Training in progress, epoch 0
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.99,
|
| 3 |
+
"total_flos": 3.3497451642252165e+18,
|
| 4 |
+
"train_loss": 0.5573047858584188,
|
| 5 |
+
"train_runtime": 2153.5042,
|
| 6 |
+
"train_samples_per_second": 62.689,
|
| 7 |
+
"train_steps_per_second": 0.489
|
| 8 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 110367448
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27b9f1f7bb1403785b476f43a32fd83b1ee073ce0a37d2d8fe21ba5ae1a2f85c
|
| 3 |
size 110367448
|
runs/Apr07_09-39-35_3c5a76d36ea5/events.out.tfevents.1712485033.3c5a76d36ea5.215.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bdb335c0f3dd1b08140a7cad3a4c0e7c4f4dc779a1730148c8c8b74bd9356ce
|
| 3 |
+
size 13005
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.99,
|
| 3 |
+
"total_flos": 3.3497451642252165e+18,
|
| 4 |
+
"train_loss": 0.5573047858584188,
|
| 5 |
+
"train_runtime": 2153.5042,
|
| 6 |
+
"train_samples_per_second": 62.689,
|
| 7 |
+
"train_steps_per_second": 0.489
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,792 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9738,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-1053",
|
| 4 |
+
"epoch": 2.9936034115138592,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1053,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"grad_norm": 5.0049214363098145,
|
| 14 |
+
"learning_rate": 4.716981132075472e-06,
|
| 15 |
+
"loss": 2.3305,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06,
|
| 20 |
+
"grad_norm": 5.08838415145874,
|
| 21 |
+
"learning_rate": 9.433962264150944e-06,
|
| 22 |
+
"loss": 2.2965,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09,
|
| 27 |
+
"grad_norm": 6.870338439941406,
|
| 28 |
+
"learning_rate": 1.4150943396226415e-05,
|
| 29 |
+
"loss": 2.1733,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11,
|
| 34 |
+
"grad_norm": 6.763277530670166,
|
| 35 |
+
"learning_rate": 1.8867924528301888e-05,
|
| 36 |
+
"loss": 1.9936,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.14,
|
| 41 |
+
"grad_norm": 7.859236717224121,
|
| 42 |
+
"learning_rate": 2.358490566037736e-05,
|
| 43 |
+
"loss": 1.7305,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.17,
|
| 48 |
+
"grad_norm": 19.806703567504883,
|
| 49 |
+
"learning_rate": 2.830188679245283e-05,
|
| 50 |
+
"loss": 1.4335,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2,
|
| 55 |
+
"grad_norm": 17.265830993652344,
|
| 56 |
+
"learning_rate": 3.30188679245283e-05,
|
| 57 |
+
"loss": 1.2302,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.23,
|
| 62 |
+
"grad_norm": 10.891965866088867,
|
| 63 |
+
"learning_rate": 3.7735849056603776e-05,
|
| 64 |
+
"loss": 1.0525,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.26,
|
| 69 |
+
"grad_norm": 12.098174095153809,
|
| 70 |
+
"learning_rate": 4.245283018867925e-05,
|
| 71 |
+
"loss": 0.9623,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.28,
|
| 76 |
+
"grad_norm": 15.227210998535156,
|
| 77 |
+
"learning_rate": 4.716981132075472e-05,
|
| 78 |
+
"loss": 0.9037,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.31,
|
| 83 |
+
"grad_norm": 10.019371032714844,
|
| 84 |
+
"learning_rate": 4.978880675818374e-05,
|
| 85 |
+
"loss": 0.8151,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.34,
|
| 90 |
+
"grad_norm": 9.77025318145752,
|
| 91 |
+
"learning_rate": 4.9260823653643085e-05,
|
| 92 |
+
"loss": 0.7041,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.37,
|
| 97 |
+
"grad_norm": 13.53752326965332,
|
| 98 |
+
"learning_rate": 4.8732840549102435e-05,
|
| 99 |
+
"loss": 0.697,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4,
|
| 104 |
+
"grad_norm": 13.287837028503418,
|
| 105 |
+
"learning_rate": 4.820485744456177e-05,
|
| 106 |
+
"loss": 0.7026,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.43,
|
| 111 |
+
"grad_norm": 12.724205017089844,
|
| 112 |
+
"learning_rate": 4.767687434002112e-05,
|
| 113 |
+
"loss": 0.6429,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.45,
|
| 118 |
+
"grad_norm": 12.56704330444336,
|
| 119 |
+
"learning_rate": 4.7148891235480466e-05,
|
| 120 |
+
"loss": 0.6533,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.48,
|
| 125 |
+
"grad_norm": 13.585038185119629,
|
| 126 |
+
"learning_rate": 4.662090813093981e-05,
|
| 127 |
+
"loss": 0.6176,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.51,
|
| 132 |
+
"grad_norm": 9.79969596862793,
|
| 133 |
+
"learning_rate": 4.609292502639916e-05,
|
| 134 |
+
"loss": 0.611,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.54,
|
| 139 |
+
"grad_norm": 16.163864135742188,
|
| 140 |
+
"learning_rate": 4.55649419218585e-05,
|
| 141 |
+
"loss": 0.6085,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.57,
|
| 146 |
+
"grad_norm": 14.128802299499512,
|
| 147 |
+
"learning_rate": 4.503695881731785e-05,
|
| 148 |
+
"loss": 0.6155,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6,
|
| 153 |
+
"grad_norm": 11.335792541503906,
|
| 154 |
+
"learning_rate": 4.45089757127772e-05,
|
| 155 |
+
"loss": 0.6233,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.63,
|
| 160 |
+
"grad_norm": 10.187098503112793,
|
| 161 |
+
"learning_rate": 4.398099260823654e-05,
|
| 162 |
+
"loss": 0.5814,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.65,
|
| 167 |
+
"grad_norm": 8.651871681213379,
|
| 168 |
+
"learning_rate": 4.3453009503695884e-05,
|
| 169 |
+
"loss": 0.537,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.68,
|
| 174 |
+
"grad_norm": 12.85684585571289,
|
| 175 |
+
"learning_rate": 4.292502639915523e-05,
|
| 176 |
+
"loss": 0.553,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.71,
|
| 181 |
+
"grad_norm": 11.812779426574707,
|
| 182 |
+
"learning_rate": 4.239704329461457e-05,
|
| 183 |
+
"loss": 0.5439,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.74,
|
| 188 |
+
"grad_norm": 12.811570167541504,
|
| 189 |
+
"learning_rate": 4.186906019007392e-05,
|
| 190 |
+
"loss": 0.5379,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.77,
|
| 195 |
+
"grad_norm": 7.732424259185791,
|
| 196 |
+
"learning_rate": 4.1341077085533265e-05,
|
| 197 |
+
"loss": 0.5575,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8,
|
| 202 |
+
"grad_norm": 15.759660720825195,
|
| 203 |
+
"learning_rate": 4.081309398099261e-05,
|
| 204 |
+
"loss": 0.491,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.82,
|
| 209 |
+
"grad_norm": 11.291988372802734,
|
| 210 |
+
"learning_rate": 4.028511087645195e-05,
|
| 211 |
+
"loss": 0.5135,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.85,
|
| 216 |
+
"grad_norm": 11.919870376586914,
|
| 217 |
+
"learning_rate": 3.97571277719113e-05,
|
| 218 |
+
"loss": 0.4864,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.88,
|
| 223 |
+
"grad_norm": 11.872406959533691,
|
| 224 |
+
"learning_rate": 3.9229144667370646e-05,
|
| 225 |
+
"loss": 0.4965,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.91,
|
| 230 |
+
"grad_norm": 10.008932113647461,
|
| 231 |
+
"learning_rate": 3.870116156282999e-05,
|
| 232 |
+
"loss": 0.5295,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.94,
|
| 237 |
+
"grad_norm": 8.790617942810059,
|
| 238 |
+
"learning_rate": 3.817317845828934e-05,
|
| 239 |
+
"loss": 0.5108,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.97,
|
| 244 |
+
"grad_norm": 8.419012069702148,
|
| 245 |
+
"learning_rate": 3.764519535374868e-05,
|
| 246 |
+
"loss": 0.483,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0,
|
| 251 |
+
"grad_norm": 10.467920303344727,
|
| 252 |
+
"learning_rate": 3.711721224920803e-05,
|
| 253 |
+
"loss": 0.5161,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0,
|
| 258 |
+
"eval_accuracy": 0.9576,
|
| 259 |
+
"eval_loss": 0.12418650835752487,
|
| 260 |
+
"eval_runtime": 62.9479,
|
| 261 |
+
"eval_samples_per_second": 79.431,
|
| 262 |
+
"eval_steps_per_second": 2.494,
|
| 263 |
+
"step": 351
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.02,
|
| 267 |
+
"grad_norm": 10.26750373840332,
|
| 268 |
+
"learning_rate": 3.658922914466738e-05,
|
| 269 |
+
"loss": 0.4429,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.05,
|
| 274 |
+
"grad_norm": 8.352232933044434,
|
| 275 |
+
"learning_rate": 3.6061246040126714e-05,
|
| 276 |
+
"loss": 0.4627,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.08,
|
| 281 |
+
"grad_norm": 8.920843124389648,
|
| 282 |
+
"learning_rate": 3.5533262935586064e-05,
|
| 283 |
+
"loss": 0.5083,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.11,
|
| 288 |
+
"grad_norm": 13.514233589172363,
|
| 289 |
+
"learning_rate": 3.500527983104541e-05,
|
| 290 |
+
"loss": 0.4609,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.14,
|
| 295 |
+
"grad_norm": 11.564621925354004,
|
| 296 |
+
"learning_rate": 3.447729672650475e-05,
|
| 297 |
+
"loss": 0.4598,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.17,
|
| 302 |
+
"grad_norm": 11.314290046691895,
|
| 303 |
+
"learning_rate": 3.3949313621964095e-05,
|
| 304 |
+
"loss": 0.4097,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.19,
|
| 309 |
+
"grad_norm": 8.917102813720703,
|
| 310 |
+
"learning_rate": 3.3421330517423445e-05,
|
| 311 |
+
"loss": 0.4289,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.22,
|
| 316 |
+
"grad_norm": 10.122758865356445,
|
| 317 |
+
"learning_rate": 3.289334741288279e-05,
|
| 318 |
+
"loss": 0.4413,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.25,
|
| 323 |
+
"grad_norm": 13.456986427307129,
|
| 324 |
+
"learning_rate": 3.236536430834213e-05,
|
| 325 |
+
"loss": 0.4601,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.28,
|
| 330 |
+
"grad_norm": 8.762724876403809,
|
| 331 |
+
"learning_rate": 3.183738120380148e-05,
|
| 332 |
+
"loss": 0.4447,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.31,
|
| 337 |
+
"grad_norm": 8.276448249816895,
|
| 338 |
+
"learning_rate": 3.130939809926082e-05,
|
| 339 |
+
"loss": 0.4482,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.34,
|
| 344 |
+
"grad_norm": 10.242718696594238,
|
| 345 |
+
"learning_rate": 3.078141499472017e-05,
|
| 346 |
+
"loss": 0.4513,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.36,
|
| 351 |
+
"grad_norm": 8.58539867401123,
|
| 352 |
+
"learning_rate": 3.0253431890179517e-05,
|
| 353 |
+
"loss": 0.4348,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.39,
|
| 358 |
+
"grad_norm": 11.27660846710205,
|
| 359 |
+
"learning_rate": 2.972544878563886e-05,
|
| 360 |
+
"loss": 0.4532,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.42,
|
| 365 |
+
"grad_norm": 8.342896461486816,
|
| 366 |
+
"learning_rate": 2.9197465681098207e-05,
|
| 367 |
+
"loss": 0.4256,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.45,
|
| 372 |
+
"grad_norm": 11.779293060302734,
|
| 373 |
+
"learning_rate": 2.8669482576557548e-05,
|
| 374 |
+
"loss": 0.4486,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.48,
|
| 379 |
+
"grad_norm": 8.870821952819824,
|
| 380 |
+
"learning_rate": 2.8141499472016898e-05,
|
| 381 |
+
"loss": 0.4073,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.51,
|
| 386 |
+
"grad_norm": 10.537642478942871,
|
| 387 |
+
"learning_rate": 2.7613516367476245e-05,
|
| 388 |
+
"loss": 0.4492,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.54,
|
| 393 |
+
"grad_norm": 8.932671546936035,
|
| 394 |
+
"learning_rate": 2.7085533262935585e-05,
|
| 395 |
+
"loss": 0.4127,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.56,
|
| 400 |
+
"grad_norm": 11.623788833618164,
|
| 401 |
+
"learning_rate": 2.6557550158394935e-05,
|
| 402 |
+
"loss": 0.481,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.59,
|
| 407 |
+
"grad_norm": 6.948116302490234,
|
| 408 |
+
"learning_rate": 2.6029567053854276e-05,
|
| 409 |
+
"loss": 0.4122,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.62,
|
| 414 |
+
"grad_norm": 9.48883056640625,
|
| 415 |
+
"learning_rate": 2.5501583949313622e-05,
|
| 416 |
+
"loss": 0.4364,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.65,
|
| 421 |
+
"grad_norm": 9.34141731262207,
|
| 422 |
+
"learning_rate": 2.497360084477297e-05,
|
| 423 |
+
"loss": 0.4163,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.68,
|
| 428 |
+
"grad_norm": 10.4086275100708,
|
| 429 |
+
"learning_rate": 2.4445617740232313e-05,
|
| 430 |
+
"loss": 0.4138,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.71,
|
| 435 |
+
"grad_norm": 6.621527671813965,
|
| 436 |
+
"learning_rate": 2.391763463569166e-05,
|
| 437 |
+
"loss": 0.4253,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.73,
|
| 442 |
+
"grad_norm": 9.12219524383545,
|
| 443 |
+
"learning_rate": 2.3389651531151003e-05,
|
| 444 |
+
"loss": 0.4129,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.76,
|
| 449 |
+
"grad_norm": 9.151737213134766,
|
| 450 |
+
"learning_rate": 2.286166842661035e-05,
|
| 451 |
+
"loss": 0.4341,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.79,
|
| 456 |
+
"grad_norm": 11.085309028625488,
|
| 457 |
+
"learning_rate": 2.2333685322069694e-05,
|
| 458 |
+
"loss": 0.4137,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.82,
|
| 463 |
+
"grad_norm": 8.884044647216797,
|
| 464 |
+
"learning_rate": 2.180570221752904e-05,
|
| 465 |
+
"loss": 0.3768,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.85,
|
| 470 |
+
"grad_norm": 9.140060424804688,
|
| 471 |
+
"learning_rate": 2.1277719112988384e-05,
|
| 472 |
+
"loss": 0.4154,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.88,
|
| 477 |
+
"grad_norm": 10.162819862365723,
|
| 478 |
+
"learning_rate": 2.074973600844773e-05,
|
| 479 |
+
"loss": 0.413,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.9,
|
| 484 |
+
"grad_norm": 8.661388397216797,
|
| 485 |
+
"learning_rate": 2.0221752903907075e-05,
|
| 486 |
+
"loss": 0.3951,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.93,
|
| 491 |
+
"grad_norm": 12.425116539001465,
|
| 492 |
+
"learning_rate": 1.9693769799366422e-05,
|
| 493 |
+
"loss": 0.4274,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.96,
|
| 498 |
+
"grad_norm": 10.024824142456055,
|
| 499 |
+
"learning_rate": 1.9165786694825765e-05,
|
| 500 |
+
"loss": 0.3992,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.99,
|
| 505 |
+
"grad_norm": 8.589755058288574,
|
| 506 |
+
"learning_rate": 1.863780359028511e-05,
|
| 507 |
+
"loss": 0.3889,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.0,
|
| 512 |
+
"eval_accuracy": 0.9668,
|
| 513 |
+
"eval_loss": 0.09703098982572556,
|
| 514 |
+
"eval_runtime": 62.5186,
|
| 515 |
+
"eval_samples_per_second": 79.976,
|
| 516 |
+
"eval_steps_per_second": 2.511,
|
| 517 |
+
"step": 703
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.02,
|
| 521 |
+
"grad_norm": 9.084307670593262,
|
| 522 |
+
"learning_rate": 1.810982048574446e-05,
|
| 523 |
+
"loss": 0.3593,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.05,
|
| 528 |
+
"grad_norm": 8.581757545471191,
|
| 529 |
+
"learning_rate": 1.7581837381203803e-05,
|
| 530 |
+
"loss": 0.3759,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.08,
|
| 535 |
+
"grad_norm": 7.489439487457275,
|
| 536 |
+
"learning_rate": 1.7053854276663146e-05,
|
| 537 |
+
"loss": 0.3639,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.1,
|
| 542 |
+
"grad_norm": 7.076328277587891,
|
| 543 |
+
"learning_rate": 1.6525871172122493e-05,
|
| 544 |
+
"loss": 0.3823,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.13,
|
| 549 |
+
"grad_norm": 7.6811442375183105,
|
| 550 |
+
"learning_rate": 1.5997888067581837e-05,
|
| 551 |
+
"loss": 0.377,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.16,
|
| 556 |
+
"grad_norm": 8.395451545715332,
|
| 557 |
+
"learning_rate": 1.5469904963041184e-05,
|
| 558 |
+
"loss": 0.3762,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.19,
|
| 563 |
+
"grad_norm": 10.683780670166016,
|
| 564 |
+
"learning_rate": 1.4941921858500529e-05,
|
| 565 |
+
"loss": 0.381,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.22,
|
| 570 |
+
"grad_norm": 9.537813186645508,
|
| 571 |
+
"learning_rate": 1.4413938753959874e-05,
|
| 572 |
+
"loss": 0.3947,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.25,
|
| 577 |
+
"grad_norm": 7.627201557159424,
|
| 578 |
+
"learning_rate": 1.388595564941922e-05,
|
| 579 |
+
"loss": 0.3918,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.27,
|
| 584 |
+
"grad_norm": 10.89393424987793,
|
| 585 |
+
"learning_rate": 1.3357972544878563e-05,
|
| 586 |
+
"loss": 0.4007,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.3,
|
| 591 |
+
"grad_norm": 11.160614967346191,
|
| 592 |
+
"learning_rate": 1.2829989440337912e-05,
|
| 593 |
+
"loss": 0.3564,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.33,
|
| 598 |
+
"grad_norm": 9.594536781311035,
|
| 599 |
+
"learning_rate": 1.2302006335797255e-05,
|
| 600 |
+
"loss": 0.3642,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.36,
|
| 605 |
+
"grad_norm": 10.04773998260498,
|
| 606 |
+
"learning_rate": 1.17740232312566e-05,
|
| 607 |
+
"loss": 0.3751,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.39,
|
| 612 |
+
"grad_norm": 7.632706642150879,
|
| 613 |
+
"learning_rate": 1.1246040126715946e-05,
|
| 614 |
+
"loss": 0.3802,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.42,
|
| 619 |
+
"grad_norm": 7.586816310882568,
|
| 620 |
+
"learning_rate": 1.0718057022175291e-05,
|
| 621 |
+
"loss": 0.3842,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.44,
|
| 626 |
+
"grad_norm": 10.282231330871582,
|
| 627 |
+
"learning_rate": 1.0190073917634636e-05,
|
| 628 |
+
"loss": 0.3813,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.47,
|
| 633 |
+
"grad_norm": 9.297379493713379,
|
| 634 |
+
"learning_rate": 9.662090813093982e-06,
|
| 635 |
+
"loss": 0.3579,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.5,
|
| 640 |
+
"grad_norm": 10.13786792755127,
|
| 641 |
+
"learning_rate": 9.134107708553327e-06,
|
| 642 |
+
"loss": 0.3521,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.53,
|
| 647 |
+
"grad_norm": 11.372087478637695,
|
| 648 |
+
"learning_rate": 8.606124604012672e-06,
|
| 649 |
+
"loss": 0.3649,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.56,
|
| 654 |
+
"grad_norm": 10.296939849853516,
|
| 655 |
+
"learning_rate": 8.078141499472017e-06,
|
| 656 |
+
"loss": 0.4226,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.59,
|
| 661 |
+
"grad_norm": 12.837827682495117,
|
| 662 |
+
"learning_rate": 7.5501583949313625e-06,
|
| 663 |
+
"loss": 0.3595,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.62,
|
| 668 |
+
"grad_norm": 12.584282875061035,
|
| 669 |
+
"learning_rate": 7.022175290390708e-06,
|
| 670 |
+
"loss": 0.3474,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.64,
|
| 675 |
+
"grad_norm": 8.62096118927002,
|
| 676 |
+
"learning_rate": 6.494192185850054e-06,
|
| 677 |
+
"loss": 0.3698,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.67,
|
| 682 |
+
"grad_norm": 7.057333946228027,
|
| 683 |
+
"learning_rate": 5.966209081309398e-06,
|
| 684 |
+
"loss": 0.3093,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.7,
|
| 689 |
+
"grad_norm": 9.453765869140625,
|
| 690 |
+
"learning_rate": 5.438225976768744e-06,
|
| 691 |
+
"loss": 0.3887,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.73,
|
| 696 |
+
"grad_norm": 5.629605293273926,
|
| 697 |
+
"learning_rate": 4.910242872228089e-06,
|
| 698 |
+
"loss": 0.3508,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.76,
|
| 703 |
+
"grad_norm": 10.514436721801758,
|
| 704 |
+
"learning_rate": 4.382259767687434e-06,
|
| 705 |
+
"loss": 0.3404,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.79,
|
| 710 |
+
"grad_norm": 11.674103736877441,
|
| 711 |
+
"learning_rate": 3.854276663146779e-06,
|
| 712 |
+
"loss": 0.317,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.81,
|
| 717 |
+
"grad_norm": 8.024503707885742,
|
| 718 |
+
"learning_rate": 3.326293558606125e-06,
|
| 719 |
+
"loss": 0.3626,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.84,
|
| 724 |
+
"grad_norm": 10.710383415222168,
|
| 725 |
+
"learning_rate": 2.79831045406547e-06,
|
| 726 |
+
"loss": 0.321,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.87,
|
| 731 |
+
"grad_norm": 10.499631881713867,
|
| 732 |
+
"learning_rate": 2.2703273495248154e-06,
|
| 733 |
+
"loss": 0.3551,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.9,
|
| 738 |
+
"grad_norm": 8.840764999389648,
|
| 739 |
+
"learning_rate": 1.7423442449841606e-06,
|
| 740 |
+
"loss": 0.3597,
|
| 741 |
+
"step": 1020
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.93,
|
| 745 |
+
"grad_norm": 9.39499568939209,
|
| 746 |
+
"learning_rate": 1.2143611404435059e-06,
|
| 747 |
+
"loss": 0.3473,
|
| 748 |
+
"step": 1030
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.96,
|
| 752 |
+
"grad_norm": 8.382072448730469,
|
| 753 |
+
"learning_rate": 6.863780359028511e-07,
|
| 754 |
+
"loss": 0.3382,
|
| 755 |
+
"step": 1040
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.99,
|
| 759 |
+
"grad_norm": 8.499634742736816,
|
| 760 |
+
"learning_rate": 1.5839493136219642e-07,
|
| 761 |
+
"loss": 0.3445,
|
| 762 |
+
"step": 1050
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.99,
|
| 766 |
+
"eval_accuracy": 0.9738,
|
| 767 |
+
"eval_loss": 0.08292504400014877,
|
| 768 |
+
"eval_runtime": 62.7202,
|
| 769 |
+
"eval_samples_per_second": 79.719,
|
| 770 |
+
"eval_steps_per_second": 2.503,
|
| 771 |
+
"step": 1053
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 2.99,
|
| 775 |
+
"step": 1053,
|
| 776 |
+
"total_flos": 3.3497451642252165e+18,
|
| 777 |
+
"train_loss": 0.5573047858584188,
|
| 778 |
+
"train_runtime": 2153.5042,
|
| 779 |
+
"train_samples_per_second": 62.689,
|
| 780 |
+
"train_steps_per_second": 0.489
|
| 781 |
+
}
|
| 782 |
+
],
|
| 783 |
+
"logging_steps": 10,
|
| 784 |
+
"max_steps": 1053,
|
| 785 |
+
"num_input_tokens_seen": 0,
|
| 786 |
+
"num_train_epochs": 3,
|
| 787 |
+
"save_steps": 500,
|
| 788 |
+
"total_flos": 3.3497451642252165e+18,
|
| 789 |
+
"train_batch_size": 32,
|
| 790 |
+
"trial_name": null,
|
| 791 |
+
"trial_params": null
|
| 792 |
+
}
|