End of training
Browse files
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 28.
|
| 6 |
-
"eval_samples_per_second": 5.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.17,
|
| 3 |
+
"eval_accuracy": 0.8516129032258064,
|
| 4 |
+
"eval_loss": 0.6185442209243774,
|
| 5 |
+
"eval_runtime": 28.1289,
|
| 6 |
+
"eval_samples_per_second": 5.51,
|
| 7 |
+
"eval_steps_per_second": 1.386
|
| 8 |
}
|
runs/May24_04-01-25_d951cd11c68c/events.out.tfevents.1716524277.d951cd11c68c.34.1
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b9e829316cd6256fb76eb813630c9f1e6458c019496a19a9aa23b4fbe0c27331
|
| 3 |
+
size 734
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 28.
|
| 6 |
-
"eval_samples_per_second": 5.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.17,
|
| 3 |
+
"eval_accuracy": 0.8516129032258064,
|
| 4 |
+
"eval_loss": 0.6185442209243774,
|
| 5 |
+
"eval_runtime": 28.1289,
|
| 6 |
+
"eval_samples_per_second": 5.51,
|
| 7 |
+
"eval_steps_per_second": 1.386
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,170 +1,416 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 4.
|
| 36 |
-
"loss":
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss":
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"eval_accuracy": 0.
|
| 63 |
-
"eval_loss":
|
| 64 |
-
"eval_runtime":
|
| 65 |
-
"eval_samples_per_second": 5.
|
| 66 |
-
"eval_steps_per_second": 1.
|
| 67 |
"step": 75
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch": 1.
|
| 71 |
-
"grad_norm":
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss":
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 1.
|
| 78 |
-
"grad_norm":
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss":
|
| 81 |
"step": 90
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 1.
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate":
|
| 87 |
-
"loss":
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"epoch": 1.
|
| 92 |
-
"grad_norm":
|
| 93 |
-
"learning_rate":
|
| 94 |
-
"loss": 0.
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch": 1.
|
| 99 |
-
"grad_norm":
|
| 100 |
-
"learning_rate":
|
| 101 |
-
"loss": 0.
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch": 1.
|
| 106 |
-
"grad_norm":
|
| 107 |
-
"learning_rate":
|
| 108 |
-
"loss": 0.
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
-
"epoch": 1.
|
| 113 |
-
"grad_norm":
|
| 114 |
-
"learning_rate": 3.
|
| 115 |
-
"loss": 0.
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 1.
|
| 120 |
-
"grad_norm":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss":
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 1.
|
| 127 |
-
"eval_accuracy": 0.
|
| 128 |
-
"eval_loss": 0.
|
| 129 |
-
"eval_runtime": 12.
|
| 130 |
-
"eval_samples_per_second": 5.
|
| 131 |
-
"eval_steps_per_second": 1.
|
| 132 |
"step": 150
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"
|
| 137 |
-
"
|
| 138 |
-
"
|
| 139 |
-
"
|
| 140 |
-
"train_samples_per_second": 2.577,
|
| 141 |
-
"train_steps_per_second": 0.644
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"
|
| 146 |
-
"
|
| 147 |
-
"
|
| 148 |
-
"
|
| 149 |
-
"eval_steps_per_second": 1.363,
|
| 150 |
-
"step": 150
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch":
|
| 154 |
-
"
|
| 155 |
-
"
|
| 156 |
-
"
|
| 157 |
-
"
|
| 158 |
-
|
| 159 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 160 |
}
|
| 161 |
],
|
| 162 |
"logging_steps": 10,
|
| 163 |
-
"max_steps":
|
| 164 |
"num_input_tokens_seen": 0,
|
| 165 |
"num_train_epochs": 9223372036854775807,
|
| 166 |
"save_steps": 500,
|
| 167 |
-
"total_flos":
|
| 168 |
"train_batch_size": 4,
|
| 169 |
"trial_name": null,
|
| 170 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9714285714285714,
|
| 3 |
+
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-375",
|
| 4 |
+
"epoch": 5.166666666666667,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 450,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 9.09181022644043,
|
| 14 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 15 |
+
"loss": 2.3453,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 9.389509201049805,
|
| 21 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 22 |
+
"loss": 2.3958,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.07,
|
| 27 |
+
"grad_norm": 9.11400032043457,
|
| 28 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 29 |
+
"loss": 2.2492,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.09,
|
| 34 |
+
"grad_norm": 9.454550743103027,
|
| 35 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 36 |
+
"loss": 2.1811,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.11,
|
| 41 |
+
"grad_norm": 12.871296882629395,
|
| 42 |
+
"learning_rate": 4.938271604938271e-05,
|
| 43 |
+
"loss": 2.1261,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.13,
|
| 48 |
+
"grad_norm": 9.507197380065918,
|
| 49 |
+
"learning_rate": 4.814814814814815e-05,
|
| 50 |
+
"loss": 2.0149,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.16,
|
| 55 |
+
"grad_norm": 12.240191459655762,
|
| 56 |
+
"learning_rate": 4.691358024691358e-05,
|
| 57 |
+
"loss": 1.9248,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.17,
|
| 62 |
+
"eval_accuracy": 0.3142857142857143,
|
| 63 |
+
"eval_loss": 1.6910260915756226,
|
| 64 |
+
"eval_runtime": 13.1007,
|
| 65 |
+
"eval_samples_per_second": 5.343,
|
| 66 |
+
"eval_steps_per_second": 1.374,
|
| 67 |
"step": 75
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"epoch": 1.01,
|
| 71 |
+
"grad_norm": 13.400834083557129,
|
| 72 |
+
"learning_rate": 4.567901234567901e-05,
|
| 73 |
+
"loss": 1.4961,
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 1.03,
|
| 78 |
+
"grad_norm": 12.832711219787598,
|
| 79 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 80 |
+
"loss": 1.4544,
|
| 81 |
"step": 90
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.06,
|
| 85 |
+
"grad_norm": 11.878938674926758,
|
| 86 |
+
"learning_rate": 4.3209876543209875e-05,
|
| 87 |
+
"loss": 1.3226,
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
+
"epoch": 1.08,
|
| 92 |
+
"grad_norm": 8.267742156982422,
|
| 93 |
+
"learning_rate": 4.197530864197531e-05,
|
| 94 |
+
"loss": 0.9212,
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
+
"epoch": 1.1,
|
| 99 |
+
"grad_norm": 15.088278770446777,
|
| 100 |
+
"learning_rate": 4.074074074074074e-05,
|
| 101 |
+
"loss": 0.8928,
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 1.12,
|
| 106 |
+
"grad_norm": 14.688928604125977,
|
| 107 |
+
"learning_rate": 3.950617283950617e-05,
|
| 108 |
+
"loss": 0.6026,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 1.14,
|
| 113 |
+
"grad_norm": 10.090428352355957,
|
| 114 |
+
"learning_rate": 3.82716049382716e-05,
|
| 115 |
+
"loss": 0.7323,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 1.17,
|
| 120 |
+
"grad_norm": 28.017093658447266,
|
| 121 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 122 |
+
"loss": 1.1184,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 1.17,
|
| 127 |
+
"eval_accuracy": 0.7,
|
| 128 |
+
"eval_loss": 0.8519060611724854,
|
| 129 |
+
"eval_runtime": 12.8721,
|
| 130 |
+
"eval_samples_per_second": 5.438,
|
| 131 |
+
"eval_steps_per_second": 1.398,
|
| 132 |
"step": 150
|
| 133 |
},
|
| 134 |
{
|
| 135 |
+
"epoch": 2.02,
|
| 136 |
+
"grad_norm": 7.826801776885986,
|
| 137 |
+
"learning_rate": 3.580246913580247e-05,
|
| 138 |
+
"loss": 0.6328,
|
| 139 |
+
"step": 160
|
|
|
|
|
|
|
| 140 |
},
|
| 141 |
{
|
| 142 |
+
"epoch": 2.04,
|
| 143 |
+
"grad_norm": 17.076839447021484,
|
| 144 |
+
"learning_rate": 3.45679012345679e-05,
|
| 145 |
+
"loss": 0.5843,
|
| 146 |
+
"step": 170
|
|
|
|
|
|
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 2.07,
|
| 150 |
+
"grad_norm": 8.304590225219727,
|
| 151 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 152 |
+
"loss": 0.7969,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.09,
|
| 157 |
+
"grad_norm": 7.5533223152160645,
|
| 158 |
+
"learning_rate": 3.209876543209876e-05,
|
| 159 |
+
"loss": 0.5387,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.11,
|
| 164 |
+
"grad_norm": 3.7337615489959717,
|
| 165 |
+
"learning_rate": 3.08641975308642e-05,
|
| 166 |
+
"loss": 0.5673,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.13,
|
| 171 |
+
"grad_norm": 23.18366050720215,
|
| 172 |
+
"learning_rate": 2.962962962962963e-05,
|
| 173 |
+
"loss": 0.5191,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.16,
|
| 178 |
+
"grad_norm": 11.092031478881836,
|
| 179 |
+
"learning_rate": 2.839506172839506e-05,
|
| 180 |
+
"loss": 0.3505,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.17,
|
| 185 |
+
"eval_accuracy": 0.7857142857142857,
|
| 186 |
+
"eval_loss": 0.5482387542724609,
|
| 187 |
+
"eval_runtime": 12.8506,
|
| 188 |
+
"eval_samples_per_second": 5.447,
|
| 189 |
+
"eval_steps_per_second": 1.401,
|
| 190 |
+
"step": 225
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.01,
|
| 194 |
+
"grad_norm": 1.397813320159912,
|
| 195 |
+
"learning_rate": 2.7160493827160493e-05,
|
| 196 |
+
"loss": 0.2652,
|
| 197 |
+
"step": 230
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.03,
|
| 201 |
+
"grad_norm": 2.130690574645996,
|
| 202 |
+
"learning_rate": 2.5925925925925925e-05,
|
| 203 |
+
"loss": 0.1691,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.06,
|
| 208 |
+
"grad_norm": 7.323586940765381,
|
| 209 |
+
"learning_rate": 2.4691358024691357e-05,
|
| 210 |
+
"loss": 0.1669,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.08,
|
| 215 |
+
"grad_norm": 5.903147220611572,
|
| 216 |
+
"learning_rate": 2.345679012345679e-05,
|
| 217 |
+
"loss": 0.324,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.1,
|
| 222 |
+
"grad_norm": 24.850133895874023,
|
| 223 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 224 |
+
"loss": 0.4857,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.12,
|
| 229 |
+
"grad_norm": 1.1520411968231201,
|
| 230 |
+
"learning_rate": 2.0987654320987655e-05,
|
| 231 |
+
"loss": 0.231,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.14,
|
| 236 |
+
"grad_norm": 0.6035730838775635,
|
| 237 |
+
"learning_rate": 1.9753086419753087e-05,
|
| 238 |
+
"loss": 0.3852,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.17,
|
| 243 |
+
"grad_norm": 18.224578857421875,
|
| 244 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 245 |
+
"loss": 0.2939,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.17,
|
| 250 |
+
"eval_accuracy": 0.8285714285714286,
|
| 251 |
+
"eval_loss": 0.5074731111526489,
|
| 252 |
+
"eval_runtime": 12.6199,
|
| 253 |
+
"eval_samples_per_second": 5.547,
|
| 254 |
+
"eval_steps_per_second": 1.426,
|
| 255 |
+
"step": 300
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.02,
|
| 259 |
+
"grad_norm": 4.320995330810547,
|
| 260 |
+
"learning_rate": 1.728395061728395e-05,
|
| 261 |
+
"loss": 0.1894,
|
| 262 |
+
"step": 310
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 4.04,
|
| 266 |
+
"grad_norm": 4.220107555389404,
|
| 267 |
+
"learning_rate": 1.604938271604938e-05,
|
| 268 |
+
"loss": 0.2997,
|
| 269 |
+
"step": 320
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 4.07,
|
| 273 |
+
"grad_norm": 0.14674872159957886,
|
| 274 |
+
"learning_rate": 1.4814814814814815e-05,
|
| 275 |
+
"loss": 0.2471,
|
| 276 |
+
"step": 330
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.09,
|
| 280 |
+
"grad_norm": 1.210436224937439,
|
| 281 |
+
"learning_rate": 1.3580246913580247e-05,
|
| 282 |
+
"loss": 0.1045,
|
| 283 |
+
"step": 340
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.11,
|
| 287 |
+
"grad_norm": 0.28732016682624817,
|
| 288 |
+
"learning_rate": 1.2345679012345678e-05,
|
| 289 |
+
"loss": 0.0638,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.13,
|
| 294 |
+
"grad_norm": 0.1985640674829483,
|
| 295 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 296 |
+
"loss": 0.1419,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.16,
|
| 301 |
+
"grad_norm": 0.1826096773147583,
|
| 302 |
+
"learning_rate": 9.876543209876543e-06,
|
| 303 |
+
"loss": 0.0447,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.17,
|
| 308 |
+
"eval_accuracy": 0.9714285714285714,
|
| 309 |
+
"eval_loss": 0.1740979254245758,
|
| 310 |
+
"eval_runtime": 12.7121,
|
| 311 |
+
"eval_samples_per_second": 5.507,
|
| 312 |
+
"eval_steps_per_second": 1.416,
|
| 313 |
+
"step": 375
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.01,
|
| 317 |
+
"grad_norm": 0.4490085244178772,
|
| 318 |
+
"learning_rate": 8.641975308641975e-06,
|
| 319 |
+
"loss": 0.0338,
|
| 320 |
+
"step": 380
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.03,
|
| 324 |
+
"grad_norm": 0.615115761756897,
|
| 325 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 326 |
+
"loss": 0.1366,
|
| 327 |
+
"step": 390
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.06,
|
| 331 |
+
"grad_norm": 0.19469819962978363,
|
| 332 |
+
"learning_rate": 6.172839506172839e-06,
|
| 333 |
+
"loss": 0.2805,
|
| 334 |
+
"step": 400
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 5.08,
|
| 338 |
+
"grad_norm": 0.10871770232915878,
|
| 339 |
+
"learning_rate": 4.938271604938272e-06,
|
| 340 |
+
"loss": 0.1858,
|
| 341 |
+
"step": 410
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 5.1,
|
| 345 |
+
"grad_norm": 1.5411165952682495,
|
| 346 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 347 |
+
"loss": 0.0189,
|
| 348 |
+
"step": 420
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.12,
|
| 352 |
+
"grad_norm": 10.748401641845703,
|
| 353 |
+
"learning_rate": 2.469135802469136e-06,
|
| 354 |
+
"loss": 0.0882,
|
| 355 |
+
"step": 430
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 5.14,
|
| 359 |
+
"grad_norm": 0.13508407771587372,
|
| 360 |
+
"learning_rate": 1.234567901234568e-06,
|
| 361 |
+
"loss": 0.1669,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.17,
|
| 366 |
+
"grad_norm": 6.92973518371582,
|
| 367 |
+
"learning_rate": 0.0,
|
| 368 |
+
"loss": 0.0643,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.17,
|
| 373 |
+
"eval_accuracy": 0.9571428571428572,
|
| 374 |
+
"eval_loss": 0.1370178759098053,
|
| 375 |
+
"eval_runtime": 12.713,
|
| 376 |
+
"eval_samples_per_second": 5.506,
|
| 377 |
+
"eval_steps_per_second": 1.416,
|
| 378 |
+
"step": 450
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 5.17,
|
| 382 |
+
"step": 450,
|
| 383 |
+
"total_flos": 2.243076282187776e+18,
|
| 384 |
+
"train_loss": 0.7145391458272934,
|
| 385 |
+
"train_runtime": 727.1717,
|
| 386 |
+
"train_samples_per_second": 2.475,
|
| 387 |
+
"train_steps_per_second": 0.619
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 5.17,
|
| 391 |
+
"eval_accuracy": 0.8516129032258064,
|
| 392 |
+
"eval_loss": 0.6185441613197327,
|
| 393 |
+
"eval_runtime": 29.0205,
|
| 394 |
+
"eval_samples_per_second": 5.341,
|
| 395 |
+
"eval_steps_per_second": 1.344,
|
| 396 |
+
"step": 450
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 5.17,
|
| 400 |
+
"eval_accuracy": 0.8516129032258064,
|
| 401 |
+
"eval_loss": 0.6185442209243774,
|
| 402 |
+
"eval_runtime": 28.1289,
|
| 403 |
+
"eval_samples_per_second": 5.51,
|
| 404 |
+
"eval_steps_per_second": 1.386,
|
| 405 |
+
"step": 450
|
| 406 |
}
|
| 407 |
],
|
| 408 |
"logging_steps": 10,
|
| 409 |
+
"max_steps": 450,
|
| 410 |
"num_input_tokens_seen": 0,
|
| 411 |
"num_train_epochs": 9223372036854775807,
|
| 412 |
"save_steps": 500,
|
| 413 |
+
"total_flos": 2.243076282187776e+18,
|
| 414 |
"train_batch_size": 4,
|
| 415 |
"trial_name": null,
|
| 416 |
"trial_params": null
|