Mihir Tripathy commited on
Commit
990128b
β€’
1 Parent(s): 7589f35

added vitmae models

Browse files
{checkpoint β†’ old_13M}/.gitattributes RENAMED
File without changes
{checkpoint β†’ old_13M}/config.json RENAMED
File without changes
{checkpoint β†’ old_13M}/optimizer.pt RENAMED
File without changes
{checkpoint β†’ old_13M}/pytorch_model.bin RENAMED
File without changes
{checkpoint β†’ old_13M}/rng_state.pth RENAMED
File without changes
{checkpoint β†’ old_13M}/scheduler.pt RENAMED
File without changes
{checkpoint β†’ old_13M}/trainer_state.json RENAMED
File without changes
{checkpoint β†’ old_13M}/training_args.bin RENAMED
File without changes
vitmae_111M/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/gpfs/gibbs/pi/dijk/BrainLM_runs/2023-11-16-12_05_00/checkpoint-900",
3
+ "architectures": [
4
+ "ViTMAEForPreTraining"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "decoder_hidden_size": 512,
8
+ "decoder_intermediate_size": 2048,
9
+ "decoder_num_attention_heads": 16,
10
+ "decoder_num_hidden_layers": 8,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.0,
13
+ "hidden_size": 768,
14
+ "image_size": [
15
+ 432,
16
+ 432
17
+ ],
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "layer_norm_eps": 1e-12,
21
+ "mask_ratio": 0.75,
22
+ "model_type": "vit_mae",
23
+ "norm_pix_loss": false,
24
+ "num_attention_heads": 12,
25
+ "num_channels": 3,
26
+ "num_hidden_layers": 12,
27
+ "patch_size": 16,
28
+ "qkv_bias": true,
29
+ "torch_dtype": "float32",
30
+ "transformers_version": "4.28.0"
31
+ }
vitmae_111M/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01eb309324822c7f97d7fbc58656abcab0d8d7ef5208d901f6783a95a7ce6be1
3
+ size 893440890
vitmae_111M/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25f5b52178f7409dee4fed2fa3dd18c6692d0764d96996ff0ee4be740520fcd6
3
+ size 450475202
vitmae_111M/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b24ac6de8ab41eb80a5d47d400cd9eb2c25a220d43926e7d672a46d2eb467a3b
3
+ size 14244
vitmae_111M/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d78b6f0b4efdd54634970fdb145978ff7d80f35a6b8a9bd537ef08492f5ba90e
3
+ size 988
vitmae_111M/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b89122e73508ffdb5ac677595c6c25a7904c60895d91c44c1d3585c7d5b865f
3
+ size 1064
vitmae_111M/trainer_state.json ADDED
@@ -0,0 +1,1614 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.01686818338930607,
3
+ "best_model_checkpoint": "/gpfs/gibbs/pi/dijk/BrainLM_runs/2023-11-17-17_20_00/checkpoint-3900",
4
+ "epoch": 19.711664482306684,
5
+ "global_step": 4700,
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,
12
+ "learning_rate": 2.100840336134454e-05,
13
+ "loss": 0.0209,
14
+ "step": 50
15
+ },
16
+ {
17
+ "epoch": 0,
18
+ "eval_loss": 0.020767875015735626,
19
+ "eval_mae": 0.11019120365381241,
20
+ "eval_mse": 0.020768500864505768,
21
+ "eval_r2": 0.06256083852108052,
22
+ "eval_runtime": 126.6828,
23
+ "eval_samples_per_second": 3.157,
24
+ "eval_steps_per_second": 0.197,
25
+ "step": 50
26
+ },
27
+ {
28
+ "epoch": 0,
29
+ "learning_rate": 4.201680672268908e-05,
30
+ "loss": 0.02,
31
+ "step": 100
32
+ },
33
+ {
34
+ "epoch": 0,
35
+ "eval_loss": 0.020464899018406868,
36
+ "eval_mae": 0.10934258252382278,
37
+ "eval_mse": 0.02046414092183113,
38
+ "eval_r2": 0.0822418520285656,
39
+ "eval_runtime": 190.4097,
40
+ "eval_samples_per_second": 2.101,
41
+ "eval_steps_per_second": 0.131,
42
+ "step": 100
43
+ },
44
+ {
45
+ "epoch": 0,
46
+ "learning_rate": 6.302521008403361e-05,
47
+ "loss": 0.0196,
48
+ "step": 150
49
+ },
50
+ {
51
+ "epoch": 0,
52
+ "eval_loss": 0.02013571746647358,
53
+ "eval_mae": 0.10845349729061127,
54
+ "eval_mse": 0.020132917910814285,
55
+ "eval_r2": 0.09748306648229965,
56
+ "eval_runtime": 60.1594,
57
+ "eval_samples_per_second": 6.649,
58
+ "eval_steps_per_second": 0.416,
59
+ "step": 150
60
+ },
61
+ {
62
+ "epoch": 0,
63
+ "learning_rate": 8.403361344537815e-05,
64
+ "loss": 0.0193,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0,
69
+ "eval_loss": 0.019823560491204262,
70
+ "eval_mae": 0.10756562650203705,
71
+ "eval_mse": 0.019823383539915085,
72
+ "eval_r2": 0.11099549304277823,
73
+ "eval_runtime": 162.8488,
74
+ "eval_samples_per_second": 2.456,
75
+ "eval_steps_per_second": 0.154,
76
+ "step": 200
77
+ },
78
+ {
79
+ "epoch": 1,
80
+ "learning_rate": 9.999826244478293e-05,
81
+ "loss": 0.0191,
82
+ "step": 250
83
+ },
84
+ {
85
+ "epoch": 1,
86
+ "eval_loss": 0.019535023719072342,
87
+ "eval_mae": 0.10683062672615051,
88
+ "eval_mse": 0.01953577809035778,
89
+ "eval_r2": 0.12390402334988537,
90
+ "eval_runtime": 61.7292,
91
+ "eval_samples_per_second": 6.48,
92
+ "eval_steps_per_second": 0.405,
93
+ "step": 250
94
+ },
95
+ {
96
+ "epoch": 1,
97
+ "learning_rate": 9.995362383107962e-05,
98
+ "loss": 0.0188,
99
+ "step": 300
100
+ },
101
+ {
102
+ "epoch": 1,
103
+ "eval_loss": 0.019223904237151146,
104
+ "eval_mae": 0.10594414174556732,
105
+ "eval_mse": 0.019219543784856796,
106
+ "eval_r2": 0.1355137630125941,
107
+ "eval_runtime": 62.8456,
108
+ "eval_samples_per_second": 6.365,
109
+ "eval_steps_per_second": 0.398,
110
+ "step": 300
111
+ },
112
+ {
113
+ "epoch": 1,
114
+ "learning_rate": 9.98487151097676e-05,
115
+ "loss": 0.0185,
116
+ "step": 350
117
+ },
118
+ {
119
+ "epoch": 1,
120
+ "eval_loss": 0.018948782235383987,
121
+ "eval_mae": 0.10521063953638077,
122
+ "eval_mse": 0.018949836492538452,
123
+ "eval_r2": 0.1472992781550666,
124
+ "eval_runtime": 62.8686,
125
+ "eval_samples_per_second": 6.362,
126
+ "eval_steps_per_second": 0.398,
127
+ "step": 350
128
+ },
129
+ {
130
+ "epoch": 1,
131
+ "learning_rate": 9.968366285544619e-05,
132
+ "loss": 0.0182,
133
+ "step": 400
134
+ },
135
+ {
136
+ "epoch": 1,
137
+ "eval_loss": 0.01880819909274578,
138
+ "eval_mae": 0.10481799393892288,
139
+ "eval_mse": 0.01880812644958496,
140
+ "eval_r2": 0.15779620957023455,
141
+ "eval_runtime": 57.2628,
142
+ "eval_samples_per_second": 6.985,
143
+ "eval_steps_per_second": 0.437,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 1,
148
+ "learning_rate": 9.945866620716411e-05,
149
+ "loss": 0.0181,
150
+ "step": 450
151
+ },
152
+ {
153
+ "epoch": 1,
154
+ "eval_loss": 0.018690017983317375,
155
+ "eval_mae": 0.10444584488868713,
156
+ "eval_mse": 0.01869109459221363,
157
+ "eval_r2": 0.16425393659247034,
158
+ "eval_runtime": 58.6008,
159
+ "eval_samples_per_second": 6.826,
160
+ "eval_steps_per_second": 0.427,
161
+ "step": 450
162
+ },
163
+ {
164
+ "epoch": 2,
165
+ "learning_rate": 9.917399662815391e-05,
166
+ "loss": 0.0179,
167
+ "step": 500
168
+ },
169
+ {
170
+ "epoch": 2,
171
+ "eval_loss": 0.018527410924434662,
172
+ "eval_mae": 0.10401275008916855,
173
+ "eval_mse": 0.018530065193772316,
174
+ "eval_r2": 0.16775397819003413,
175
+ "eval_runtime": 56.6595,
176
+ "eval_samples_per_second": 7.06,
177
+ "eval_steps_per_second": 0.441,
178
+ "step": 500
179
+ },
180
+ {
181
+ "epoch": 2,
182
+ "learning_rate": 9.882999757830588e-05,
183
+ "loss": 0.0179,
184
+ "step": 550
185
+ },
186
+ {
187
+ "epoch": 2,
188
+ "eval_loss": 0.018433190882205963,
189
+ "eval_mae": 0.10371904820203781,
190
+ "eval_mse": 0.018430989235639572,
191
+ "eval_r2": 0.17258796261060627,
192
+ "eval_runtime": 61.318,
193
+ "eval_samples_per_second": 6.523,
194
+ "eval_steps_per_second": 0.408,
195
+ "step": 550
196
+ },
197
+ {
198
+ "epoch": 2,
199
+ "learning_rate": 9.842708409977635e-05,
200
+ "loss": 0.0178,
201
+ "step": 600
202
+ },
203
+ {
204
+ "epoch": 2,
205
+ "eval_loss": 0.018353702500462532,
206
+ "eval_mae": 0.1034744456410408,
207
+ "eval_mse": 0.018353240564465523,
208
+ "eval_r2": 0.17714196201923582,
209
+ "eval_runtime": 64.7666,
210
+ "eval_samples_per_second": 6.176,
211
+ "eval_steps_per_second": 0.386,
212
+ "step": 600
213
+ },
214
+ {
215
+ "epoch": 2,
216
+ "learning_rate": 9.796574231623055e-05,
217
+ "loss": 0.0177,
218
+ "step": 650
219
+ },
220
+ {
221
+ "epoch": 2,
222
+ "eval_loss": 0.018298938870429993,
223
+ "eval_mae": 0.10329218953847885,
224
+ "eval_mse": 0.01829737052321434,
225
+ "eval_r2": 0.18098740062966168,
226
+ "eval_runtime": 58.7251,
227
+ "eval_samples_per_second": 6.811,
228
+ "eval_steps_per_second": 0.426,
229
+ "step": 650
230
+ },
231
+ {
232
+ "epoch": 2,
233
+ "learning_rate": 9.744652884632406e-05,
234
+ "loss": 0.0176,
235
+ "step": 700
236
+ },
237
+ {
238
+ "epoch": 2,
239
+ "eval_loss": 0.01817883551120758,
240
+ "eval_mae": 0.10299910604953766,
241
+ "eval_mse": 0.018180398270487785,
242
+ "eval_r2": 0.1854382026037652,
243
+ "eval_runtime": 64.0748,
244
+ "eval_samples_per_second": 6.243,
245
+ "eval_steps_per_second": 0.39,
246
+ "step": 700
247
+ },
248
+ {
249
+ "epoch": 3,
250
+ "learning_rate": 9.687007013213063e-05,
251
+ "loss": 0.0175,
252
+ "step": 750
253
+ },
254
+ {
255
+ "epoch": 3,
256
+ "eval_loss": 0.018099399283528328,
257
+ "eval_mae": 0.10274745523929596,
258
+ "eval_mse": 0.018099674955010414,
259
+ "eval_r2": 0.18714389318808033,
260
+ "eval_runtime": 57.0327,
261
+ "eval_samples_per_second": 7.014,
262
+ "eval_steps_per_second": 0.438,
263
+ "step": 750
264
+ },
265
+ {
266
+ "epoch": 3,
267
+ "learning_rate": 9.623706168332645e-05,
268
+ "loss": 0.0174,
269
+ "step": 800
270
+ },
271
+ {
272
+ "epoch": 3,
273
+ "eval_loss": 0.01805899292230606,
274
+ "eval_mae": 0.1026226207613945,
275
+ "eval_mse": 0.018060266971588135,
276
+ "eval_r2": 0.19156849973920198,
277
+ "eval_runtime": 58.7166,
278
+ "eval_samples_per_second": 6.812,
279
+ "eval_steps_per_second": 0.426,
280
+ "step": 800
281
+ },
282
+ {
283
+ "epoch": 3,
284
+ "learning_rate": 9.554826723804303e-05,
285
+ "loss": 0.0174,
286
+ "step": 850
287
+ },
288
+ {
289
+ "epoch": 3,
290
+ "eval_loss": 0.01803247630596161,
291
+ "eval_mae": 0.10254286229610443,
292
+ "eval_mse": 0.018035681918263435,
293
+ "eval_r2": 0.18969497919656642,
294
+ "eval_runtime": 60.3899,
295
+ "eval_samples_per_second": 6.624,
296
+ "eval_steps_per_second": 0.414,
297
+ "step": 850
298
+ },
299
+ {
300
+ "epoch": 3,
301
+ "learning_rate": 9.480451784140091e-05,
302
+ "loss": 0.0173,
303
+ "step": 900
304
+ },
305
+ {
306
+ "epoch": 3,
307
+ "eval_loss": 0.017897402867674828,
308
+ "eval_mae": 0.1021430492401123,
309
+ "eval_mse": 0.017896411940455437,
310
+ "eval_r2": 0.1964586734458611,
311
+ "eval_runtime": 62.8605,
312
+ "eval_samples_per_second": 6.363,
313
+ "eval_steps_per_second": 0.398,
314
+ "step": 900
315
+ },
316
+ {
317
+ "epoch": 3,
318
+ "learning_rate": 9.400671084283607e-05,
319
+ "loss": 0.0173,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 3,
324
+ "eval_loss": 0.017886007204651833,
325
+ "eval_mae": 0.10208527743816376,
326
+ "eval_mse": 0.01788811758160591,
327
+ "eval_r2": 0.19957007249531145,
328
+ "eval_runtime": 60.1438,
329
+ "eval_samples_per_second": 6.651,
330
+ "eval_steps_per_second": 0.416,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 4,
335
+ "learning_rate": 9.315580881342876e-05,
336
+ "loss": 0.0172,
337
+ "step": 1000
338
+ },
339
+ {
340
+ "epoch": 4,
341
+ "eval_loss": 0.017858153209090233,
342
+ "eval_mae": 0.10202876478433609,
343
+ "eval_mse": 0.017859304323792458,
344
+ "eval_r2": 0.19855005202715592,
345
+ "eval_runtime": 58.8873,
346
+ "eval_samples_per_second": 6.793,
347
+ "eval_steps_per_second": 0.425,
348
+ "step": 1000
349
+ },
350
+ {
351
+ "epoch": 4,
352
+ "learning_rate": 9.225283838454111e-05,
353
+ "loss": 0.0172,
354
+ "step": 1050
355
+ },
356
+ {
357
+ "epoch": 4,
358
+ "eval_loss": 0.017775438725948334,
359
+ "eval_mae": 0.10180441290140152,
360
+ "eval_mse": 0.017775651067495346,
361
+ "eval_r2": 0.20128072921203966,
362
+ "eval_runtime": 65.7432,
363
+ "eval_samples_per_second": 6.084,
364
+ "eval_steps_per_second": 0.38,
365
+ "step": 1050
366
+ },
367
+ {
368
+ "epoch": 4,
369
+ "learning_rate": 9.129888900916456e-05,
370
+ "loss": 0.0171,
371
+ "step": 1100
372
+ },
373
+ {
374
+ "epoch": 4,
375
+ "eval_loss": 0.01773221045732498,
376
+ "eval_mae": 0.10164433717727661,
377
+ "eval_mse": 0.017730284482240677,
378
+ "eval_r2": 0.20249305414753993,
379
+ "eval_runtime": 56.7262,
380
+ "eval_samples_per_second": 7.051,
381
+ "eval_steps_per_second": 0.441,
382
+ "step": 1100
383
+ },
384
+ {
385
+ "epoch": 4,
386
+ "learning_rate": 9.029511164747175e-05,
387
+ "loss": 0.0172,
388
+ "step": 1150
389
+ },
390
+ {
391
+ "epoch": 4,
392
+ "eval_loss": 0.017706887796521187,
393
+ "eval_mae": 0.10156488418579102,
394
+ "eval_mse": 0.017707258462905884,
395
+ "eval_r2": 0.2078045399984002,
396
+ "eval_runtime": 58.2849,
397
+ "eval_samples_per_second": 6.863,
398
+ "eval_steps_per_second": 0.429,
399
+ "step": 1150
400
+ },
401
+ {
402
+ "epoch": 5,
403
+ "learning_rate": 8.924271737815854e-05,
404
+ "loss": 0.0171,
405
+ "step": 1200
406
+ },
407
+ {
408
+ "epoch": 5,
409
+ "eval_loss": 0.017714276909828186,
410
+ "eval_mae": 0.10158883035182953,
411
+ "eval_mse": 0.017715107649564743,
412
+ "eval_r2": 0.2074889762440164,
413
+ "eval_runtime": 57.6803,
414
+ "eval_samples_per_second": 6.935,
415
+ "eval_steps_per_second": 0.433,
416
+ "step": 1200
417
+ },
418
+ {
419
+ "epoch": 5,
420
+ "learning_rate": 8.814297593725199e-05,
421
+ "loss": 0.0171,
422
+ "step": 1250
423
+ },
424
+ {
425
+ "epoch": 5,
426
+ "eval_loss": 0.017648547887802124,
427
+ "eval_mae": 0.10141555964946747,
428
+ "eval_mse": 0.017654363065958023,
429
+ "eval_r2": 0.20722885865364904,
430
+ "eval_runtime": 64.9526,
431
+ "eval_samples_per_second": 6.158,
432
+ "eval_steps_per_second": 0.385,
433
+ "step": 1250
434
+ },
435
+ {
436
+ "epoch": 5,
437
+ "learning_rate": 8.699721418614673e-05,
438
+ "loss": 0.017,
439
+ "step": 1300
440
+ },
441
+ {
442
+ "epoch": 5,
443
+ "eval_loss": 0.017601532861590385,
444
+ "eval_mae": 0.10125792026519775,
445
+ "eval_mse": 0.017598113045096397,
446
+ "eval_r2": 0.20836808697871778,
447
+ "eval_runtime": 57.4819,
448
+ "eval_samples_per_second": 6.959,
449
+ "eval_steps_per_second": 0.435,
450
+ "step": 1300
451
+ },
452
+ {
453
+ "epoch": 5,
454
+ "learning_rate": 8.580681451071866e-05,
455
+ "loss": 0.017,
456
+ "step": 1350
457
+ },
458
+ {
459
+ "epoch": 5,
460
+ "eval_loss": 0.017580362036824226,
461
+ "eval_mae": 0.10121016204357147,
462
+ "eval_mse": 0.017584411427378654,
463
+ "eval_r2": 0.21084466622121223,
464
+ "eval_runtime": 60.2943,
465
+ "eval_samples_per_second": 6.634,
466
+ "eval_steps_per_second": 0.415,
467
+ "step": 1350
468
+ },
469
+ {
470
+ "epoch": 5,
471
+ "learning_rate": 8.457321315344694e-05,
472
+ "loss": 0.017,
473
+ "step": 1400
474
+ },
475
+ {
476
+ "epoch": 5,
477
+ "eval_loss": 0.017519284039735794,
478
+ "eval_mae": 0.10104001313447952,
479
+ "eval_mse": 0.01752064935863018,
480
+ "eval_r2": 0.2120289842294325,
481
+ "eval_runtime": 60.1697,
482
+ "eval_samples_per_second": 6.648,
483
+ "eval_steps_per_second": 0.415,
484
+ "step": 1400
485
+ },
486
+ {
487
+ "epoch": 6,
488
+ "learning_rate": 8.329789848055704e-05,
489
+ "loss": 0.0168,
490
+ "step": 1450
491
+ },
492
+ {
493
+ "epoch": 6,
494
+ "eval_loss": 0.017550144344568253,
495
+ "eval_mae": 0.10109207779169083,
496
+ "eval_mse": 0.017550604417920113,
497
+ "eval_r2": 0.2133444473175402,
498
+ "eval_runtime": 61.3969,
499
+ "eval_samples_per_second": 6.515,
500
+ "eval_steps_per_second": 0.407,
501
+ "step": 1450
502
+ },
503
+ {
504
+ "epoch": 6,
505
+ "learning_rate": 8.198240918627524e-05,
506
+ "loss": 0.0169,
507
+ "step": 1500
508
+ },
509
+ {
510
+ "epoch": 6,
511
+ "eval_loss": 0.01755845546722412,
512
+ "eval_mae": 0.10110194236040115,
513
+ "eval_mse": 0.017556479200720787,
514
+ "eval_r2": 0.21656763083400632,
515
+ "eval_runtime": 59.5515,
516
+ "eval_samples_per_second": 6.717,
517
+ "eval_steps_per_second": 0.42,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 6,
522
+ "learning_rate": 8.062833243636134e-05,
523
+ "loss": 0.0169,
524
+ "step": 1550
525
+ },
526
+ {
527
+ "epoch": 6,
528
+ "eval_loss": 0.017458565533161163,
529
+ "eval_mae": 0.10081926733255386,
530
+ "eval_mse": 0.017458630725741386,
531
+ "eval_r2": 0.21455437331377625,
532
+ "eval_runtime": 60.6637,
533
+ "eval_samples_per_second": 6.594,
534
+ "eval_steps_per_second": 0.412,
535
+ "step": 1550
536
+ },
537
+ {
538
+ "epoch": 6,
539
+ "learning_rate": 7.923730195315962e-05,
540
+ "loss": 0.0169,
541
+ "step": 1600
542
+ },
543
+ {
544
+ "epoch": 6,
545
+ "eval_loss": 0.017479494214057922,
546
+ "eval_mae": 0.10087893158197403,
547
+ "eval_mse": 0.017478736117482185,
548
+ "eval_r2": 0.2180120287564331,
549
+ "eval_runtime": 60.772,
550
+ "eval_samples_per_second": 6.582,
551
+ "eval_steps_per_second": 0.411,
552
+ "step": 1600
553
+ },
554
+ {
555
+ "epoch": 6,
556
+ "learning_rate": 7.781099604447794e-05,
557
+ "loss": 0.0168,
558
+ "step": 1650
559
+ },
560
+ {
561
+ "epoch": 6,
562
+ "eval_loss": 0.01743287593126297,
563
+ "eval_mae": 0.10074793547391891,
564
+ "eval_mse": 0.017432522028684616,
565
+ "eval_r2": 0.2161429733042286,
566
+ "eval_runtime": 64.79,
567
+ "eval_samples_per_second": 6.174,
568
+ "eval_steps_per_second": 0.386,
569
+ "step": 1650
570
+ },
571
+ {
572
+ "epoch": 7,
573
+ "learning_rate": 7.635113557867395e-05,
574
+ "loss": 0.0169,
575
+ "step": 1700
576
+ },
577
+ {
578
+ "epoch": 7,
579
+ "eval_loss": 0.017418913543224335,
580
+ "eval_mae": 0.10069431364536285,
581
+ "eval_mse": 0.01741931587457657,
582
+ "eval_r2": 0.2169098601070394,
583
+ "eval_runtime": 60.5486,
584
+ "eval_samples_per_second": 6.606,
585
+ "eval_steps_per_second": 0.413,
586
+ "step": 1700
587
+ },
588
+ {
589
+ "epoch": 7,
590
+ "learning_rate": 7.485948190839077e-05,
591
+ "loss": 0.0168,
592
+ "step": 1750
593
+ },
594
+ {
595
+ "epoch": 7,
596
+ "eval_loss": 0.01736665517091751,
597
+ "eval_mae": 0.1005297377705574,
598
+ "eval_mse": 0.01736704632639885,
599
+ "eval_r2": 0.21929555782464094,
600
+ "eval_runtime": 58.4317,
601
+ "eval_samples_per_second": 6.846,
602
+ "eval_steps_per_second": 0.428,
603
+ "step": 1750
604
+ },
605
+ {
606
+ "epoch": 7,
607
+ "learning_rate": 7.333783474544758e-05,
608
+ "loss": 0.0168,
609
+ "step": 1800
610
+ },
611
+ {
612
+ "epoch": 7,
613
+ "eval_loss": 0.017343418672680855,
614
+ "eval_mae": 0.10048038512468338,
615
+ "eval_mse": 0.017341628670692444,
616
+ "eval_r2": 0.21864552418706762,
617
+ "eval_runtime": 60.3965,
618
+ "eval_samples_per_second": 6.623,
619
+ "eval_steps_per_second": 0.414,
620
+ "step": 1800
621
+ },
622
+ {
623
+ "epoch": 7,
624
+ "learning_rate": 7.178802998944933e-05,
625
+ "loss": 0.0166,
626
+ "step": 1850
627
+ },
628
+ {
629
+ "epoch": 7,
630
+ "eval_loss": 0.017404844984412193,
631
+ "eval_mae": 0.10062351077795029,
632
+ "eval_mse": 0.017404979094862938,
633
+ "eval_r2": 0.22093177958560895,
634
+ "eval_runtime": 56.7185,
635
+ "eval_samples_per_second": 7.052,
636
+ "eval_steps_per_second": 0.441,
637
+ "step": 1850
638
+ },
639
+ {
640
+ "epoch": 7,
641
+ "learning_rate": 7.021193751273462e-05,
642
+ "loss": 0.0167,
643
+ "step": 1900
644
+ },
645
+ {
646
+ "epoch": 7,
647
+ "eval_loss": 0.017301246523857117,
648
+ "eval_mae": 0.10035215318202972,
649
+ "eval_mse": 0.017301900312304497,
650
+ "eval_r2": 0.22134026003758744,
651
+ "eval_runtime": 57.7642,
652
+ "eval_samples_per_second": 6.925,
653
+ "eval_steps_per_second": 0.433,
654
+ "step": 1900
655
+ },
656
+ {
657
+ "epoch": 8,
658
+ "learning_rate": 6.86114589043352e-05,
659
+ "loss": 0.0167,
660
+ "step": 1950
661
+ },
662
+ {
663
+ "epoch": 8,
664
+ "eval_loss": 0.01728537492454052,
665
+ "eval_mae": 0.10028935223817825,
666
+ "eval_mse": 0.017283864319324493,
667
+ "eval_r2": 0.2218958493073836,
668
+ "eval_runtime": 61.7673,
669
+ "eval_samples_per_second": 6.476,
670
+ "eval_steps_per_second": 0.405,
671
+ "step": 1950
672
+ },
673
+ {
674
+ "epoch": 8,
675
+ "learning_rate": 6.698852517566836e-05,
676
+ "loss": 0.0167,
677
+ "step": 2000
678
+ },
679
+ {
680
+ "epoch": 8,
681
+ "eval_loss": 0.0173500906676054,
682
+ "eval_mae": 0.10045037418603897,
683
+ "eval_mse": 0.017350492998957634,
684
+ "eval_r2": 0.222062406323153,
685
+ "eval_runtime": 63.3674,
686
+ "eval_samples_per_second": 6.312,
687
+ "eval_steps_per_second": 0.395,
688
+ "step": 2000
689
+ },
690
+ {
691
+ "epoch": 8,
692
+ "learning_rate": 6.534509443073072e-05,
693
+ "loss": 0.0166,
694
+ "step": 2050
695
+ },
696
+ {
697
+ "epoch": 8,
698
+ "eval_loss": 0.017278773710131645,
699
+ "eval_mae": 0.10025625675916672,
700
+ "eval_mse": 0.017278244718909264,
701
+ "eval_r2": 0.22539687501717476,
702
+ "eval_runtime": 57.2192,
703
+ "eval_samples_per_second": 6.991,
704
+ "eval_steps_per_second": 0.437,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 8,
709
+ "learning_rate": 6.368314950360415e-05,
710
+ "loss": 0.0166,
711
+ "step": 2100
712
+ },
713
+ {
714
+ "epoch": 8,
715
+ "eval_loss": 0.017245473340153694,
716
+ "eval_mae": 0.10017417371273041,
717
+ "eval_mse": 0.017246615141630173,
718
+ "eval_r2": 0.22572849115543347,
719
+ "eval_runtime": 59.9387,
720
+ "eval_samples_per_second": 6.673,
721
+ "eval_steps_per_second": 0.417,
722
+ "step": 2100
723
+ },
724
+ {
725
+ "epoch": 9,
726
+ "learning_rate": 6.200469556612435e-05,
727
+ "loss": 0.0166,
728
+ "step": 2150
729
+ },
730
+ {
731
+ "epoch": 9,
732
+ "eval_loss": 0.017270274460315704,
733
+ "eval_mae": 0.10024827718734741,
734
+ "eval_mse": 0.017270220443606377,
735
+ "eval_r2": 0.22628471179764975,
736
+ "eval_runtime": 63.0094,
737
+ "eval_samples_per_second": 6.348,
738
+ "eval_steps_per_second": 0.397,
739
+ "step": 2150
740
+ },
741
+ {
742
+ "epoch": 9,
743
+ "learning_rate": 6.031175770859848e-05,
744
+ "loss": 0.0166,
745
+ "step": 2200
746
+ },
747
+ {
748
+ "epoch": 9,
749
+ "eval_loss": 0.017215639352798462,
750
+ "eval_mae": 0.10008691996335983,
751
+ "eval_mse": 0.017215678468346596,
752
+ "eval_r2": 0.22768342129753605,
753
+ "eval_runtime": 65.042,
754
+ "eval_samples_per_second": 6.15,
755
+ "eval_steps_per_second": 0.384,
756
+ "step": 2200
757
+ },
758
+ {
759
+ "epoch": 9,
760
+ "learning_rate": 5.8606378496490735e-05,
761
+ "loss": 0.0166,
762
+ "step": 2250
763
+ },
764
+ {
765
+ "epoch": 9,
766
+ "eval_loss": 0.01721777580678463,
767
+ "eval_mae": 0.10010144114494324,
768
+ "eval_mse": 0.017218952998518944,
769
+ "eval_r2": 0.22907452898565506,
770
+ "eval_runtime": 65.4931,
771
+ "eval_samples_per_second": 6.108,
772
+ "eval_steps_per_second": 0.382,
773
+ "step": 2250
774
+ },
775
+ {
776
+ "epoch": 9,
777
+ "learning_rate": 5.6890615506023705e-05,
778
+ "loss": 0.0166,
779
+ "step": 2300
780
+ },
781
+ {
782
+ "epoch": 9,
783
+ "eval_loss": 0.017191417515277863,
784
+ "eval_mae": 0.10001853853464127,
785
+ "eval_mse": 0.01718985289335251,
786
+ "eval_r2": 0.2292284056000078,
787
+ "eval_runtime": 64.9938,
788
+ "eval_samples_per_second": 6.154,
789
+ "eval_steps_per_second": 0.385,
790
+ "step": 2300
791
+ },
792
+ {
793
+ "epoch": 9,
794
+ "learning_rate": 5.5166538841669025e-05,
795
+ "loss": 0.0165,
796
+ "step": 2350
797
+ },
798
+ {
799
+ "epoch": 9,
800
+ "eval_loss": 0.017166707664728165,
801
+ "eval_mae": 0.09992814064025879,
802
+ "eval_mse": 0.017164934426546097,
803
+ "eval_r2": 0.23049413465235846,
804
+ "eval_runtime": 63.4634,
805
+ "eval_samples_per_second": 6.303,
806
+ "eval_steps_per_second": 0.394,
807
+ "step": 2350
808
+ },
809
+ {
810
+ "epoch": 10,
811
+ "learning_rate": 5.343622863852232e-05,
812
+ "loss": 0.0166,
813
+ "step": 2400
814
+ },
815
+ {
816
+ "epoch": 10,
817
+ "eval_loss": 0.017118161544203758,
818
+ "eval_mae": 0.0998067557811737,
819
+ "eval_mse": 0.017116105183959007,
820
+ "eval_r2": 0.23199928273221992,
821
+ "eval_runtime": 58.0985,
822
+ "eval_samples_per_second": 6.885,
823
+ "eval_steps_per_second": 0.43,
824
+ "step": 2400
825
+ },
826
+ {
827
+ "epoch": 10,
828
+ "learning_rate": 5.170177255257618e-05,
829
+ "loss": 0.0165,
830
+ "step": 2450
831
+ },
832
+ {
833
+ "epoch": 10,
834
+ "eval_loss": 0.017119385302066803,
835
+ "eval_mae": 0.09977566450834274,
836
+ "eval_mse": 0.017119543626904488,
837
+ "eval_r2": 0.22948023375281368,
838
+ "eval_runtime": 58.2148,
839
+ "eval_samples_per_second": 6.871,
840
+ "eval_steps_per_second": 0.429,
841
+ "step": 2450
842
+ },
843
+ {
844
+ "epoch": 10,
845
+ "learning_rate": 4.996526324191872e-05,
846
+ "loss": 0.0164,
847
+ "step": 2500
848
+ },
849
+ {
850
+ "epoch": 10,
851
+ "eval_loss": 0.017105862498283386,
852
+ "eval_mae": 0.09975843131542206,
853
+ "eval_mse": 0.017107795923948288,
854
+ "eval_r2": 0.23160311839973746,
855
+ "eval_runtime": 58.2306,
856
+ "eval_samples_per_second": 6.869,
857
+ "eval_steps_per_second": 0.429,
858
+ "step": 2500
859
+ },
860
+ {
861
+ "epoch": 10,
862
+ "learning_rate": 4.822879584189731e-05,
863
+ "loss": 0.0165,
864
+ "step": 2550
865
+ },
866
+ {
867
+ "epoch": 10,
868
+ "eval_loss": 0.017133817076683044,
869
+ "eval_mae": 0.09982656687498093,
870
+ "eval_mse": 0.01713182032108307,
871
+ "eval_r2": 0.23212640913876725,
872
+ "eval_runtime": 60.6469,
873
+ "eval_samples_per_second": 6.596,
874
+ "eval_steps_per_second": 0.412,
875
+ "step": 2550
876
+ },
877
+ {
878
+ "epoch": 10,
879
+ "learning_rate": 4.6494465437293225e-05,
880
+ "loss": 0.0165,
881
+ "step": 2600
882
+ },
883
+ {
884
+ "epoch": 10,
885
+ "eval_loss": 0.01714298315346241,
886
+ "eval_mae": 0.0998401865363121,
887
+ "eval_mse": 0.017137613147497177,
888
+ "eval_r2": 0.2327516850849627,
889
+ "eval_runtime": 58.7954,
890
+ "eval_samples_per_second": 6.803,
891
+ "eval_steps_per_second": 0.425,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 11,
896
+ "learning_rate": 4.476436453455742e-05,
897
+ "loss": 0.0164,
898
+ "step": 2650
899
+ },
900
+ {
901
+ "epoch": 11,
902
+ "eval_loss": 0.017095215618610382,
903
+ "eval_mae": 0.09970412403345108,
904
+ "eval_mse": 0.01709286868572235,
905
+ "eval_r2": 0.23346266146762296,
906
+ "eval_runtime": 57.7966,
907
+ "eval_samples_per_second": 6.921,
908
+ "eval_steps_per_second": 0.433,
909
+ "step": 2650
910
+ },
911
+ {
912
+ "epoch": 11,
913
+ "learning_rate": 4.3040580537157024e-05,
914
+ "loss": 0.0165,
915
+ "step": 2700
916
+ },
917
+ {
918
+ "epoch": 11,
919
+ "eval_loss": 0.017114490270614624,
920
+ "eval_mae": 0.09975600242614746,
921
+ "eval_mse": 0.01711602509021759,
922
+ "eval_r2": 0.23458319148513607,
923
+ "eval_runtime": 60.7105,
924
+ "eval_samples_per_second": 6.589,
925
+ "eval_steps_per_second": 0.412,
926
+ "step": 2700
927
+ },
928
+ {
929
+ "epoch": 11,
930
+ "learning_rate": 4.1325193227078816e-05,
931
+ "loss": 0.0164,
932
+ "step": 2750
933
+ },
934
+ {
935
+ "epoch": 11,
936
+ "eval_loss": 0.01703326217830181,
937
+ "eval_mae": 0.0995149165391922,
938
+ "eval_mse": 0.01703346148133278,
939
+ "eval_r2": 0.23379130521799119,
940
+ "eval_runtime": 56.541,
941
+ "eval_samples_per_second": 7.075,
942
+ "eval_steps_per_second": 0.442,
943
+ "step": 2750
944
+ },
945
+ {
946
+ "epoch": 11,
947
+ "learning_rate": 3.962027225552807e-05,
948
+ "loss": 0.0165,
949
+ "step": 2800
950
+ },
951
+ {
952
+ "epoch": 11,
953
+ "eval_loss": 0.017094749957323074,
954
+ "eval_mae": 0.09970033168792725,
955
+ "eval_mse": 0.017094548791646957,
956
+ "eval_r2": 0.2349240350955354,
957
+ "eval_runtime": 65.2081,
958
+ "eval_samples_per_second": 6.134,
959
+ "eval_steps_per_second": 0.383,
960
+ "step": 2800
961
+ },
962
+ {
963
+ "epoch": 11,
964
+ "learning_rate": 3.79278746458504e-05,
965
+ "loss": 0.0164,
966
+ "step": 2850
967
+ },
968
+ {
969
+ "epoch": 11,
970
+ "eval_loss": 0.017008375376462936,
971
+ "eval_mae": 0.09947662055492401,
972
+ "eval_mse": 0.01700720377266407,
973
+ "eval_r2": 0.23690627111999107,
974
+ "eval_runtime": 57.0201,
975
+ "eval_samples_per_second": 7.015,
976
+ "eval_steps_per_second": 0.438,
977
+ "step": 2850
978
+ },
979
+ {
980
+ "epoch": 12,
981
+ "learning_rate": 3.6250042311689505e-05,
982
+ "loss": 0.0164,
983
+ "step": 2900
984
+ },
985
+ {
986
+ "epoch": 12,
987
+ "eval_loss": 0.01702900230884552,
988
+ "eval_mae": 0.09952793270349503,
989
+ "eval_mse": 0.01703030802309513,
990
+ "eval_r2": 0.23451840328295703,
991
+ "eval_runtime": 59.8559,
992
+ "eval_samples_per_second": 6.683,
993
+ "eval_steps_per_second": 0.418,
994
+ "step": 2900
995
+ },
996
+ {
997
+ "epoch": 12,
998
+ "learning_rate": 3.458879959337494e-05,
999
+ "loss": 0.0163,
1000
+ "step": 2950
1001
+ },
1002
+ {
1003
+ "epoch": 12,
1004
+ "eval_loss": 0.017079642042517662,
1005
+ "eval_mae": 0.09966063499450684,
1006
+ "eval_mse": 0.017079809680581093,
1007
+ "eval_r2": 0.23587889518099114,
1008
+ "eval_runtime": 63.9836,
1009
+ "eval_samples_per_second": 6.252,
1010
+ "eval_steps_per_second": 0.391,
1011
+ "step": 2950
1012
+ },
1013
+ {
1014
+ "epoch": 12,
1015
+ "learning_rate": 3.294615081551259e-05,
1016
+ "loss": 0.0164,
1017
+ "step": 3000
1018
+ },
1019
+ {
1020
+ "epoch": 12,
1021
+ "eval_loss": 0.017014725133776665,
1022
+ "eval_mae": 0.09949040412902832,
1023
+ "eval_mse": 0.01701604574918747,
1024
+ "eval_r2": 0.2344194498588299,
1025
+ "eval_runtime": 64.9206,
1026
+ "eval_samples_per_second": 6.161,
1027
+ "eval_steps_per_second": 0.385,
1028
+ "step": 3000
1029
+ },
1030
+ {
1031
+ "epoch": 12,
1032
+ "learning_rate": 3.132407786872442e-05,
1033
+ "loss": 0.0164,
1034
+ "step": 3050
1035
+ },
1036
+ {
1037
+ "epoch": 12,
1038
+ "eval_loss": 0.017031442373991013,
1039
+ "eval_mae": 0.09950780868530273,
1040
+ "eval_mse": 0.01702970452606678,
1041
+ "eval_r2": 0.23397646038139297,
1042
+ "eval_runtime": 63.2282,
1043
+ "eval_samples_per_second": 6.326,
1044
+ "eval_steps_per_second": 0.395,
1045
+ "step": 3050
1046
+ },
1047
+ {
1048
+ "epoch": 13,
1049
+ "learning_rate": 2.9724537818455466e-05,
1050
+ "loss": 0.0163,
1051
+ "step": 3100
1052
+ },
1053
+ {
1054
+ "epoch": 13,
1055
+ "eval_loss": 0.017020680010318756,
1056
+ "eval_mae": 0.099485382437706,
1057
+ "eval_mse": 0.0170209389179945,
1058
+ "eval_r2": 0.23745581810255922,
1059
+ "eval_runtime": 57.2248,
1060
+ "eval_samples_per_second": 6.99,
1061
+ "eval_steps_per_second": 0.437,
1062
+ "step": 3100
1063
+ },
1064
+ {
1065
+ "epoch": 13,
1066
+ "learning_rate": 2.8149460543732664e-05,
1067
+ "loss": 0.0164,
1068
+ "step": 3150
1069
+ },
1070
+ {
1071
+ "epoch": 13,
1072
+ "eval_loss": 0.016971083357930183,
1073
+ "eval_mae": 0.09935550391674042,
1074
+ "eval_mse": 0.016970161348581314,
1075
+ "eval_r2": 0.23819887983596166,
1076
+ "eval_runtime": 58.3514,
1077
+ "eval_samples_per_second": 6.855,
1078
+ "eval_steps_per_second": 0.428,
1079
+ "step": 3150
1080
+ },
1081
+ {
1082
+ "epoch": 13,
1083
+ "learning_rate": 2.6600746408725063e-05,
1084
+ "loss": 0.0163,
1085
+ "step": 3200
1086
+ },
1087
+ {
1088
+ "epoch": 13,
1089
+ "eval_loss": 0.01702103763818741,
1090
+ "eval_mae": 0.09947419166564941,
1091
+ "eval_mse": 0.017017606645822525,
1092
+ "eval_r2": 0.23663776763542477,
1093
+ "eval_runtime": 65.0546,
1094
+ "eval_samples_per_second": 6.149,
1095
+ "eval_steps_per_second": 0.384,
1096
+ "step": 3200
1097
+ },
1098
+ {
1099
+ "epoch": 13,
1100
+ "learning_rate": 2.5080263969913897e-05,
1101
+ "loss": 0.0163,
1102
+ "step": 3250
1103
+ },
1104
+ {
1105
+ "epoch": 13,
1106
+ "eval_loss": 0.01702267862856388,
1107
+ "eval_mae": 0.09951028972864151,
1108
+ "eval_mse": 0.017020724713802338,
1109
+ "eval_r2": 0.23804915917081026,
1110
+ "eval_runtime": 63.3322,
1111
+ "eval_samples_per_second": 6.316,
1112
+ "eval_steps_per_second": 0.395,
1113
+ "step": 3250
1114
+ },
1115
+ {
1116
+ "epoch": 13,
1117
+ "learning_rate": 2.3589847721639617e-05,
1118
+ "loss": 0.0164,
1119
+ "step": 3300
1120
+ },
1121
+ {
1122
+ "epoch": 13,
1123
+ "eval_loss": 0.01695895753800869,
1124
+ "eval_mae": 0.09931185096502304,
1125
+ "eval_mse": 0.016959920525550842,
1126
+ "eval_r2": 0.23948235402533236,
1127
+ "eval_runtime": 62.2021,
1128
+ "eval_samples_per_second": 6.431,
1129
+ "eval_steps_per_second": 0.402,
1130
+ "step": 3300
1131
+ },
1132
+ {
1133
+ "epoch": 14,
1134
+ "learning_rate": 2.2131295882745597e-05,
1135
+ "loss": 0.0163,
1136
+ "step": 3350
1137
+ },
1138
+ {
1139
+ "epoch": 14,
1140
+ "eval_loss": 0.016953853890299797,
1141
+ "eval_mae": 0.09930908679962158,
1142
+ "eval_mse": 0.016957899555563927,
1143
+ "eval_r2": 0.2384040127996312,
1144
+ "eval_runtime": 59.6825,
1145
+ "eval_samples_per_second": 6.702,
1146
+ "eval_steps_per_second": 0.419,
1147
+ "step": 3350
1148
+ },
1149
+ {
1150
+ "epoch": 14,
1151
+ "learning_rate": 2.070636822698877e-05,
1152
+ "loss": 0.0162,
1153
+ "step": 3400
1154
+ },
1155
+ {
1156
+ "epoch": 14,
1157
+ "eval_loss": 0.016969269141554832,
1158
+ "eval_mae": 0.09930410981178284,
1159
+ "eval_mse": 0.016971617937088013,
1160
+ "eval_r2": 0.23911331324321627,
1161
+ "eval_runtime": 61.4012,
1162
+ "eval_samples_per_second": 6.515,
1163
+ "eval_steps_per_second": 0.407,
1164
+ "step": 3400
1165
+ },
1166
+ {
1167
+ "epoch": 14,
1168
+ "learning_rate": 1.9316783959835345e-05,
1169
+ "loss": 0.0164,
1170
+ "step": 3450
1171
+ },
1172
+ {
1173
+ "epoch": 14,
1174
+ "eval_loss": 0.01695878989994526,
1175
+ "eval_mae": 0.09929464012384415,
1176
+ "eval_mse": 0.016956914216279984,
1177
+ "eval_r2": 0.23904322049736015,
1178
+ "eval_runtime": 61.0952,
1179
+ "eval_samples_per_second": 6.547,
1180
+ "eval_steps_per_second": 0.409,
1181
+ "step": 3450
1182
+ },
1183
+ {
1184
+ "epoch": 14,
1185
+ "learning_rate": 1.796421964420285e-05,
1186
+ "loss": 0.0164,
1187
+ "step": 3500
1188
+ },
1189
+ {
1190
+ "epoch": 14,
1191
+ "eval_loss": 0.016925431787967682,
1192
+ "eval_mae": 0.09921905398368835,
1193
+ "eval_mse": 0.016928432509303093,
1194
+ "eval_r2": 0.23861995325924557,
1195
+ "eval_runtime": 58.2956,
1196
+ "eval_samples_per_second": 6.862,
1197
+ "eval_steps_per_second": 0.429,
1198
+ "step": 3500
1199
+ },
1200
+ {
1201
+ "epoch": 14,
1202
+ "learning_rate": 1.665030717765149e-05,
1203
+ "loss": 0.0163,
1204
+ "step": 3550
1205
+ },
1206
+ {
1207
+ "epoch": 14,
1208
+ "eval_loss": 0.016955168917775154,
1209
+ "eval_mae": 0.09928599745035172,
1210
+ "eval_mse": 0.016958734020590782,
1211
+ "eval_r2": 0.24003949016117532,
1212
+ "eval_runtime": 60.224,
1213
+ "eval_samples_per_second": 6.642,
1214
+ "eval_steps_per_second": 0.415,
1215
+ "step": 3550
1216
+ },
1217
+ {
1218
+ "epoch": 15,
1219
+ "learning_rate": 1.5376631823464953e-05,
1220
+ "loss": 0.0163,
1221
+ "step": 3600
1222
+ },
1223
+ {
1224
+ "epoch": 15,
1225
+ "eval_loss": 0.017045069485902786,
1226
+ "eval_mae": 0.09952731430530548,
1227
+ "eval_mse": 0.017046233639121056,
1228
+ "eval_r2": 0.2409846653602259,
1229
+ "eval_runtime": 56.9312,
1230
+ "eval_samples_per_second": 7.026,
1231
+ "eval_steps_per_second": 0.439,
1232
+ "step": 3600
1233
+ },
1234
+ {
1235
+ "epoch": 15,
1236
+ "learning_rate": 1.4144730297996666e-05,
1237
+ "loss": 0.0163,
1238
+ "step": 3650
1239
+ },
1240
+ {
1241
+ "epoch": 15,
1242
+ "eval_loss": 0.01690911501646042,
1243
+ "eval_mae": 0.09913720935583115,
1244
+ "eval_mse": 0.01690947264432907,
1245
+ "eval_r2": 0.24005320295619414,
1246
+ "eval_runtime": 63.4396,
1247
+ "eval_samples_per_second": 6.305,
1248
+ "eval_steps_per_second": 0.394,
1249
+ "step": 3650
1250
+ },
1251
+ {
1252
+ "epoch": 15,
1253
+ "learning_rate": 1.295608891658896e-05,
1254
+ "loss": 0.0163,
1255
+ "step": 3700
1256
+ },
1257
+ {
1258
+ "epoch": 15,
1259
+ "eval_loss": 0.016995713114738464,
1260
+ "eval_mae": 0.09939228743314743,
1261
+ "eval_mse": 0.016998106613755226,
1262
+ "eval_r2": 0.24257331961419693,
1263
+ "eval_runtime": 58.9315,
1264
+ "eval_samples_per_second": 6.788,
1265
+ "eval_steps_per_second": 0.424,
1266
+ "step": 3700
1267
+ },
1268
+ {
1269
+ "epoch": 15,
1270
+ "learning_rate": 1.1812141800301945e-05,
1271
+ "loss": 0.0163,
1272
+ "step": 3750
1273
+ },
1274
+ {
1275
+ "epoch": 15,
1276
+ "eval_loss": 0.01695888675749302,
1277
+ "eval_mae": 0.09927807003259659,
1278
+ "eval_mse": 0.016960280016064644,
1279
+ "eval_r2": 0.24255255772513262,
1280
+ "eval_runtime": 58.2565,
1281
+ "eval_samples_per_second": 6.866,
1282
+ "eval_steps_per_second": 0.429,
1283
+ "step": 3750
1284
+ },
1285
+ {
1286
+ "epoch": 15,
1287
+ "learning_rate": 1.0714269145616063e-05,
1288
+ "loss": 0.0163,
1289
+ "step": 3800
1290
+ },
1291
+ {
1292
+ "epoch": 15,
1293
+ "eval_loss": 0.016934260725975037,
1294
+ "eval_mae": 0.09918724000453949,
1295
+ "eval_mse": 0.016934024170041084,
1296
+ "eval_r2": 0.23905501031932674,
1297
+ "eval_runtime": 59.5916,
1298
+ "eval_samples_per_second": 6.712,
1299
+ "eval_steps_per_second": 0.42,
1300
+ "step": 3800
1301
+ },
1302
+ {
1303
+ "epoch": 16,
1304
+ "learning_rate": 9.663795559195733e-06,
1305
+ "loss": 0.0162,
1306
+ "step": 3850
1307
+ },
1308
+ {
1309
+ "epoch": 16,
1310
+ "eval_loss": 0.016940169036388397,
1311
+ "eval_mae": 0.09923317283391953,
1312
+ "eval_mse": 0.016941731795668602,
1313
+ "eval_r2": 0.2415171870593763,
1314
+ "eval_runtime": 59.9307,
1315
+ "eval_samples_per_second": 6.674,
1316
+ "eval_steps_per_second": 0.417,
1317
+ "step": 3850
1318
+ },
1319
+ {
1320
+ "epoch": 16,
1321
+ "learning_rate": 8.661988459723309e-06,
1322
+ "loss": 0.0163,
1323
+ "step": 3900
1324
+ },
1325
+ {
1326
+ "epoch": 16,
1327
+ "eval_loss": 0.01686818338930607,
1328
+ "eval_mae": 0.09902676939964294,
1329
+ "eval_mse": 0.01686619035899639,
1330
+ "eval_r2": 0.23881856654464984,
1331
+ "eval_runtime": 59.939,
1332
+ "eval_samples_per_second": 6.673,
1333
+ "eval_steps_per_second": 0.417,
1334
+ "step": 3900
1335
+ },
1336
+ {
1337
+ "epoch": 16,
1338
+ "learning_rate": 7.710056548731447e-06,
1339
+ "loss": 0.0164,
1340
+ "step": 3950
1341
+ },
1342
+ {
1343
+ "epoch": 16,
1344
+ "eval_loss": 0.01692971959710121,
1345
+ "eval_mae": 0.09918209910392761,
1346
+ "eval_mse": 0.01693139225244522,
1347
+ "eval_r2": 0.24229291180997825,
1348
+ "eval_runtime": 58.0894,
1349
+ "eval_samples_per_second": 6.886,
1350
+ "eval_steps_per_second": 0.43,
1351
+ "step": 3950
1352
+ },
1353
+ {
1354
+ "epoch": 16,
1355
+ "learning_rate": 6.809148352279182e-06,
1356
+ "loss": 0.0163,
1357
+ "step": 4000
1358
+ },
1359
+ {
1360
+ "epoch": 16,
1361
+ "eval_loss": 0.016949491575360298,
1362
+ "eval_mae": 0.09925098717212677,
1363
+ "eval_mse": 0.01695188321173191,
1364
+ "eval_r2": 0.23993232169758605,
1365
+ "eval_runtime": 61.4621,
1366
+ "eval_samples_per_second": 6.508,
1367
+ "eval_steps_per_second": 0.407,
1368
+ "step": 4000
1369
+ },
1370
+ {
1371
+ "epoch": 16,
1372
+ "learning_rate": 5.960350835230766e-06,
1373
+ "loss": 0.0163,
1374
+ "step": 4050
1375
+ },
1376
+ {
1377
+ "epoch": 16,
1378
+ "eval_loss": 0.016934702172875404,
1379
+ "eval_mae": 0.09920256584882736,
1380
+ "eval_mse": 0.016937121748924255,
1381
+ "eval_r2": 0.2416569624089897,
1382
+ "eval_runtime": 58.2108,
1383
+ "eval_samples_per_second": 6.872,
1384
+ "eval_steps_per_second": 0.429,
1385
+ "step": 4050
1386
+ },
1387
+ {
1388
+ "epoch": 17,
1389
+ "learning_rate": 5.164688089809444e-06,
1390
+ "loss": 0.0162,
1391
+ "step": 4100
1392
+ },
1393
+ {
1394
+ "epoch": 17,
1395
+ "eval_loss": 0.016918588429689407,
1396
+ "eval_mae": 0.09915737062692642,
1397
+ "eval_mse": 0.016918879002332687,
1398
+ "eval_r2": 0.24260718873499565,
1399
+ "eval_runtime": 56.8092,
1400
+ "eval_samples_per_second": 7.041,
1401
+ "eval_steps_per_second": 0.44,
1402
+ "step": 4100
1403
+ },
1404
+ {
1405
+ "epoch": 17,
1406
+ "learning_rate": 4.423120100008582e-06,
1407
+ "loss": 0.0163,
1408
+ "step": 4150
1409
+ },
1410
+ {
1411
+ "epoch": 17,
1412
+ "eval_loss": 0.016940301284193993,
1413
+ "eval_mae": 0.09919126331806183,
1414
+ "eval_mse": 0.016939815133810043,
1415
+ "eval_r2": 0.23998661125272824,
1416
+ "eval_runtime": 59.9186,
1417
+ "eval_samples_per_second": 6.676,
1418
+ "eval_steps_per_second": 0.417,
1419
+ "step": 4150
1420
+ },
1421
+ {
1422
+ "epoch": 17,
1423
+ "learning_rate": 3.7365415833504725e-06,
1424
+ "loss": 0.0162,
1425
+ "step": 4200
1426
+ },
1427
+ {
1428
+ "epoch": 17,
1429
+ "eval_loss": 0.016918940469622612,
1430
+ "eval_mae": 0.09915791451931,
1431
+ "eval_mse": 0.016918599605560303,
1432
+ "eval_r2": 0.23974174795237568,
1433
+ "eval_runtime": 63.0008,
1434
+ "eval_samples_per_second": 6.349,
1435
+ "eval_steps_per_second": 0.397,
1436
+ "step": 4200
1437
+ },
1438
+ {
1439
+ "epoch": 17,
1440
+ "learning_rate": 3.105780911390738e-06,
1441
+ "loss": 0.0163,
1442
+ "step": 4250
1443
+ },
1444
+ {
1445
+ "epoch": 17,
1446
+ "eval_loss": 0.016923611983656883,
1447
+ "eval_mae": 0.09916673600673676,
1448
+ "eval_mse": 0.016924140974879265,
1449
+ "eval_r2": 0.23963991013957076,
1450
+ "eval_runtime": 56.3879,
1451
+ "eval_samples_per_second": 7.094,
1452
+ "eval_steps_per_second": 0.443,
1453
+ "step": 4250
1454
+ },
1455
+ {
1456
+ "epoch": 18,
1457
+ "learning_rate": 2.5315991102703716e-06,
1458
+ "loss": 0.0163,
1459
+ "step": 4300
1460
+ },
1461
+ {
1462
+ "epoch": 18,
1463
+ "eval_loss": 0.01691032014787197,
1464
+ "eval_mae": 0.09915610402822495,
1465
+ "eval_mse": 0.01691248081624508,
1466
+ "eval_r2": 0.24413583693173724,
1467
+ "eval_runtime": 61.6302,
1468
+ "eval_samples_per_second": 6.49,
1469
+ "eval_steps_per_second": 0.406,
1470
+ "step": 4300
1471
+ },
1472
+ {
1473
+ "epoch": 18,
1474
+ "learning_rate": 2.0146889425216476e-06,
1475
+ "loss": 0.0163,
1476
+ "step": 4350
1477
+ },
1478
+ {
1479
+ "epoch": 18,
1480
+ "eval_loss": 0.016894323751330376,
1481
+ "eval_mae": 0.09909255057573318,
1482
+ "eval_mse": 0.016893018037080765,
1483
+ "eval_r2": 0.24154658591740408,
1484
+ "eval_runtime": 61.7977,
1485
+ "eval_samples_per_second": 6.473,
1486
+ "eval_steps_per_second": 0.405,
1487
+ "step": 4350
1488
+ },
1489
+ {
1490
+ "epoch": 18,
1491
+ "learning_rate": 1.555674071235358e-06,
1492
+ "loss": 0.0163,
1493
+ "step": 4400
1494
+ },
1495
+ {
1496
+ "epoch": 18,
1497
+ "eval_loss": 0.016896896064281464,
1498
+ "eval_mae": 0.09909563511610031,
1499
+ "eval_mse": 0.016896726563572884,
1500
+ "eval_r2": 0.2395997969159448,
1501
+ "eval_runtime": 63.3638,
1502
+ "eval_samples_per_second": 6.313,
1503
+ "eval_steps_per_second": 0.395,
1504
+ "step": 4400
1505
+ },
1506
+ {
1507
+ "epoch": 18,
1508
+ "learning_rate": 1.155108307598024e-06,
1509
+ "loss": 0.0163,
1510
+ "step": 4450
1511
+ },
1512
+ {
1513
+ "epoch": 18,
1514
+ "eval_loss": 0.016881046816706657,
1515
+ "eval_mae": 0.09909423440694809,
1516
+ "eval_mse": 0.016880322247743607,
1517
+ "eval_r2": 0.2425676493378176,
1518
+ "eval_runtime": 60.1702,
1519
+ "eval_samples_per_second": 6.648,
1520
+ "eval_steps_per_second": 0.415,
1521
+ "step": 4450
1522
+ },
1523
+ {
1524
+ "epoch": 18,
1525
+ "learning_rate": 8.134749427070376e-07,
1526
+ "loss": 0.0163,
1527
+ "step": 4500
1528
+ },
1529
+ {
1530
+ "epoch": 18,
1531
+ "eval_loss": 0.016920508816838264,
1532
+ "eval_mae": 0.09913913160562515,
1533
+ "eval_mse": 0.016919521614909172,
1534
+ "eval_r2": 0.240687826762218,
1535
+ "eval_runtime": 56.9763,
1536
+ "eval_samples_per_second": 7.02,
1537
+ "eval_steps_per_second": 0.439,
1538
+ "step": 4500
1539
+ },
1540
+ {
1541
+ "epoch": 19,
1542
+ "learning_rate": 5.311861644696048e-07,
1543
+ "loss": 0.0163,
1544
+ "step": 4550
1545
+ },
1546
+ {
1547
+ "epoch": 19,
1548
+ "eval_loss": 0.016956361010670662,
1549
+ "eval_mae": 0.09927552938461304,
1550
+ "eval_mse": 0.016957107931375504,
1551
+ "eval_r2": 0.23962235723125958,
1552
+ "eval_runtime": 57.5809,
1553
+ "eval_samples_per_second": 6.947,
1554
+ "eval_steps_per_second": 0.434,
1555
+ "step": 4550
1556
+ },
1557
+ {
1558
+ "epoch": 19,
1559
+ "learning_rate": 3.0858256028932776e-07,
1560
+ "loss": 0.0163,
1561
+ "step": 4600
1562
+ },
1563
+ {
1564
+ "epoch": 19,
1565
+ "eval_loss": 0.016894662752747536,
1566
+ "eval_mae": 0.09909520298242569,
1567
+ "eval_mse": 0.016894804313778877,
1568
+ "eval_r2": 0.24004589904875062,
1569
+ "eval_runtime": 58.5108,
1570
+ "eval_samples_per_second": 6.836,
1571
+ "eval_steps_per_second": 0.427,
1572
+ "step": 4600
1573
+ },
1574
+ {
1575
+ "epoch": 19,
1576
+ "learning_rate": 1.4593270614026154e-07,
1577
+ "loss": 0.0162,
1578
+ "step": 4650
1579
+ },
1580
+ {
1581
+ "epoch": 19,
1582
+ "eval_loss": 0.016891133040189743,
1583
+ "eval_mae": 0.09907051920890808,
1584
+ "eval_mse": 0.016890019178390503,
1585
+ "eval_r2": 0.24048031726771313,
1586
+ "eval_runtime": 62.2953,
1587
+ "eval_samples_per_second": 6.421,
1588
+ "eval_steps_per_second": 0.401,
1589
+ "step": 4650
1590
+ },
1591
+ {
1592
+ "epoch": 19,
1593
+ "learning_rate": 4.343284252429913e-08,
1594
+ "loss": 0.0162,
1595
+ "step": 4700
1596
+ },
1597
+ {
1598
+ "epoch": 19,
1599
+ "eval_loss": 0.016868911683559418,
1600
+ "eval_mae": 0.0990411788225174,
1601
+ "eval_mse": 0.016869619488716125,
1602
+ "eval_r2": 0.24109475503821198,
1603
+ "eval_runtime": 58.3864,
1604
+ "eval_samples_per_second": 6.851,
1605
+ "eval_steps_per_second": 0.428,
1606
+ "step": 4700
1607
+ }
1608
+ ],
1609
+ "max_steps": 4760,
1610
+ "num_train_epochs": 20,
1611
+ "total_flos": 2.0677241114343014e+20,
1612
+ "trial_name": null,
1613
+ "trial_params": null
1614
+ }
vitmae_111M/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca1fa2f0f9f90bc54e3b7413754762700c9f54a48fa5ac4947f41996883090a3
3
+ size 4600
vitmae_650M/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/gpfs/gibbs/pi/dijk/BrainLM_runs/2023-11-15-14_47_00/checkpoint-3300",
3
+ "architectures": [
4
+ "ViTMAEForPreTraining"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "decoder_hidden_size": 512,
8
+ "decoder_intermediate_size": 2048,
9
+ "decoder_num_attention_heads": 16,
10
+ "decoder_num_hidden_layers": 8,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.0,
13
+ "hidden_size": 1280,
14
+ "image_size": [
15
+ 434,
16
+ 434
17
+ ],
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 5120,
20
+ "layer_norm_eps": 1e-12,
21
+ "mask_ratio": 0.75,
22
+ "model_type": "vit_mae",
23
+ "norm_pix_loss": false,
24
+ "num_attention_heads": 16,
25
+ "num_channels": 3,
26
+ "num_hidden_layers": 32,
27
+ "patch_size": 14,
28
+ "qkv_bias": true,
29
+ "torch_dtype": "float32",
30
+ "transformers_version": "4.28.0"
31
+ }
vitmae_650M/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b4c403704b4bf8d800cb5aca85db4fd52a982af0a045c9633f510dd87bc99b2
3
+ size 5253303842
vitmae_650M/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8221b7ad99f5dd42ee1f58cca4824b2904157c9e16ab2cd47cf377e8a90f75a6
3
+ size 2633576194
vitmae_650M/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ef164f3ddf0ee3f3595bb53948fa8352bf6391c8eb9b01bbd66fed8c46988da
3
+ size 14244
vitmae_650M/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15d60713e239aa461be03135bdbca99b2bfe14ea1d561b3ce05394a2a8b3b9e7
3
+ size 988
vitmae_650M/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dc1bb0e97fd476b1d9a69ca508d6cd21c01eb4920bd2c8d40259d50b0e95e6f
3
+ size 1064
vitmae_650M/trainer_state.json ADDED
@@ -0,0 +1,1580 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.016573524102568626,
3
+ "best_model_checkpoint": "/gpfs/gibbs/pi/dijk/BrainLM_runs/2023-11-17-17_37_00/checkpoint-4400",
4
+ "epoch": 19.292899505226252,
5
+ "global_step": 4600,
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,
12
+ "learning_rate": 2.100840336134454e-05,
13
+ "loss": 0.0205,
14
+ "step": 50
15
+ },
16
+ {
17
+ "epoch": 0,
18
+ "eval_loss": 0.020380878821015358,
19
+ "eval_mae": 0.10914861410856247,
20
+ "eval_mse": 0.020377445966005325,
21
+ "eval_r2": 0.08699230219766863,
22
+ "eval_runtime": 65.4495,
23
+ "eval_samples_per_second": 6.112,
24
+ "eval_steps_per_second": 3.056,
25
+ "step": 50
26
+ },
27
+ {
28
+ "epoch": 0,
29
+ "learning_rate": 4.201680672268908e-05,
30
+ "loss": 0.0195,
31
+ "step": 100
32
+ },
33
+ {
34
+ "epoch": 0,
35
+ "eval_loss": 0.01993260346353054,
36
+ "eval_mae": 0.10798273980617523,
37
+ "eval_mse": 0.019940046593546867,
38
+ "eval_r2": 0.10634720488402916,
39
+ "eval_runtime": 82.6009,
40
+ "eval_samples_per_second": 4.843,
41
+ "eval_steps_per_second": 2.421,
42
+ "step": 100
43
+ },
44
+ {
45
+ "epoch": 0,
46
+ "learning_rate": 6.302521008403361e-05,
47
+ "loss": 0.0191,
48
+ "step": 150
49
+ },
50
+ {
51
+ "epoch": 0,
52
+ "eval_loss": 0.01966927945613861,
53
+ "eval_mae": 0.10720709711313248,
54
+ "eval_mse": 0.019672850146889687,
55
+ "eval_r2": 0.1188939908583877,
56
+ "eval_runtime": 273.6405,
57
+ "eval_samples_per_second": 1.462,
58
+ "eval_steps_per_second": 0.731,
59
+ "step": 150
60
+ },
61
+ {
62
+ "epoch": 0,
63
+ "learning_rate": 8.403361344537815e-05,
64
+ "loss": 0.0189,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0,
69
+ "eval_loss": 0.019402366131544113,
70
+ "eval_mae": 0.10650211572647095,
71
+ "eval_mse": 0.01940837688744068,
72
+ "eval_r2": 0.1273264644968536,
73
+ "eval_runtime": 65.4954,
74
+ "eval_samples_per_second": 6.107,
75
+ "eval_steps_per_second": 3.054,
76
+ "step": 200
77
+ },
78
+ {
79
+ "epoch": 1,
80
+ "learning_rate": 9.999826244478293e-05,
81
+ "loss": 0.0187,
82
+ "step": 250
83
+ },
84
+ {
85
+ "epoch": 1,
86
+ "eval_loss": 0.019103730097413063,
87
+ "eval_mae": 0.10563357919454575,
88
+ "eval_mse": 0.01910446584224701,
89
+ "eval_r2": 0.14288129156056273,
90
+ "eval_runtime": 67.6226,
91
+ "eval_samples_per_second": 5.915,
92
+ "eval_steps_per_second": 2.958,
93
+ "step": 250
94
+ },
95
+ {
96
+ "epoch": 1,
97
+ "learning_rate": 9.995362383107962e-05,
98
+ "loss": 0.0183,
99
+ "step": 300
100
+ },
101
+ {
102
+ "epoch": 1,
103
+ "eval_loss": 0.0188433900475502,
104
+ "eval_mae": 0.1049303263425827,
105
+ "eval_mse": 0.018846124410629272,
106
+ "eval_r2": 0.15557466279282905,
107
+ "eval_runtime": 63.0108,
108
+ "eval_samples_per_second": 6.348,
109
+ "eval_steps_per_second": 3.174,
110
+ "step": 300
111
+ },
112
+ {
113
+ "epoch": 1,
114
+ "learning_rate": 9.98487151097676e-05,
115
+ "loss": 0.0181,
116
+ "step": 350
117
+ },
118
+ {
119
+ "epoch": 1,
120
+ "eval_loss": 0.018534274771809578,
121
+ "eval_mae": 0.10398274660110474,
122
+ "eval_mse": 0.018525807186961174,
123
+ "eval_r2": 0.16440368152546336,
124
+ "eval_runtime": 63.6126,
125
+ "eval_samples_per_second": 6.288,
126
+ "eval_steps_per_second": 3.144,
127
+ "step": 350
128
+ },
129
+ {
130
+ "epoch": 1,
131
+ "learning_rate": 9.968366285544619e-05,
132
+ "loss": 0.0179,
133
+ "step": 400
134
+ },
135
+ {
136
+ "epoch": 1,
137
+ "eval_loss": 0.01846296340227127,
138
+ "eval_mae": 0.10383900254964828,
139
+ "eval_mse": 0.018469005823135376,
140
+ "eval_r2": 0.17126507792798895,
141
+ "eval_runtime": 64.4338,
142
+ "eval_samples_per_second": 6.208,
143
+ "eval_steps_per_second": 3.104,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 1,
148
+ "learning_rate": 9.945866620716411e-05,
149
+ "loss": 0.0177,
150
+ "step": 450
151
+ },
152
+ {
153
+ "epoch": 1,
154
+ "eval_loss": 0.01835346780717373,
155
+ "eval_mae": 0.10348159074783325,
156
+ "eval_mse": 0.018356231972575188,
157
+ "eval_r2": 0.1812631242910432,
158
+ "eval_runtime": 63.7952,
159
+ "eval_samples_per_second": 6.27,
160
+ "eval_steps_per_second": 3.135,
161
+ "step": 450
162
+ },
163
+ {
164
+ "epoch": 2,
165
+ "learning_rate": 9.917399662815391e-05,
166
+ "loss": 0.0176,
167
+ "step": 500
168
+ },
169
+ {
170
+ "epoch": 2,
171
+ "eval_loss": 0.018152762204408646,
172
+ "eval_mae": 0.10292253643274307,
173
+ "eval_mse": 0.018158329650759697,
174
+ "eval_r2": 0.18455260254083927,
175
+ "eval_runtime": 62.2123,
176
+ "eval_samples_per_second": 6.43,
177
+ "eval_steps_per_second": 3.215,
178
+ "step": 500
179
+ },
180
+ {
181
+ "epoch": 2,
182
+ "learning_rate": 9.882999757830588e-05,
183
+ "loss": 0.0175,
184
+ "step": 550
185
+ },
186
+ {
187
+ "epoch": 2,
188
+ "eval_loss": 0.018096117302775383,
189
+ "eval_mae": 0.10270749032497406,
190
+ "eval_mse": 0.01808706857264042,
191
+ "eval_r2": 0.1878013710016545,
192
+ "eval_runtime": 62.45,
193
+ "eval_samples_per_second": 6.405,
194
+ "eval_steps_per_second": 3.203,
195
+ "step": 550
196
+ },
197
+ {
198
+ "epoch": 2,
199
+ "learning_rate": 9.842708409977635e-05,
200
+ "loss": 0.0174,
201
+ "step": 600
202
+ },
203
+ {
204
+ "epoch": 2,
205
+ "eval_loss": 0.018007792532444,
206
+ "eval_mae": 0.10250235348939896,
207
+ "eval_mse": 0.01801513321697712,
208
+ "eval_r2": 0.19156642781886424,
209
+ "eval_runtime": 62.617,
210
+ "eval_samples_per_second": 6.388,
211
+ "eval_steps_per_second": 3.194,
212
+ "step": 600
213
+ },
214
+ {
215
+ "epoch": 2,
216
+ "learning_rate": 9.796574231623055e-05,
217
+ "loss": 0.0173,
218
+ "step": 650
219
+ },
220
+ {
221
+ "epoch": 2,
222
+ "eval_loss": 0.017887134104967117,
223
+ "eval_mae": 0.10214251279830933,
224
+ "eval_mse": 0.017883572727441788,
225
+ "eval_r2": 0.19610529281414957,
226
+ "eval_runtime": 63.0647,
227
+ "eval_samples_per_second": 6.343,
228
+ "eval_steps_per_second": 3.171,
229
+ "step": 650
230
+ },
231
+ {
232
+ "epoch": 2,
233
+ "learning_rate": 9.744652884632406e-05,
234
+ "loss": 0.0173,
235
+ "step": 700
236
+ },
237
+ {
238
+ "epoch": 2,
239
+ "eval_loss": 0.017843402922153473,
240
+ "eval_mae": 0.10201350599527359,
241
+ "eval_mse": 0.017849572002887726,
242
+ "eval_r2": 0.20036411183239422,
243
+ "eval_runtime": 63.1964,
244
+ "eval_samples_per_second": 6.329,
245
+ "eval_steps_per_second": 3.165,
246
+ "step": 700
247
+ },
248
+ {
249
+ "epoch": 3,
250
+ "learning_rate": 9.687007013213063e-05,
251
+ "loss": 0.0172,
252
+ "step": 750
253
+ },
254
+ {
255
+ "epoch": 3,
256
+ "eval_loss": 0.017804041504859924,
257
+ "eval_mae": 0.10189322382211685,
258
+ "eval_mse": 0.01781976968050003,
259
+ "eval_r2": 0.20562580407293896,
260
+ "eval_runtime": 63.7696,
261
+ "eval_samples_per_second": 6.273,
262
+ "eval_steps_per_second": 3.136,
263
+ "step": 750
264
+ },
265
+ {
266
+ "epoch": 3,
267
+ "learning_rate": 9.623706168332645e-05,
268
+ "loss": 0.0171,
269
+ "step": 800
270
+ },
271
+ {
272
+ "epoch": 3,
273
+ "eval_loss": 0.017708342522382736,
274
+ "eval_mae": 0.10164093226194382,
275
+ "eval_mse": 0.017712706699967384,
276
+ "eval_r2": 0.20602592052107038,
277
+ "eval_runtime": 63.456,
278
+ "eval_samples_per_second": 6.304,
279
+ "eval_steps_per_second": 3.152,
280
+ "step": 800
281
+ },
282
+ {
283
+ "epoch": 3,
284
+ "learning_rate": 9.554826723804303e-05,
285
+ "loss": 0.017,
286
+ "step": 850
287
+ },
288
+ {
289
+ "epoch": 3,
290
+ "eval_loss": 0.017668385058641434,
291
+ "eval_mae": 0.10145936906337738,
292
+ "eval_mse": 0.017669973894953728,
293
+ "eval_r2": 0.20682701060142172,
294
+ "eval_runtime": 63.5861,
295
+ "eval_samples_per_second": 6.291,
296
+ "eval_steps_per_second": 3.145,
297
+ "step": 850
298
+ },
299
+ {
300
+ "epoch": 3,
301
+ "learning_rate": 9.480451784140091e-05,
302
+ "loss": 0.017,
303
+ "step": 900
304
+ },
305
+ {
306
+ "epoch": 3,
307
+ "eval_loss": 0.017553946003317833,
308
+ "eval_mae": 0.10114699602127075,
309
+ "eval_mse": 0.01755087450146675,
310
+ "eval_r2": 0.2111549450823803,
311
+ "eval_runtime": 62.7965,
312
+ "eval_samples_per_second": 6.37,
313
+ "eval_steps_per_second": 3.185,
314
+ "step": 900
315
+ },
316
+ {
317
+ "epoch": 3,
318
+ "learning_rate": 9.400671084283607e-05,
319
+ "loss": 0.017,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 3,
324
+ "eval_loss": 0.017535727471113205,
325
+ "eval_mae": 0.10109320282936096,
326
+ "eval_mse": 0.017534563317894936,
327
+ "eval_r2": 0.21152867082641214,
328
+ "eval_runtime": 62.081,
329
+ "eval_samples_per_second": 6.443,
330
+ "eval_steps_per_second": 3.222,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 4,
335
+ "learning_rate": 9.315580881342876e-05,
336
+ "loss": 0.0168,
337
+ "step": 1000
338
+ },
339
+ {
340
+ "epoch": 4,
341
+ "eval_loss": 0.01751740463078022,
342
+ "eval_mae": 0.10102058947086334,
343
+ "eval_mse": 0.017521008849143982,
344
+ "eval_r2": 0.2173146917939317,
345
+ "eval_runtime": 63.7702,
346
+ "eval_samples_per_second": 6.273,
347
+ "eval_steps_per_second": 3.136,
348
+ "step": 1000
349
+ },
350
+ {
351
+ "epoch": 4,
352
+ "learning_rate": 9.225283838454111e-05,
353
+ "loss": 0.0169,
354
+ "step": 1050
355
+ },
356
+ {
357
+ "epoch": 4,
358
+ "eval_loss": 0.017473317682743073,
359
+ "eval_mae": 0.1008896455168724,
360
+ "eval_mse": 0.0174697358161211,
361
+ "eval_r2": 0.21592913051237428,
362
+ "eval_runtime": 63.0763,
363
+ "eval_samples_per_second": 6.342,
364
+ "eval_steps_per_second": 3.171,
365
+ "step": 1050
366
+ },
367
+ {
368
+ "epoch": 4,
369
+ "learning_rate": 9.129888900916456e-05,
370
+ "loss": 0.0168,
371
+ "step": 1100
372
+ },
373
+ {
374
+ "epoch": 4,
375
+ "eval_loss": 0.01740197278559208,
376
+ "eval_mae": 0.10066132992506027,
377
+ "eval_mse": 0.017395442351698875,
378
+ "eval_r2": 0.21965390849250066,
379
+ "eval_runtime": 62.679,
380
+ "eval_samples_per_second": 6.382,
381
+ "eval_steps_per_second": 3.191,
382
+ "step": 1100
383
+ },
384
+ {
385
+ "epoch": 4,
386
+ "learning_rate": 9.029511164747175e-05,
387
+ "loss": 0.0168,
388
+ "step": 1150
389
+ },
390
+ {
391
+ "epoch": 4,
392
+ "eval_loss": 0.01737840846180916,
393
+ "eval_mae": 0.100642628967762,
394
+ "eval_mse": 0.01738792657852173,
395
+ "eval_r2": 0.2177673870807011,
396
+ "eval_runtime": 63.0635,
397
+ "eval_samples_per_second": 6.343,
398
+ "eval_steps_per_second": 3.171,
399
+ "step": 1150
400
+ },
401
+ {
402
+ "epoch": 5,
403
+ "learning_rate": 8.924271737815854e-05,
404
+ "loss": 0.0167,
405
+ "step": 1200
406
+ },
407
+ {
408
+ "epoch": 5,
409
+ "eval_loss": 0.017356639727950096,
410
+ "eval_mae": 0.10054408758878708,
411
+ "eval_mse": 0.01736092008650303,
412
+ "eval_r2": 0.22268234003478382,
413
+ "eval_runtime": 62.6903,
414
+ "eval_samples_per_second": 6.381,
415
+ "eval_steps_per_second": 3.19,
416
+ "step": 1200
417
+ },
418
+ {
419
+ "epoch": 5,
420
+ "learning_rate": 8.814297593725199e-05,
421
+ "loss": 0.0167,
422
+ "step": 1250
423
+ },
424
+ {
425
+ "epoch": 5,
426
+ "eval_loss": 0.017370322719216347,
427
+ "eval_mae": 0.10060014575719833,
428
+ "eval_mse": 0.017372848466038704,
429
+ "eval_r2": 0.22218442610232647,
430
+ "eval_runtime": 62.9578,
431
+ "eval_samples_per_second": 6.353,
432
+ "eval_steps_per_second": 3.177,
433
+ "step": 1250
434
+ },
435
+ {
436
+ "epoch": 5,
437
+ "learning_rate": 8.699721418614673e-05,
438
+ "loss": 0.0167,
439
+ "step": 1300
440
+ },
441
+ {
442
+ "epoch": 5,
443
+ "eval_loss": 0.017255190759897232,
444
+ "eval_mae": 0.10025202482938766,
445
+ "eval_mse": 0.0172572061419487,
446
+ "eval_r2": 0.2245981157395297,
447
+ "eval_runtime": 62.725,
448
+ "eval_samples_per_second": 6.377,
449
+ "eval_steps_per_second": 3.189,
450
+ "step": 1300
451
+ },
452
+ {
453
+ "epoch": 5,
454
+ "learning_rate": 8.580681451071866e-05,
455
+ "loss": 0.0166,
456
+ "step": 1350
457
+ },
458
+ {
459
+ "epoch": 5,
460
+ "eval_loss": 0.017206793650984764,
461
+ "eval_mae": 0.10011155903339386,
462
+ "eval_mse": 0.017209110781550407,
463
+ "eval_r2": 0.22857719354252215,
464
+ "eval_runtime": 63.8964,
465
+ "eval_samples_per_second": 6.26,
466
+ "eval_steps_per_second": 3.13,
467
+ "step": 1350
468
+ },
469
+ {
470
+ "epoch": 5,
471
+ "learning_rate": 8.457321315344694e-05,
472
+ "loss": 0.0167,
473
+ "step": 1400
474
+ },
475
+ {
476
+ "epoch": 5,
477
+ "eval_loss": 0.017166156321763992,
478
+ "eval_mae": 0.10000505298376083,
479
+ "eval_mse": 0.017173225060105324,
480
+ "eval_r2": 0.22773855218146166,
481
+ "eval_runtime": 63.4268,
482
+ "eval_samples_per_second": 6.306,
483
+ "eval_steps_per_second": 3.153,
484
+ "step": 1400
485
+ },
486
+ {
487
+ "epoch": 6,
488
+ "learning_rate": 8.329789848055704e-05,
489
+ "loss": 0.0164,
490
+ "step": 1450
491
+ },
492
+ {
493
+ "epoch": 6,
494
+ "eval_loss": 0.01718437485396862,
495
+ "eval_mae": 0.10002291202545166,
496
+ "eval_mse": 0.01718369498848915,
497
+ "eval_r2": 0.2300388059901679,
498
+ "eval_runtime": 63.8338,
499
+ "eval_samples_per_second": 6.266,
500
+ "eval_steps_per_second": 3.133,
501
+ "step": 1450
502
+ },
503
+ {
504
+ "epoch": 6,
505
+ "learning_rate": 8.198240918627524e-05,
506
+ "loss": 0.0166,
507
+ "step": 1500
508
+ },
509
+ {
510
+ "epoch": 6,
511
+ "eval_loss": 0.017157739028334618,
512
+ "eval_mae": 0.09996546804904938,
513
+ "eval_mse": 0.017160937190055847,
514
+ "eval_r2": 0.23035895673172035,
515
+ "eval_runtime": 62.0856,
516
+ "eval_samples_per_second": 6.443,
517
+ "eval_steps_per_second": 3.221,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 6,
522
+ "learning_rate": 8.062833243636134e-05,
523
+ "loss": 0.0165,
524
+ "step": 1550
525
+ },
526
+ {
527
+ "epoch": 6,
528
+ "eval_loss": 0.01712297834455967,
529
+ "eval_mae": 0.09981416165828705,
530
+ "eval_mse": 0.017125777900218964,
531
+ "eval_r2": 0.23338275672251196,
532
+ "eval_runtime": 62.5642,
533
+ "eval_samples_per_second": 6.393,
534
+ "eval_steps_per_second": 3.197,
535
+ "step": 1550
536
+ },
537
+ {
538
+ "epoch": 6,
539
+ "learning_rate": 7.923730195315962e-05,
540
+ "loss": 0.0165,
541
+ "step": 1600
542
+ },
543
+ {
544
+ "epoch": 6,
545
+ "eval_loss": 0.01711028628051281,
546
+ "eval_mae": 0.09980826824903488,
547
+ "eval_mse": 0.01711735874414444,
548
+ "eval_r2": 0.23415986485617102,
549
+ "eval_runtime": 63.2895,
550
+ "eval_samples_per_second": 6.32,
551
+ "eval_steps_per_second": 3.16,
552
+ "step": 1600
553
+ },
554
+ {
555
+ "epoch": 6,
556
+ "learning_rate": 7.781099604447794e-05,
557
+ "loss": 0.0164,
558
+ "step": 1650
559
+ },
560
+ {
561
+ "epoch": 6,
562
+ "eval_loss": 0.017070267349481583,
563
+ "eval_mae": 0.09970243275165558,
564
+ "eval_mse": 0.01707725040614605,
565
+ "eval_r2": 0.23408428218933097,
566
+ "eval_runtime": 62.7311,
567
+ "eval_samples_per_second": 6.376,
568
+ "eval_steps_per_second": 3.188,
569
+ "step": 1650
570
+ },
571
+ {
572
+ "epoch": 7,
573
+ "learning_rate": 7.635113557867395e-05,
574
+ "loss": 0.0165,
575
+ "step": 1700
576
+ },
577
+ {
578
+ "epoch": 7,
579
+ "eval_loss": 0.017109189182519913,
580
+ "eval_mae": 0.09978970885276794,
581
+ "eval_mse": 0.017108654603362083,
582
+ "eval_r2": 0.23454243286934806,
583
+ "eval_runtime": 63.1248,
584
+ "eval_samples_per_second": 6.337,
585
+ "eval_steps_per_second": 3.168,
586
+ "step": 1700
587
+ },
588
+ {
589
+ "epoch": 7,
590
+ "learning_rate": 7.485948190839077e-05,
591
+ "loss": 0.0164,
592
+ "step": 1750
593
+ },
594
+ {
595
+ "epoch": 7,
596
+ "eval_loss": 0.01705513708293438,
597
+ "eval_mae": 0.09960421919822693,
598
+ "eval_mse": 0.017058243975043297,
599
+ "eval_r2": 0.2349071550024685,
600
+ "eval_runtime": 62.8704,
601
+ "eval_samples_per_second": 6.362,
602
+ "eval_steps_per_second": 3.181,
603
+ "step": 1750
604
+ },
605
+ {
606
+ "epoch": 7,
607
+ "learning_rate": 7.333783474544758e-05,
608
+ "loss": 0.0165,
609
+ "step": 1800
610
+ },
611
+ {
612
+ "epoch": 7,
613
+ "eval_loss": 0.01703445240855217,
614
+ "eval_mae": 0.09956526756286621,
615
+ "eval_mse": 0.017035936936736107,
616
+ "eval_r2": 0.23700014890629983,
617
+ "eval_runtime": 62.8756,
618
+ "eval_samples_per_second": 6.362,
619
+ "eval_steps_per_second": 3.181,
620
+ "step": 1800
621
+ },
622
+ {
623
+ "epoch": 7,
624
+ "learning_rate": 7.178802998944933e-05,
625
+ "loss": 0.0163,
626
+ "step": 1850
627
+ },
628
+ {
629
+ "epoch": 7,
630
+ "eval_loss": 0.017014818266034126,
631
+ "eval_mae": 0.09949322044849396,
632
+ "eval_mse": 0.01701194979250431,
633
+ "eval_r2": 0.23927528453479352,
634
+ "eval_runtime": 62.7426,
635
+ "eval_samples_per_second": 6.375,
636
+ "eval_steps_per_second": 3.188,
637
+ "step": 1850
638
+ },
639
+ {
640
+ "epoch": 7,
641
+ "learning_rate": 7.021193751273462e-05,
642
+ "loss": 0.0163,
643
+ "step": 1900
644
+ },
645
+ {
646
+ "epoch": 7,
647
+ "eval_loss": 0.016955234110355377,
648
+ "eval_mae": 0.09930498152971268,
649
+ "eval_mse": 0.01695319451391697,
650
+ "eval_r2": 0.2370158653005695,
651
+ "eval_runtime": 63.4621,
652
+ "eval_samples_per_second": 6.303,
653
+ "eval_steps_per_second": 3.151,
654
+ "step": 1900
655
+ },
656
+ {
657
+ "epoch": 8,
658
+ "learning_rate": 6.86114589043352e-05,
659
+ "loss": 0.0163,
660
+ "step": 1950
661
+ },
662
+ {
663
+ "epoch": 8,
664
+ "eval_loss": 0.016957027837634087,
665
+ "eval_mae": 0.09929080307483673,
666
+ "eval_mse": 0.016954666003584862,
667
+ "eval_r2": 0.23837789423341782,
668
+ "eval_runtime": 63.6096,
669
+ "eval_samples_per_second": 6.288,
670
+ "eval_steps_per_second": 3.144,
671
+ "step": 1950
672
+ },
673
+ {
674
+ "epoch": 8,
675
+ "learning_rate": 6.698852517566836e-05,
676
+ "loss": 0.0163,
677
+ "step": 2000
678
+ },
679
+ {
680
+ "epoch": 8,
681
+ "eval_loss": 0.016924967989325523,
682
+ "eval_mae": 0.09920791536569595,
683
+ "eval_mse": 0.016921618953347206,
684
+ "eval_r2": 0.23921405049323863,
685
+ "eval_runtime": 62.3232,
686
+ "eval_samples_per_second": 6.418,
687
+ "eval_steps_per_second": 3.209,
688
+ "step": 2000
689
+ },
690
+ {
691
+ "epoch": 8,
692
+ "learning_rate": 6.534509443073072e-05,
693
+ "loss": 0.0163,
694
+ "step": 2050
695
+ },
696
+ {
697
+ "epoch": 8,
698
+ "eval_loss": 0.016955628991127014,
699
+ "eval_mae": 0.09933258593082428,
700
+ "eval_mse": 0.01695682480931282,
701
+ "eval_r2": 0.24050878076217264,
702
+ "eval_runtime": 62.7525,
703
+ "eval_samples_per_second": 6.374,
704
+ "eval_steps_per_second": 3.187,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 8,
709
+ "learning_rate": 6.368314950360415e-05,
710
+ "loss": 0.0163,
711
+ "step": 2100
712
+ },
713
+ {
714
+ "epoch": 8,
715
+ "eval_loss": 0.016915885731577873,
716
+ "eval_mae": 0.09923145920038223,
717
+ "eval_mse": 0.01692046783864498,
718
+ "eval_r2": 0.24116580712888347,
719
+ "eval_runtime": 62.4103,
720
+ "eval_samples_per_second": 6.409,
721
+ "eval_steps_per_second": 3.205,
722
+ "step": 2100
723
+ },
724
+ {
725
+ "epoch": 9,
726
+ "learning_rate": 6.200469556612435e-05,
727
+ "loss": 0.0163,
728
+ "step": 2150
729
+ },
730
+ {
731
+ "epoch": 9,
732
+ "eval_loss": 0.016883673146367073,
733
+ "eval_mae": 0.09907432645559311,
734
+ "eval_mse": 0.016882291063666344,
735
+ "eval_r2": 0.23994247077801656,
736
+ "eval_runtime": 63.1914,
737
+ "eval_samples_per_second": 6.33,
738
+ "eval_steps_per_second": 3.165,
739
+ "step": 2150
740
+ },
741
+ {
742
+ "epoch": 9,
743
+ "learning_rate": 6.031175770859848e-05,
744
+ "loss": 0.0163,
745
+ "step": 2200
746
+ },
747
+ {
748
+ "epoch": 9,
749
+ "eval_loss": 0.01691693253815174,
750
+ "eval_mae": 0.09921909868717194,
751
+ "eval_mse": 0.01691514253616333,
752
+ "eval_r2": 0.24448073571582285,
753
+ "eval_runtime": 63.0869,
754
+ "eval_samples_per_second": 6.34,
755
+ "eval_steps_per_second": 3.17,
756
+ "step": 2200
757
+ },
758
+ {
759
+ "epoch": 9,
760
+ "learning_rate": 5.8606378496490735e-05,
761
+ "loss": 0.0163,
762
+ "step": 2250
763
+ },
764
+ {
765
+ "epoch": 9,
766
+ "eval_loss": 0.01691095158457756,
767
+ "eval_mae": 0.09917566925287247,
768
+ "eval_mse": 0.016911856830120087,
769
+ "eval_r2": 0.24367263560029062,
770
+ "eval_runtime": 62.6214,
771
+ "eval_samples_per_second": 6.388,
772
+ "eval_steps_per_second": 3.194,
773
+ "step": 2250
774
+ },
775
+ {
776
+ "epoch": 9,
777
+ "learning_rate": 5.6890615506023705e-05,
778
+ "loss": 0.0162,
779
+ "step": 2300
780
+ },
781
+ {
782
+ "epoch": 9,
783
+ "eval_loss": 0.016852255910634995,
784
+ "eval_mae": 0.09895263612270355,
785
+ "eval_mse": 0.016846586018800735,
786
+ "eval_r2": 0.24440525606317043,
787
+ "eval_runtime": 63.9638,
788
+ "eval_samples_per_second": 6.254,
789
+ "eval_steps_per_second": 3.127,
790
+ "step": 2300
791
+ },
792
+ {
793
+ "epoch": 9,
794
+ "learning_rate": 5.5166538841669025e-05,
795
+ "loss": 0.0161,
796
+ "step": 2350
797
+ },
798
+ {
799
+ "epoch": 9,
800
+ "eval_loss": 0.016831671819090843,
801
+ "eval_mae": 0.09894772619009018,
802
+ "eval_mse": 0.01683351770043373,
803
+ "eval_r2": 0.24440384235611556,
804
+ "eval_runtime": 63.0545,
805
+ "eval_samples_per_second": 6.344,
806
+ "eval_steps_per_second": 3.172,
807
+ "step": 2350
808
+ },
809
+ {
810
+ "epoch": 10,
811
+ "learning_rate": 5.343622863852232e-05,
812
+ "loss": 0.0163,
813
+ "step": 2400
814
+ },
815
+ {
816
+ "epoch": 10,
817
+ "eval_loss": 0.016827262938022614,
818
+ "eval_mae": 0.09892594069242477,
819
+ "eval_mse": 0.01683083176612854,
820
+ "eval_r2": 0.24909316086739675,
821
+ "eval_runtime": 62.71,
822
+ "eval_samples_per_second": 6.379,
823
+ "eval_steps_per_second": 3.189,
824
+ "step": 2400
825
+ },
826
+ {
827
+ "epoch": 10,
828
+ "learning_rate": 5.170177255257618e-05,
829
+ "loss": 0.0162,
830
+ "step": 2450
831
+ },
832
+ {
833
+ "epoch": 10,
834
+ "eval_loss": 0.016822684556245804,
835
+ "eval_mae": 0.09890511631965637,
836
+ "eval_mse": 0.016828058287501335,
837
+ "eval_r2": 0.24537013068417224,
838
+ "eval_runtime": 62.313,
839
+ "eval_samples_per_second": 6.419,
840
+ "eval_steps_per_second": 3.21,
841
+ "step": 2450
842
+ },
843
+ {
844
+ "epoch": 10,
845
+ "learning_rate": 4.996526324191872e-05,
846
+ "loss": 0.0161,
847
+ "step": 2500
848
+ },
849
+ {
850
+ "epoch": 10,
851
+ "eval_loss": 0.016806134954094887,
852
+ "eval_mae": 0.09885768592357635,
853
+ "eval_mse": 0.01680225133895874,
854
+ "eval_r2": 0.24621643781805225,
855
+ "eval_runtime": 62.9889,
856
+ "eval_samples_per_second": 6.35,
857
+ "eval_steps_per_second": 3.175,
858
+ "step": 2500
859
+ },
860
+ {
861
+ "epoch": 10,
862
+ "learning_rate": 4.822879584189731e-05,
863
+ "loss": 0.0162,
864
+ "step": 2550
865
+ },
866
+ {
867
+ "epoch": 10,
868
+ "eval_loss": 0.01680067367851734,
869
+ "eval_mae": 0.09885375201702118,
870
+ "eval_mse": 0.01680714637041092,
871
+ "eval_r2": 0.2480926793721172,
872
+ "eval_runtime": 62.8794,
873
+ "eval_samples_per_second": 6.361,
874
+ "eval_steps_per_second": 3.181,
875
+ "step": 2550
876
+ },
877
+ {
878
+ "epoch": 10,
879
+ "learning_rate": 4.6494465437293225e-05,
880
+ "loss": 0.0162,
881
+ "step": 2600
882
+ },
883
+ {
884
+ "epoch": 10,
885
+ "eval_loss": 0.01678573526442051,
886
+ "eval_mae": 0.09876307845115662,
887
+ "eval_mse": 0.01678406447172165,
888
+ "eval_r2": 0.24622674800267175,
889
+ "eval_runtime": 63.1145,
890
+ "eval_samples_per_second": 6.338,
891
+ "eval_steps_per_second": 3.169,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 11,
896
+ "learning_rate": 4.476436453455742e-05,
897
+ "loss": 0.0161,
898
+ "step": 2650
899
+ },
900
+ {
901
+ "epoch": 11,
902
+ "eval_loss": 0.016767781227827072,
903
+ "eval_mae": 0.09874103963375092,
904
+ "eval_mse": 0.01676834560930729,
905
+ "eval_r2": 0.24893675048568276,
906
+ "eval_runtime": 63.6881,
907
+ "eval_samples_per_second": 6.281,
908
+ "eval_steps_per_second": 3.14,
909
+ "step": 2650
910
+ },
911
+ {
912
+ "epoch": 11,
913
+ "learning_rate": 4.3040580537157024e-05,
914
+ "loss": 0.0161,
915
+ "step": 2700
916
+ },
917
+ {
918
+ "epoch": 11,
919
+ "eval_loss": 0.016766654327511787,
920
+ "eval_mae": 0.0987214520573616,
921
+ "eval_mse": 0.016764981672167778,
922
+ "eval_r2": 0.24921689372938038,
923
+ "eval_runtime": 62.2087,
924
+ "eval_samples_per_second": 6.43,
925
+ "eval_steps_per_second": 3.215,
926
+ "step": 2700
927
+ },
928
+ {
929
+ "epoch": 11,
930
+ "learning_rate": 4.1325193227078816e-05,
931
+ "loss": 0.0161,
932
+ "step": 2750
933
+ },
934
+ {
935
+ "epoch": 11,
936
+ "eval_loss": 0.01673816703259945,
937
+ "eval_mae": 0.09863594174385071,
938
+ "eval_mse": 0.01673576422035694,
939
+ "eval_r2": 0.24761288350255228,
940
+ "eval_runtime": 62.9694,
941
+ "eval_samples_per_second": 6.352,
942
+ "eval_steps_per_second": 3.176,
943
+ "step": 2750
944
+ },
945
+ {
946
+ "epoch": 11,
947
+ "learning_rate": 3.962027225552807e-05,
948
+ "loss": 0.0162,
949
+ "step": 2800
950
+ },
951
+ {
952
+ "epoch": 11,
953
+ "eval_loss": 0.016739826649427414,
954
+ "eval_mae": 0.09863518178462982,
955
+ "eval_mse": 0.016731785610318184,
956
+ "eval_r2": 0.24922910335616988,
957
+ "eval_runtime": 62.5788,
958
+ "eval_samples_per_second": 6.392,
959
+ "eval_steps_per_second": 3.196,
960
+ "step": 2800
961
+ },
962
+ {
963
+ "epoch": 11,
964
+ "learning_rate": 3.79278746458504e-05,
965
+ "loss": 0.0161,
966
+ "step": 2850
967
+ },
968
+ {
969
+ "epoch": 11,
970
+ "eval_loss": 0.016736237332224846,
971
+ "eval_mae": 0.09862860292196274,
972
+ "eval_mse": 0.016738129779696465,
973
+ "eval_r2": 0.24936630700011952,
974
+ "eval_runtime": 63.2169,
975
+ "eval_samples_per_second": 6.327,
976
+ "eval_steps_per_second": 3.164,
977
+ "step": 2850
978
+ },
979
+ {
980
+ "epoch": 12,
981
+ "learning_rate": 3.6250042311689505e-05,
982
+ "loss": 0.0161,
983
+ "step": 2900
984
+ },
985
+ {
986
+ "epoch": 12,
987
+ "eval_loss": 0.016720617190003395,
988
+ "eval_mae": 0.09856829792261124,
989
+ "eval_mse": 0.016721663996577263,
990
+ "eval_r2": 0.24943757367634767,
991
+ "eval_runtime": 62.9559,
992
+ "eval_samples_per_second": 6.354,
993
+ "eval_steps_per_second": 3.177,
994
+ "step": 2900
995
+ },
996
+ {
997
+ "epoch": 12,
998
+ "learning_rate": 3.458879959337494e-05,
999
+ "loss": 0.016,
1000
+ "step": 2950
1001
+ },
1002
+ {
1003
+ "epoch": 12,
1004
+ "eval_loss": 0.01672213524580002,
1005
+ "eval_mae": 0.0985676720738411,
1006
+ "eval_mse": 0.016720319166779518,
1007
+ "eval_r2": 0.24865355220765406,
1008
+ "eval_runtime": 63.4117,
1009
+ "eval_samples_per_second": 6.308,
1010
+ "eval_steps_per_second": 3.154,
1011
+ "step": 2950
1012
+ },
1013
+ {
1014
+ "epoch": 12,
1015
+ "learning_rate": 3.294615081551259e-05,
1016
+ "loss": 0.0161,
1017
+ "step": 3000
1018
+ },
1019
+ {
1020
+ "epoch": 12,
1021
+ "eval_loss": 0.016707362607121468,
1022
+ "eval_mae": 0.09850940108299255,
1023
+ "eval_mse": 0.016702940687537193,
1024
+ "eval_r2": 0.24840998553438287,
1025
+ "eval_runtime": 62.8197,
1026
+ "eval_samples_per_second": 6.367,
1027
+ "eval_steps_per_second": 3.184,
1028
+ "step": 3000
1029
+ },
1030
+ {
1031
+ "epoch": 12,
1032
+ "learning_rate": 3.132407786872442e-05,
1033
+ "loss": 0.0161,
1034
+ "step": 3050
1035
+ },
1036
+ {
1037
+ "epoch": 12,
1038
+ "eval_loss": 0.01671912521123886,
1039
+ "eval_mae": 0.09860337525606155,
1040
+ "eval_mse": 0.016718650236725807,
1041
+ "eval_r2": 0.2525614324396165,
1042
+ "eval_runtime": 62.3061,
1043
+ "eval_samples_per_second": 6.42,
1044
+ "eval_steps_per_second": 3.21,
1045
+ "step": 3050
1046
+ },
1047
+ {
1048
+ "epoch": 13,
1049
+ "learning_rate": 2.9724537818455466e-05,
1050
+ "loss": 0.016,
1051
+ "step": 3100
1052
+ },
1053
+ {
1054
+ "epoch": 13,
1055
+ "eval_loss": 0.016661079600453377,
1056
+ "eval_mae": 0.09843301773071289,
1057
+ "eval_mse": 0.016662681475281715,
1058
+ "eval_r2": 0.25208811381954643,
1059
+ "eval_runtime": 62.656,
1060
+ "eval_samples_per_second": 6.384,
1061
+ "eval_steps_per_second": 3.192,
1062
+ "step": 3100
1063
+ },
1064
+ {
1065
+ "epoch": 13,
1066
+ "learning_rate": 2.8149460543732664e-05,
1067
+ "loss": 0.0161,
1068
+ "step": 3150
1069
+ },
1070
+ {
1071
+ "epoch": 13,
1072
+ "eval_loss": 0.016676336526870728,
1073
+ "eval_mae": 0.09842842072248459,
1074
+ "eval_mse": 0.01667998544871807,
1075
+ "eval_r2": 0.25096189530992574,
1076
+ "eval_runtime": 61.8687,
1077
+ "eval_samples_per_second": 6.465,
1078
+ "eval_steps_per_second": 3.233,
1079
+ "step": 3150
1080
+ },
1081
+ {
1082
+ "epoch": 13,
1083
+ "learning_rate": 2.6600746408725063e-05,
1084
+ "loss": 0.0161,
1085
+ "step": 3200
1086
+ },
1087
+ {
1088
+ "epoch": 13,
1089
+ "eval_loss": 0.016732489690184593,
1090
+ "eval_mae": 0.09858258813619614,
1091
+ "eval_mse": 0.016734851524233818,
1092
+ "eval_r2": 0.25186837331344525,
1093
+ "eval_runtime": 62.9675,
1094
+ "eval_samples_per_second": 6.352,
1095
+ "eval_steps_per_second": 3.176,
1096
+ "step": 3200
1097
+ },
1098
+ {
1099
+ "epoch": 13,
1100
+ "learning_rate": 2.5080263969913897e-05,
1101
+ "loss": 0.016,
1102
+ "step": 3250
1103
+ },
1104
+ {
1105
+ "epoch": 13,
1106
+ "eval_loss": 0.016638994216918945,
1107
+ "eval_mae": 0.09835705161094666,
1108
+ "eval_mse": 0.016642747446894646,
1109
+ "eval_r2": 0.25120772331092656,
1110
+ "eval_runtime": 62.2252,
1111
+ "eval_samples_per_second": 6.428,
1112
+ "eval_steps_per_second": 3.214,
1113
+ "step": 3250
1114
+ },
1115
+ {
1116
+ "epoch": 13,
1117
+ "learning_rate": 2.3589847721639617e-05,
1118
+ "loss": 0.0161,
1119
+ "step": 3300
1120
+ },
1121
+ {
1122
+ "epoch": 13,
1123
+ "eval_loss": 0.01668979413807392,
1124
+ "eval_mae": 0.09847620874643326,
1125
+ "eval_mse": 0.016691412776708603,
1126
+ "eval_r2": 0.2516642585572433,
1127
+ "eval_runtime": 63.1219,
1128
+ "eval_samples_per_second": 6.337,
1129
+ "eval_steps_per_second": 3.168,
1130
+ "step": 3300
1131
+ },
1132
+ {
1133
+ "epoch": 14,
1134
+ "learning_rate": 2.2131295882745597e-05,
1135
+ "loss": 0.016,
1136
+ "step": 3350
1137
+ },
1138
+ {
1139
+ "epoch": 14,
1140
+ "eval_loss": 0.016662921756505966,
1141
+ "eval_mae": 0.09842050820589066,
1142
+ "eval_mse": 0.016670849174261093,
1143
+ "eval_r2": 0.2505160489332994,
1144
+ "eval_runtime": 62.776,
1145
+ "eval_samples_per_second": 6.372,
1146
+ "eval_steps_per_second": 3.186,
1147
+ "step": 3350
1148
+ },
1149
+ {
1150
+ "epoch": 14,
1151
+ "learning_rate": 2.070636822698877e-05,
1152
+ "loss": 0.0159,
1153
+ "step": 3400
1154
+ },
1155
+ {
1156
+ "epoch": 14,
1157
+ "eval_loss": 0.01665448024868965,
1158
+ "eval_mae": 0.09836740046739578,
1159
+ "eval_mse": 0.016657505184412003,
1160
+ "eval_r2": 0.25081546773511,
1161
+ "eval_runtime": 62.5067,
1162
+ "eval_samples_per_second": 6.399,
1163
+ "eval_steps_per_second": 3.2,
1164
+ "step": 3400
1165
+ },
1166
+ {
1167
+ "epoch": 14,
1168
+ "learning_rate": 1.9316783959835345e-05,
1169
+ "loss": 0.016,
1170
+ "step": 3450
1171
+ },
1172
+ {
1173
+ "epoch": 14,
1174
+ "eval_loss": 0.01669839210808277,
1175
+ "eval_mae": 0.09849409759044647,
1176
+ "eval_mse": 0.01669597439467907,
1177
+ "eval_r2": 0.25497888003639246,
1178
+ "eval_runtime": 63.4933,
1179
+ "eval_samples_per_second": 6.3,
1180
+ "eval_steps_per_second": 3.15,
1181
+ "step": 3450
1182
+ },
1183
+ {
1184
+ "epoch": 14,
1185
+ "learning_rate": 1.796421964420285e-05,
1186
+ "loss": 0.0161,
1187
+ "step": 3500
1188
+ },
1189
+ {
1190
+ "epoch": 14,
1191
+ "eval_loss": 0.01662326790392399,
1192
+ "eval_mae": 0.0982794538140297,
1193
+ "eval_mse": 0.016625171527266502,
1194
+ "eval_r2": 0.2520529598686879,
1195
+ "eval_runtime": 62.7103,
1196
+ "eval_samples_per_second": 6.379,
1197
+ "eval_steps_per_second": 3.189,
1198
+ "step": 3500
1199
+ },
1200
+ {
1201
+ "epoch": 14,
1202
+ "learning_rate": 1.665030717765149e-05,
1203
+ "loss": 0.016,
1204
+ "step": 3550
1205
+ },
1206
+ {
1207
+ "epoch": 14,
1208
+ "eval_loss": 0.016656357795000076,
1209
+ "eval_mae": 0.09835183620452881,
1210
+ "eval_mse": 0.016651729121804237,
1211
+ "eval_r2": 0.2521458867843698,
1212
+ "eval_runtime": 64.4578,
1213
+ "eval_samples_per_second": 6.206,
1214
+ "eval_steps_per_second": 3.103,
1215
+ "step": 3550
1216
+ },
1217
+ {
1218
+ "epoch": 15,
1219
+ "learning_rate": 1.5376631823464953e-05,
1220
+ "loss": 0.0161,
1221
+ "step": 3600
1222
+ },
1223
+ {
1224
+ "epoch": 15,
1225
+ "eval_loss": 0.01665300317108631,
1226
+ "eval_mae": 0.09836214780807495,
1227
+ "eval_mse": 0.01665370911359787,
1228
+ "eval_r2": 0.25267425736469795,
1229
+ "eval_runtime": 63.0441,
1230
+ "eval_samples_per_second": 6.345,
1231
+ "eval_steps_per_second": 3.172,
1232
+ "step": 3600
1233
+ },
1234
+ {
1235
+ "epoch": 15,
1236
+ "learning_rate": 1.4144730297996666e-05,
1237
+ "loss": 0.016,
1238
+ "step": 3650
1239
+ },
1240
+ {
1241
+ "epoch": 15,
1242
+ "eval_loss": 0.01662749983370304,
1243
+ "eval_mae": 0.09827445447444916,
1244
+ "eval_mse": 0.016627401113510132,
1245
+ "eval_r2": 0.2526889218862315,
1246
+ "eval_runtime": 62.6614,
1247
+ "eval_samples_per_second": 6.384,
1248
+ "eval_steps_per_second": 3.192,
1249
+ "step": 3650
1250
+ },
1251
+ {
1252
+ "epoch": 15,
1253
+ "learning_rate": 1.295608891658896e-05,
1254
+ "loss": 0.016,
1255
+ "step": 3700
1256
+ },
1257
+ {
1258
+ "epoch": 15,
1259
+ "eval_loss": 0.016592269763350487,
1260
+ "eval_mae": 0.09820396453142166,
1261
+ "eval_mse": 0.0165996253490448,
1262
+ "eval_r2": 0.25472996642841683,
1263
+ "eval_runtime": 63.4328,
1264
+ "eval_samples_per_second": 6.306,
1265
+ "eval_steps_per_second": 3.153,
1266
+ "step": 3700
1267
+ },
1268
+ {
1269
+ "epoch": 15,
1270
+ "learning_rate": 1.1812141800301945e-05,
1271
+ "loss": 0.016,
1272
+ "step": 3750
1273
+ },
1274
+ {
1275
+ "epoch": 15,
1276
+ "eval_loss": 0.016636423766613007,
1277
+ "eval_mae": 0.0983065739274025,
1278
+ "eval_mse": 0.016633499413728714,
1279
+ "eval_r2": 0.25374046203309764,
1280
+ "eval_runtime": 62.8839,
1281
+ "eval_samples_per_second": 6.361,
1282
+ "eval_steps_per_second": 3.18,
1283
+ "step": 3750
1284
+ },
1285
+ {
1286
+ "epoch": 15,
1287
+ "learning_rate": 1.0714269145616063e-05,
1288
+ "loss": 0.0159,
1289
+ "step": 3800
1290
+ },
1291
+ {
1292
+ "epoch": 15,
1293
+ "eval_loss": 0.016631022095680237,
1294
+ "eval_mae": 0.09828473627567291,
1295
+ "eval_mse": 0.01662875898182392,
1296
+ "eval_r2": 0.25143078561422905,
1297
+ "eval_runtime": 62.4886,
1298
+ "eval_samples_per_second": 6.401,
1299
+ "eval_steps_per_second": 3.201,
1300
+ "step": 3800
1301
+ },
1302
+ {
1303
+ "epoch": 16,
1304
+ "learning_rate": 9.663795559195733e-06,
1305
+ "loss": 0.0159,
1306
+ "step": 3850
1307
+ },
1308
+ {
1309
+ "epoch": 16,
1310
+ "eval_loss": 0.016640575602650642,
1311
+ "eval_mae": 0.09832051396369934,
1312
+ "eval_mse": 0.016639817506074905,
1313
+ "eval_r2": 0.25420519067693936,
1314
+ "eval_runtime": 63.7918,
1315
+ "eval_samples_per_second": 6.27,
1316
+ "eval_steps_per_second": 3.135,
1317
+ "step": 3850
1318
+ },
1319
+ {
1320
+ "epoch": 16,
1321
+ "learning_rate": 8.661988459723309e-06,
1322
+ "loss": 0.016,
1323
+ "step": 3900
1324
+ },
1325
+ {
1326
+ "epoch": 16,
1327
+ "eval_loss": 0.016633687540888786,
1328
+ "eval_mae": 0.09830807894468307,
1329
+ "eval_mse": 0.0166340135037899,
1330
+ "eval_r2": 0.2559419583101087,
1331
+ "eval_runtime": 64.9995,
1332
+ "eval_samples_per_second": 6.154,
1333
+ "eval_steps_per_second": 3.077,
1334
+ "step": 3900
1335
+ },
1336
+ {
1337
+ "epoch": 16,
1338
+ "learning_rate": 7.710056548731447e-06,
1339
+ "loss": 0.016,
1340
+ "step": 3950
1341
+ },
1342
+ {
1343
+ "epoch": 16,
1344
+ "eval_loss": 0.01665404625236988,
1345
+ "eval_mae": 0.09835316985845566,
1346
+ "eval_mse": 0.01665414310991764,
1347
+ "eval_r2": 0.25389154611800946,
1348
+ "eval_runtime": 63.3284,
1349
+ "eval_samples_per_second": 6.316,
1350
+ "eval_steps_per_second": 3.158,
1351
+ "step": 3950
1352
+ },
1353
+ {
1354
+ "epoch": 16,
1355
+ "learning_rate": 6.809148352279182e-06,
1356
+ "loss": 0.016,
1357
+ "step": 4000
1358
+ },
1359
+ {
1360
+ "epoch": 16,
1361
+ "eval_loss": 0.016643621027469635,
1362
+ "eval_mae": 0.09831266850233078,
1363
+ "eval_mse": 0.0166462492197752,
1364
+ "eval_r2": 0.253610910132235,
1365
+ "eval_runtime": 62.8474,
1366
+ "eval_samples_per_second": 6.365,
1367
+ "eval_steps_per_second": 3.182,
1368
+ "step": 4000
1369
+ },
1370
+ {
1371
+ "epoch": 16,
1372
+ "learning_rate": 5.960350835230766e-06,
1373
+ "loss": 0.0159,
1374
+ "step": 4050
1375
+ },
1376
+ {
1377
+ "epoch": 16,
1378
+ "eval_loss": 0.016641885042190552,
1379
+ "eval_mae": 0.09831728786230087,
1380
+ "eval_mse": 0.01664014533162117,
1381
+ "eval_r2": 0.2530172455633002,
1382
+ "eval_runtime": 62.9766,
1383
+ "eval_samples_per_second": 6.352,
1384
+ "eval_steps_per_second": 3.176,
1385
+ "step": 4050
1386
+ },
1387
+ {
1388
+ "epoch": 17,
1389
+ "learning_rate": 5.164688089809444e-06,
1390
+ "loss": 0.0159,
1391
+ "step": 4100
1392
+ },
1393
+ {
1394
+ "epoch": 17,
1395
+ "eval_loss": 0.016620321199297905,
1396
+ "eval_mae": 0.09825479984283447,
1397
+ "eval_mse": 0.016620123758912086,
1398
+ "eval_r2": 0.2543216647578763,
1399
+ "eval_runtime": 63.7821,
1400
+ "eval_samples_per_second": 6.271,
1401
+ "eval_steps_per_second": 3.136,
1402
+ "step": 4100
1403
+ },
1404
+ {
1405
+ "epoch": 17,
1406
+ "learning_rate": 4.423120100008582e-06,
1407
+ "loss": 0.016,
1408
+ "step": 4150
1409
+ },
1410
+ {
1411
+ "epoch": 17,
1412
+ "eval_loss": 0.016626114025712013,
1413
+ "eval_mae": 0.0982520654797554,
1414
+ "eval_mse": 0.01661662384867668,
1415
+ "eval_r2": 0.25688829356860643,
1416
+ "eval_runtime": 63.3768,
1417
+ "eval_samples_per_second": 6.311,
1418
+ "eval_steps_per_second": 3.156,
1419
+ "step": 4150
1420
+ },
1421
+ {
1422
+ "epoch": 17,
1423
+ "learning_rate": 3.7365415833504725e-06,
1424
+ "loss": 0.0159,
1425
+ "step": 4200
1426
+ },
1427
+ {
1428
+ "epoch": 17,
1429
+ "eval_loss": 0.016621023416519165,
1430
+ "eval_mae": 0.0982469841837883,
1431
+ "eval_mse": 0.01662403903901577,
1432
+ "eval_r2": 0.2562592876194061,
1433
+ "eval_runtime": 63.3616,
1434
+ "eval_samples_per_second": 6.313,
1435
+ "eval_steps_per_second": 3.156,
1436
+ "step": 4200
1437
+ },
1438
+ {
1439
+ "epoch": 17,
1440
+ "learning_rate": 3.105780911390738e-06,
1441
+ "loss": 0.016,
1442
+ "step": 4250
1443
+ },
1444
+ {
1445
+ "epoch": 17,
1446
+ "eval_loss": 0.01661110669374466,
1447
+ "eval_mae": 0.09821216762065887,
1448
+ "eval_mse": 0.016603710129857063,
1449
+ "eval_r2": 0.2534045631002596,
1450
+ "eval_runtime": 63.6674,
1451
+ "eval_samples_per_second": 6.283,
1452
+ "eval_steps_per_second": 3.141,
1453
+ "step": 4250
1454
+ },
1455
+ {
1456
+ "epoch": 18,
1457
+ "learning_rate": 2.5315991102703716e-06,
1458
+ "loss": 0.016,
1459
+ "step": 4300
1460
+ },
1461
+ {
1462
+ "epoch": 18,
1463
+ "eval_loss": 0.016594773158431053,
1464
+ "eval_mae": 0.09816820174455643,
1465
+ "eval_mse": 0.016600053757429123,
1466
+ "eval_r2": 0.25425809369382535,
1467
+ "eval_runtime": 63.3127,
1468
+ "eval_samples_per_second": 6.318,
1469
+ "eval_steps_per_second": 3.159,
1470
+ "step": 4300
1471
+ },
1472
+ {
1473
+ "epoch": 18,
1474
+ "learning_rate": 2.0146889425216476e-06,
1475
+ "loss": 0.016,
1476
+ "step": 4350
1477
+ },
1478
+ {
1479
+ "epoch": 18,
1480
+ "eval_loss": 0.016598813235759735,
1481
+ "eval_mae": 0.09816797077655792,
1482
+ "eval_mse": 0.016591567546129227,
1483
+ "eval_r2": 0.2551125072294508,
1484
+ "eval_runtime": 61.9837,
1485
+ "eval_samples_per_second": 6.453,
1486
+ "eval_steps_per_second": 3.227,
1487
+ "step": 4350
1488
+ },
1489
+ {
1490
+ "epoch": 18,
1491
+ "learning_rate": 1.555674071235358e-06,
1492
+ "loss": 0.0159,
1493
+ "step": 4400
1494
+ },
1495
+ {
1496
+ "epoch": 18,
1497
+ "eval_loss": 0.016573524102568626,
1498
+ "eval_mae": 0.09814300388097763,
1499
+ "eval_mse": 0.016571756452322006,
1500
+ "eval_r2": 0.25463921169715287,
1501
+ "eval_runtime": 63.1303,
1502
+ "eval_samples_per_second": 6.336,
1503
+ "eval_steps_per_second": 3.168,
1504
+ "step": 4400
1505
+ },
1506
+ {
1507
+ "epoch": 18,
1508
+ "learning_rate": 1.155108307598024e-06,
1509
+ "loss": 0.016,
1510
+ "step": 4450
1511
+ },
1512
+ {
1513
+ "epoch": 18,
1514
+ "eval_loss": 0.016580162569880486,
1515
+ "eval_mae": 0.09815159440040588,
1516
+ "eval_mse": 0.01658167876303196,
1517
+ "eval_r2": 0.25566179703610914,
1518
+ "eval_runtime": 63.3862,
1519
+ "eval_samples_per_second": 6.311,
1520
+ "eval_steps_per_second": 3.155,
1521
+ "step": 4450
1522
+ },
1523
+ {
1524
+ "epoch": 18,
1525
+ "learning_rate": 8.134749427070376e-07,
1526
+ "loss": 0.016,
1527
+ "step": 4500
1528
+ },
1529
+ {
1530
+ "epoch": 18,
1531
+ "eval_loss": 0.016592783853411674,
1532
+ "eval_mae": 0.0981706902384758,
1533
+ "eval_mse": 0.016597216948866844,
1534
+ "eval_r2": 0.25482598937850753,
1535
+ "eval_runtime": 63.3827,
1536
+ "eval_samples_per_second": 6.311,
1537
+ "eval_steps_per_second": 3.155,
1538
+ "step": 4500
1539
+ },
1540
+ {
1541
+ "epoch": 19,
1542
+ "learning_rate": 5.311861644696048e-07,
1543
+ "loss": 0.0159,
1544
+ "step": 4550
1545
+ },
1546
+ {
1547
+ "epoch": 19,
1548
+ "eval_loss": 0.01663564145565033,
1549
+ "eval_mae": 0.09826894104480743,
1550
+ "eval_mse": 0.0166276004165411,
1551
+ "eval_r2": 0.2547617221900892,
1552
+ "eval_runtime": 62.9726,
1553
+ "eval_samples_per_second": 6.352,
1554
+ "eval_steps_per_second": 3.176,
1555
+ "step": 4550
1556
+ },
1557
+ {
1558
+ "epoch": 19,
1559
+ "learning_rate": 3.0858256028932776e-07,
1560
+ "loss": 0.016,
1561
+ "step": 4600
1562
+ },
1563
+ {
1564
+ "epoch": 19,
1565
+ "eval_loss": 0.016616160050034523,
1566
+ "eval_mae": 0.09824874997138977,
1567
+ "eval_mse": 0.016616739332675934,
1568
+ "eval_r2": 0.2550670819722727,
1569
+ "eval_runtime": 62.541,
1570
+ "eval_samples_per_second": 6.396,
1571
+ "eval_steps_per_second": 3.198,
1572
+ "step": 4600
1573
+ }
1574
+ ],
1575
+ "max_steps": 4760,
1576
+ "num_train_epochs": 20,
1577
+ "total_flos": 1.1833548390119886e+21,
1578
+ "trial_name": null,
1579
+ "trial_params": null
1580
+ }
vitmae_650M/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dabe900a5dac07f1090e31642262ba5bfd2e2d8732a1bd9d8dc5569c5c5ad991
3
+ size 4600