End of training
Browse files- README.md +6 -2
- all_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +564 -84
README.md
CHANGED
|
@@ -40,8 +40,12 @@ The following hyperparameters were used during training:
|
|
| 40 |
- total_train_batch_size: 64
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
-
- lr_scheduler_warmup_steps:
|
| 44 |
-
- training_steps:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
|
| 46 |
### Framework versions
|
| 47 |
|
|
|
|
| 40 |
- total_train_batch_size: 64
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
+
- lr_scheduler_warmup_steps: 5
|
| 44 |
+
- training_steps: 500
|
| 45 |
+
|
| 46 |
+
### Training results
|
| 47 |
+
|
| 48 |
+
|
| 49 |
|
| 50 |
### Framework versions
|
| 51 |
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.56,
|
| 3 |
+
"total_flos": 1.3764276825690931e+17,
|
| 4 |
+
"train_loss": 0.8930109763145446,
|
| 5 |
+
"train_runtime": 25972.5576,
|
| 6 |
+
"train_samples_per_second": 1.232,
|
| 7 |
+
"train_steps_per_second": 0.019
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.56,
|
| 3 |
+
"total_flos": 1.3764276825690931e+17,
|
| 4 |
+
"train_loss": 0.8930109763145446,
|
| 5 |
+
"train_runtime": 25972.5576,
|
| 6 |
+
"train_samples_per_second": 1.232,
|
| 7 |
+
"train_steps_per_second": 0.019
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,148 +1,628 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 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 |
-
"learning_rate": 0.
|
| 14 |
-
"loss":
|
| 15 |
-
"step":
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 0.
|
| 20 |
-
"loss":
|
| 21 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 0.
|
| 26 |
-
"loss":
|
| 27 |
-
"step":
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
"learning_rate": 0.00017777777777777779,
|
| 32 |
-
"loss":
|
| 33 |
-
"step":
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 0.
|
| 38 |
-
"loss":
|
| 39 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
"learning_rate": 0.00015555555555555556,
|
| 44 |
-
"loss":
|
| 45 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss":
|
| 51 |
-
"step":
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
"learning_rate": 0.00013333333333333334,
|
| 56 |
-
"loss":
|
| 57 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 0.
|
| 62 |
-
"loss":
|
| 63 |
-
"step":
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
"learning_rate": 0.00011111111111111112,
|
| 68 |
-
"loss":
|
| 69 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss":
|
| 75 |
-
"step":
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
"learning_rate": 8.888888888888889e-05,
|
| 80 |
-
"loss":
|
| 81 |
-
"step":
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
-
"step":
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
"learning_rate": 6.666666666666667e-05,
|
| 92 |
-
"loss":
|
| 93 |
-
"step":
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss":
|
| 99 |
-
"step":
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
"learning_rate": 4.4444444444444447e-05,
|
| 104 |
-
"loss":
|
| 105 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss":
|
| 111 |
-
"step":
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 115 |
"learning_rate": 2.2222222222222223e-05,
|
| 116 |
-
"loss":
|
| 117 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate": 1.
|
| 122 |
-
"loss":
|
| 123 |
-
"step":
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 127 |
"learning_rate": 0.0,
|
| 128 |
-
"loss":
|
| 129 |
-
"step":
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"step":
|
| 134 |
-
"total_flos":
|
| 135 |
-
"train_loss":
|
| 136 |
-
"train_runtime":
|
| 137 |
-
"train_samples_per_second":
|
| 138 |
-
"train_steps_per_second": 0.
|
| 139 |
}
|
| 140 |
],
|
| 141 |
-
"logging_steps":
|
| 142 |
-
"max_steps":
|
| 143 |
-
"num_train_epochs":
|
| 144 |
"save_steps": 500,
|
| 145 |
-
"total_flos":
|
| 146 |
"trial_name": null,
|
| 147 |
"trial_params": null
|
| 148 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.558846453624318,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 500,
|
| 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 |
+
"learning_rate": 0.0002,
|
| 14 |
+
"loss": 1.9726,
|
| 15 |
+
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 0.000197979797979798,
|
| 20 |
+
"loss": 1.3188,
|
| 21 |
+
"step": 10
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.05,
|
| 25 |
+
"learning_rate": 0.00019595959595959596,
|
| 26 |
+
"loss": 1.1313,
|
| 27 |
+
"step": 15
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.06,
|
| 31 |
+
"learning_rate": 0.00019393939393939395,
|
| 32 |
+
"loss": 1.0928,
|
| 33 |
+
"step": 20
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.08,
|
| 37 |
+
"learning_rate": 0.00019191919191919191,
|
| 38 |
+
"loss": 1.0552,
|
| 39 |
+
"step": 25
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.09,
|
| 43 |
+
"learning_rate": 0.0001898989898989899,
|
| 44 |
+
"loss": 1.0644,
|
| 45 |
+
"step": 30
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11,
|
| 49 |
+
"learning_rate": 0.0001878787878787879,
|
| 50 |
+
"loss": 1.0013,
|
| 51 |
+
"step": 35
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"learning_rate": 0.00018585858585858586,
|
| 56 |
+
"loss": 1.0065,
|
| 57 |
+
"step": 40
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.14,
|
| 61 |
+
"learning_rate": 0.00018383838383838384,
|
| 62 |
+
"loss": 1.0054,
|
| 63 |
+
"step": 45
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.16,
|
| 67 |
+
"learning_rate": 0.00018181818181818183,
|
| 68 |
+
"loss": 0.9734,
|
| 69 |
+
"step": 50
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.17,
|
| 73 |
+
"learning_rate": 0.0001797979797979798,
|
| 74 |
+
"loss": 0.9654,
|
| 75 |
+
"step": 55
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.19,
|
| 79 |
"learning_rate": 0.00017777777777777779,
|
| 80 |
+
"loss": 0.9673,
|
| 81 |
+
"step": 60
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.2,
|
| 85 |
+
"learning_rate": 0.00017575757575757578,
|
| 86 |
+
"loss": 0.9584,
|
| 87 |
+
"step": 65
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22,
|
| 91 |
+
"learning_rate": 0.00017373737373737377,
|
| 92 |
+
"loss": 0.9694,
|
| 93 |
+
"step": 70
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"learning_rate": 0.00017171717171717173,
|
| 98 |
+
"loss": 0.9773,
|
| 99 |
+
"step": 75
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.25,
|
| 103 |
+
"learning_rate": 0.00016969696969696972,
|
| 104 |
+
"loss": 0.9531,
|
| 105 |
+
"step": 80
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.27,
|
| 109 |
+
"learning_rate": 0.00016767676767676768,
|
| 110 |
+
"loss": 0.9775,
|
| 111 |
+
"step": 85
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.28,
|
| 115 |
+
"learning_rate": 0.00016565656565656567,
|
| 116 |
+
"loss": 0.9261,
|
| 117 |
+
"step": 90
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.3,
|
| 121 |
+
"learning_rate": 0.00016363636363636366,
|
| 122 |
+
"loss": 0.9715,
|
| 123 |
+
"step": 95
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.31,
|
| 127 |
+
"learning_rate": 0.00016161616161616162,
|
| 128 |
+
"loss": 0.9392,
|
| 129 |
+
"step": 100
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33,
|
| 133 |
+
"learning_rate": 0.0001595959595959596,
|
| 134 |
+
"loss": 0.9722,
|
| 135 |
+
"step": 105
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34,
|
| 139 |
+
"learning_rate": 0.00015757575757575757,
|
| 140 |
+
"loss": 0.9546,
|
| 141 |
+
"step": 110
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.36,
|
| 145 |
"learning_rate": 0.00015555555555555556,
|
| 146 |
+
"loss": 0.9177,
|
| 147 |
+
"step": 115
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.37,
|
| 151 |
+
"learning_rate": 0.00015353535353535353,
|
| 152 |
+
"loss": 0.8817,
|
| 153 |
+
"step": 120
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.39,
|
| 157 |
+
"learning_rate": 0.00015151515151515152,
|
| 158 |
+
"loss": 0.9303,
|
| 159 |
+
"step": 125
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.41,
|
| 163 |
+
"learning_rate": 0.0001494949494949495,
|
| 164 |
+
"loss": 0.9237,
|
| 165 |
+
"step": 130
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.42,
|
| 169 |
+
"learning_rate": 0.00014747474747474747,
|
| 170 |
+
"loss": 0.9136,
|
| 171 |
+
"step": 135
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.44,
|
| 175 |
+
"learning_rate": 0.00014545454545454546,
|
| 176 |
+
"loss": 0.9011,
|
| 177 |
+
"step": 140
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.45,
|
| 181 |
+
"learning_rate": 0.00014343434343434342,
|
| 182 |
+
"loss": 0.9071,
|
| 183 |
+
"step": 145
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.47,
|
| 187 |
+
"learning_rate": 0.0001414141414141414,
|
| 188 |
+
"loss": 0.9315,
|
| 189 |
+
"step": 150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.48,
|
| 193 |
+
"learning_rate": 0.0001393939393939394,
|
| 194 |
+
"loss": 0.8887,
|
| 195 |
+
"step": 155
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.5,
|
| 199 |
+
"learning_rate": 0.0001373737373737374,
|
| 200 |
+
"loss": 0.9104,
|
| 201 |
+
"step": 160
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.51,
|
| 205 |
+
"learning_rate": 0.00013535353535353538,
|
| 206 |
+
"loss": 0.9178,
|
| 207 |
+
"step": 165
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.53,
|
| 211 |
"learning_rate": 0.00013333333333333334,
|
| 212 |
+
"loss": 0.9236,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.55,
|
| 217 |
+
"learning_rate": 0.00013131313131313133,
|
| 218 |
+
"loss": 0.908,
|
| 219 |
+
"step": 175
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.56,
|
| 223 |
+
"learning_rate": 0.00012929292929292932,
|
| 224 |
+
"loss": 0.9268,
|
| 225 |
+
"step": 180
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.58,
|
| 229 |
+
"learning_rate": 0.00012727272727272728,
|
| 230 |
+
"loss": 0.8726,
|
| 231 |
+
"step": 185
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.59,
|
| 235 |
+
"learning_rate": 0.00012525252525252527,
|
| 236 |
+
"loss": 0.9305,
|
| 237 |
+
"step": 190
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.61,
|
| 241 |
+
"learning_rate": 0.00012323232323232323,
|
| 242 |
+
"loss": 0.9196,
|
| 243 |
+
"step": 195
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.62,
|
| 247 |
+
"learning_rate": 0.00012121212121212122,
|
| 248 |
+
"loss": 0.8803,
|
| 249 |
+
"step": 200
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.64,
|
| 253 |
+
"learning_rate": 0.00011919191919191919,
|
| 254 |
+
"loss": 0.8714,
|
| 255 |
+
"step": 205
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.65,
|
| 259 |
+
"learning_rate": 0.00011717171717171717,
|
| 260 |
+
"loss": 0.903,
|
| 261 |
+
"step": 210
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.67,
|
| 265 |
+
"learning_rate": 0.00011515151515151516,
|
| 266 |
+
"loss": 0.8812,
|
| 267 |
+
"step": 215
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 0.69,
|
| 271 |
+
"learning_rate": 0.00011313131313131313,
|
| 272 |
+
"loss": 0.8887,
|
| 273 |
+
"step": 220
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.7,
|
| 277 |
"learning_rate": 0.00011111111111111112,
|
| 278 |
+
"loss": 0.8708,
|
| 279 |
+
"step": 225
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.72,
|
| 283 |
+
"learning_rate": 0.00010909090909090909,
|
| 284 |
+
"loss": 0.8859,
|
| 285 |
+
"step": 230
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.73,
|
| 289 |
+
"learning_rate": 0.00010707070707070708,
|
| 290 |
+
"loss": 0.922,
|
| 291 |
+
"step": 235
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.75,
|
| 295 |
+
"learning_rate": 0.00010505050505050507,
|
| 296 |
+
"loss": 0.8849,
|
| 297 |
+
"step": 240
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.76,
|
| 301 |
+
"learning_rate": 0.00010303030303030303,
|
| 302 |
+
"loss": 0.8729,
|
| 303 |
+
"step": 245
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.78,
|
| 307 |
+
"learning_rate": 0.00010101010101010102,
|
| 308 |
+
"loss": 0.8851,
|
| 309 |
+
"step": 250
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.8,
|
| 313 |
+
"learning_rate": 9.8989898989899e-05,
|
| 314 |
+
"loss": 0.9178,
|
| 315 |
+
"step": 255
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.81,
|
| 319 |
+
"learning_rate": 9.696969696969698e-05,
|
| 320 |
+
"loss": 0.8472,
|
| 321 |
+
"step": 260
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.83,
|
| 325 |
+
"learning_rate": 9.494949494949495e-05,
|
| 326 |
+
"loss": 0.8691,
|
| 327 |
+
"step": 265
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.84,
|
| 331 |
+
"learning_rate": 9.292929292929293e-05,
|
| 332 |
+
"loss": 0.8663,
|
| 333 |
+
"step": 270
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 0.86,
|
| 337 |
+
"learning_rate": 9.090909090909092e-05,
|
| 338 |
+
"loss": 0.8658,
|
| 339 |
+
"step": 275
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.87,
|
| 343 |
"learning_rate": 8.888888888888889e-05,
|
| 344 |
+
"loss": 0.8641,
|
| 345 |
+
"step": 280
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.89,
|
| 349 |
+
"learning_rate": 8.686868686868688e-05,
|
| 350 |
+
"loss": 0.8331,
|
| 351 |
+
"step": 285
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 0.9,
|
| 355 |
+
"learning_rate": 8.484848484848486e-05,
|
| 356 |
+
"loss": 0.8843,
|
| 357 |
+
"step": 290
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.92,
|
| 361 |
+
"learning_rate": 8.282828282828283e-05,
|
| 362 |
+
"loss": 0.8753,
|
| 363 |
+
"step": 295
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.94,
|
| 367 |
+
"learning_rate": 8.080808080808081e-05,
|
| 368 |
+
"loss": 0.8589,
|
| 369 |
+
"step": 300
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.95,
|
| 373 |
+
"learning_rate": 7.878787878787879e-05,
|
| 374 |
+
"loss": 0.8395,
|
| 375 |
+
"step": 305
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.97,
|
| 379 |
+
"learning_rate": 7.676767676767676e-05,
|
| 380 |
+
"loss": 0.852,
|
| 381 |
+
"step": 310
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.98,
|
| 385 |
+
"learning_rate": 7.474747474747475e-05,
|
| 386 |
+
"loss": 0.8413,
|
| 387 |
+
"step": 315
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.0,
|
| 391 |
+
"learning_rate": 7.272727272727273e-05,
|
| 392 |
+
"loss": 0.866,
|
| 393 |
+
"step": 320
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.01,
|
| 397 |
+
"learning_rate": 7.07070707070707e-05,
|
| 398 |
+
"loss": 0.8244,
|
| 399 |
+
"step": 325
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.03,
|
| 403 |
+
"learning_rate": 6.86868686868687e-05,
|
| 404 |
+
"loss": 0.79,
|
| 405 |
+
"step": 330
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.04,
|
| 409 |
"learning_rate": 6.666666666666667e-05,
|
| 410 |
+
"loss": 0.7924,
|
| 411 |
+
"step": 335
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 1.06,
|
| 415 |
+
"learning_rate": 6.464646464646466e-05,
|
| 416 |
+
"loss": 0.8056,
|
| 417 |
+
"step": 340
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 1.08,
|
| 421 |
+
"learning_rate": 6.262626262626264e-05,
|
| 422 |
+
"loss": 0.7935,
|
| 423 |
+
"step": 345
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.09,
|
| 427 |
+
"learning_rate": 6.060606060606061e-05,
|
| 428 |
+
"loss": 0.8212,
|
| 429 |
+
"step": 350
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.11,
|
| 433 |
+
"learning_rate": 5.858585858585859e-05,
|
| 434 |
+
"loss": 0.7993,
|
| 435 |
+
"step": 355
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.12,
|
| 439 |
+
"learning_rate": 5.6565656565656563e-05,
|
| 440 |
+
"loss": 0.8074,
|
| 441 |
+
"step": 360
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.14,
|
| 445 |
+
"learning_rate": 5.4545454545454546e-05,
|
| 446 |
+
"loss": 0.8426,
|
| 447 |
+
"step": 365
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.15,
|
| 451 |
+
"learning_rate": 5.2525252525252536e-05,
|
| 452 |
+
"loss": 0.796,
|
| 453 |
+
"step": 370
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.17,
|
| 457 |
+
"learning_rate": 5.050505050505051e-05,
|
| 458 |
+
"loss": 0.783,
|
| 459 |
+
"step": 375
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.18,
|
| 463 |
+
"learning_rate": 4.848484848484849e-05,
|
| 464 |
+
"loss": 0.801,
|
| 465 |
+
"step": 380
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2,
|
| 469 |
+
"learning_rate": 4.6464646464646464e-05,
|
| 470 |
+
"loss": 0.8028,
|
| 471 |
+
"step": 385
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.22,
|
| 475 |
"learning_rate": 4.4444444444444447e-05,
|
| 476 |
+
"loss": 0.7935,
|
| 477 |
+
"step": 390
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.23,
|
| 481 |
+
"learning_rate": 4.242424242424243e-05,
|
| 482 |
+
"loss": 0.7957,
|
| 483 |
+
"step": 395
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.25,
|
| 487 |
+
"learning_rate": 4.0404040404040405e-05,
|
| 488 |
+
"loss": 0.8139,
|
| 489 |
+
"step": 400
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.26,
|
| 493 |
+
"learning_rate": 3.838383838383838e-05,
|
| 494 |
+
"loss": 0.7718,
|
| 495 |
+
"step": 405
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.28,
|
| 499 |
+
"learning_rate": 3.6363636363636364e-05,
|
| 500 |
+
"loss": 0.8142,
|
| 501 |
+
"step": 410
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.29,
|
| 505 |
+
"learning_rate": 3.434343434343435e-05,
|
| 506 |
+
"loss": 0.8083,
|
| 507 |
+
"step": 415
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.31,
|
| 511 |
+
"learning_rate": 3.232323232323233e-05,
|
| 512 |
+
"loss": 0.7984,
|
| 513 |
+
"step": 420
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.33,
|
| 517 |
+
"learning_rate": 3.0303030303030306e-05,
|
| 518 |
+
"loss": 0.7968,
|
| 519 |
+
"step": 425
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.34,
|
| 523 |
+
"learning_rate": 2.8282828282828282e-05,
|
| 524 |
+
"loss": 0.7945,
|
| 525 |
+
"step": 430
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 1.36,
|
| 529 |
+
"learning_rate": 2.6262626262626268e-05,
|
| 530 |
+
"loss": 0.803,
|
| 531 |
+
"step": 435
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 1.37,
|
| 535 |
+
"learning_rate": 2.4242424242424244e-05,
|
| 536 |
+
"loss": 0.8167,
|
| 537 |
+
"step": 440
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.39,
|
| 541 |
"learning_rate": 2.2222222222222223e-05,
|
| 542 |
+
"loss": 0.7801,
|
| 543 |
+
"step": 445
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.4,
|
| 547 |
+
"learning_rate": 2.0202020202020203e-05,
|
| 548 |
+
"loss": 0.8059,
|
| 549 |
+
"step": 450
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.42,
|
| 553 |
+
"learning_rate": 1.8181818181818182e-05,
|
| 554 |
+
"loss": 0.7927,
|
| 555 |
+
"step": 455
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 1.43,
|
| 559 |
+
"learning_rate": 1.6161616161616165e-05,
|
| 560 |
+
"loss": 0.7962,
|
| 561 |
+
"step": 460
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 1.45,
|
| 565 |
+
"learning_rate": 1.4141414141414141e-05,
|
| 566 |
+
"loss": 0.7893,
|
| 567 |
+
"step": 465
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 1.47,
|
| 571 |
+
"learning_rate": 1.2121212121212122e-05,
|
| 572 |
+
"loss": 0.7965,
|
| 573 |
+
"step": 470
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 1.48,
|
| 577 |
+
"learning_rate": 1.0101010101010101e-05,
|
| 578 |
+
"loss": 0.7817,
|
| 579 |
+
"step": 475
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.5,
|
| 583 |
+
"learning_rate": 8.080808080808082e-06,
|
| 584 |
+
"loss": 0.8159,
|
| 585 |
+
"step": 480
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.51,
|
| 589 |
+
"learning_rate": 6.060606060606061e-06,
|
| 590 |
+
"loss": 0.7822,
|
| 591 |
+
"step": 485
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.53,
|
| 595 |
+
"learning_rate": 4.040404040404041e-06,
|
| 596 |
+
"loss": 0.7937,
|
| 597 |
+
"step": 490
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.54,
|
| 601 |
+
"learning_rate": 2.0202020202020206e-06,
|
| 602 |
+
"loss": 0.8044,
|
| 603 |
+
"step": 495
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 1.56,
|
| 607 |
"learning_rate": 0.0,
|
| 608 |
+
"loss": 0.8113,
|
| 609 |
+
"step": 500
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 1.56,
|
| 613 |
+
"step": 500,
|
| 614 |
+
"total_flos": 1.3764276825690931e+17,
|
| 615 |
+
"train_loss": 0.8930109763145446,
|
| 616 |
+
"train_runtime": 25972.5576,
|
| 617 |
+
"train_samples_per_second": 1.232,
|
| 618 |
+
"train_steps_per_second": 0.019
|
| 619 |
}
|
| 620 |
],
|
| 621 |
+
"logging_steps": 5,
|
| 622 |
+
"max_steps": 500,
|
| 623 |
+
"num_train_epochs": 2,
|
| 624 |
"save_steps": 500,
|
| 625 |
+
"total_flos": 1.3764276825690931e+17,
|
| 626 |
"trial_name": null,
|
| 627 |
"trial_params": null
|
| 628 |
}
|