End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9936034115138592,
|
| 3 |
+
"eval_accuracy": 0.9704,
|
| 4 |
+
"eval_loss": 0.09273476898670197,
|
| 5 |
+
"eval_runtime": 31.5905,
|
| 6 |
+
"eval_samples_per_second": 158.276,
|
| 7 |
+
"eval_steps_per_second": 4.97,
|
| 8 |
+
"total_flos": 3.3497451642252165e+18,
|
| 9 |
+
"train_loss": 0.5471523041285907,
|
| 10 |
+
"train_runtime": 2023.4436,
|
| 11 |
+
"train_samples_per_second": 66.718,
|
| 12 |
+
"train_steps_per_second": 0.52
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9936034115138592,
|
| 3 |
+
"eval_accuracy": 0.9704,
|
| 4 |
+
"eval_loss": 0.09273476898670197,
|
| 5 |
+
"eval_runtime": 31.5905,
|
| 6 |
+
"eval_samples_per_second": 158.276,
|
| 7 |
+
"eval_steps_per_second": 4.97
|
| 8 |
+
}
|
runs/May06_04-32-17_5819780813d0/events.out.tfevents.1714972009.5819780813d0.502.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:958ad5a566a02c7f7b75b292a68607247492d1e172bc04dfb13a0756803c2b58
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.9936034115138592,
|
| 3 |
+
"total_flos": 3.3497451642252165e+18,
|
| 4 |
+
"train_loss": 0.5471523041285907,
|
| 5 |
+
"train_runtime": 2023.4436,
|
| 6 |
+
"train_samples_per_second": 66.718,
|
| 7 |
+
"train_steps_per_second": 0.52
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,792 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9704,
|
| 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": 0.028429282160625444,
|
| 13 |
+
"grad_norm": 5.52622652053833,
|
| 14 |
+
"learning_rate": 4.716981132075472e-06,
|
| 15 |
+
"loss": 2.3451,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05685856432125089,
|
| 20 |
+
"grad_norm": 5.207578182220459,
|
| 21 |
+
"learning_rate": 9.433962264150944e-06,
|
| 22 |
+
"loss": 2.2737,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08528784648187633,
|
| 27 |
+
"grad_norm": 6.15794563293457,
|
| 28 |
+
"learning_rate": 1.4150943396226415e-05,
|
| 29 |
+
"loss": 2.1375,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11371712864250177,
|
| 34 |
+
"grad_norm": 7.397068023681641,
|
| 35 |
+
"learning_rate": 1.8867924528301888e-05,
|
| 36 |
+
"loss": 1.9859,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.14214641080312723,
|
| 41 |
+
"grad_norm": 11.379995346069336,
|
| 42 |
+
"learning_rate": 2.358490566037736e-05,
|
| 43 |
+
"loss": 1.7134,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.17057569296375266,
|
| 48 |
+
"grad_norm": 17.0799560546875,
|
| 49 |
+
"learning_rate": 2.830188679245283e-05,
|
| 50 |
+
"loss": 1.3856,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19900497512437812,
|
| 55 |
+
"grad_norm": 16.41306495666504,
|
| 56 |
+
"learning_rate": 3.30188679245283e-05,
|
| 57 |
+
"loss": 1.1398,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.22743425728500355,
|
| 62 |
+
"grad_norm": 15.939576148986816,
|
| 63 |
+
"learning_rate": 3.7735849056603776e-05,
|
| 64 |
+
"loss": 1.0317,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.255863539445629,
|
| 69 |
+
"grad_norm": 10.1390380859375,
|
| 70 |
+
"learning_rate": 4.245283018867925e-05,
|
| 71 |
+
"loss": 0.9435,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.28429282160625446,
|
| 76 |
+
"grad_norm": 20.10552406311035,
|
| 77 |
+
"learning_rate": 4.716981132075472e-05,
|
| 78 |
+
"loss": 0.8801,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.31272210376687987,
|
| 83 |
+
"grad_norm": 12.014751434326172,
|
| 84 |
+
"learning_rate": 4.978880675818374e-05,
|
| 85 |
+
"loss": 0.7564,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3411513859275053,
|
| 90 |
+
"grad_norm": 14.332877159118652,
|
| 91 |
+
"learning_rate": 4.9260823653643085e-05,
|
| 92 |
+
"loss": 0.6985,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3695806680881308,
|
| 97 |
+
"grad_norm": 11.214934349060059,
|
| 98 |
+
"learning_rate": 4.8732840549102435e-05,
|
| 99 |
+
"loss": 0.6713,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.39800995024875624,
|
| 104 |
+
"grad_norm": 11.159377098083496,
|
| 105 |
+
"learning_rate": 4.820485744456177e-05,
|
| 106 |
+
"loss": 0.6749,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.42643923240938164,
|
| 111 |
+
"grad_norm": 9.342412948608398,
|
| 112 |
+
"learning_rate": 4.767687434002112e-05,
|
| 113 |
+
"loss": 0.6158,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4548685145700071,
|
| 118 |
+
"grad_norm": 9.084425926208496,
|
| 119 |
+
"learning_rate": 4.7148891235480466e-05,
|
| 120 |
+
"loss": 0.6129,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.48329779673063256,
|
| 125 |
+
"grad_norm": 11.425093650817871,
|
| 126 |
+
"learning_rate": 4.662090813093981e-05,
|
| 127 |
+
"loss": 0.6107,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.511727078891258,
|
| 132 |
+
"grad_norm": 17.25963020324707,
|
| 133 |
+
"learning_rate": 4.609292502639916e-05,
|
| 134 |
+
"loss": 0.583,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5401563610518835,
|
| 139 |
+
"grad_norm": 9.029903411865234,
|
| 140 |
+
"learning_rate": 4.55649419218585e-05,
|
| 141 |
+
"loss": 0.5778,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5685856432125089,
|
| 146 |
+
"grad_norm": 11.78687858581543,
|
| 147 |
+
"learning_rate": 4.503695881731785e-05,
|
| 148 |
+
"loss": 0.6027,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5970149253731343,
|
| 153 |
+
"grad_norm": 10.835305213928223,
|
| 154 |
+
"learning_rate": 4.45089757127772e-05,
|
| 155 |
+
"loss": 0.5801,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6254442075337597,
|
| 160 |
+
"grad_norm": 9.951684951782227,
|
| 161 |
+
"learning_rate": 4.398099260823654e-05,
|
| 162 |
+
"loss": 0.5272,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6538734896943852,
|
| 167 |
+
"grad_norm": 14.171173095703125,
|
| 168 |
+
"learning_rate": 4.3453009503695884e-05,
|
| 169 |
+
"loss": 0.5197,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6823027718550106,
|
| 174 |
+
"grad_norm": 13.092684745788574,
|
| 175 |
+
"learning_rate": 4.292502639915523e-05,
|
| 176 |
+
"loss": 0.5628,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7107320540156361,
|
| 181 |
+
"grad_norm": 10.938400268554688,
|
| 182 |
+
"learning_rate": 4.239704329461457e-05,
|
| 183 |
+
"loss": 0.5621,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7391613361762616,
|
| 188 |
+
"grad_norm": 7.700541019439697,
|
| 189 |
+
"learning_rate": 4.186906019007392e-05,
|
| 190 |
+
"loss": 0.5223,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.767590618336887,
|
| 195 |
+
"grad_norm": 14.169013023376465,
|
| 196 |
+
"learning_rate": 4.1341077085533265e-05,
|
| 197 |
+
"loss": 0.5656,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7960199004975125,
|
| 202 |
+
"grad_norm": 9.239885330200195,
|
| 203 |
+
"learning_rate": 4.081309398099261e-05,
|
| 204 |
+
"loss": 0.4799,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8244491826581379,
|
| 209 |
+
"grad_norm": 12.713397979736328,
|
| 210 |
+
"learning_rate": 4.028511087645195e-05,
|
| 211 |
+
"loss": 0.4996,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.8528784648187633,
|
| 216 |
+
"grad_norm": 11.337251663208008,
|
| 217 |
+
"learning_rate": 3.97571277719113e-05,
|
| 218 |
+
"loss": 0.4954,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8813077469793887,
|
| 223 |
+
"grad_norm": 11.150245666503906,
|
| 224 |
+
"learning_rate": 3.9229144667370646e-05,
|
| 225 |
+
"loss": 0.4993,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9097370291400142,
|
| 230 |
+
"grad_norm": 10.354619026184082,
|
| 231 |
+
"learning_rate": 3.870116156282999e-05,
|
| 232 |
+
"loss": 0.4576,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9381663113006397,
|
| 237 |
+
"grad_norm": 9.165169715881348,
|
| 238 |
+
"learning_rate": 3.817317845828934e-05,
|
| 239 |
+
"loss": 0.4748,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.9665955934612651,
|
| 244 |
+
"grad_norm": 8.658796310424805,
|
| 245 |
+
"learning_rate": 3.764519535374868e-05,
|
| 246 |
+
"loss": 0.4578,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9950248756218906,
|
| 251 |
+
"grad_norm": 11.11921215057373,
|
| 252 |
+
"learning_rate": 3.711721224920803e-05,
|
| 253 |
+
"loss": 0.502,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.997867803837953,
|
| 258 |
+
"eval_accuracy": 0.9496,
|
| 259 |
+
"eval_loss": 0.15361838042736053,
|
| 260 |
+
"eval_runtime": 30.4943,
|
| 261 |
+
"eval_samples_per_second": 163.965,
|
| 262 |
+
"eval_steps_per_second": 5.148,
|
| 263 |
+
"step": 351
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.023454157782516,
|
| 267 |
+
"grad_norm": 10.098979949951172,
|
| 268 |
+
"learning_rate": 3.658922914466738e-05,
|
| 269 |
+
"loss": 0.4338,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.0518834399431414,
|
| 274 |
+
"grad_norm": 9.755829811096191,
|
| 275 |
+
"learning_rate": 3.6061246040126714e-05,
|
| 276 |
+
"loss": 0.4701,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.080312722103767,
|
| 281 |
+
"grad_norm": 7.626587867736816,
|
| 282 |
+
"learning_rate": 3.5533262935586064e-05,
|
| 283 |
+
"loss": 0.4872,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.1087420042643923,
|
| 288 |
+
"grad_norm": 8.716501235961914,
|
| 289 |
+
"learning_rate": 3.500527983104541e-05,
|
| 290 |
+
"loss": 0.4495,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.1371712864250179,
|
| 295 |
+
"grad_norm": 10.042838096618652,
|
| 296 |
+
"learning_rate": 3.447729672650475e-05,
|
| 297 |
+
"loss": 0.4688,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.1656005685856432,
|
| 302 |
+
"grad_norm": 9.901670455932617,
|
| 303 |
+
"learning_rate": 3.3949313621964095e-05,
|
| 304 |
+
"loss": 0.4572,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.1940298507462686,
|
| 309 |
+
"grad_norm": 12.942643165588379,
|
| 310 |
+
"learning_rate": 3.3421330517423445e-05,
|
| 311 |
+
"loss": 0.4174,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.2224591329068941,
|
| 316 |
+
"grad_norm": 8.440229415893555,
|
| 317 |
+
"learning_rate": 3.289334741288279e-05,
|
| 318 |
+
"loss": 0.4433,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.2508884150675195,
|
| 323 |
+
"grad_norm": 8.764989852905273,
|
| 324 |
+
"learning_rate": 3.236536430834213e-05,
|
| 325 |
+
"loss": 0.4542,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.279317697228145,
|
| 330 |
+
"grad_norm": 11.38581657409668,
|
| 331 |
+
"learning_rate": 3.183738120380148e-05,
|
| 332 |
+
"loss": 0.4038,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.3077469793887704,
|
| 337 |
+
"grad_norm": 10.445256233215332,
|
| 338 |
+
"learning_rate": 3.130939809926082e-05,
|
| 339 |
+
"loss": 0.4418,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.336176261549396,
|
| 344 |
+
"grad_norm": 12.250975608825684,
|
| 345 |
+
"learning_rate": 3.078141499472017e-05,
|
| 346 |
+
"loss": 0.4717,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.3646055437100213,
|
| 351 |
+
"grad_norm": 11.0380859375,
|
| 352 |
+
"learning_rate": 3.0253431890179517e-05,
|
| 353 |
+
"loss": 0.421,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.3930348258706466,
|
| 358 |
+
"grad_norm": 12.513349533081055,
|
| 359 |
+
"learning_rate": 2.972544878563886e-05,
|
| 360 |
+
"loss": 0.4719,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.4214641080312722,
|
| 365 |
+
"grad_norm": 11.536993980407715,
|
| 366 |
+
"learning_rate": 2.9197465681098207e-05,
|
| 367 |
+
"loss": 0.4042,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.4498933901918978,
|
| 372 |
+
"grad_norm": 7.970879077911377,
|
| 373 |
+
"learning_rate": 2.8669482576557548e-05,
|
| 374 |
+
"loss": 0.415,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.4783226723525231,
|
| 379 |
+
"grad_norm": 10.883822441101074,
|
| 380 |
+
"learning_rate": 2.8141499472016898e-05,
|
| 381 |
+
"loss": 0.4353,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.5067519545131485,
|
| 386 |
+
"grad_norm": 14.611568450927734,
|
| 387 |
+
"learning_rate": 2.7613516367476245e-05,
|
| 388 |
+
"loss": 0.4385,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.535181236673774,
|
| 393 |
+
"grad_norm": 9.424297332763672,
|
| 394 |
+
"learning_rate": 2.7085533262935585e-05,
|
| 395 |
+
"loss": 0.4712,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.5636105188343994,
|
| 400 |
+
"grad_norm": 9.887263298034668,
|
| 401 |
+
"learning_rate": 2.6557550158394935e-05,
|
| 402 |
+
"loss": 0.4291,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.5920398009950247,
|
| 407 |
+
"grad_norm": 8.692890167236328,
|
| 408 |
+
"learning_rate": 2.6029567053854276e-05,
|
| 409 |
+
"loss": 0.3915,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.6204690831556503,
|
| 414 |
+
"grad_norm": 10.517526626586914,
|
| 415 |
+
"learning_rate": 2.5501583949313622e-05,
|
| 416 |
+
"loss": 0.4303,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.6488983653162759,
|
| 421 |
+
"grad_norm": 8.96032428741455,
|
| 422 |
+
"learning_rate": 2.497360084477297e-05,
|
| 423 |
+
"loss": 0.43,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.6773276474769012,
|
| 428 |
+
"grad_norm": 8.543038368225098,
|
| 429 |
+
"learning_rate": 2.4445617740232313e-05,
|
| 430 |
+
"loss": 0.4177,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.7057569296375266,
|
| 435 |
+
"grad_norm": 10.253332138061523,
|
| 436 |
+
"learning_rate": 2.391763463569166e-05,
|
| 437 |
+
"loss": 0.4033,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.7341862117981521,
|
| 442 |
+
"grad_norm": 9.046222686767578,
|
| 443 |
+
"learning_rate": 2.3389651531151003e-05,
|
| 444 |
+
"loss": 0.4031,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.7626154939587777,
|
| 449 |
+
"grad_norm": 8.473087310791016,
|
| 450 |
+
"learning_rate": 2.286166842661035e-05,
|
| 451 |
+
"loss": 0.4012,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.7910447761194028,
|
| 456 |
+
"grad_norm": 9.883254051208496,
|
| 457 |
+
"learning_rate": 2.2333685322069694e-05,
|
| 458 |
+
"loss": 0.3694,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.8194740582800284,
|
| 463 |
+
"grad_norm": 10.291656494140625,
|
| 464 |
+
"learning_rate": 2.180570221752904e-05,
|
| 465 |
+
"loss": 0.3938,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.847903340440654,
|
| 470 |
+
"grad_norm": 9.967813491821289,
|
| 471 |
+
"learning_rate": 2.1277719112988384e-05,
|
| 472 |
+
"loss": 0.4355,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.8763326226012793,
|
| 477 |
+
"grad_norm": 11.451550483703613,
|
| 478 |
+
"learning_rate": 2.074973600844773e-05,
|
| 479 |
+
"loss": 0.4208,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.9047619047619047,
|
| 484 |
+
"grad_norm": 6.351629257202148,
|
| 485 |
+
"learning_rate": 2.0221752903907075e-05,
|
| 486 |
+
"loss": 0.4091,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.9331911869225302,
|
| 491 |
+
"grad_norm": 11.102736473083496,
|
| 492 |
+
"learning_rate": 1.9693769799366422e-05,
|
| 493 |
+
"loss": 0.3957,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.9616204690831558,
|
| 498 |
+
"grad_norm": 7.058555603027344,
|
| 499 |
+
"learning_rate": 1.9165786694825765e-05,
|
| 500 |
+
"loss": 0.4503,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.9900497512437811,
|
| 505 |
+
"grad_norm": 8.091329574584961,
|
| 506 |
+
"learning_rate": 1.863780359028511e-05,
|
| 507 |
+
"loss": 0.3782,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.9985785358919688,
|
| 512 |
+
"eval_accuracy": 0.9656,
|
| 513 |
+
"eval_loss": 0.10786179453134537,
|
| 514 |
+
"eval_runtime": 30.9276,
|
| 515 |
+
"eval_samples_per_second": 161.668,
|
| 516 |
+
"eval_steps_per_second": 5.076,
|
| 517 |
+
"step": 703
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.0184790334044065,
|
| 521 |
+
"grad_norm": 8.83482551574707,
|
| 522 |
+
"learning_rate": 1.810982048574446e-05,
|
| 523 |
+
"loss": 0.3938,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.046908315565032,
|
| 528 |
+
"grad_norm": 8.442475318908691,
|
| 529 |
+
"learning_rate": 1.7581837381203803e-05,
|
| 530 |
+
"loss": 0.379,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.0753375977256576,
|
| 535 |
+
"grad_norm": 12.47827434539795,
|
| 536 |
+
"learning_rate": 1.7053854276663146e-05,
|
| 537 |
+
"loss": 0.3684,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.1037668798862827,
|
| 542 |
+
"grad_norm": 8.256750106811523,
|
| 543 |
+
"learning_rate": 1.6525871172122493e-05,
|
| 544 |
+
"loss": 0.3741,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.1321961620469083,
|
| 549 |
+
"grad_norm": 6.83493185043335,
|
| 550 |
+
"learning_rate": 1.5997888067581837e-05,
|
| 551 |
+
"loss": 0.361,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.160625444207534,
|
| 556 |
+
"grad_norm": 7.63658332824707,
|
| 557 |
+
"learning_rate": 1.5469904963041184e-05,
|
| 558 |
+
"loss": 0.3199,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.189054726368159,
|
| 563 |
+
"grad_norm": 10.230888366699219,
|
| 564 |
+
"learning_rate": 1.4941921858500529e-05,
|
| 565 |
+
"loss": 0.3495,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.2174840085287846,
|
| 570 |
+
"grad_norm": 9.374876022338867,
|
| 571 |
+
"learning_rate": 1.4413938753959874e-05,
|
| 572 |
+
"loss": 0.3852,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.24591329068941,
|
| 577 |
+
"grad_norm": 6.269606590270996,
|
| 578 |
+
"learning_rate": 1.388595564941922e-05,
|
| 579 |
+
"loss": 0.3398,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.2743425728500357,
|
| 584 |
+
"grad_norm": 10.073060989379883,
|
| 585 |
+
"learning_rate": 1.3357972544878563e-05,
|
| 586 |
+
"loss": 0.3876,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.302771855010661,
|
| 591 |
+
"grad_norm": 9.419556617736816,
|
| 592 |
+
"learning_rate": 1.2829989440337912e-05,
|
| 593 |
+
"loss": 0.3517,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.3312011371712864,
|
| 598 |
+
"grad_norm": 10.461442947387695,
|
| 599 |
+
"learning_rate": 1.2302006335797255e-05,
|
| 600 |
+
"loss": 0.367,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.359630419331912,
|
| 605 |
+
"grad_norm": 7.293799877166748,
|
| 606 |
+
"learning_rate": 1.17740232312566e-05,
|
| 607 |
+
"loss": 0.3402,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.388059701492537,
|
| 612 |
+
"grad_norm": 6.696635723114014,
|
| 613 |
+
"learning_rate": 1.1246040126715946e-05,
|
| 614 |
+
"loss": 0.3578,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.4164889836531627,
|
| 619 |
+
"grad_norm": 8.069903373718262,
|
| 620 |
+
"learning_rate": 1.0718057022175291e-05,
|
| 621 |
+
"loss": 0.3638,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.4449182658137882,
|
| 626 |
+
"grad_norm": 8.903417587280273,
|
| 627 |
+
"learning_rate": 1.0190073917634636e-05,
|
| 628 |
+
"loss": 0.3618,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.473347547974414,
|
| 633 |
+
"grad_norm": 10.900948524475098,
|
| 634 |
+
"learning_rate": 9.662090813093982e-06,
|
| 635 |
+
"loss": 0.3302,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.501776830135039,
|
| 640 |
+
"grad_norm": 7.725928783416748,
|
| 641 |
+
"learning_rate": 9.134107708553327e-06,
|
| 642 |
+
"loss": 0.3483,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.5302061122956645,
|
| 647 |
+
"grad_norm": 7.297862529754639,
|
| 648 |
+
"learning_rate": 8.606124604012672e-06,
|
| 649 |
+
"loss": 0.4057,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.55863539445629,
|
| 654 |
+
"grad_norm": 9.81568717956543,
|
| 655 |
+
"learning_rate": 8.078141499472017e-06,
|
| 656 |
+
"loss": 0.3726,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.587064676616915,
|
| 661 |
+
"grad_norm": 11.697639465332031,
|
| 662 |
+
"learning_rate": 7.5501583949313625e-06,
|
| 663 |
+
"loss": 0.3695,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.6154939587775408,
|
| 668 |
+
"grad_norm": 8.077279090881348,
|
| 669 |
+
"learning_rate": 7.022175290390708e-06,
|
| 670 |
+
"loss": 0.3723,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.6439232409381663,
|
| 675 |
+
"grad_norm": 8.867374420166016,
|
| 676 |
+
"learning_rate": 6.494192185850054e-06,
|
| 677 |
+
"loss": 0.3785,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.672352523098792,
|
| 682 |
+
"grad_norm": 6.055303573608398,
|
| 683 |
+
"learning_rate": 5.966209081309398e-06,
|
| 684 |
+
"loss": 0.3597,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.7007818052594175,
|
| 689 |
+
"grad_norm": 7.143568992614746,
|
| 690 |
+
"learning_rate": 5.438225976768744e-06,
|
| 691 |
+
"loss": 0.3668,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.7292110874200426,
|
| 696 |
+
"grad_norm": 7.816047668457031,
|
| 697 |
+
"learning_rate": 4.910242872228089e-06,
|
| 698 |
+
"loss": 0.3296,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.757640369580668,
|
| 703 |
+
"grad_norm": 8.865290641784668,
|
| 704 |
+
"learning_rate": 4.382259767687434e-06,
|
| 705 |
+
"loss": 0.3502,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.7860696517412933,
|
| 710 |
+
"grad_norm": 12.268588066101074,
|
| 711 |
+
"learning_rate": 3.854276663146779e-06,
|
| 712 |
+
"loss": 0.34,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.814498933901919,
|
| 717 |
+
"grad_norm": 11.221428871154785,
|
| 718 |
+
"learning_rate": 3.326293558606125e-06,
|
| 719 |
+
"loss": 0.359,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.8429282160625444,
|
| 724 |
+
"grad_norm": 7.872012138366699,
|
| 725 |
+
"learning_rate": 2.79831045406547e-06,
|
| 726 |
+
"loss": 0.3483,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.87135749822317,
|
| 731 |
+
"grad_norm": 7.7068257331848145,
|
| 732 |
+
"learning_rate": 2.2703273495248154e-06,
|
| 733 |
+
"loss": 0.3468,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.8997867803837956,
|
| 738 |
+
"grad_norm": 7.7084503173828125,
|
| 739 |
+
"learning_rate": 1.7423442449841606e-06,
|
| 740 |
+
"loss": 0.3782,
|
| 741 |
+
"step": 1020
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.9282160625444207,
|
| 745 |
+
"grad_norm": 11.527771949768066,
|
| 746 |
+
"learning_rate": 1.2143611404435059e-06,
|
| 747 |
+
"loss": 0.3441,
|
| 748 |
+
"step": 1030
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.9566453447050463,
|
| 752 |
+
"grad_norm": 8.335259437561035,
|
| 753 |
+
"learning_rate": 6.863780359028511e-07,
|
| 754 |
+
"loss": 0.3625,
|
| 755 |
+
"step": 1040
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.9850746268656714,
|
| 759 |
+
"grad_norm": 8.47817611694336,
|
| 760 |
+
"learning_rate": 1.5839493136219642e-07,
|
| 761 |
+
"loss": 0.3119,
|
| 762 |
+
"step": 1050
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.9936034115138592,
|
| 766 |
+
"eval_accuracy": 0.9704,
|
| 767 |
+
"eval_loss": 0.09273476898670197,
|
| 768 |
+
"eval_runtime": 31.2742,
|
| 769 |
+
"eval_samples_per_second": 159.876,
|
| 770 |
+
"eval_steps_per_second": 5.02,
|
| 771 |
+
"step": 1053
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 2.9936034115138592,
|
| 775 |
+
"step": 1053,
|
| 776 |
+
"total_flos": 3.3497451642252165e+18,
|
| 777 |
+
"train_loss": 0.5471523041285907,
|
| 778 |
+
"train_runtime": 2023.4436,
|
| 779 |
+
"train_samples_per_second": 66.718,
|
| 780 |
+
"train_steps_per_second": 0.52
|
| 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 |
+
}
|