End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.1,
|
| 3 |
+
"eval_accuracy": 0.8502994011976048,
|
| 4 |
+
"eval_loss": 0.6543134450912476,
|
| 5 |
+
"eval_runtime": 119.6956,
|
| 6 |
+
"eval_samples_per_second": 4.186,
|
| 7 |
+
"eval_steps_per_second": 0.526
|
| 8 |
+
}
|
runs/May26_05-44-34_e54c14a66b63/events.out.tfevents.1748246726.e54c14a66b63.2476.1
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12c313847dc678add0e6340d243661b895b4f8e6a0c74c00852b20ce6316468f
|
| 3 |
+
size 734
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.1,
|
| 3 |
+
"eval_accuracy": 0.8502994011976048,
|
| 4 |
+
"eval_loss": 0.6543134450912476,
|
| 5 |
+
"eval_runtime": 119.6956,
|
| 6 |
+
"eval_samples_per_second": 4.186,
|
| 7 |
+
"eval_steps_per_second": 0.526
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,911 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 576,
|
| 3 |
+
"best_metric": 0.8495934959349594,
|
| 4 |
+
"best_model_checkpoint": "ALL_RGBCROP_Aug16F-8B16F/checkpoint-576",
|
| 5 |
+
"epoch": 3.1,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1152,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.003472222222222222,
|
| 14 |
+
"grad_norm": 4.709033489227295,
|
| 15 |
+
"learning_rate": 1.5625e-06,
|
| 16 |
+
"loss": 0.712,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.006944444444444444,
|
| 21 |
+
"grad_norm": 5.104576587677002,
|
| 22 |
+
"learning_rate": 3.2986111111111115e-06,
|
| 23 |
+
"loss": 0.7204,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.010416666666666666,
|
| 28 |
+
"grad_norm": 4.670226573944092,
|
| 29 |
+
"learning_rate": 5.034722222222222e-06,
|
| 30 |
+
"loss": 0.7194,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.013888888888888888,
|
| 35 |
+
"grad_norm": 5.235384941101074,
|
| 36 |
+
"learning_rate": 6.770833333333333e-06,
|
| 37 |
+
"loss": 0.6691,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.017361111111111112,
|
| 42 |
+
"grad_norm": 5.082781791687012,
|
| 43 |
+
"learning_rate": 8.506944444444445e-06,
|
| 44 |
+
"loss": 0.6538,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.020833333333333332,
|
| 49 |
+
"grad_norm": 4.47755765914917,
|
| 50 |
+
"learning_rate": 1.0243055555555556e-05,
|
| 51 |
+
"loss": 0.6051,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.024305555555555556,
|
| 56 |
+
"grad_norm": 7.663668632507324,
|
| 57 |
+
"learning_rate": 1.1979166666666667e-05,
|
| 58 |
+
"loss": 0.6125,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.027777777777777776,
|
| 63 |
+
"grad_norm": 6.4383978843688965,
|
| 64 |
+
"learning_rate": 1.371527777777778e-05,
|
| 65 |
+
"loss": 0.5258,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.03125,
|
| 70 |
+
"grad_norm": 6.019410610198975,
|
| 71 |
+
"learning_rate": 1.545138888888889e-05,
|
| 72 |
+
"loss": 0.532,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.034722222222222224,
|
| 77 |
+
"grad_norm": 6.127612590789795,
|
| 78 |
+
"learning_rate": 1.71875e-05,
|
| 79 |
+
"loss": 0.4173,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.03819444444444445,
|
| 84 |
+
"grad_norm": 6.65797233581543,
|
| 85 |
+
"learning_rate": 1.8923611111111112e-05,
|
| 86 |
+
"loss": 0.4882,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.041666666666666664,
|
| 91 |
+
"grad_norm": 8.373489379882812,
|
| 92 |
+
"learning_rate": 2.0659722222222223e-05,
|
| 93 |
+
"loss": 0.4912,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.04513888888888889,
|
| 98 |
+
"grad_norm": 5.64392614364624,
|
| 99 |
+
"learning_rate": 2.2395833333333337e-05,
|
| 100 |
+
"loss": 0.3324,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04861111111111111,
|
| 105 |
+
"grad_norm": 6.752640724182129,
|
| 106 |
+
"learning_rate": 2.4131944444444448e-05,
|
| 107 |
+
"loss": 0.3179,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.052083333333333336,
|
| 112 |
+
"grad_norm": 6.095239162445068,
|
| 113 |
+
"learning_rate": 2.5868055555555558e-05,
|
| 114 |
+
"loss": 0.2183,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.05555555555555555,
|
| 119 |
+
"grad_norm": 11.378706932067871,
|
| 120 |
+
"learning_rate": 2.760416666666667e-05,
|
| 121 |
+
"loss": 0.2803,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.059027777777777776,
|
| 126 |
+
"grad_norm": 3.435918092727661,
|
| 127 |
+
"learning_rate": 2.934027777777778e-05,
|
| 128 |
+
"loss": 0.2328,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.0625,
|
| 133 |
+
"grad_norm": 9.458333015441895,
|
| 134 |
+
"learning_rate": 3.107638888888889e-05,
|
| 135 |
+
"loss": 0.2167,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.06597222222222222,
|
| 140 |
+
"grad_norm": 36.986854553222656,
|
| 141 |
+
"learning_rate": 3.2812500000000005e-05,
|
| 142 |
+
"loss": 0.2547,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06944444444444445,
|
| 147 |
+
"grad_norm": 9.727249145507812,
|
| 148 |
+
"learning_rate": 3.454861111111111e-05,
|
| 149 |
+
"loss": 0.2211,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.07291666666666667,
|
| 154 |
+
"grad_norm": 8.362494468688965,
|
| 155 |
+
"learning_rate": 3.628472222222222e-05,
|
| 156 |
+
"loss": 0.2203,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.0763888888888889,
|
| 161 |
+
"grad_norm": 8.82453727722168,
|
| 162 |
+
"learning_rate": 3.8020833333333334e-05,
|
| 163 |
+
"loss": 0.1637,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.0798611111111111,
|
| 168 |
+
"grad_norm": 4.613597393035889,
|
| 169 |
+
"learning_rate": 3.975694444444444e-05,
|
| 170 |
+
"loss": 0.3848,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.08333333333333333,
|
| 175 |
+
"grad_norm": 12.759675025939941,
|
| 176 |
+
"learning_rate": 4.149305555555556e-05,
|
| 177 |
+
"loss": 0.2376,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.08680555555555555,
|
| 182 |
+
"grad_norm": 1.1166067123413086,
|
| 183 |
+
"learning_rate": 4.322916666666667e-05,
|
| 184 |
+
"loss": 0.2498,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.09027777777777778,
|
| 189 |
+
"grad_norm": 6.368658542633057,
|
| 190 |
+
"learning_rate": 4.4965277777777784e-05,
|
| 191 |
+
"loss": 0.155,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.09375,
|
| 196 |
+
"grad_norm": 28.25998306274414,
|
| 197 |
+
"learning_rate": 4.670138888888889e-05,
|
| 198 |
+
"loss": 0.3182,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.09722222222222222,
|
| 203 |
+
"grad_norm": 9.867981910705566,
|
| 204 |
+
"learning_rate": 4.8437500000000005e-05,
|
| 205 |
+
"loss": 0.3887,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.1,
|
| 210 |
+
"eval_accuracy": 0.8109756097560976,
|
| 211 |
+
"eval_loss": 0.6568520665168762,
|
| 212 |
+
"eval_runtime": 553.1651,
|
| 213 |
+
"eval_samples_per_second": 0.889,
|
| 214 |
+
"eval_steps_per_second": 0.112,
|
| 215 |
+
"step": 288
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.0006944444444446,
|
| 219 |
+
"grad_norm": 11.57461929321289,
|
| 220 |
+
"learning_rate": 4.9980709876543215e-05,
|
| 221 |
+
"loss": 0.1709,
|
| 222 |
+
"step": 290
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.0041666666666667,
|
| 226 |
+
"grad_norm": 15.742400169372559,
|
| 227 |
+
"learning_rate": 4.978780864197531e-05,
|
| 228 |
+
"loss": 0.1815,
|
| 229 |
+
"step": 300
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.007638888888889,
|
| 233 |
+
"grad_norm": 0.1386760175228119,
|
| 234 |
+
"learning_rate": 4.959490740740741e-05,
|
| 235 |
+
"loss": 0.0484,
|
| 236 |
+
"step": 310
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.011111111111111,
|
| 240 |
+
"grad_norm": 0.07092873752117157,
|
| 241 |
+
"learning_rate": 4.940200617283951e-05,
|
| 242 |
+
"loss": 0.0473,
|
| 243 |
+
"step": 320
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.0145833333333334,
|
| 247 |
+
"grad_norm": 9.820779800415039,
|
| 248 |
+
"learning_rate": 4.92091049382716e-05,
|
| 249 |
+
"loss": 0.1945,
|
| 250 |
+
"step": 330
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.0180555555555555,
|
| 254 |
+
"grad_norm": 18.48552131652832,
|
| 255 |
+
"learning_rate": 4.9016203703703705e-05,
|
| 256 |
+
"loss": 0.0859,
|
| 257 |
+
"step": 340
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.0215277777777778,
|
| 261 |
+
"grad_norm": 0.23785443603992462,
|
| 262 |
+
"learning_rate": 4.882330246913581e-05,
|
| 263 |
+
"loss": 0.2122,
|
| 264 |
+
"step": 350
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.025,
|
| 268 |
+
"grad_norm": 0.36233341693878174,
|
| 269 |
+
"learning_rate": 4.86304012345679e-05,
|
| 270 |
+
"loss": 0.066,
|
| 271 |
+
"step": 360
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.0284722222222222,
|
| 275 |
+
"grad_norm": 1.8070086240768433,
|
| 276 |
+
"learning_rate": 4.8437500000000005e-05,
|
| 277 |
+
"loss": 0.0808,
|
| 278 |
+
"step": 370
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.0319444444444446,
|
| 282 |
+
"grad_norm": 6.958967685699463,
|
| 283 |
+
"learning_rate": 4.82445987654321e-05,
|
| 284 |
+
"loss": 0.155,
|
| 285 |
+
"step": 380
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.0354166666666667,
|
| 289 |
+
"grad_norm": 7.139225959777832,
|
| 290 |
+
"learning_rate": 4.8051697530864196e-05,
|
| 291 |
+
"loss": 0.0155,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.038888888888889,
|
| 296 |
+
"grad_norm": 35.38884353637695,
|
| 297 |
+
"learning_rate": 4.78587962962963e-05,
|
| 298 |
+
"loss": 0.1301,
|
| 299 |
+
"step": 400
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.042361111111111,
|
| 303 |
+
"grad_norm": 0.03199063986539841,
|
| 304 |
+
"learning_rate": 4.76658950617284e-05,
|
| 305 |
+
"loss": 0.269,
|
| 306 |
+
"step": 410
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.0458333333333334,
|
| 310 |
+
"grad_norm": 18.229887008666992,
|
| 311 |
+
"learning_rate": 4.7472993827160496e-05,
|
| 312 |
+
"loss": 0.2097,
|
| 313 |
+
"step": 420
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.0493055555555555,
|
| 317 |
+
"grad_norm": 14.07356071472168,
|
| 318 |
+
"learning_rate": 4.72800925925926e-05,
|
| 319 |
+
"loss": 0.2068,
|
| 320 |
+
"step": 430
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.0527777777777778,
|
| 324 |
+
"grad_norm": 12.532169342041016,
|
| 325 |
+
"learning_rate": 4.708719135802469e-05,
|
| 326 |
+
"loss": 0.0579,
|
| 327 |
+
"step": 440
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.05625,
|
| 331 |
+
"grad_norm": 0.1333385556936264,
|
| 332 |
+
"learning_rate": 4.689429012345679e-05,
|
| 333 |
+
"loss": 0.0325,
|
| 334 |
+
"step": 450
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 1.0597222222222222,
|
| 338 |
+
"grad_norm": 0.01660008542239666,
|
| 339 |
+
"learning_rate": 4.670138888888889e-05,
|
| 340 |
+
"loss": 0.0696,
|
| 341 |
+
"step": 460
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.0631944444444446,
|
| 345 |
+
"grad_norm": 6.17806339263916,
|
| 346 |
+
"learning_rate": 4.650848765432099e-05,
|
| 347 |
+
"loss": 0.0105,
|
| 348 |
+
"step": 470
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.0666666666666667,
|
| 352 |
+
"grad_norm": 15.657581329345703,
|
| 353 |
+
"learning_rate": 4.631558641975309e-05,
|
| 354 |
+
"loss": 0.1608,
|
| 355 |
+
"step": 480
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 1.070138888888889,
|
| 359 |
+
"grad_norm": 0.2093486487865448,
|
| 360 |
+
"learning_rate": 4.612268518518519e-05,
|
| 361 |
+
"loss": 0.1097,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.073611111111111,
|
| 366 |
+
"grad_norm": 0.8768461346626282,
|
| 367 |
+
"learning_rate": 4.5929783950617286e-05,
|
| 368 |
+
"loss": 0.006,
|
| 369 |
+
"step": 500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.0770833333333334,
|
| 373 |
+
"grad_norm": 0.07378463447093964,
|
| 374 |
+
"learning_rate": 4.573688271604938e-05,
|
| 375 |
+
"loss": 0.1396,
|
| 376 |
+
"step": 510
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.0805555555555555,
|
| 380 |
+
"grad_norm": 0.10317941755056381,
|
| 381 |
+
"learning_rate": 4.5543981481481484e-05,
|
| 382 |
+
"loss": 0.104,
|
| 383 |
+
"step": 520
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.0840277777777778,
|
| 387 |
+
"grad_norm": 30.956636428833008,
|
| 388 |
+
"learning_rate": 4.535108024691358e-05,
|
| 389 |
+
"loss": 0.1163,
|
| 390 |
+
"step": 530
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.0875,
|
| 394 |
+
"grad_norm": 0.4652070701122284,
|
| 395 |
+
"learning_rate": 4.515817901234568e-05,
|
| 396 |
+
"loss": 0.0798,
|
| 397 |
+
"step": 540
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.0909722222222222,
|
| 401 |
+
"grad_norm": 0.08043529838323593,
|
| 402 |
+
"learning_rate": 4.4965277777777784e-05,
|
| 403 |
+
"loss": 0.0093,
|
| 404 |
+
"step": 550
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.0944444444444446,
|
| 408 |
+
"grad_norm": 0.0071389698423445225,
|
| 409 |
+
"learning_rate": 4.477237654320988e-05,
|
| 410 |
+
"loss": 0.0612,
|
| 411 |
+
"step": 560
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.0979166666666667,
|
| 415 |
+
"grad_norm": 29.914186477661133,
|
| 416 |
+
"learning_rate": 4.4579475308641974e-05,
|
| 417 |
+
"loss": 0.0345,
|
| 418 |
+
"step": 570
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.1,
|
| 422 |
+
"eval_accuracy": 0.8495934959349594,
|
| 423 |
+
"eval_loss": 0.8915717005729675,
|
| 424 |
+
"eval_runtime": 117.2393,
|
| 425 |
+
"eval_samples_per_second": 4.197,
|
| 426 |
+
"eval_steps_per_second": 0.529,
|
| 427 |
+
"step": 576
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 2.001388888888889,
|
| 431 |
+
"grad_norm": 0.0068489667028188705,
|
| 432 |
+
"learning_rate": 4.4386574074074077e-05,
|
| 433 |
+
"loss": 0.0045,
|
| 434 |
+
"step": 580
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.004861111111111,
|
| 438 |
+
"grad_norm": 0.01323405746370554,
|
| 439 |
+
"learning_rate": 4.419367283950617e-05,
|
| 440 |
+
"loss": 0.0197,
|
| 441 |
+
"step": 590
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0083333333333333,
|
| 445 |
+
"grad_norm": 7.431529998779297,
|
| 446 |
+
"learning_rate": 4.4000771604938274e-05,
|
| 447 |
+
"loss": 0.0124,
|
| 448 |
+
"step": 600
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 2.0118055555555556,
|
| 452 |
+
"grad_norm": 0.0020631514489650726,
|
| 453 |
+
"learning_rate": 4.3807870370370376e-05,
|
| 454 |
+
"loss": 0.0004,
|
| 455 |
+
"step": 610
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 2.015277777777778,
|
| 459 |
+
"grad_norm": 0.0061363764107227325,
|
| 460 |
+
"learning_rate": 4.361496913580247e-05,
|
| 461 |
+
"loss": 0.0008,
|
| 462 |
+
"step": 620
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 2.01875,
|
| 466 |
+
"grad_norm": 0.005492759868502617,
|
| 467 |
+
"learning_rate": 4.342206790123457e-05,
|
| 468 |
+
"loss": 0.0134,
|
| 469 |
+
"step": 630
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.022222222222222,
|
| 473 |
+
"grad_norm": 0.008418605662882328,
|
| 474 |
+
"learning_rate": 4.322916666666667e-05,
|
| 475 |
+
"loss": 0.0012,
|
| 476 |
+
"step": 640
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.0256944444444445,
|
| 480 |
+
"grad_norm": 0.23562128841876984,
|
| 481 |
+
"learning_rate": 4.3036265432098765e-05,
|
| 482 |
+
"loss": 0.0248,
|
| 483 |
+
"step": 650
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 2.029166666666667,
|
| 487 |
+
"grad_norm": 0.0033202702179551125,
|
| 488 |
+
"learning_rate": 4.284336419753087e-05,
|
| 489 |
+
"loss": 0.0005,
|
| 490 |
+
"step": 660
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 2.032638888888889,
|
| 494 |
+
"grad_norm": 0.011993384920060635,
|
| 495 |
+
"learning_rate": 4.265046296296297e-05,
|
| 496 |
+
"loss": 0.0003,
|
| 497 |
+
"step": 670
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.036111111111111,
|
| 501 |
+
"grad_norm": 0.02084469608962536,
|
| 502 |
+
"learning_rate": 4.2457561728395065e-05,
|
| 503 |
+
"loss": 0.001,
|
| 504 |
+
"step": 680
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 2.0395833333333333,
|
| 508 |
+
"grad_norm": 0.0016855947906151414,
|
| 509 |
+
"learning_rate": 4.226466049382716e-05,
|
| 510 |
+
"loss": 0.0004,
|
| 511 |
+
"step": 690
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 2.0430555555555556,
|
| 515 |
+
"grad_norm": 0.016722600907087326,
|
| 516 |
+
"learning_rate": 4.207175925925926e-05,
|
| 517 |
+
"loss": 0.0005,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.046527777777778,
|
| 522 |
+
"grad_norm": 0.013101249001920223,
|
| 523 |
+
"learning_rate": 4.187885802469136e-05,
|
| 524 |
+
"loss": 0.0077,
|
| 525 |
+
"step": 710
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 2.05,
|
| 529 |
+
"grad_norm": 0.004911151714622974,
|
| 530 |
+
"learning_rate": 4.168595679012346e-05,
|
| 531 |
+
"loss": 0.0088,
|
| 532 |
+
"step": 720
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 2.053472222222222,
|
| 536 |
+
"grad_norm": 0.0035500626545399427,
|
| 537 |
+
"learning_rate": 4.149305555555556e-05,
|
| 538 |
+
"loss": 0.0004,
|
| 539 |
+
"step": 730
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 2.0569444444444445,
|
| 543 |
+
"grad_norm": 0.003059967188164592,
|
| 544 |
+
"learning_rate": 4.130015432098766e-05,
|
| 545 |
+
"loss": 0.0009,
|
| 546 |
+
"step": 740
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 2.060416666666667,
|
| 550 |
+
"grad_norm": 2.3829972743988037,
|
| 551 |
+
"learning_rate": 4.110725308641975e-05,
|
| 552 |
+
"loss": 0.0007,
|
| 553 |
+
"step": 750
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.063888888888889,
|
| 557 |
+
"grad_norm": 0.0023200404830276966,
|
| 558 |
+
"learning_rate": 4.0914351851851855e-05,
|
| 559 |
+
"loss": 0.0479,
|
| 560 |
+
"step": 760
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.067361111111111,
|
| 564 |
+
"grad_norm": 0.0032272222451865673,
|
| 565 |
+
"learning_rate": 4.072145061728395e-05,
|
| 566 |
+
"loss": 0.0001,
|
| 567 |
+
"step": 770
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.0708333333333333,
|
| 571 |
+
"grad_norm": 0.0031554449815303087,
|
| 572 |
+
"learning_rate": 4.052854938271605e-05,
|
| 573 |
+
"loss": 0.0102,
|
| 574 |
+
"step": 780
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 2.0743055555555556,
|
| 578 |
+
"grad_norm": 0.6856481432914734,
|
| 579 |
+
"learning_rate": 4.033564814814815e-05,
|
| 580 |
+
"loss": 0.0003,
|
| 581 |
+
"step": 790
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 2.077777777777778,
|
| 585 |
+
"grad_norm": 0.12966661155223846,
|
| 586 |
+
"learning_rate": 4.014274691358025e-05,
|
| 587 |
+
"loss": 0.0002,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 2.08125,
|
| 592 |
+
"grad_norm": 0.005961376242339611,
|
| 593 |
+
"learning_rate": 3.9949845679012346e-05,
|
| 594 |
+
"loss": 0.1566,
|
| 595 |
+
"step": 810
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 2.084722222222222,
|
| 599 |
+
"grad_norm": 0.0027580575551837683,
|
| 600 |
+
"learning_rate": 3.975694444444444e-05,
|
| 601 |
+
"loss": 0.0522,
|
| 602 |
+
"step": 820
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.0881944444444445,
|
| 606 |
+
"grad_norm": 0.006767706014215946,
|
| 607 |
+
"learning_rate": 3.956404320987654e-05,
|
| 608 |
+
"loss": 0.0002,
|
| 609 |
+
"step": 830
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.091666666666667,
|
| 613 |
+
"grad_norm": 0.0025393294636160135,
|
| 614 |
+
"learning_rate": 3.9371141975308645e-05,
|
| 615 |
+
"loss": 0.0002,
|
| 616 |
+
"step": 840
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 2.095138888888889,
|
| 620 |
+
"grad_norm": 0.0009938559960573912,
|
| 621 |
+
"learning_rate": 3.917824074074074e-05,
|
| 622 |
+
"loss": 0.0001,
|
| 623 |
+
"step": 850
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 2.098611111111111,
|
| 627 |
+
"grad_norm": 0.001436732243746519,
|
| 628 |
+
"learning_rate": 3.898533950617284e-05,
|
| 629 |
+
"loss": 0.0117,
|
| 630 |
+
"step": 860
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.1,
|
| 634 |
+
"eval_accuracy": 0.8373983739837398,
|
| 635 |
+
"eval_loss": 1.0666677951812744,
|
| 636 |
+
"eval_runtime": 120.4352,
|
| 637 |
+
"eval_samples_per_second": 4.085,
|
| 638 |
+
"eval_steps_per_second": 0.515,
|
| 639 |
+
"step": 864
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.002083333333333,
|
| 643 |
+
"grad_norm": 0.03854840621352196,
|
| 644 |
+
"learning_rate": 3.879243827160494e-05,
|
| 645 |
+
"loss": 0.0022,
|
| 646 |
+
"step": 870
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.0055555555555555,
|
| 650 |
+
"grad_norm": 0.0014166207984089851,
|
| 651 |
+
"learning_rate": 3.8599537037037034e-05,
|
| 652 |
+
"loss": 0.0001,
|
| 653 |
+
"step": 880
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.009027777777778,
|
| 657 |
+
"grad_norm": 0.0014319154433906078,
|
| 658 |
+
"learning_rate": 3.8406635802469136e-05,
|
| 659 |
+
"loss": 0.0001,
|
| 660 |
+
"step": 890
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.0125,
|
| 664 |
+
"grad_norm": 0.0013770597288385034,
|
| 665 |
+
"learning_rate": 3.821373456790124e-05,
|
| 666 |
+
"loss": 0.0003,
|
| 667 |
+
"step": 900
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.015972222222222,
|
| 671 |
+
"grad_norm": 0.006136466283351183,
|
| 672 |
+
"learning_rate": 3.8020833333333334e-05,
|
| 673 |
+
"loss": 0.0001,
|
| 674 |
+
"step": 910
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.0194444444444444,
|
| 678 |
+
"grad_norm": 0.0012927391799166799,
|
| 679 |
+
"learning_rate": 3.7827932098765436e-05,
|
| 680 |
+
"loss": 0.0001,
|
| 681 |
+
"step": 920
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0229166666666667,
|
| 685 |
+
"grad_norm": 0.0009090897510759532,
|
| 686 |
+
"learning_rate": 3.763503086419753e-05,
|
| 687 |
+
"loss": 0.0001,
|
| 688 |
+
"step": 930
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.026388888888889,
|
| 692 |
+
"grad_norm": 0.007072359789162874,
|
| 693 |
+
"learning_rate": 3.744212962962963e-05,
|
| 694 |
+
"loss": 0.0001,
|
| 695 |
+
"step": 940
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.029861111111111,
|
| 699 |
+
"grad_norm": 0.000913340481929481,
|
| 700 |
+
"learning_rate": 3.724922839506173e-05,
|
| 701 |
+
"loss": 0.0001,
|
| 702 |
+
"step": 950
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.033333333333333,
|
| 706 |
+
"grad_norm": 0.0012041399022564292,
|
| 707 |
+
"learning_rate": 3.705632716049383e-05,
|
| 708 |
+
"loss": 0.0001,
|
| 709 |
+
"step": 960
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.0368055555555555,
|
| 713 |
+
"grad_norm": 0.0008073053904809058,
|
| 714 |
+
"learning_rate": 3.6863425925925926e-05,
|
| 715 |
+
"loss": 0.0,
|
| 716 |
+
"step": 970
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.040277777777778,
|
| 720 |
+
"grad_norm": 0.0016801038291305304,
|
| 721 |
+
"learning_rate": 3.667052469135803e-05,
|
| 722 |
+
"loss": 0.0002,
|
| 723 |
+
"step": 980
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.04375,
|
| 727 |
+
"grad_norm": 2.7720425128936768,
|
| 728 |
+
"learning_rate": 3.647762345679013e-05,
|
| 729 |
+
"loss": 0.0095,
|
| 730 |
+
"step": 990
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.047222222222222,
|
| 734 |
+
"grad_norm": 0.0007318072021007538,
|
| 735 |
+
"learning_rate": 3.628472222222222e-05,
|
| 736 |
+
"loss": 0.0001,
|
| 737 |
+
"step": 1000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.0506944444444444,
|
| 741 |
+
"grad_norm": 0.0022503375075757504,
|
| 742 |
+
"learning_rate": 3.609182098765432e-05,
|
| 743 |
+
"loss": 0.0005,
|
| 744 |
+
"step": 1010
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.0541666666666667,
|
| 748 |
+
"grad_norm": 0.0006915331468917429,
|
| 749 |
+
"learning_rate": 3.5898919753086424e-05,
|
| 750 |
+
"loss": 0.0001,
|
| 751 |
+
"step": 1020
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.057638888888889,
|
| 755 |
+
"grad_norm": 0.0023397556506097317,
|
| 756 |
+
"learning_rate": 3.570601851851852e-05,
|
| 757 |
+
"loss": 0.0002,
|
| 758 |
+
"step": 1030
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.061111111111111,
|
| 762 |
+
"grad_norm": 0.0009047880303114653,
|
| 763 |
+
"learning_rate": 3.551311728395062e-05,
|
| 764 |
+
"loss": 0.0028,
|
| 765 |
+
"step": 1040
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.064583333333333,
|
| 769 |
+
"grad_norm": 0.003123590722680092,
|
| 770 |
+
"learning_rate": 3.532021604938272e-05,
|
| 771 |
+
"loss": 0.0,
|
| 772 |
+
"step": 1050
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.0680555555555555,
|
| 776 |
+
"grad_norm": 0.003919905982911587,
|
| 777 |
+
"learning_rate": 3.512731481481481e-05,
|
| 778 |
+
"loss": 0.0002,
|
| 779 |
+
"step": 1060
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.071527777777778,
|
| 783 |
+
"grad_norm": 0.001332888612523675,
|
| 784 |
+
"learning_rate": 3.4934413580246915e-05,
|
| 785 |
+
"loss": 0.003,
|
| 786 |
+
"step": 1070
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.075,
|
| 790 |
+
"grad_norm": 0.000786742486525327,
|
| 791 |
+
"learning_rate": 3.474151234567901e-05,
|
| 792 |
+
"loss": 0.0219,
|
| 793 |
+
"step": 1080
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.078472222222222,
|
| 797 |
+
"grad_norm": 0.0009092329419218004,
|
| 798 |
+
"learning_rate": 3.454861111111111e-05,
|
| 799 |
+
"loss": 0.0001,
|
| 800 |
+
"step": 1090
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.0819444444444444,
|
| 804 |
+
"grad_norm": 0.022490588948130608,
|
| 805 |
+
"learning_rate": 3.4355709876543214e-05,
|
| 806 |
+
"loss": 0.0001,
|
| 807 |
+
"step": 1100
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.0854166666666667,
|
| 811 |
+
"grad_norm": 0.0010289773344993591,
|
| 812 |
+
"learning_rate": 3.416280864197531e-05,
|
| 813 |
+
"loss": 0.0,
|
| 814 |
+
"step": 1110
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.088888888888889,
|
| 818 |
+
"grad_norm": 0.03139325976371765,
|
| 819 |
+
"learning_rate": 3.396990740740741e-05,
|
| 820 |
+
"loss": 0.0,
|
| 821 |
+
"step": 1120
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.092361111111111,
|
| 825 |
+
"grad_norm": 1.5481621026992798,
|
| 826 |
+
"learning_rate": 3.377700617283951e-05,
|
| 827 |
+
"loss": 0.0007,
|
| 828 |
+
"step": 1130
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.095833333333333,
|
| 832 |
+
"grad_norm": 0.4089558720588684,
|
| 833 |
+
"learning_rate": 3.35841049382716e-05,
|
| 834 |
+
"loss": 0.0017,
|
| 835 |
+
"step": 1140
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.0993055555555555,
|
| 839 |
+
"grad_norm": 0.024518415331840515,
|
| 840 |
+
"learning_rate": 3.3391203703703705e-05,
|
| 841 |
+
"loss": 0.0001,
|
| 842 |
+
"step": 1150
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.1,
|
| 846 |
+
"eval_accuracy": 0.7764227642276422,
|
| 847 |
+
"eval_loss": 1.549495816230774,
|
| 848 |
+
"eval_runtime": 119.7871,
|
| 849 |
+
"eval_samples_per_second": 4.107,
|
| 850 |
+
"eval_steps_per_second": 0.518,
|
| 851 |
+
"step": 1152
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.1,
|
| 855 |
+
"step": 1152,
|
| 856 |
+
"total_flos": 1.1483731256571593e+19,
|
| 857 |
+
"train_loss": 0.12870611023062553,
|
| 858 |
+
"train_runtime": 7836.725,
|
| 859 |
+
"train_samples_per_second": 2.94,
|
| 860 |
+
"train_steps_per_second": 0.368
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 3.1,
|
| 864 |
+
"eval_accuracy": 0.8502994011976048,
|
| 865 |
+
"eval_loss": 0.6543134450912476,
|
| 866 |
+
"eval_runtime": 569.5516,
|
| 867 |
+
"eval_samples_per_second": 0.88,
|
| 868 |
+
"eval_steps_per_second": 0.111,
|
| 869 |
+
"step": 1152
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.1,
|
| 873 |
+
"eval_accuracy": 0.8502994011976048,
|
| 874 |
+
"eval_loss": 0.6543134450912476,
|
| 875 |
+
"eval_runtime": 119.6956,
|
| 876 |
+
"eval_samples_per_second": 4.186,
|
| 877 |
+
"eval_steps_per_second": 0.526,
|
| 878 |
+
"step": 1152
|
| 879 |
+
}
|
| 880 |
+
],
|
| 881 |
+
"logging_steps": 10,
|
| 882 |
+
"max_steps": 2880,
|
| 883 |
+
"num_input_tokens_seen": 0,
|
| 884 |
+
"num_train_epochs": 9223372036854775807,
|
| 885 |
+
"save_steps": 500,
|
| 886 |
+
"stateful_callbacks": {
|
| 887 |
+
"EarlyStoppingCallback": {
|
| 888 |
+
"args": {
|
| 889 |
+
"early_stopping_patience": 2,
|
| 890 |
+
"early_stopping_threshold": 0.0
|
| 891 |
+
},
|
| 892 |
+
"attributes": {
|
| 893 |
+
"early_stopping_patience_counter": 2
|
| 894 |
+
}
|
| 895 |
+
},
|
| 896 |
+
"TrainerControl": {
|
| 897 |
+
"args": {
|
| 898 |
+
"should_epoch_stop": false,
|
| 899 |
+
"should_evaluate": false,
|
| 900 |
+
"should_log": false,
|
| 901 |
+
"should_save": true,
|
| 902 |
+
"should_training_stop": true
|
| 903 |
+
},
|
| 904 |
+
"attributes": {}
|
| 905 |
+
}
|
| 906 |
+
},
|
| 907 |
+
"total_flos": 1.1483731256571593e+19,
|
| 908 |
+
"train_batch_size": 8,
|
| 909 |
+
"trial_name": null,
|
| 910 |
+
"trial_params": null
|
| 911 |
+
}
|