  "should_log": false,
733
  "should_save": true,
734
+ "should_training_stop": true
735
  },
736
  "attributes": {}
737
  }
738
  },
739
  "total_flos": 0.0,
740
+ "train_batch_size": 48,
741
  "trial_name": null,
742
  "trial_params": null
743
  }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:783b80d81286a5b55670d0b48c8fb948472e733210763a68cc74c07e6c359f64
3
+ size 5304