rshrott commited on
Commit
ec1fd10
1 Parent(s): 8bfa53c

🍻 cheers

Browse files
README.md CHANGED
@@ -2,6 +2,7 @@
2
  license: apache-2.0
3
  base_model: google/vit-base-patch16-224-in21k
4
  tags:
 
5
  - generated_from_trainer
6
  model-index:
7
  - name: ryan_model3272024
@@ -13,12 +14,12 @@ should probably proofread and complete it, then remove this comment. -->
13
 
14
  # ryan_model3272024
15
 
16
- This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the None dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 1.2907
19
- - Na Accuracy: 0.938
20
- - Ordinal Accuracy: 0.604
21
- - Ordinal Mae: 45.8334
22
 
23
  ## Model description
24
 
 
2
  license: apache-2.0
3
  base_model: google/vit-base-patch16-224-in21k
4
  tags:
5
+ - image-classification
6
  - generated_from_trainer
7
  model-index:
8
  - name: ryan_model3272024
 
14
 
15
  # ryan_model3272024
16
 
17
+ This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the beans dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 1.0327
20
+ - Na Accuracy: 0.956
21
+ - Ordinal Accuracy: 0.568
22
+ - Ordinal Mae: 56.3484
23
 
24
  ## Model description
25
 
all_results.json CHANGED
@@ -1,14 +1,14 @@
1
  {
2
- "epoch": 4.0,
3
- "eval_loss": 1.8227792978286743,
4
- "eval_na_accuracy": 0.8,
5
- "eval_ordinal_accuracy": 0.5,
6
- "eval_ordinal_mae": 8.665261681005358,
7
- "eval_runtime": 1.3319,
8
- "eval_samples_per_second": 7.508,
9
- "eval_steps_per_second": 1.502,
10
- "train_loss": 1.6283700466156006,
11
- "train_runtime": 32.1209,
12
- "train_samples_per_second": 12.453,
13
- "train_steps_per_second": 0.872
14
  }
 
1
  {
2
+ "epoch": 3.83,
3
+ "eval_loss": 1.0327022075653076,
4
+ "eval_na_accuracy": 0.956,
5
+ "eval_ordinal_accuracy": 0.568,
6
+ "eval_ordinal_mae": 56.34840202213265,
7
+ "eval_runtime": 20.4149,
8
+ "eval_samples_per_second": 24.492,
9
+ "eval_steps_per_second": 3.086,
10
+ "train_loss": 0.6305676238735517,
11
+ "train_runtime": 1636.5913,
12
+ "train_samples_per_second": 12.221,
13
+ "train_steps_per_second": 0.765
14
  }
eval_results.json CHANGED
@@ -1,10 +1,10 @@
1
  {
2
- "epoch": 4.0,
3
- "eval_loss": 1.8227792978286743,
4
- "eval_na_accuracy": 0.8,
5
- "eval_ordinal_accuracy": 0.5,
6
- "eval_ordinal_mae": 8.665261681005358,
7
- "eval_runtime": 1.3319,
8
- "eval_samples_per_second": 7.508,
9
- "eval_steps_per_second": 1.502
10
  }
 
1
  {
2
+ "epoch": 3.83,
3
+ "eval_loss": 1.0327022075653076,
4
+ "eval_na_accuracy": 0.956,
5
+ "eval_ordinal_accuracy": 0.568,
6
+ "eval_ordinal_mae": 56.34840202213265,
7
+ "eval_runtime": 20.4149,
8
+ "eval_samples_per_second": 24.492,
9
+ "eval_steps_per_second": 3.086
10
  }
runs/Mar27_15-52-23_ryanserver/events.out.tfevents.1711570860.ryanserver.25227.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5255e5c83d495937a64e35fc9ee5b6a6cf7c8721a14a86cf099a7c3429ad11b
3
+ size 529
train_results.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "epoch": 4.0,
3
- "train_loss": 1.6283700466156006,
4
- "train_runtime": 32.1209,
5
- "train_samples_per_second": 12.453,
6
- "train_steps_per_second": 0.872
7
  }
 
1
  {
2
+ "epoch": 3.83,
3
+ "train_loss": 0.6305676238735517,
4
+ "train_runtime": 1636.5913,
5
+ "train_samples_per_second": 12.221,
6
+ "train_steps_per_second": 0.765
7
  }
trainer_state.json CHANGED
@@ -1,43 +1,1001 @@
1
  {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 4.0,
5
  "eval_steps": 100,
6
- "global_step": 28,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 1.43,
13
- "grad_norm": 1.993449091911316,
14
- "learning_rate": 0.00012857142857142858,
15
- "loss": 1.9695,
16
  "step": 10
17
  },
18
  {
19
- "epoch": 2.86,
20
- "grad_norm": 1.4520915746688843,
21
- "learning_rate": 5.714285714285714e-05,
22
- "loss": 1.524,
23
  "step": 20
24
  },
25
  {
26
- "epoch": 4.0,
27
- "step": 28,
28
- "total_flos": 3.0997907103744e+16,
29
- "train_loss": 1.6283700466156006,
30
- "train_runtime": 32.1209,
31
- "train_samples_per_second": 12.453,
32
- "train_steps_per_second": 0.872
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  }
34
  ],
35
  "logging_steps": 10,
36
- "max_steps": 28,
37
  "num_input_tokens_seen": 0,
38
  "num_train_epochs": 4,
39
  "save_steps": 100,
40
- "total_flos": 3.0997907103744e+16,
41
  "train_batch_size": 16,
42
  "trial_name": null,
43
  "trial_params": null
 
1
  {
2
+ "best_metric": 1.0327022075653076,
3
+ "best_model_checkpoint": "./ryan_model3272024/checkpoint-500",
4
+ "epoch": 3.8338658146964857,
5
  "eval_steps": 100,
6
+ "global_step": 1200,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.03,
13
+ "grad_norm": 2.4977924823760986,
14
+ "learning_rate": 0.00019840255591054313,
15
+ "loss": 1.9988,
16
  "step": 10
17
  },
18
  {
19
+ "epoch": 0.06,
20
+ "grad_norm": 2.3356950283050537,
21
+ "learning_rate": 0.00019680511182108628,
22
+ "loss": 1.6981,
23
  "step": 20
24
  },
25
  {
26
+ "epoch": 0.1,
27
+ "grad_norm": 2.1436116695404053,
28
+ "learning_rate": 0.0001952076677316294,
29
+ "loss": 1.6585,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.13,
34
+ "grad_norm": 1.9430339336395264,
35
+ "learning_rate": 0.00019361022364217253,
36
+ "loss": 1.6929,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.16,
41
+ "grad_norm": 1.7540597915649414,
42
+ "learning_rate": 0.00019201277955271565,
43
+ "loss": 1.4822,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.19,
48
+ "grad_norm": 2.9073445796966553,
49
+ "learning_rate": 0.0001904153354632588,
50
+ "loss": 1.5968,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.22,
55
+ "grad_norm": 1.6934313774108887,
56
+ "learning_rate": 0.00018881789137380192,
57
+ "loss": 1.4684,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.26,
62
+ "grad_norm": 3.451613426208496,
63
+ "learning_rate": 0.00018722044728434505,
64
+ "loss": 1.4425,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.29,
69
+ "grad_norm": 1.747863531112671,
70
+ "learning_rate": 0.0001856230031948882,
71
+ "loss": 1.3641,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.32,
76
+ "grad_norm": 2.263026714324951,
77
+ "learning_rate": 0.00018402555910543132,
78
+ "loss": 1.3578,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.32,
83
+ "eval_loss": 1.2057933807373047,
84
+ "eval_na_accuracy": 0.944,
85
+ "eval_ordinal_accuracy": 0.54,
86
+ "eval_ordinal_mae": 90.9845891143661,
87
+ "eval_runtime": 46.445,
88
+ "eval_samples_per_second": 10.765,
89
+ "eval_steps_per_second": 1.356,
90
+ "step": 100
91
+ },
92
+ {
93
+ "epoch": 0.35,
94
+ "grad_norm": 2.7085518836975098,
95
+ "learning_rate": 0.00018242811501597444,
96
+ "loss": 1.3463,
97
+ "step": 110
98
+ },
99
+ {
100
+ "epoch": 0.38,
101
+ "grad_norm": 2.2408878803253174,
102
+ "learning_rate": 0.00018083067092651756,
103
+ "loss": 1.1704,
104
+ "step": 120
105
+ },
106
+ {
107
+ "epoch": 0.42,
108
+ "grad_norm": 2.1588001251220703,
109
+ "learning_rate": 0.00017923322683706071,
110
+ "loss": 1.24,
111
+ "step": 130
112
+ },
113
+ {
114
+ "epoch": 0.45,
115
+ "grad_norm": 2.582127094268799,
116
+ "learning_rate": 0.00017763578274760384,
117
+ "loss": 1.2735,
118
+ "step": 140
119
+ },
120
+ {
121
+ "epoch": 0.48,
122
+ "grad_norm": 1.9934537410736084,
123
+ "learning_rate": 0.000176038338658147,
124
+ "loss": 1.344,
125
+ "step": 150
126
+ },
127
+ {
128
+ "epoch": 0.51,
129
+ "grad_norm": 2.1636180877685547,
130
+ "learning_rate": 0.0001744408945686901,
131
+ "loss": 1.1532,
132
+ "step": 160
133
+ },
134
+ {
135
+ "epoch": 0.54,
136
+ "grad_norm": 3.9455337524414062,
137
+ "learning_rate": 0.00017284345047923323,
138
+ "loss": 1.1231,
139
+ "step": 170
140
+ },
141
+ {
142
+ "epoch": 0.58,
143
+ "grad_norm": 3.46232271194458,
144
+ "learning_rate": 0.00017124600638977638,
145
+ "loss": 1.0945,
146
+ "step": 180
147
+ },
148
+ {
149
+ "epoch": 0.61,
150
+ "grad_norm": 2.913996696472168,
151
+ "learning_rate": 0.00016964856230031948,
152
+ "loss": 1.3168,
153
+ "step": 190
154
+ },
155
+ {
156
+ "epoch": 0.64,
157
+ "grad_norm": 2.16892671585083,
158
+ "learning_rate": 0.00016805111821086263,
159
+ "loss": 1.089,
160
+ "step": 200
161
+ },
162
+ {
163
+ "epoch": 0.64,
164
+ "eval_loss": 1.0986785888671875,
165
+ "eval_na_accuracy": 0.95,
166
+ "eval_ordinal_accuracy": 0.548,
167
+ "eval_ordinal_mae": 120.70968111835792,
168
+ "eval_runtime": 20.2841,
169
+ "eval_samples_per_second": 24.65,
170
+ "eval_steps_per_second": 3.106,
171
+ "step": 200
172
+ },
173
+ {
174
+ "epoch": 0.67,
175
+ "grad_norm": 2.9793784618377686,
176
+ "learning_rate": 0.00016645367412140575,
177
+ "loss": 1.1788,
178
+ "step": 210
179
+ },
180
+ {
181
+ "epoch": 0.7,
182
+ "grad_norm": 2.662822723388672,
183
+ "learning_rate": 0.0001648562300319489,
184
+ "loss": 1.0067,
185
+ "step": 220
186
+ },
187
+ {
188
+ "epoch": 0.73,
189
+ "grad_norm": 2.3921396732330322,
190
+ "learning_rate": 0.00016325878594249202,
191
+ "loss": 1.2542,
192
+ "step": 230
193
+ },
194
+ {
195
+ "epoch": 0.77,
196
+ "grad_norm": 2.670137882232666,
197
+ "learning_rate": 0.00016166134185303515,
198
+ "loss": 1.1284,
199
+ "step": 240
200
+ },
201
+ {
202
+ "epoch": 0.8,
203
+ "grad_norm": 3.212069511413574,
204
+ "learning_rate": 0.0001600638977635783,
205
+ "loss": 1.3407,
206
+ "step": 250
207
+ },
208
+ {
209
+ "epoch": 0.83,
210
+ "grad_norm": 2.284740447998047,
211
+ "learning_rate": 0.00015846645367412142,
212
+ "loss": 1.1876,
213
+ "step": 260
214
+ },
215
+ {
216
+ "epoch": 0.86,
217
+ "grad_norm": 2.98988938331604,
218
+ "learning_rate": 0.00015686900958466454,
219
+ "loss": 1.0993,
220
+ "step": 270
221
+ },
222
+ {
223
+ "epoch": 0.89,
224
+ "grad_norm": 3.687721014022827,
225
+ "learning_rate": 0.00015527156549520767,
226
+ "loss": 1.1273,
227
+ "step": 280
228
+ },
229
+ {
230
+ "epoch": 0.93,
231
+ "grad_norm": 2.0068211555480957,
232
+ "learning_rate": 0.00015367412140575082,
233
+ "loss": 1.0204,
234
+ "step": 290
235
+ },
236
+ {
237
+ "epoch": 0.96,
238
+ "grad_norm": 2.719209909439087,
239
+ "learning_rate": 0.00015207667731629394,
240
+ "loss": 0.924,
241
+ "step": 300
242
+ },
243
+ {
244
+ "epoch": 0.96,
245
+ "eval_loss": 1.083807110786438,
246
+ "eval_na_accuracy": 0.946,
247
+ "eval_ordinal_accuracy": 0.568,
248
+ "eval_ordinal_mae": 76.79816162186209,
249
+ "eval_runtime": 20.3196,
250
+ "eval_samples_per_second": 24.607,
251
+ "eval_steps_per_second": 3.1,
252
+ "step": 300
253
+ },
254
+ {
255
+ "epoch": 0.99,
256
+ "grad_norm": 6.174163341522217,
257
+ "learning_rate": 0.00015063897763578277,
258
+ "loss": 1.0079,
259
+ "step": 310
260
+ },
261
+ {
262
+ "epoch": 1.02,
263
+ "grad_norm": 1.8171716928482056,
264
+ "learning_rate": 0.0001490415335463259,
265
+ "loss": 0.9573,
266
+ "step": 320
267
+ },
268
+ {
269
+ "epoch": 1.05,
270
+ "grad_norm": 1.7882392406463623,
271
+ "learning_rate": 0.00014744408945686902,
272
+ "loss": 0.8008,
273
+ "step": 330
274
+ },
275
+ {
276
+ "epoch": 1.09,
277
+ "grad_norm": 2.8711063861846924,
278
+ "learning_rate": 0.00014584664536741214,
279
+ "loss": 0.7867,
280
+ "step": 340
281
+ },
282
+ {
283
+ "epoch": 1.12,
284
+ "grad_norm": 3.8461153507232666,
285
+ "learning_rate": 0.00014424920127795526,
286
+ "loss": 0.7375,
287
+ "step": 350
288
+ },
289
+ {
290
+ "epoch": 1.15,
291
+ "grad_norm": 3.599330186843872,
292
+ "learning_rate": 0.0001426517571884984,
293
+ "loss": 0.8057,
294
+ "step": 360
295
+ },
296
+ {
297
+ "epoch": 1.18,
298
+ "grad_norm": 1.9864487648010254,
299
+ "learning_rate": 0.00014105431309904153,
300
+ "loss": 0.7876,
301
+ "step": 370
302
+ },
303
+ {
304
+ "epoch": 1.21,
305
+ "grad_norm": 3.377739191055298,
306
+ "learning_rate": 0.00013945686900958468,
307
+ "loss": 0.7835,
308
+ "step": 380
309
+ },
310
+ {
311
+ "epoch": 1.25,
312
+ "grad_norm": 3.5138072967529297,
313
+ "learning_rate": 0.0001378594249201278,
314
+ "loss": 0.8265,
315
+ "step": 390
316
+ },
317
+ {
318
+ "epoch": 1.28,
319
+ "grad_norm": 3.5262341499328613,
320
+ "learning_rate": 0.00013626198083067093,
321
+ "loss": 0.694,
322
+ "step": 400
323
+ },
324
+ {
325
+ "epoch": 1.28,
326
+ "eval_loss": 1.068002700805664,
327
+ "eval_na_accuracy": 0.942,
328
+ "eval_ordinal_accuracy": 0.556,
329
+ "eval_ordinal_mae": 105.63124973691814,
330
+ "eval_runtime": 19.9749,
331
+ "eval_samples_per_second": 25.031,
332
+ "eval_steps_per_second": 3.154,
333
+ "step": 400
334
+ },
335
+ {
336
+ "epoch": 1.31,
337
+ "grad_norm": 2.814100980758667,
338
+ "learning_rate": 0.00013466453674121405,
339
+ "loss": 0.7246,
340
+ "step": 410
341
+ },
342
+ {
343
+ "epoch": 1.34,
344
+ "grad_norm": 1.9095064401626587,
345
+ "learning_rate": 0.00013306709265175718,
346
+ "loss": 0.768,
347
+ "step": 420
348
+ },
349
+ {
350
+ "epoch": 1.37,
351
+ "grad_norm": 2.849485158920288,
352
+ "learning_rate": 0.00013146964856230033,
353
+ "loss": 0.6669,
354
+ "step": 430
355
+ },
356
+ {
357
+ "epoch": 1.41,
358
+ "grad_norm": 2.2935235500335693,
359
+ "learning_rate": 0.00012987220447284345,
360
+ "loss": 0.6708,
361
+ "step": 440
362
+ },
363
+ {
364
+ "epoch": 1.44,
365
+ "grad_norm": 2.6063146591186523,
366
+ "learning_rate": 0.0001282747603833866,
367
+ "loss": 0.7481,
368
+ "step": 450
369
+ },
370
+ {
371
+ "epoch": 1.47,
372
+ "grad_norm": 2.329845428466797,
373
+ "learning_rate": 0.00012667731629392972,
374
+ "loss": 0.7341,
375
+ "step": 460
376
+ },
377
+ {
378
+ "epoch": 1.5,
379
+ "grad_norm": 4.184568881988525,
380
+ "learning_rate": 0.00012507987220447287,
381
+ "loss": 0.785,
382
+ "step": 470
383
+ },
384
+ {
385
+ "epoch": 1.53,
386
+ "grad_norm": 2.142719268798828,
387
+ "learning_rate": 0.00012348242811501597,
388
+ "loss": 0.702,
389
+ "step": 480
390
+ },
391
+ {
392
+ "epoch": 1.57,
393
+ "grad_norm": 2.7038168907165527,
394
+ "learning_rate": 0.0001218849840255591,
395
+ "loss": 0.8301,
396
+ "step": 490
397
+ },
398
+ {
399
+ "epoch": 1.6,
400
+ "grad_norm": 3.093721389770508,
401
+ "learning_rate": 0.00012028753993610224,
402
+ "loss": 0.7739,
403
+ "step": 500
404
+ },
405
+ {
406
+ "epoch": 1.6,
407
+ "eval_loss": 1.0327022075653076,
408
+ "eval_na_accuracy": 0.956,
409
+ "eval_ordinal_accuracy": 0.568,
410
+ "eval_ordinal_mae": 56.34840202213265,
411
+ "eval_runtime": 20.2051,
412
+ "eval_samples_per_second": 24.746,
413
+ "eval_steps_per_second": 3.118,
414
+ "step": 500
415
+ },
416
+ {
417
+ "epoch": 1.63,
418
+ "grad_norm": 3.727419853210449,
419
+ "learning_rate": 0.00011869009584664536,
420
+ "loss": 0.7536,
421
+ "step": 510
422
+ },
423
+ {
424
+ "epoch": 1.66,
425
+ "grad_norm": 3.3148770332336426,
426
+ "learning_rate": 0.00011709265175718851,
427
+ "loss": 0.7675,
428
+ "step": 520
429
+ },
430
+ {
431
+ "epoch": 1.69,
432
+ "grad_norm": 3.7952725887298584,
433
+ "learning_rate": 0.00011549520766773163,
434
+ "loss": 0.7543,
435
+ "step": 530
436
+ },
437
+ {
438
+ "epoch": 1.73,
439
+ "grad_norm": 3.9252896308898926,
440
+ "learning_rate": 0.00011389776357827477,
441
+ "loss": 0.7754,
442
+ "step": 540
443
+ },
444
+ {
445
+ "epoch": 1.76,
446
+ "grad_norm": 3.045405387878418,
447
+ "learning_rate": 0.0001123003194888179,
448
+ "loss": 0.5713,
449
+ "step": 550
450
+ },
451
+ {
452
+ "epoch": 1.79,
453
+ "grad_norm": 1.8544007539749146,
454
+ "learning_rate": 0.00011070287539936102,
455
+ "loss": 0.5592,
456
+ "step": 560
457
+ },
458
+ {
459
+ "epoch": 1.82,
460
+ "grad_norm": 3.626079797744751,
461
+ "learning_rate": 0.00010910543130990417,
462
+ "loss": 0.6631,
463
+ "step": 570
464
+ },
465
+ {
466
+ "epoch": 1.85,
467
+ "grad_norm": 4.355627536773682,
468
+ "learning_rate": 0.00010750798722044728,
469
+ "loss": 0.5981,
470
+ "step": 580
471
+ },
472
+ {
473
+ "epoch": 1.88,
474
+ "grad_norm": 4.025274276733398,
475
+ "learning_rate": 0.00010591054313099043,
476
+ "loss": 0.7375,
477
+ "step": 590
478
+ },
479
+ {
480
+ "epoch": 1.92,
481
+ "grad_norm": 2.506023645401001,
482
+ "learning_rate": 0.00010431309904153355,
483
+ "loss": 0.5935,
484
+ "step": 600
485
+ },
486
+ {
487
+ "epoch": 1.92,
488
+ "eval_loss": 1.0479341745376587,
489
+ "eval_na_accuracy": 0.932,
490
+ "eval_ordinal_accuracy": 0.598,
491
+ "eval_ordinal_mae": 50.45202396943141,
492
+ "eval_runtime": 20.1021,
493
+ "eval_samples_per_second": 24.873,
494
+ "eval_steps_per_second": 3.134,
495
+ "step": 600
496
+ },
497
+ {
498
+ "epoch": 1.95,
499
+ "grad_norm": 4.494864463806152,
500
+ "learning_rate": 0.00010271565495207669,
501
+ "loss": 0.651,
502
+ "step": 610
503
+ },
504
+ {
505
+ "epoch": 1.98,
506
+ "grad_norm": 4.595503807067871,
507
+ "learning_rate": 0.00010111821086261981,
508
+ "loss": 0.6743,
509
+ "step": 620
510
+ },
511
+ {
512
+ "epoch": 2.01,
513
+ "grad_norm": 3.0126616954803467,
514
+ "learning_rate": 9.952076677316294e-05,
515
+ "loss": 0.5441,
516
+ "step": 630
517
+ },
518
+ {
519
+ "epoch": 2.04,
520
+ "grad_norm": 2.2599124908447266,
521
+ "learning_rate": 9.792332268370608e-05,
522
+ "loss": 0.3816,
523
+ "step": 640
524
+ },
525
+ {
526
+ "epoch": 2.08,
527
+ "grad_norm": 1.8868279457092285,
528
+ "learning_rate": 9.63258785942492e-05,
529
+ "loss": 0.4175,
530
+ "step": 650
531
+ },
532
+ {
533
+ "epoch": 2.11,
534
+ "grad_norm": 2.2594425678253174,
535
+ "learning_rate": 9.472843450479234e-05,
536
+ "loss": 0.3208,
537
+ "step": 660
538
+ },
539
+ {
540
+ "epoch": 2.14,
541
+ "grad_norm": 1.4436126947402954,
542
+ "learning_rate": 9.313099041533548e-05,
543
+ "loss": 0.2459,
544
+ "step": 670
545
+ },
546
+ {
547
+ "epoch": 2.17,
548
+ "grad_norm": 2.018686532974243,
549
+ "learning_rate": 9.15335463258786e-05,
550
+ "loss": 0.2556,
551
+ "step": 680
552
+ },
553
+ {
554
+ "epoch": 2.2,
555
+ "grad_norm": 3.8970091342926025,
556
+ "learning_rate": 8.993610223642172e-05,
557
+ "loss": 0.2912,
558
+ "step": 690
559
+ },
560
+ {
561
+ "epoch": 2.24,
562
+ "grad_norm": 7.18654727935791,
563
+ "learning_rate": 8.833865814696486e-05,
564
+ "loss": 0.3525,
565
+ "step": 700
566
+ },
567
+ {
568
+ "epoch": 2.24,
569
+ "eval_loss": 1.1914987564086914,
570
+ "eval_na_accuracy": 0.94,
571
+ "eval_ordinal_accuracy": 0.578,
572
+ "eval_ordinal_mae": 68.50985479798727,
573
+ "eval_runtime": 20.035,
574
+ "eval_samples_per_second": 24.956,
575
+ "eval_steps_per_second": 3.145,
576
+ "step": 700
577
+ },
578
+ {
579
+ "epoch": 2.27,
580
+ "grad_norm": 3.8287854194641113,
581
+ "learning_rate": 8.6741214057508e-05,
582
+ "loss": 0.3517,
583
+ "step": 710
584
+ },
585
+ {
586
+ "epoch": 2.3,
587
+ "grad_norm": 4.602190971374512,
588
+ "learning_rate": 8.514376996805112e-05,
589
+ "loss": 0.3595,
590
+ "step": 720
591
+ },
592
+ {
593
+ "epoch": 2.33,
594
+ "grad_norm": 4.8052978515625,
595
+ "learning_rate": 8.354632587859425e-05,
596
+ "loss": 0.3686,
597
+ "step": 730
598
+ },
599
+ {
600
+ "epoch": 2.36,
601
+ "grad_norm": 1.684538722038269,
602
+ "learning_rate": 8.194888178913739e-05,
603
+ "loss": 0.3499,
604
+ "step": 740
605
+ },
606
+ {
607
+ "epoch": 2.4,
608
+ "grad_norm": 2.9729549884796143,
609
+ "learning_rate": 8.035143769968051e-05,
610
+ "loss": 0.3116,
611
+ "step": 750
612
+ },
613
+ {
614
+ "epoch": 2.43,
615
+ "grad_norm": 4.066229343414307,
616
+ "learning_rate": 7.875399361022364e-05,
617
+ "loss": 0.301,
618
+ "step": 760
619
+ },
620
+ {
621
+ "epoch": 2.46,
622
+ "grad_norm": 4.297142505645752,
623
+ "learning_rate": 7.715654952076677e-05,
624
+ "loss": 0.3501,
625
+ "step": 770
626
+ },
627
+ {
628
+ "epoch": 2.49,
629
+ "grad_norm": 3.189189910888672,
630
+ "learning_rate": 7.555910543130991e-05,
631
+ "loss": 0.4095,
632
+ "step": 780
633
+ },
634
+ {
635
+ "epoch": 2.52,
636
+ "grad_norm": 4.354482173919678,
637
+ "learning_rate": 7.396166134185304e-05,
638
+ "loss": 0.2547,
639
+ "step": 790
640
+ },
641
+ {
642
+ "epoch": 2.56,
643
+ "grad_norm": 2.2913010120391846,
644
+ "learning_rate": 7.236421725239617e-05,
645
+ "loss": 0.2385,
646
+ "step": 800
647
+ },
648
+ {
649
+ "epoch": 2.56,
650
+ "eval_loss": 1.1303021907806396,
651
+ "eval_na_accuracy": 0.948,
652
+ "eval_ordinal_accuracy": 0.586,
653
+ "eval_ordinal_mae": 43.02211729006702,
654
+ "eval_runtime": 20.2963,
655
+ "eval_samples_per_second": 24.635,
656
+ "eval_steps_per_second": 3.104,
657
+ "step": 800
658
+ },
659
+ {
660
+ "epoch": 2.59,
661
+ "grad_norm": 4.907588481903076,
662
+ "learning_rate": 7.07667731629393e-05,
663
+ "loss": 0.2578,
664
+ "step": 810
665
+ },
666
+ {
667
+ "epoch": 2.62,
668
+ "grad_norm": 5.770906925201416,
669
+ "learning_rate": 6.916932907348244e-05,
670
+ "loss": 0.3875,
671
+ "step": 820
672
+ },
673
+ {
674
+ "epoch": 2.65,
675
+ "grad_norm": 3.4392714500427246,
676
+ "learning_rate": 6.757188498402556e-05,
677
+ "loss": 0.3435,
678
+ "step": 830
679
+ },
680
+ {
681
+ "epoch": 2.68,
682
+ "grad_norm": 1.785447359085083,
683
+ "learning_rate": 6.597444089456869e-05,
684
+ "loss": 0.2681,
685
+ "step": 840
686
+ },
687
+ {
688
+ "epoch": 2.72,
689
+ "grad_norm": 3.527557373046875,
690
+ "learning_rate": 6.437699680511182e-05,
691
+ "loss": 0.3525,
692
+ "step": 850
693
+ },
694
+ {
695
+ "epoch": 2.75,
696
+ "grad_norm": 0.4556908905506134,
697
+ "learning_rate": 6.277955271565496e-05,
698
+ "loss": 0.2462,
699
+ "step": 860
700
+ },
701
+ {
702
+ "epoch": 2.78,
703
+ "grad_norm": 2.7756271362304688,
704
+ "learning_rate": 6.118210862619808e-05,
705
+ "loss": 0.3107,
706
+ "step": 870
707
+ },
708
+ {
709
+ "epoch": 2.81,
710
+ "grad_norm": 4.53245735168457,
711
+ "learning_rate": 5.958466453674122e-05,
712
+ "loss": 0.319,
713
+ "step": 880
714
+ },
715
+ {
716
+ "epoch": 2.84,
717
+ "grad_norm": 1.7556302547454834,
718
+ "learning_rate": 5.7987220447284354e-05,
719
+ "loss": 0.2754,
720
+ "step": 890
721
+ },
722
+ {
723
+ "epoch": 2.88,
724
+ "grad_norm": 4.295087814331055,
725
+ "learning_rate": 5.6389776357827484e-05,
726
+ "loss": 0.3423,
727
+ "step": 900
728
+ },
729
+ {
730
+ "epoch": 2.88,
731
+ "eval_loss": 1.1767209768295288,
732
+ "eval_na_accuracy": 0.94,
733
+ "eval_ordinal_accuracy": 0.604,
734
+ "eval_ordinal_mae": 72.14369884862379,
735
+ "eval_runtime": 19.1966,
736
+ "eval_samples_per_second": 26.046,
737
+ "eval_steps_per_second": 3.282,
738
+ "step": 900
739
+ },
740
+ {
741
+ "epoch": 2.91,
742
+ "grad_norm": 3.532813549041748,
743
+ "learning_rate": 5.479233226837061e-05,
744
+ "loss": 0.2459,
745
+ "step": 910
746
+ },
747
+ {
748
+ "epoch": 2.94,
749
+ "grad_norm": 2.6231563091278076,
750
+ "learning_rate": 5.3194888178913736e-05,
751
+ "loss": 0.2713,
752
+ "step": 920
753
+ },
754
+ {
755
+ "epoch": 2.97,
756
+ "grad_norm": 1.7921823263168335,
757
+ "learning_rate": 5.159744408945687e-05,
758
+ "loss": 0.1915,
759
+ "step": 930
760
+ },
761
+ {
762
+ "epoch": 3.0,
763
+ "grad_norm": 0.42470985651016235,
764
+ "learning_rate": 5e-05,
765
+ "loss": 0.2758,
766
+ "step": 940
767
+ },
768
+ {
769
+ "epoch": 3.04,
770
+ "grad_norm": 2.1311330795288086,
771
+ "learning_rate": 4.840255591054313e-05,
772
+ "loss": 0.1447,
773
+ "step": 950
774
+ },
775
+ {
776
+ "epoch": 3.07,
777
+ "grad_norm": 2.371814489364624,
778
+ "learning_rate": 4.680511182108626e-05,
779
+ "loss": 0.0985,
780
+ "step": 960
781
+ },
782
+ {
783
+ "epoch": 3.1,
784
+ "grad_norm": 0.3315708637237549,
785
+ "learning_rate": 4.520766773162939e-05,
786
+ "loss": 0.0887,
787
+ "step": 970
788
+ },
789
+ {
790
+ "epoch": 3.13,
791
+ "grad_norm": 5.331906795501709,
792
+ "learning_rate": 4.361022364217253e-05,
793
+ "loss": 0.1125,
794
+ "step": 980
795
+ },
796
+ {
797
+ "epoch": 3.16,
798
+ "grad_norm": 1.143639326095581,
799
+ "learning_rate": 4.201277955271566e-05,
800
+ "loss": 0.1029,
801
+ "step": 990
802
+ },
803
+ {
804
+ "epoch": 3.19,
805
+ "grad_norm": 0.28226566314697266,
806
+ "learning_rate": 4.041533546325879e-05,
807
+ "loss": 0.0674,
808
+ "step": 1000
809
+ },
810
+ {
811
+ "epoch": 3.19,
812
+ "eval_loss": 1.2294036149978638,
813
+ "eval_na_accuracy": 0.938,
814
+ "eval_ordinal_accuracy": 0.606,
815
+ "eval_ordinal_mae": 28.07016432112176,
816
+ "eval_runtime": 20.2957,
817
+ "eval_samples_per_second": 24.636,
818
+ "eval_steps_per_second": 3.104,
819
+ "step": 1000
820
+ },
821
+ {
822
+ "epoch": 3.23,
823
+ "grad_norm": 0.1891467422246933,
824
+ "learning_rate": 3.8817891373801916e-05,
825
+ "loss": 0.0911,
826
+ "step": 1010
827
+ },
828
+ {
829
+ "epoch": 3.26,
830
+ "grad_norm": 0.14374223351478577,
831
+ "learning_rate": 3.722044728434505e-05,
832
+ "loss": 0.1007,
833
+ "step": 1020
834
+ },
835
+ {
836
+ "epoch": 3.29,
837
+ "grad_norm": 0.22984538972377777,
838
+ "learning_rate": 3.562300319488818e-05,
839
+ "loss": 0.0686,
840
+ "step": 1030
841
+ },
842
+ {
843
+ "epoch": 3.32,
844
+ "grad_norm": 6.988112926483154,
845
+ "learning_rate": 3.402555910543131e-05,
846
+ "loss": 0.0882,
847
+ "step": 1040
848
+ },
849
+ {
850
+ "epoch": 3.35,
851
+ "grad_norm": 0.12949563562870026,
852
+ "learning_rate": 3.242811501597444e-05,
853
+ "loss": 0.1098,
854
+ "step": 1050
855
+ },
856
+ {
857
+ "epoch": 3.39,
858
+ "grad_norm": 0.40354716777801514,
859
+ "learning_rate": 3.083067092651757e-05,
860
+ "loss": 0.0809,
861
+ "step": 1060
862
+ },
863
+ {
864
+ "epoch": 3.42,
865
+ "grad_norm": 0.6469861268997192,
866
+ "learning_rate": 2.9233226837060707e-05,
867
+ "loss": 0.0998,
868
+ "step": 1070
869
+ },
870
+ {
871
+ "epoch": 3.45,
872
+ "grad_norm": 2.1838839054107666,
873
+ "learning_rate": 2.7635782747603834e-05,
874
+ "loss": 0.0923,
875
+ "step": 1080
876
+ },
877
+ {
878
+ "epoch": 3.48,
879
+ "grad_norm": 1.0453554391860962,
880
+ "learning_rate": 2.6038338658146967e-05,
881
+ "loss": 0.0669,
882
+ "step": 1090
883
+ },
884
+ {
885
+ "epoch": 3.51,
886
+ "grad_norm": 0.6662182807922363,
887
+ "learning_rate": 2.44408945686901e-05,
888
+ "loss": 0.1206,
889
+ "step": 1100
890
+ },
891
+ {
892
+ "epoch": 3.51,
893
+ "eval_loss": 1.2335716485977173,
894
+ "eval_na_accuracy": 0.938,
895
+ "eval_ordinal_accuracy": 0.616,
896
+ "eval_ordinal_mae": 65.07940023336745,
897
+ "eval_runtime": 20.0687,
898
+ "eval_samples_per_second": 24.914,
899
+ "eval_steps_per_second": 3.139,
900
+ "step": 1100
901
+ },
902
+ {
903
+ "epoch": 3.55,
904
+ "grad_norm": 3.52840256690979,
905
+ "learning_rate": 2.284345047923323e-05,
906
+ "loss": 0.1402,
907
+ "step": 1110
908
+ },
909
+ {
910
+ "epoch": 3.58,
911
+ "grad_norm": 0.1548507809638977,
912
+ "learning_rate": 2.124600638977636e-05,
913
+ "loss": 0.1038,
914
+ "step": 1120
915
+ },
916
+ {
917
+ "epoch": 3.61,
918
+ "grad_norm": 3.753204584121704,
919
+ "learning_rate": 1.964856230031949e-05,
920
+ "loss": 0.1237,
921
+ "step": 1130
922
+ },
923
+ {
924
+ "epoch": 3.64,
925
+ "grad_norm": 2.5928821563720703,
926
+ "learning_rate": 1.805111821086262e-05,
927
+ "loss": 0.1653,
928
+ "step": 1140
929
+ },
930
+ {
931
+ "epoch": 3.67,
932
+ "grad_norm": 0.26777195930480957,
933
+ "learning_rate": 1.645367412140575e-05,
934
+ "loss": 0.0924,
935
+ "step": 1150
936
+ },
937
+ {
938
+ "epoch": 3.71,
939
+ "grad_norm": 0.11366520822048187,
940
+ "learning_rate": 1.485623003194888e-05,
941
+ "loss": 0.0376,
942
+ "step": 1160
943
+ },
944
+ {
945
+ "epoch": 3.74,
946
+ "grad_norm": 2.3529770374298096,
947
+ "learning_rate": 1.3258785942492014e-05,
948
+ "loss": 0.1034,
949
+ "step": 1170
950
+ },
951
+ {
952
+ "epoch": 3.77,
953
+ "grad_norm": 0.16830460727214813,
954
+ "learning_rate": 1.1661341853035145e-05,
955
+ "loss": 0.107,
956
+ "step": 1180
957
+ },
958
+ {
959
+ "epoch": 3.8,
960
+ "grad_norm": 0.9607967734336853,
961
+ "learning_rate": 1.0063897763578276e-05,
962
+ "loss": 0.0751,
963
+ "step": 1190
964
+ },
965
+ {
966
+ "epoch": 3.83,
967
+ "grad_norm": 7.122812271118164,
968
+ "learning_rate": 8.466453674121406e-06,
969
+ "loss": 0.1261,
970
+ "step": 1200
971
+ },
972
+ {
973
+ "epoch": 3.83,
974
+ "eval_loss": 1.2907226085662842,
975
+ "eval_na_accuracy": 0.938,
976
+ "eval_ordinal_accuracy": 0.604,
977
+ "eval_ordinal_mae": 45.83343084012577,
978
+ "eval_runtime": 20.016,
979
+ "eval_samples_per_second": 24.98,
980
+ "eval_steps_per_second": 3.147,
981
+ "step": 1200
982
+ },
983
+ {
984
+ "epoch": 3.83,
985
+ "step": 1200,
986
+ "total_flos": 1.4860396665534874e+18,
987
+ "train_loss": 0.6305676238735517,
988
+ "train_runtime": 1636.5913,
989
+ "train_samples_per_second": 12.221,
990
+ "train_steps_per_second": 0.765
991
  }
992
  ],
993
  "logging_steps": 10,
994
+ "max_steps": 1252,
995
  "num_input_tokens_seen": 0,
996
  "num_train_epochs": 4,
997
  "save_steps": 100,
998
+ "total_flos": 1.4860396665534874e+18,
999
  "train_batch_size": 16,
1000
  "trial_name": null,
1001
  "trial_params": null