carted-nlp/categorization-finetuned-20220721-164940-pruned-20220803-184651
Browse files- all_results.json +15 -0
- pytorch_model.bin +1 -1
- recipe.yaml +27 -0
- test_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +489 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"eval_accuracy": 0.8760195530726257,
|
| 4 |
+
"eval_f1": 0.8750784966730759,
|
| 5 |
+
"eval_loss": 0.46725794672966003,
|
| 6 |
+
"eval_runtime": 42.6171,
|
| 7 |
+
"eval_samples_per_second": 3360.157,
|
| 8 |
+
"eval_steps_per_second": 70.019,
|
| 9 |
+
"test_samples": 143200,
|
| 10 |
+
"train_loss": 0.4135634679527994,
|
| 11 |
+
"train_runtime": 19096.9194,
|
| 12 |
+
"train_samples": 1138117,
|
| 13 |
+
"train_samples_per_second": 893.953,
|
| 14 |
+
"train_steps_per_second": 3.103
|
| 15 |
+
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 488228637
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da5f963c6319f63a4c53614c92da3838876574304c3d7339fd3508e5b5a49766
|
| 3 |
size 488228637
|
recipe.yaml
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version: 1.1.0
|
| 2 |
+
|
| 3 |
+
__metadata__:
|
| 4 |
+
per_device_train_batch_size: 48
|
| 5 |
+
per_device_eval_batch_size: 48
|
| 6 |
+
fp16: False
|
| 7 |
+
framework_metadata:
|
| 8 |
+
python_version: 3.8.13
|
| 9 |
+
sparseml_version: 1.0.1
|
| 10 |
+
torch_version: 1.9.1+cu111
|
| 11 |
+
|
| 12 |
+
modifiers:
|
| 13 |
+
- !GMPruningModifier
|
| 14 |
+
end_epoch: 11
|
| 15 |
+
final_sparsity: 0.7
|
| 16 |
+
global_sparsity: False
|
| 17 |
+
init_sparsity: 0.0
|
| 18 |
+
inter_func: cubic
|
| 19 |
+
leave_enabled: True
|
| 20 |
+
mask_type: unstructured
|
| 21 |
+
params: ['re:roberta.encoder.layer.*.attention.self.query.weight', 're:roberta.encoder.layer.*.attention.self.key.weight', 're:roberta.encoder.layer.*.attention.self.value.weight', 're:roberta.encoder.layer.*.attention.output.dense.weight', 're:roberta.encoder.layer.*.intermediate.dense.weight', 're:roberta.encoder.layer.*.output.dense.weight', 're:classifier.dense.weight']
|
| 22 |
+
start_epoch: 2
|
| 23 |
+
update_frequency: 0.05
|
| 24 |
+
|
| 25 |
+
- !EpochRangeModifier
|
| 26 |
+
end_epoch: 15
|
| 27 |
+
start_epoch: 0.0
|
test_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"eval_accuracy": 0.8760195530726257,
|
| 4 |
+
"eval_f1": 0.8750784966730759,
|
| 5 |
+
"eval_loss": 0.46725794672966003,
|
| 6 |
+
"eval_runtime": 42.6171,
|
| 7 |
+
"eval_samples_per_second": 3360.157,
|
| 8 |
+
"eval_steps_per_second": 70.019,
|
| 9 |
+
"test_samples": 143200
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"train_loss": 0.4135634679527994,
|
| 4 |
+
"train_runtime": 19096.9194,
|
| 5 |
+
"train_samples": 1138117,
|
| 6 |
+
"train_samples_per_second": 893.953,
|
| 7 |
+
"train_steps_per_second": 3.103
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,489 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 14.999789127409219,
|
| 5 |
+
"global_step": 59265,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.51,
|
| 12 |
+
"learning_rate": 6.994913523145705e-06,
|
| 13 |
+
"loss": 0.3404,
|
| 14 |
+
"step": 2000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.51,
|
| 18 |
+
"eval_accuracy": 0.887192844025166,
|
| 19 |
+
"eval_f1": 0.886488607488543,
|
| 20 |
+
"eval_loss": 0.4329048991203308,
|
| 21 |
+
"eval_runtime": 42.8766,
|
| 22 |
+
"eval_samples_per_second": 3340.027,
|
| 23 |
+
"eval_steps_per_second": 69.595,
|
| 24 |
+
"step": 2000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.01,
|
| 28 |
+
"learning_rate": 6.954310370186513e-06,
|
| 29 |
+
"loss": 0.3433,
|
| 30 |
+
"step": 4000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.01,
|
| 34 |
+
"eval_accuracy": 0.8883170750441662,
|
| 35 |
+
"eval_f1": 0.8876090017723502,
|
| 36 |
+
"eval_loss": 0.4280129075050354,
|
| 37 |
+
"eval_runtime": 46.562,
|
| 38 |
+
"eval_samples_per_second": 3075.664,
|
| 39 |
+
"eval_steps_per_second": 64.087,
|
| 40 |
+
"step": 4000
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 1.52,
|
| 44 |
+
"learning_rate": 6.873575782958537e-06,
|
| 45 |
+
"loss": 0.3281,
|
| 46 |
+
"step": 6000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.52,
|
| 50 |
+
"eval_accuracy": 0.8889804411733899,
|
| 51 |
+
"eval_f1": 0.8882523897529724,
|
| 52 |
+
"eval_loss": 0.43023422360420227,
|
| 53 |
+
"eval_runtime": 45.3226,
|
| 54 |
+
"eval_samples_per_second": 3159.767,
|
| 55 |
+
"eval_steps_per_second": 65.839,
|
| 56 |
+
"step": 6000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 2.02,
|
| 60 |
+
"learning_rate": 6.753647718516397e-06,
|
| 61 |
+
"loss": 0.331,
|
| 62 |
+
"step": 8000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 2.02,
|
| 66 |
+
"eval_accuracy": 0.8891340628033154,
|
| 67 |
+
"eval_f1": 0.8885291871869458,
|
| 68 |
+
"eval_loss": 0.42653217911720276,
|
| 69 |
+
"eval_runtime": 54.726,
|
| 70 |
+
"eval_samples_per_second": 2616.839,
|
| 71 |
+
"eval_steps_per_second": 54.526,
|
| 72 |
+
"step": 8000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 2.53,
|
| 76 |
+
"learning_rate": 6.5959194752957715e-06,
|
| 77 |
+
"loss": 0.3224,
|
| 78 |
+
"step": 10000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.53,
|
| 82 |
+
"eval_accuracy": 0.8881355222087998,
|
| 83 |
+
"eval_f1": 0.8873919365623191,
|
| 84 |
+
"eval_loss": 0.43002641201019287,
|
| 85 |
+
"eval_runtime": 54.91,
|
| 86 |
+
"eval_samples_per_second": 2608.066,
|
| 87 |
+
"eval_steps_per_second": 54.343,
|
| 88 |
+
"step": 10000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 3.04,
|
| 92 |
+
"learning_rate": 6.402223506072121e-06,
|
| 93 |
+
"loss": 0.3361,
|
| 94 |
+
"step": 12000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 3.04,
|
| 98 |
+
"eval_accuracy": 0.8889175959611477,
|
| 99 |
+
"eval_f1": 0.8882468792028413,
|
| 100 |
+
"eval_loss": 0.42911502718925476,
|
| 101 |
+
"eval_runtime": 50.9856,
|
| 102 |
+
"eval_samples_per_second": 2808.81,
|
| 103 |
+
"eval_steps_per_second": 58.526,
|
| 104 |
+
"step": 12000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.54,
|
| 108 |
+
"learning_rate": 6.1748101289188055e-06,
|
| 109 |
+
"loss": 0.3323,
|
| 110 |
+
"step": 14000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.54,
|
| 114 |
+
"eval_accuracy": 0.8877584509353462,
|
| 115 |
+
"eval_f1": 0.8871109456457335,
|
| 116 |
+
"eval_loss": 0.4336757957935333,
|
| 117 |
+
"eval_runtime": 51.7236,
|
| 118 |
+
"eval_samples_per_second": 2768.739,
|
| 119 |
+
"eval_steps_per_second": 57.691,
|
| 120 |
+
"step": 14000
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 4.05,
|
| 124 |
+
"learning_rate": 5.916321383496139e-06,
|
| 125 |
+
"loss": 0.3556,
|
| 126 |
+
"step": 16000
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 4.05,
|
| 130 |
+
"eval_accuracy": 0.8856845589313521,
|
| 131 |
+
"eval_f1": 0.8851065217265167,
|
| 132 |
+
"eval_loss": 0.4344990849494934,
|
| 133 |
+
"eval_runtime": 50.059,
|
| 134 |
+
"eval_samples_per_second": 2860.803,
|
| 135 |
+
"eval_steps_per_second": 59.61,
|
| 136 |
+
"step": 16000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 4.56,
|
| 140 |
+
"learning_rate": 5.629760336403372e-06,
|
| 141 |
+
"loss": 0.3663,
|
| 142 |
+
"step": 18000
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 4.56,
|
| 146 |
+
"eval_accuracy": 0.8836455809341591,
|
| 147 |
+
"eval_f1": 0.8828440694403326,
|
| 148 |
+
"eval_loss": 0.44171223044395447,
|
| 149 |
+
"eval_runtime": 52.3029,
|
| 150 |
+
"eval_samples_per_second": 2738.072,
|
| 151 |
+
"eval_steps_per_second": 57.052,
|
| 152 |
+
"step": 18000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 5.06,
|
| 156 |
+
"learning_rate": 5.318456192197413e-06,
|
| 157 |
+
"loss": 0.3902,
|
| 158 |
+
"step": 20000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 5.06,
|
| 162 |
+
"eval_accuracy": 0.8789252072146304,
|
| 163 |
+
"eval_f1": 0.8780871328644805,
|
| 164 |
+
"eval_loss": 0.4555477797985077,
|
| 165 |
+
"eval_runtime": 50.7366,
|
| 166 |
+
"eval_samples_per_second": 2822.599,
|
| 167 |
+
"eval_steps_per_second": 58.814,
|
| 168 |
+
"step": 20000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 5.57,
|
| 172 |
+
"learning_rate": 4.986025615410962e-06,
|
| 173 |
+
"loss": 0.4036,
|
| 174 |
+
"step": 22000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 5.57,
|
| 178 |
+
"eval_accuracy": 0.8788483963996676,
|
| 179 |
+
"eval_f1": 0.877917228234661,
|
| 180 |
+
"eval_loss": 0.4555710554122925,
|
| 181 |
+
"eval_runtime": 49.8006,
|
| 182 |
+
"eval_samples_per_second": 2875.647,
|
| 183 |
+
"eval_steps_per_second": 59.919,
|
| 184 |
+
"step": 22000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 6.07,
|
| 188 |
+
"learning_rate": 4.636330712922542e-06,
|
| 189 |
+
"loss": 0.4305,
|
| 190 |
+
"step": 24000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 6.07,
|
| 194 |
+
"eval_accuracy": 0.8750637180624122,
|
| 195 |
+
"eval_f1": 0.8741648169470942,
|
| 196 |
+
"eval_loss": 0.46971216797828674,
|
| 197 |
+
"eval_runtime": 50.8439,
|
| 198 |
+
"eval_samples_per_second": 2816.642,
|
| 199 |
+
"eval_steps_per_second": 58.689,
|
| 200 |
+
"step": 24000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 6.58,
|
| 204 |
+
"learning_rate": 4.273434164830173e-06,
|
| 205 |
+
"loss": 0.4501,
|
| 206 |
+
"step": 26000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 6.58,
|
| 210 |
+
"eval_accuracy": 0.8737788826121263,
|
| 211 |
+
"eval_f1": 0.8725169321502271,
|
| 212 |
+
"eval_loss": 0.4762924611568451,
|
| 213 |
+
"eval_runtime": 50.5449,
|
| 214 |
+
"eval_samples_per_second": 2833.301,
|
| 215 |
+
"eval_steps_per_second": 59.037,
|
| 216 |
+
"step": 26000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 7.09,
|
| 220 |
+
"learning_rate": 3.901552025108571e-06,
|
| 221 |
+
"loss": 0.4733,
|
| 222 |
+
"step": 28000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 7.09,
|
| 226 |
+
"eval_accuracy": 0.8710486072802687,
|
| 227 |
+
"eval_f1": 0.8700483733487925,
|
| 228 |
+
"eval_loss": 0.48572415113449097,
|
| 229 |
+
"eval_runtime": 51.8675,
|
| 230 |
+
"eval_samples_per_second": 2761.054,
|
| 231 |
+
"eval_steps_per_second": 57.531,
|
| 232 |
+
"step": 28000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 7.59,
|
| 236 |
+
"learning_rate": 3.5250047404016715e-06,
|
| 237 |
+
"loss": 0.4851,
|
| 238 |
+
"step": 30000
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 7.59,
|
| 242 |
+
"eval_accuracy": 0.8705109315755295,
|
| 243 |
+
"eval_f1": 0.8695447653344123,
|
| 244 |
+
"eval_loss": 0.4862508177757263,
|
| 245 |
+
"eval_runtime": 50.7033,
|
| 246 |
+
"eval_samples_per_second": 2824.454,
|
| 247 |
+
"eval_steps_per_second": 58.852,
|
| 248 |
+
"step": 30000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 8.1,
|
| 252 |
+
"learning_rate": 3.1481669560035742e-06,
|
| 253 |
+
"loss": 0.4846,
|
| 254 |
+
"step": 32000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 8.1,
|
| 258 |
+
"eval_accuracy": 0.8708251576367407,
|
| 259 |
+
"eval_f1": 0.869757227158951,
|
| 260 |
+
"eval_loss": 0.48489654064178467,
|
| 261 |
+
"eval_runtime": 50.2252,
|
| 262 |
+
"eval_samples_per_second": 2851.338,
|
| 263 |
+
"eval_steps_per_second": 59.412,
|
| 264 |
+
"step": 32000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 8.61,
|
| 268 |
+
"learning_rate": 2.775416692171258e-06,
|
| 269 |
+
"loss": 0.4856,
|
| 270 |
+
"step": 34000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 8.61,
|
| 274 |
+
"eval_accuracy": 0.870664553205455,
|
| 275 |
+
"eval_f1": 0.8694665590028052,
|
| 276 |
+
"eval_loss": 0.4835050404071808,
|
| 277 |
+
"eval_runtime": 50.112,
|
| 278 |
+
"eval_samples_per_second": 2857.776,
|
| 279 |
+
"eval_steps_per_second": 59.547,
|
| 280 |
+
"step": 34000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 9.11,
|
| 284 |
+
"learning_rate": 2.411084481227692e-06,
|
| 285 |
+
"loss": 0.4774,
|
| 286 |
+
"step": 36000
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 9.11,
|
| 290 |
+
"eval_accuracy": 0.8719214574502999,
|
| 291 |
+
"eval_f1": 0.870818507712314,
|
| 292 |
+
"eval_loss": 0.47967976331710815,
|
| 293 |
+
"eval_runtime": 51.643,
|
| 294 |
+
"eval_samples_per_second": 2773.058,
|
| 295 |
+
"eval_steps_per_second": 57.781,
|
| 296 |
+
"step": 36000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 9.62,
|
| 300 |
+
"learning_rate": 2.059403056369544e-06,
|
| 301 |
+
"loss": 0.4635,
|
| 302 |
+
"step": 38000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 9.62,
|
| 306 |
+
"eval_accuracy": 0.8728222388257721,
|
| 307 |
+
"eval_f1": 0.8716593868274884,
|
| 308 |
+
"eval_loss": 0.47762736678123474,
|
| 309 |
+
"eval_runtime": 49.924,
|
| 310 |
+
"eval_samples_per_second": 2868.539,
|
| 311 |
+
"eval_steps_per_second": 59.771,
|
| 312 |
+
"step": 38000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 10.12,
|
| 316 |
+
"learning_rate": 1.7244581766840846e-06,
|
| 317 |
+
"loss": 0.4561,
|
| 318 |
+
"step": 40000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 10.12,
|
| 322 |
+
"eval_accuracy": 0.8738906074338904,
|
| 323 |
+
"eval_f1": 0.8728714900247343,
|
| 324 |
+
"eval_loss": 0.4745886027812958,
|
| 325 |
+
"eval_runtime": 52.1859,
|
| 326 |
+
"eval_samples_per_second": 2744.206,
|
| 327 |
+
"eval_steps_per_second": 57.18,
|
| 328 |
+
"step": 40000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 10.63,
|
| 332 |
+
"learning_rate": 1.4101411596796607e-06,
|
| 333 |
+
"loss": 0.4475,
|
| 334 |
+
"step": 42000
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 10.63,
|
| 338 |
+
"eval_accuracy": 0.8749380276379278,
|
| 339 |
+
"eval_f1": 0.8739688992879336,
|
| 340 |
+
"eval_loss": 0.4705161154270172,
|
| 341 |
+
"eval_runtime": 52.4489,
|
| 342 |
+
"eval_samples_per_second": 2730.449,
|
| 343 |
+
"eval_steps_per_second": 56.893,
|
| 344 |
+
"step": 42000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 11.14,
|
| 348 |
+
"learning_rate": 1.1201036727965609e-06,
|
| 349 |
+
"loss": 0.4413,
|
| 350 |
+
"step": 44000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 11.14,
|
| 354 |
+
"eval_accuracy": 0.8754268237331453,
|
| 355 |
+
"eval_f1": 0.8743869003587098,
|
| 356 |
+
"eval_loss": 0.46908074617385864,
|
| 357 |
+
"eval_runtime": 50.2149,
|
| 358 |
+
"eval_samples_per_second": 2851.921,
|
| 359 |
+
"eval_steps_per_second": 59.425,
|
| 360 |
+
"step": 44000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 11.64,
|
| 364 |
+
"learning_rate": 8.577153091208042e-07,
|
| 365 |
+
"loss": 0.4389,
|
| 366 |
+
"step": 46000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 11.64,
|
| 370 |
+
"eval_accuracy": 0.8759644994378845,
|
| 371 |
+
"eval_f1": 0.875008769768229,
|
| 372 |
+
"eval_loss": 0.46789219975471497,
|
| 373 |
+
"eval_runtime": 50.3984,
|
| 374 |
+
"eval_samples_per_second": 2841.539,
|
| 375 |
+
"eval_steps_per_second": 59.208,
|
| 376 |
+
"step": 46000
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 12.15,
|
| 380 |
+
"learning_rate": 6.260244401770287e-07,
|
| 381 |
+
"loss": 0.4361,
|
| 382 |
+
"step": 48000
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 12.15,
|
| 386 |
+
"eval_accuracy": 0.875894671424282,
|
| 387 |
+
"eval_f1": 0.8749081220424403,
|
| 388 |
+
"eval_loss": 0.46769237518310547,
|
| 389 |
+
"eval_runtime": 50.042,
|
| 390 |
+
"eval_samples_per_second": 2861.778,
|
| 391 |
+
"eval_steps_per_second": 59.63,
|
| 392 |
+
"step": 48000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 12.65,
|
| 396 |
+
"learning_rate": 4.2772280060435426e-07,
|
| 397 |
+
"loss": 0.4362,
|
| 398 |
+
"step": 50000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 12.65,
|
| 402 |
+
"eval_accuracy": 0.876257777095015,
|
| 403 |
+
"eval_f1": 0.875265867499827,
|
| 404 |
+
"eval_loss": 0.4671792984008789,
|
| 405 |
+
"eval_runtime": 52.5275,
|
| 406 |
+
"eval_samples_per_second": 2726.361,
|
| 407 |
+
"eval_steps_per_second": 56.808,
|
| 408 |
+
"step": 50000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 13.16,
|
| 412 |
+
"learning_rate": 2.6511421616278837e-07,
|
| 413 |
+
"loss": 0.4309,
|
| 414 |
+
"step": 52000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 13.16,
|
| 418 |
+
"eval_accuracy": 0.87611812106781,
|
| 419 |
+
"eval_f1": 0.8751424993393537,
|
| 420 |
+
"eval_loss": 0.4671061038970947,
|
| 421 |
+
"eval_runtime": 51.1386,
|
| 422 |
+
"eval_samples_per_second": 2800.408,
|
| 423 |
+
"eval_steps_per_second": 58.351,
|
| 424 |
+
"step": 52000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 13.67,
|
| 428 |
+
"learning_rate": 1.400878383814987e-07,
|
| 429 |
+
"loss": 0.4316,
|
| 430 |
+
"step": 54000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 13.67,
|
| 434 |
+
"eval_accuracy": 0.8764183815263007,
|
| 435 |
+
"eval_f1": 0.8754265388560385,
|
| 436 |
+
"eval_loss": 0.4670025110244751,
|
| 437 |
+
"eval_runtime": 51.6718,
|
| 438 |
+
"eval_samples_per_second": 2771.511,
|
| 439 |
+
"eval_steps_per_second": 57.749,
|
| 440 |
+
"step": 54000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 14.17,
|
| 444 |
+
"learning_rate": 5.409619680302491e-08,
|
| 445 |
+
"loss": 0.4321,
|
| 446 |
+
"step": 56000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 14.17,
|
| 450 |
+
"eval_accuracy": 0.8764393299303814,
|
| 451 |
+
"eval_f1": 0.8754605862048482,
|
| 452 |
+
"eval_loss": 0.4668178856372833,
|
| 453 |
+
"eval_runtime": 50.8951,
|
| 454 |
+
"eval_samples_per_second": 2813.807,
|
| 455 |
+
"eval_steps_per_second": 58.63,
|
| 456 |
+
"step": 56000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 14.68,
|
| 460 |
+
"learning_rate": 8.138323807738212e-09,
|
| 461 |
+
"loss": 0.4311,
|
| 462 |
+
"step": 58000
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 14.68,
|
| 466 |
+
"eval_accuracy": 0.8763904503208597,
|
| 467 |
+
"eval_f1": 0.8754165344722714,
|
| 468 |
+
"eval_loss": 0.46675533056259155,
|
| 469 |
+
"eval_runtime": 50.6992,
|
| 470 |
+
"eval_samples_per_second": 2824.68,
|
| 471 |
+
"eval_steps_per_second": 58.857,
|
| 472 |
+
"step": 58000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 15.0,
|
| 476 |
+
"step": 59265,
|
| 477 |
+
"total_flos": 3.822729571998278e+17,
|
| 478 |
+
"train_loss": 0.4135634679527994,
|
| 479 |
+
"train_runtime": 19096.9194,
|
| 480 |
+
"train_samples_per_second": 893.953,
|
| 481 |
+
"train_steps_per_second": 3.103
|
| 482 |
+
}
|
| 483 |
+
],
|
| 484 |
+
"max_steps": 59265,
|
| 485 |
+
"num_train_epochs": 15,
|
| 486 |
+
"total_flos": 3.822729571998278e+17,
|
| 487 |
+
"trial_name": null,
|
| 488 |
+
"trial_params": null
|
| 489 |
+
}
|