End of training
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +782 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 1.0028942094237696e+19,
|
| 4 |
+
"train_loss": 0.6026930379867553,
|
| 5 |
+
"train_runtime": 35487.7868,
|
| 6 |
+
"train_samples_per_second": 3.607,
|
| 7 |
+
"train_steps_per_second": 0.225
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 1.0028942094237696e+19,
|
| 4 |
+
"train_loss": 0.6026930379867553,
|
| 5 |
+
"train_runtime": 35487.7868,
|
| 6 |
+
"train_samples_per_second": 3.607,
|
| 7 |
+
"train_steps_per_second": 0.225
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,782 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.51348477602005,
|
| 3 |
+
"best_model_checkpoint": "./training_output/swin-base-patch4-window7-224_11092024\\checkpoint-8000",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.25,
|
| 13 |
+
"grad_norm": 4.37050724029541,
|
| 14 |
+
"learning_rate": 0.00029624999999999996,
|
| 15 |
+
"loss": 1.6781,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.5,
|
| 20 |
+
"grad_norm": 3.8442654609680176,
|
| 21 |
+
"learning_rate": 0.00029249999999999995,
|
| 22 |
+
"loss": 1.2684,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.75,
|
| 27 |
+
"grad_norm": 3.500859498977661,
|
| 28 |
+
"learning_rate": 0.00028875,
|
| 29 |
+
"loss": 1.0998,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.0,
|
| 34 |
+
"grad_norm": 5.394394874572754,
|
| 35 |
+
"learning_rate": 0.000285,
|
| 36 |
+
"loss": 1.0366,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.0,
|
| 41 |
+
"eval_accuracy": 0.72,
|
| 42 |
+
"eval_loss": 0.9471067190170288,
|
| 43 |
+
"eval_runtime": 191.6064,
|
| 44 |
+
"eval_samples_per_second": 4.175,
|
| 45 |
+
"eval_steps_per_second": 0.522,
|
| 46 |
+
"step": 400
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.25,
|
| 50 |
+
"grad_norm": 5.127144813537598,
|
| 51 |
+
"learning_rate": 0.00028125,
|
| 52 |
+
"loss": 0.9272,
|
| 53 |
+
"step": 500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.5,
|
| 57 |
+
"grad_norm": 3.240920305252075,
|
| 58 |
+
"learning_rate": 0.00027749999999999997,
|
| 59 |
+
"loss": 0.8831,
|
| 60 |
+
"step": 600
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.75,
|
| 64 |
+
"grad_norm": 2.9190714359283447,
|
| 65 |
+
"learning_rate": 0.00027374999999999996,
|
| 66 |
+
"loss": 0.8501,
|
| 67 |
+
"step": 700
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 2.0,
|
| 71 |
+
"grad_norm": 4.049932956695557,
|
| 72 |
+
"learning_rate": 0.00027,
|
| 73 |
+
"loss": 0.8257,
|
| 74 |
+
"step": 800
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.0,
|
| 78 |
+
"eval_accuracy": 0.75375,
|
| 79 |
+
"eval_loss": 0.7888995409011841,
|
| 80 |
+
"eval_runtime": 190.7471,
|
| 81 |
+
"eval_samples_per_second": 4.194,
|
| 82 |
+
"eval_steps_per_second": 0.524,
|
| 83 |
+
"step": 800
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 2.25,
|
| 87 |
+
"grad_norm": 2.773448944091797,
|
| 88 |
+
"learning_rate": 0.00026624999999999994,
|
| 89 |
+
"loss": 0.804,
|
| 90 |
+
"step": 900
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.5,
|
| 94 |
+
"grad_norm": 4.934525489807129,
|
| 95 |
+
"learning_rate": 0.0002625,
|
| 96 |
+
"loss": 0.7622,
|
| 97 |
+
"step": 1000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.75,
|
| 101 |
+
"grad_norm": 2.755113363265991,
|
| 102 |
+
"learning_rate": 0.00025875,
|
| 103 |
+
"loss": 0.7747,
|
| 104 |
+
"step": 1100
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.0,
|
| 108 |
+
"grad_norm": 2.5829150676727295,
|
| 109 |
+
"learning_rate": 0.00025499999999999996,
|
| 110 |
+
"loss": 0.7119,
|
| 111 |
+
"step": 1200
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 3.0,
|
| 115 |
+
"eval_accuracy": 0.7775,
|
| 116 |
+
"eval_loss": 0.723233699798584,
|
| 117 |
+
"eval_runtime": 192.2538,
|
| 118 |
+
"eval_samples_per_second": 4.161,
|
| 119 |
+
"eval_steps_per_second": 0.52,
|
| 120 |
+
"step": 1200
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 3.25,
|
| 124 |
+
"grad_norm": 3.0522494316101074,
|
| 125 |
+
"learning_rate": 0.00025125,
|
| 126 |
+
"loss": 0.7059,
|
| 127 |
+
"step": 1300
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.5,
|
| 131 |
+
"grad_norm": 2.336416482925415,
|
| 132 |
+
"learning_rate": 0.00024749999999999994,
|
| 133 |
+
"loss": 0.6803,
|
| 134 |
+
"step": 1400
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.75,
|
| 138 |
+
"grad_norm": 2.500382900238037,
|
| 139 |
+
"learning_rate": 0.00024375,
|
| 140 |
+
"loss": 0.6944,
|
| 141 |
+
"step": 1500
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 4.0,
|
| 145 |
+
"grad_norm": 2.7047812938690186,
|
| 146 |
+
"learning_rate": 0.00023999999999999998,
|
| 147 |
+
"loss": 0.6969,
|
| 148 |
+
"step": 1600
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 4.0,
|
| 152 |
+
"eval_accuracy": 0.78375,
|
| 153 |
+
"eval_loss": 0.673926830291748,
|
| 154 |
+
"eval_runtime": 188.8251,
|
| 155 |
+
"eval_samples_per_second": 4.237,
|
| 156 |
+
"eval_steps_per_second": 0.53,
|
| 157 |
+
"step": 1600
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 4.25,
|
| 161 |
+
"grad_norm": 1.99785315990448,
|
| 162 |
+
"learning_rate": 0.00023624999999999997,
|
| 163 |
+
"loss": 0.6529,
|
| 164 |
+
"step": 1700
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.5,
|
| 168 |
+
"grad_norm": 2.6651065349578857,
|
| 169 |
+
"learning_rate": 0.00023249999999999999,
|
| 170 |
+
"loss": 0.6514,
|
| 171 |
+
"step": 1800
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 4.75,
|
| 175 |
+
"grad_norm": 2.7539191246032715,
|
| 176 |
+
"learning_rate": 0.00022874999999999998,
|
| 177 |
+
"loss": 0.6553,
|
| 178 |
+
"step": 1900
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 5.0,
|
| 182 |
+
"grad_norm": 2.1046321392059326,
|
| 183 |
+
"learning_rate": 0.000225,
|
| 184 |
+
"loss": 0.6487,
|
| 185 |
+
"step": 2000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.0,
|
| 189 |
+
"eval_accuracy": 0.78625,
|
| 190 |
+
"eval_loss": 0.6371492743492126,
|
| 191 |
+
"eval_runtime": 187.904,
|
| 192 |
+
"eval_samples_per_second": 4.257,
|
| 193 |
+
"eval_steps_per_second": 0.532,
|
| 194 |
+
"step": 2000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 5.25,
|
| 198 |
+
"grad_norm": 3.1880531311035156,
|
| 199 |
+
"learning_rate": 0.00022124999999999998,
|
| 200 |
+
"loss": 0.6217,
|
| 201 |
+
"step": 2100
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 5.5,
|
| 205 |
+
"grad_norm": 2.7244873046875,
|
| 206 |
+
"learning_rate": 0.00021749999999999997,
|
| 207 |
+
"loss": 0.6168,
|
| 208 |
+
"step": 2200
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.75,
|
| 212 |
+
"grad_norm": 3.3323047161102295,
|
| 213 |
+
"learning_rate": 0.00021375,
|
| 214 |
+
"loss": 0.6135,
|
| 215 |
+
"step": 2300
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 6.0,
|
| 219 |
+
"grad_norm": 2.5545437335968018,
|
| 220 |
+
"learning_rate": 0.00020999999999999998,
|
| 221 |
+
"loss": 0.5956,
|
| 222 |
+
"step": 2400
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 6.0,
|
| 226 |
+
"eval_accuracy": 0.78875,
|
| 227 |
+
"eval_loss": 0.6197744607925415,
|
| 228 |
+
"eval_runtime": 190.0271,
|
| 229 |
+
"eval_samples_per_second": 4.21,
|
| 230 |
+
"eval_steps_per_second": 0.526,
|
| 231 |
+
"step": 2400
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 6.25,
|
| 235 |
+
"grad_norm": 2.105663299560547,
|
| 236 |
+
"learning_rate": 0.00020624999999999997,
|
| 237 |
+
"loss": 0.5865,
|
| 238 |
+
"step": 2500
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 6.5,
|
| 242 |
+
"grad_norm": 3.24285626411438,
|
| 243 |
+
"learning_rate": 0.0002025,
|
| 244 |
+
"loss": 0.5933,
|
| 245 |
+
"step": 2600
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 6.75,
|
| 249 |
+
"grad_norm": 2.4483067989349365,
|
| 250 |
+
"learning_rate": 0.00019874999999999998,
|
| 251 |
+
"loss": 0.6034,
|
| 252 |
+
"step": 2700
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 7.0,
|
| 256 |
+
"grad_norm": 3.4036154747009277,
|
| 257 |
+
"learning_rate": 0.000195,
|
| 258 |
+
"loss": 0.5604,
|
| 259 |
+
"step": 2800
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 7.0,
|
| 263 |
+
"eval_accuracy": 0.8025,
|
| 264 |
+
"eval_loss": 0.5941271781921387,
|
| 265 |
+
"eval_runtime": 188.5123,
|
| 266 |
+
"eval_samples_per_second": 4.244,
|
| 267 |
+
"eval_steps_per_second": 0.53,
|
| 268 |
+
"step": 2800
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 7.25,
|
| 272 |
+
"grad_norm": 2.846735954284668,
|
| 273 |
+
"learning_rate": 0.00019124999999999996,
|
| 274 |
+
"loss": 0.5863,
|
| 275 |
+
"step": 2900
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 7.5,
|
| 279 |
+
"grad_norm": 2.242002248764038,
|
| 280 |
+
"learning_rate": 0.00018749999999999998,
|
| 281 |
+
"loss": 0.5762,
|
| 282 |
+
"step": 3000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 7.75,
|
| 286 |
+
"grad_norm": 3.0385332107543945,
|
| 287 |
+
"learning_rate": 0.00018375,
|
| 288 |
+
"loss": 0.5552,
|
| 289 |
+
"step": 3100
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 8.0,
|
| 293 |
+
"grad_norm": 4.019121170043945,
|
| 294 |
+
"learning_rate": 0.00017999999999999998,
|
| 295 |
+
"loss": 0.5732,
|
| 296 |
+
"step": 3200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 8.0,
|
| 300 |
+
"eval_accuracy": 0.795,
|
| 301 |
+
"eval_loss": 0.5866623520851135,
|
| 302 |
+
"eval_runtime": 189.5806,
|
| 303 |
+
"eval_samples_per_second": 4.22,
|
| 304 |
+
"eval_steps_per_second": 0.527,
|
| 305 |
+
"step": 3200
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 8.25,
|
| 309 |
+
"grad_norm": 2.7213943004608154,
|
| 310 |
+
"learning_rate": 0.00017625,
|
| 311 |
+
"loss": 0.5913,
|
| 312 |
+
"step": 3300
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 8.5,
|
| 316 |
+
"grad_norm": 2.490701675415039,
|
| 317 |
+
"learning_rate": 0.00017249999999999996,
|
| 318 |
+
"loss": 0.5469,
|
| 319 |
+
"step": 3400
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 8.75,
|
| 323 |
+
"grad_norm": 2.9301528930664062,
|
| 324 |
+
"learning_rate": 0.00016874999999999998,
|
| 325 |
+
"loss": 0.5188,
|
| 326 |
+
"step": 3500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 9.0,
|
| 330 |
+
"grad_norm": 3.5669283866882324,
|
| 331 |
+
"learning_rate": 0.000165,
|
| 332 |
+
"loss": 0.5578,
|
| 333 |
+
"step": 3600
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 9.0,
|
| 337 |
+
"eval_accuracy": 0.8025,
|
| 338 |
+
"eval_loss": 0.570527195930481,
|
| 339 |
+
"eval_runtime": 187.5953,
|
| 340 |
+
"eval_samples_per_second": 4.265,
|
| 341 |
+
"eval_steps_per_second": 0.533,
|
| 342 |
+
"step": 3600
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 9.25,
|
| 346 |
+
"grad_norm": 2.9957895278930664,
|
| 347 |
+
"learning_rate": 0.00016125,
|
| 348 |
+
"loss": 0.532,
|
| 349 |
+
"step": 3700
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 9.5,
|
| 353 |
+
"grad_norm": 3.6296515464782715,
|
| 354 |
+
"learning_rate": 0.00015749999999999998,
|
| 355 |
+
"loss": 0.5359,
|
| 356 |
+
"step": 3800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.75,
|
| 360 |
+
"grad_norm": 2.170017957687378,
|
| 361 |
+
"learning_rate": 0.00015374999999999997,
|
| 362 |
+
"loss": 0.5334,
|
| 363 |
+
"step": 3900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 10.0,
|
| 367 |
+
"grad_norm": 2.3260116577148438,
|
| 368 |
+
"learning_rate": 0.00015,
|
| 369 |
+
"loss": 0.5449,
|
| 370 |
+
"step": 4000
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 10.0,
|
| 374 |
+
"eval_accuracy": 0.81125,
|
| 375 |
+
"eval_loss": 0.5574603080749512,
|
| 376 |
+
"eval_runtime": 186.6415,
|
| 377 |
+
"eval_samples_per_second": 4.286,
|
| 378 |
+
"eval_steps_per_second": 0.536,
|
| 379 |
+
"step": 4000
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 10.25,
|
| 383 |
+
"grad_norm": 3.2211897373199463,
|
| 384 |
+
"learning_rate": 0.00014624999999999998,
|
| 385 |
+
"loss": 0.5333,
|
| 386 |
+
"step": 4100
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 10.5,
|
| 390 |
+
"grad_norm": 1.6050548553466797,
|
| 391 |
+
"learning_rate": 0.0001425,
|
| 392 |
+
"loss": 0.5206,
|
| 393 |
+
"step": 4200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 10.75,
|
| 397 |
+
"grad_norm": 3.3255255222320557,
|
| 398 |
+
"learning_rate": 0.00013874999999999998,
|
| 399 |
+
"loss": 0.4932,
|
| 400 |
+
"step": 4300
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 11.0,
|
| 404 |
+
"grad_norm": 2.705442190170288,
|
| 405 |
+
"learning_rate": 0.000135,
|
| 406 |
+
"loss": 0.5419,
|
| 407 |
+
"step": 4400
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 11.0,
|
| 411 |
+
"eval_accuracy": 0.82125,
|
| 412 |
+
"eval_loss": 0.5505005121231079,
|
| 413 |
+
"eval_runtime": 186.7804,
|
| 414 |
+
"eval_samples_per_second": 4.283,
|
| 415 |
+
"eval_steps_per_second": 0.535,
|
| 416 |
+
"step": 4400
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 11.25,
|
| 420 |
+
"grad_norm": 2.7391810417175293,
|
| 421 |
+
"learning_rate": 0.00013125,
|
| 422 |
+
"loss": 0.4945,
|
| 423 |
+
"step": 4500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 11.5,
|
| 427 |
+
"grad_norm": 2.746691942214966,
|
| 428 |
+
"learning_rate": 0.00012749999999999998,
|
| 429 |
+
"loss": 0.5196,
|
| 430 |
+
"step": 4600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 11.75,
|
| 434 |
+
"grad_norm": 3.134870767593384,
|
| 435 |
+
"learning_rate": 0.00012374999999999997,
|
| 436 |
+
"loss": 0.548,
|
| 437 |
+
"step": 4700
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 12.0,
|
| 441 |
+
"grad_norm": 3.3967835903167725,
|
| 442 |
+
"learning_rate": 0.00011999999999999999,
|
| 443 |
+
"loss": 0.5086,
|
| 444 |
+
"step": 4800
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 12.0,
|
| 448 |
+
"eval_accuracy": 0.82125,
|
| 449 |
+
"eval_loss": 0.5385071039199829,
|
| 450 |
+
"eval_runtime": 187.7731,
|
| 451 |
+
"eval_samples_per_second": 4.26,
|
| 452 |
+
"eval_steps_per_second": 0.533,
|
| 453 |
+
"step": 4800
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 12.25,
|
| 457 |
+
"grad_norm": 2.0934133529663086,
|
| 458 |
+
"learning_rate": 0.00011624999999999999,
|
| 459 |
+
"loss": 0.5059,
|
| 460 |
+
"step": 4900
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 12.5,
|
| 464 |
+
"grad_norm": 3.039590835571289,
|
| 465 |
+
"learning_rate": 0.0001125,
|
| 466 |
+
"loss": 0.5043,
|
| 467 |
+
"step": 5000
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 12.75,
|
| 471 |
+
"grad_norm": 2.2998437881469727,
|
| 472 |
+
"learning_rate": 0.00010874999999999999,
|
| 473 |
+
"loss": 0.5084,
|
| 474 |
+
"step": 5100
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 13.0,
|
| 478 |
+
"grad_norm": 1.5604509115219116,
|
| 479 |
+
"learning_rate": 0.00010499999999999999,
|
| 480 |
+
"loss": 0.4929,
|
| 481 |
+
"step": 5200
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 13.0,
|
| 485 |
+
"eval_accuracy": 0.82125,
|
| 486 |
+
"eval_loss": 0.5340179204940796,
|
| 487 |
+
"eval_runtime": 189.2687,
|
| 488 |
+
"eval_samples_per_second": 4.227,
|
| 489 |
+
"eval_steps_per_second": 0.528,
|
| 490 |
+
"step": 5200
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 13.25,
|
| 494 |
+
"grad_norm": 4.638809680938721,
|
| 495 |
+
"learning_rate": 0.00010125,
|
| 496 |
+
"loss": 0.5198,
|
| 497 |
+
"step": 5300
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 13.5,
|
| 501 |
+
"grad_norm": 2.361114263534546,
|
| 502 |
+
"learning_rate": 9.75e-05,
|
| 503 |
+
"loss": 0.5016,
|
| 504 |
+
"step": 5400
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 13.75,
|
| 508 |
+
"grad_norm": 1.9207624197006226,
|
| 509 |
+
"learning_rate": 9.374999999999999e-05,
|
| 510 |
+
"loss": 0.5064,
|
| 511 |
+
"step": 5500
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 14.0,
|
| 515 |
+
"grad_norm": 2.6521153450012207,
|
| 516 |
+
"learning_rate": 8.999999999999999e-05,
|
| 517 |
+
"loss": 0.4701,
|
| 518 |
+
"step": 5600
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 14.0,
|
| 522 |
+
"eval_accuracy": 0.81875,
|
| 523 |
+
"eval_loss": 0.5296584963798523,
|
| 524 |
+
"eval_runtime": 188.8651,
|
| 525 |
+
"eval_samples_per_second": 4.236,
|
| 526 |
+
"eval_steps_per_second": 0.529,
|
| 527 |
+
"step": 5600
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 14.25,
|
| 531 |
+
"grad_norm": 1.9743889570236206,
|
| 532 |
+
"learning_rate": 8.624999999999998e-05,
|
| 533 |
+
"loss": 0.5078,
|
| 534 |
+
"step": 5700
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 14.5,
|
| 538 |
+
"grad_norm": 2.232062578201294,
|
| 539 |
+
"learning_rate": 8.25e-05,
|
| 540 |
+
"loss": 0.4723,
|
| 541 |
+
"step": 5800
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.75,
|
| 545 |
+
"grad_norm": 3.261901617050171,
|
| 546 |
+
"learning_rate": 7.874999999999999e-05,
|
| 547 |
+
"loss": 0.5011,
|
| 548 |
+
"step": 5900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 15.0,
|
| 552 |
+
"grad_norm": 2.4302258491516113,
|
| 553 |
+
"learning_rate": 7.5e-05,
|
| 554 |
+
"loss": 0.4803,
|
| 555 |
+
"step": 6000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 15.0,
|
| 559 |
+
"eval_accuracy": 0.8225,
|
| 560 |
+
"eval_loss": 0.5240346789360046,
|
| 561 |
+
"eval_runtime": 189.3391,
|
| 562 |
+
"eval_samples_per_second": 4.225,
|
| 563 |
+
"eval_steps_per_second": 0.528,
|
| 564 |
+
"step": 6000
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 15.25,
|
| 568 |
+
"grad_norm": 2.341607093811035,
|
| 569 |
+
"learning_rate": 7.125e-05,
|
| 570 |
+
"loss": 0.4791,
|
| 571 |
+
"step": 6100
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 15.5,
|
| 575 |
+
"grad_norm": 2.3899717330932617,
|
| 576 |
+
"learning_rate": 6.75e-05,
|
| 577 |
+
"loss": 0.4872,
|
| 578 |
+
"step": 6200
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 15.75,
|
| 582 |
+
"grad_norm": 2.253009557723999,
|
| 583 |
+
"learning_rate": 6.374999999999999e-05,
|
| 584 |
+
"loss": 0.5044,
|
| 585 |
+
"step": 6300
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 16.0,
|
| 589 |
+
"grad_norm": 3.2920048236846924,
|
| 590 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 591 |
+
"loss": 0.4988,
|
| 592 |
+
"step": 6400
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 16.0,
|
| 596 |
+
"eval_accuracy": 0.83,
|
| 597 |
+
"eval_loss": 0.5196738243103027,
|
| 598 |
+
"eval_runtime": 189.4702,
|
| 599 |
+
"eval_samples_per_second": 4.222,
|
| 600 |
+
"eval_steps_per_second": 0.528,
|
| 601 |
+
"step": 6400
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 16.25,
|
| 605 |
+
"grad_norm": 4.486973762512207,
|
| 606 |
+
"learning_rate": 5.625e-05,
|
| 607 |
+
"loss": 0.4831,
|
| 608 |
+
"step": 6500
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 16.5,
|
| 612 |
+
"grad_norm": 1.689079999923706,
|
| 613 |
+
"learning_rate": 5.2499999999999995e-05,
|
| 614 |
+
"loss": 0.4961,
|
| 615 |
+
"step": 6600
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 16.75,
|
| 619 |
+
"grad_norm": 2.2517282962799072,
|
| 620 |
+
"learning_rate": 4.875e-05,
|
| 621 |
+
"loss": 0.489,
|
| 622 |
+
"step": 6700
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 17.0,
|
| 626 |
+
"grad_norm": 2.318455219268799,
|
| 627 |
+
"learning_rate": 4.4999999999999996e-05,
|
| 628 |
+
"loss": 0.4842,
|
| 629 |
+
"step": 6800
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 17.0,
|
| 633 |
+
"eval_accuracy": 0.83125,
|
| 634 |
+
"eval_loss": 0.5164589881896973,
|
| 635 |
+
"eval_runtime": 188.3077,
|
| 636 |
+
"eval_samples_per_second": 4.248,
|
| 637 |
+
"eval_steps_per_second": 0.531,
|
| 638 |
+
"step": 6800
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 17.25,
|
| 642 |
+
"grad_norm": 3.113399028778076,
|
| 643 |
+
"learning_rate": 4.125e-05,
|
| 644 |
+
"loss": 0.4676,
|
| 645 |
+
"step": 6900
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 17.5,
|
| 649 |
+
"grad_norm": 2.0549120903015137,
|
| 650 |
+
"learning_rate": 3.75e-05,
|
| 651 |
+
"loss": 0.4679,
|
| 652 |
+
"step": 7000
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 17.75,
|
| 656 |
+
"grad_norm": 2.0566205978393555,
|
| 657 |
+
"learning_rate": 3.375e-05,
|
| 658 |
+
"loss": 0.4988,
|
| 659 |
+
"step": 7100
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 18.0,
|
| 663 |
+
"grad_norm": 2.392482280731201,
|
| 664 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 665 |
+
"loss": 0.4917,
|
| 666 |
+
"step": 7200
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 18.0,
|
| 670 |
+
"eval_accuracy": 0.83125,
|
| 671 |
+
"eval_loss": 0.5148141384124756,
|
| 672 |
+
"eval_runtime": 188.4138,
|
| 673 |
+
"eval_samples_per_second": 4.246,
|
| 674 |
+
"eval_steps_per_second": 0.531,
|
| 675 |
+
"step": 7200
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 18.25,
|
| 679 |
+
"grad_norm": 1.7607569694519043,
|
| 680 |
+
"learning_rate": 2.6249999999999998e-05,
|
| 681 |
+
"loss": 0.475,
|
| 682 |
+
"step": 7300
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 18.5,
|
| 686 |
+
"grad_norm": 2.870453357696533,
|
| 687 |
+
"learning_rate": 2.2499999999999998e-05,
|
| 688 |
+
"loss": 0.469,
|
| 689 |
+
"step": 7400
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 18.75,
|
| 693 |
+
"grad_norm": 2.928583860397339,
|
| 694 |
+
"learning_rate": 1.875e-05,
|
| 695 |
+
"loss": 0.4691,
|
| 696 |
+
"step": 7500
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 19.0,
|
| 700 |
+
"grad_norm": 3.0488061904907227,
|
| 701 |
+
"learning_rate": 1.4999999999999999e-05,
|
| 702 |
+
"loss": 0.4734,
|
| 703 |
+
"step": 7600
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 19.0,
|
| 707 |
+
"eval_accuracy": 0.8325,
|
| 708 |
+
"eval_loss": 0.5140394568443298,
|
| 709 |
+
"eval_runtime": 187.2755,
|
| 710 |
+
"eval_samples_per_second": 4.272,
|
| 711 |
+
"eval_steps_per_second": 0.534,
|
| 712 |
+
"step": 7600
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 19.25,
|
| 716 |
+
"grad_norm": 3.021597146987915,
|
| 717 |
+
"learning_rate": 1.1249999999999999e-05,
|
| 718 |
+
"loss": 0.5122,
|
| 719 |
+
"step": 7700
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 19.5,
|
| 723 |
+
"grad_norm": 1.8236939907073975,
|
| 724 |
+
"learning_rate": 7.499999999999999e-06,
|
| 725 |
+
"loss": 0.4582,
|
| 726 |
+
"step": 7800
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 19.75,
|
| 730 |
+
"grad_norm": 1.9407800436019897,
|
| 731 |
+
"learning_rate": 3.7499999999999997e-06,
|
| 732 |
+
"loss": 0.4578,
|
| 733 |
+
"step": 7900
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 20.0,
|
| 737 |
+
"grad_norm": 2.3664169311523438,
|
| 738 |
+
"learning_rate": 0.0,
|
| 739 |
+
"loss": 0.4714,
|
| 740 |
+
"step": 8000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 20.0,
|
| 744 |
+
"eval_accuracy": 0.83375,
|
| 745 |
+
"eval_loss": 0.51348477602005,
|
| 746 |
+
"eval_runtime": 192.831,
|
| 747 |
+
"eval_samples_per_second": 4.149,
|
| 748 |
+
"eval_steps_per_second": 0.519,
|
| 749 |
+
"step": 8000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 20.0,
|
| 753 |
+
"step": 8000,
|
| 754 |
+
"total_flos": 1.0028942094237696e+19,
|
| 755 |
+
"train_loss": 0.6026930379867553,
|
| 756 |
+
"train_runtime": 35487.7868,
|
| 757 |
+
"train_samples_per_second": 3.607,
|
| 758 |
+
"train_steps_per_second": 0.225
|
| 759 |
+
}
|
| 760 |
+
],
|
| 761 |
+
"logging_steps": 100,
|
| 762 |
+
"max_steps": 8000,
|
| 763 |
+
"num_input_tokens_seen": 0,
|
| 764 |
+
"num_train_epochs": 20,
|
| 765 |
+
"save_steps": 500,
|
| 766 |
+
"stateful_callbacks": {
|
| 767 |
+
"TrainerControl": {
|
| 768 |
+
"args": {
|
| 769 |
+
"should_epoch_stop": false,
|
| 770 |
+
"should_evaluate": false,
|
| 771 |
+
"should_log": false,
|
| 772 |
+
"should_save": true,
|
| 773 |
+
"should_training_stop": true
|
| 774 |
+
},
|
| 775 |
+
"attributes": {}
|
| 776 |
+
}
|
| 777 |
+
},
|
| 778 |
+
"total_flos": 1.0028942094237696e+19,
|
| 779 |
+
"train_batch_size": 16,
|
| 780 |
+
"trial_name": null,
|
| 781 |
+
"trial_params": null
|
| 782 |
+
}
|