Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +994 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,994 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 16.932038834951456,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 136,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11650485436893204,
|
| 13 |
+
"grad_norm": 1.7820011377334595,
|
| 14 |
+
"learning_rate": 7.142857142857143e-07,
|
| 15 |
+
"loss": 0.4027,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.23300970873786409,
|
| 20 |
+
"grad_norm": 1.7389452457427979,
|
| 21 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 22 |
+
"loss": 0.402,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.34951456310679613,
|
| 27 |
+
"grad_norm": 1.74604332447052,
|
| 28 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 29 |
+
"loss": 0.3966,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.46601941747572817,
|
| 34 |
+
"grad_norm": 1.7389558553695679,
|
| 35 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 36 |
+
"loss": 0.404,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5825242718446602,
|
| 41 |
+
"grad_norm": 1.6699142456054688,
|
| 42 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 43 |
+
"loss": 0.3975,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6990291262135923,
|
| 48 |
+
"grad_norm": 1.5735969543457031,
|
| 49 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 50 |
+
"loss": 0.3923,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8155339805825242,
|
| 55 |
+
"grad_norm": 1.4590719938278198,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.405,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9320388349514563,
|
| 62 |
+
"grad_norm": 1.2758820056915283,
|
| 63 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 64 |
+
"loss": 0.3803,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.116504854368932,
|
| 69 |
+
"grad_norm": 1.3157119750976562,
|
| 70 |
+
"learning_rate": 6.4285714285714295e-06,
|
| 71 |
+
"loss": 0.7377,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.233009708737864,
|
| 76 |
+
"grad_norm": 0.6604185700416565,
|
| 77 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 78 |
+
"loss": 0.3704,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.3495145631067962,
|
| 83 |
+
"grad_norm": 0.5376127362251282,
|
| 84 |
+
"learning_rate": 7.857142857142858e-06,
|
| 85 |
+
"loss": 0.3616,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.4660194174757282,
|
| 90 |
+
"grad_norm": 1.2837574481964111,
|
| 91 |
+
"learning_rate": 8.571428571428571e-06,
|
| 92 |
+
"loss": 0.3684,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.5825242718446602,
|
| 97 |
+
"grad_norm": 1.3026747703552246,
|
| 98 |
+
"learning_rate": 9.285714285714288e-06,
|
| 99 |
+
"loss": 0.3633,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.6990291262135924,
|
| 104 |
+
"grad_norm": 1.3066328763961792,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 0.3629,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.8155339805825244,
|
| 111 |
+
"grad_norm": 1.172119379043579,
|
| 112 |
+
"learning_rate": 9.998342337571566e-06,
|
| 113 |
+
"loss": 0.3631,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.9320388349514563,
|
| 118 |
+
"grad_norm": 0.9207857251167297,
|
| 119 |
+
"learning_rate": 9.993370449424153e-06,
|
| 120 |
+
"loss": 0.3506,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.116504854368932,
|
| 125 |
+
"grad_norm": 1.1936880350112915,
|
| 126 |
+
"learning_rate": 9.985087632242634e-06,
|
| 127 |
+
"loss": 0.7052,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.233009708737864,
|
| 132 |
+
"grad_norm": 0.589175283908844,
|
| 133 |
+
"learning_rate": 9.973499378072947e-06,
|
| 134 |
+
"loss": 0.3364,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.349514563106796,
|
| 139 |
+
"grad_norm": 0.4948810040950775,
|
| 140 |
+
"learning_rate": 9.958613370680507e-06,
|
| 141 |
+
"loss": 0.3307,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.466019417475728,
|
| 146 |
+
"grad_norm": 0.4881579875946045,
|
| 147 |
+
"learning_rate": 9.940439480455386e-06,
|
| 148 |
+
"loss": 0.3345,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.58252427184466,
|
| 153 |
+
"grad_norm": 0.5249760150909424,
|
| 154 |
+
"learning_rate": 9.918989757867584e-06,
|
| 155 |
+
"loss": 0.329,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.6990291262135924,
|
| 160 |
+
"grad_norm": 0.5218749046325684,
|
| 161 |
+
"learning_rate": 9.89427842547679e-06,
|
| 162 |
+
"loss": 0.3212,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.8155339805825244,
|
| 167 |
+
"grad_norm": 0.48922351002693176,
|
| 168 |
+
"learning_rate": 9.866321868501914e-06,
|
| 169 |
+
"loss": 0.317,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.9320388349514563,
|
| 174 |
+
"grad_norm": 0.3870461881160736,
|
| 175 |
+
"learning_rate": 9.835138623956603e-06,
|
| 176 |
+
"loss": 0.3131,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.116504854368932,
|
| 181 |
+
"grad_norm": 0.33141207695007324,
|
| 182 |
+
"learning_rate": 9.80074936835801e-06,
|
| 183 |
+
"loss": 0.5829,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.233009708737864,
|
| 188 |
+
"grad_norm": 0.32673436403274536,
|
| 189 |
+
"learning_rate": 9.763176904016914e-06,
|
| 190 |
+
"loss": 0.3009,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.349514563106796,
|
| 195 |
+
"grad_norm": 0.33463019132614136,
|
| 196 |
+
"learning_rate": 9.722446143918307e-06,
|
| 197 |
+
"loss": 0.3081,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.466019417475728,
|
| 202 |
+
"grad_norm": 0.29102134704589844,
|
| 203 |
+
"learning_rate": 9.678584095202468e-06,
|
| 204 |
+
"loss": 0.2774,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.58252427184466,
|
| 209 |
+
"grad_norm": 0.31500688195228577,
|
| 210 |
+
"learning_rate": 9.631619841257477e-06,
|
| 211 |
+
"loss": 0.2803,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.6990291262135924,
|
| 216 |
+
"grad_norm": 0.30529990792274475,
|
| 217 |
+
"learning_rate": 9.581584522435025e-06,
|
| 218 |
+
"loss": 0.2804,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.8155339805825244,
|
| 223 |
+
"grad_norm": 0.28200799226760864,
|
| 224 |
+
"learning_rate": 9.528511315402358e-06,
|
| 225 |
+
"loss": 0.2687,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.9320388349514563,
|
| 230 |
+
"grad_norm": 0.27144694328308105,
|
| 231 |
+
"learning_rate": 9.472435411143979e-06,
|
| 232 |
+
"loss": 0.2722,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.116504854368932,
|
| 237 |
+
"grad_norm": 0.43679553270339966,
|
| 238 |
+
"learning_rate": 9.413393991627737e-06,
|
| 239 |
+
"loss": 0.526,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.233009708737864,
|
| 244 |
+
"grad_norm": 0.25320038199424744,
|
| 245 |
+
"learning_rate": 9.351426205150778e-06,
|
| 246 |
+
"loss": 0.2568,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 4.349514563106796,
|
| 251 |
+
"grad_norm": 0.22309613227844238,
|
| 252 |
+
"learning_rate": 9.286573140381663e-06,
|
| 253 |
+
"loss": 0.2472,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.466019417475728,
|
| 258 |
+
"grad_norm": 0.20222555100917816,
|
| 259 |
+
"learning_rate": 9.218877799115929e-06,
|
| 260 |
+
"loss": 0.2229,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.58252427184466,
|
| 265 |
+
"grad_norm": 0.21417675912380219,
|
| 266 |
+
"learning_rate": 9.148385067763094e-06,
|
| 267 |
+
"loss": 0.2326,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.699029126213592,
|
| 272 |
+
"grad_norm": 0.26747238636016846,
|
| 273 |
+
"learning_rate": 9.075141687584056e-06,
|
| 274 |
+
"loss": 0.2339,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.815533980582524,
|
| 279 |
+
"grad_norm": 0.2342718243598938,
|
| 280 |
+
"learning_rate": 8.999196223698599e-06,
|
| 281 |
+
"loss": 0.2304,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.932038834951456,
|
| 286 |
+
"grad_norm": 0.23761263489723206,
|
| 287 |
+
"learning_rate": 8.920599032883553e-06,
|
| 288 |
+
"loss": 0.2331,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 5.116504854368932,
|
| 293 |
+
"grad_norm": 0.3980836570262909,
|
| 294 |
+
"learning_rate": 8.839402230183e-06,
|
| 295 |
+
"loss": 0.4385,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.233009708737864,
|
| 300 |
+
"grad_norm": 0.22973115742206573,
|
| 301 |
+
"learning_rate": 8.755659654352599e-06,
|
| 302 |
+
"loss": 0.2075,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.349514563106796,
|
| 307 |
+
"grad_norm": 0.21187792718410492,
|
| 308 |
+
"learning_rate": 8.669426832160997e-06,
|
| 309 |
+
"loss": 0.1988,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.466019417475728,
|
| 314 |
+
"grad_norm": 0.20210106670856476,
|
| 315 |
+
"learning_rate": 8.580760941571968e-06,
|
| 316 |
+
"loss": 0.1886,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.58252427184466,
|
| 321 |
+
"grad_norm": 0.24134023487567902,
|
| 322 |
+
"learning_rate": 8.489720773831717e-06,
|
| 323 |
+
"loss": 0.1984,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.699029126213592,
|
| 328 |
+
"grad_norm": 0.22599560022354126,
|
| 329 |
+
"learning_rate": 8.396366694486466e-06,
|
| 330 |
+
"loss": 0.1767,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.815533980582524,
|
| 335 |
+
"grad_norm": 0.20777945220470428,
|
| 336 |
+
"learning_rate": 8.30076060335616e-06,
|
| 337 |
+
"loss": 0.1834,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.932038834951456,
|
| 342 |
+
"grad_norm": 0.20150279998779297,
|
| 343 |
+
"learning_rate": 8.202965893490877e-06,
|
| 344 |
+
"loss": 0.1878,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.116504854368932,
|
| 349 |
+
"grad_norm": 0.2539062798023224,
|
| 350 |
+
"learning_rate": 8.103047409137114e-06,
|
| 351 |
+
"loss": 0.3518,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.233009708737864,
|
| 356 |
+
"grad_norm": 0.1837664395570755,
|
| 357 |
+
"learning_rate": 8.001071402741843e-06,
|
| 358 |
+
"loss": 0.152,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.349514563106796,
|
| 363 |
+
"grad_norm": 0.20487381517887115,
|
| 364 |
+
"learning_rate": 7.897105491022819e-06,
|
| 365 |
+
"loss": 0.1557,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.466019417475728,
|
| 370 |
+
"grad_norm": 0.24897804856300354,
|
| 371 |
+
"learning_rate": 7.791218610134324e-06,
|
| 372 |
+
"loss": 0.154,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.58252427184466,
|
| 377 |
+
"grad_norm": 0.18702609837055206,
|
| 378 |
+
"learning_rate": 7.683480969958005e-06,
|
| 379 |
+
"loss": 0.1479,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.699029126213592,
|
| 384 |
+
"grad_norm": 0.19429221749305725,
|
| 385 |
+
"learning_rate": 7.5739640075491546e-06,
|
| 386 |
+
"loss": 0.142,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.815533980582524,
|
| 391 |
+
"grad_norm": 0.20301687717437744,
|
| 392 |
+
"learning_rate": 7.462740339769323e-06,
|
| 393 |
+
"loss": 0.1462,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.932038834951456,
|
| 398 |
+
"grad_norm": 0.17031480371952057,
|
| 399 |
+
"learning_rate": 7.349883715136601e-06,
|
| 400 |
+
"loss": 0.1417,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 7.116504854368932,
|
| 405 |
+
"grad_norm": 0.34165364503860474,
|
| 406 |
+
"learning_rate": 7.235468964925571e-06,
|
| 407 |
+
"loss": 0.2634,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.233009708737864,
|
| 412 |
+
"grad_norm": 0.17344710230827332,
|
| 413 |
+
"learning_rate": 7.119571953549305e-06,
|
| 414 |
+
"loss": 0.1224,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.349514563106796,
|
| 419 |
+
"grad_norm": 0.17058910429477692,
|
| 420 |
+
"learning_rate": 7.002269528256334e-06,
|
| 421 |
+
"loss": 0.1097,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.466019417475728,
|
| 426 |
+
"grad_norm": 0.21691343188285828,
|
| 427 |
+
"learning_rate": 6.883639468175926e-06,
|
| 428 |
+
"loss": 0.1112,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.58252427184466,
|
| 433 |
+
"grad_norm": 0.1959153115749359,
|
| 434 |
+
"learning_rate": 6.763760432745475e-06,
|
| 435 |
+
"loss": 0.0991,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.699029126213592,
|
| 440 |
+
"grad_norm": 0.15394563972949982,
|
| 441 |
+
"learning_rate": 6.6427119095541745e-06,
|
| 442 |
+
"loss": 0.1107,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.815533980582524,
|
| 447 |
+
"grad_norm": 0.18826133012771606,
|
| 448 |
+
"learning_rate": 6.520574161637591e-06,
|
| 449 |
+
"loss": 0.1086,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.932038834951456,
|
| 454 |
+
"grad_norm": 0.21221591532230377,
|
| 455 |
+
"learning_rate": 6.397428174258048e-06,
|
| 456 |
+
"loss": 0.1062,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 8.116504854368932,
|
| 461 |
+
"grad_norm": 0.3180226981639862,
|
| 462 |
+
"learning_rate": 6.273355601206143e-06,
|
| 463 |
+
"loss": 0.1849,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 8.233009708737864,
|
| 468 |
+
"grad_norm": 0.17723064124584198,
|
| 469 |
+
"learning_rate": 6.148438710658979e-06,
|
| 470 |
+
"loss": 0.0826,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 8.349514563106796,
|
| 475 |
+
"grad_norm": 0.19632180035114288,
|
| 476 |
+
"learning_rate": 6.022760330631006e-06,
|
| 477 |
+
"loss": 0.082,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.466019417475728,
|
| 482 |
+
"grad_norm": 0.22644098103046417,
|
| 483 |
+
"learning_rate": 5.896403794053679e-06,
|
| 484 |
+
"loss": 0.0773,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 8.58252427184466,
|
| 489 |
+
"grad_norm": 0.1707296371459961,
|
| 490 |
+
"learning_rate": 5.76945288352031e-06,
|
| 491 |
+
"loss": 0.078,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 8.699029126213592,
|
| 496 |
+
"grad_norm": 0.16724658012390137,
|
| 497 |
+
"learning_rate": 5.641991775732756e-06,
|
| 498 |
+
"loss": 0.0766,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 8.815533980582524,
|
| 503 |
+
"grad_norm": 0.18082226812839508,
|
| 504 |
+
"learning_rate": 5.514104985686802e-06,
|
| 505 |
+
"loss": 0.0776,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 8.932038834951456,
|
| 510 |
+
"grad_norm": 0.22430583834648132,
|
| 511 |
+
"learning_rate": 5.385877310633233e-06,
|
| 512 |
+
"loss": 0.0778,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 9.116504854368932,
|
| 517 |
+
"grad_norm": 0.17507706582546234,
|
| 518 |
+
"learning_rate": 5.257393773851733e-06,
|
| 519 |
+
"loss": 0.1409,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 9.233009708737864,
|
| 524 |
+
"grad_norm": 0.1874881386756897,
|
| 525 |
+
"learning_rate": 5.1287395682749444e-06,
|
| 526 |
+
"loss": 0.0568,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 9.349514563106796,
|
| 531 |
+
"grad_norm": 0.3130532205104828,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.0553,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 9.466019417475728,
|
| 538 |
+
"grad_norm": 0.17253145575523376,
|
| 539 |
+
"learning_rate": 4.871260431725058e-06,
|
| 540 |
+
"loss": 0.0553,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 9.58252427184466,
|
| 545 |
+
"grad_norm": 0.16302235424518585,
|
| 546 |
+
"learning_rate": 4.742606226148268e-06,
|
| 547 |
+
"loss": 0.0545,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 9.699029126213592,
|
| 552 |
+
"grad_norm": 0.18094216287136078,
|
| 553 |
+
"learning_rate": 4.614122689366769e-06,
|
| 554 |
+
"loss": 0.0548,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.815533980582524,
|
| 559 |
+
"grad_norm": 0.1743619292974472,
|
| 560 |
+
"learning_rate": 4.485895014313198e-06,
|
| 561 |
+
"loss": 0.0512,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 9.932038834951456,
|
| 566 |
+
"grad_norm": 0.16189619898796082,
|
| 567 |
+
"learning_rate": 4.358008224267245e-06,
|
| 568 |
+
"loss": 0.0496,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 10.116504854368932,
|
| 573 |
+
"grad_norm": 0.3171091675758362,
|
| 574 |
+
"learning_rate": 4.230547116479691e-06,
|
| 575 |
+
"loss": 0.1007,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.233009708737864,
|
| 580 |
+
"grad_norm": 0.154684916138649,
|
| 581 |
+
"learning_rate": 4.103596205946323e-06,
|
| 582 |
+
"loss": 0.0414,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 10.349514563106796,
|
| 587 |
+
"grad_norm": 0.1529693901538849,
|
| 588 |
+
"learning_rate": 3.977239669368998e-06,
|
| 589 |
+
"loss": 0.0378,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 10.466019417475728,
|
| 594 |
+
"grad_norm": 0.18426883220672607,
|
| 595 |
+
"learning_rate": 3.851561289341023e-06,
|
| 596 |
+
"loss": 0.0392,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 10.58252427184466,
|
| 601 |
+
"grad_norm": 0.18451394140720367,
|
| 602 |
+
"learning_rate": 3.726644398793857e-06,
|
| 603 |
+
"loss": 0.0383,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 10.699029126213592,
|
| 608 |
+
"grad_norm": 0.17146584391593933,
|
| 609 |
+
"learning_rate": 3.6025718257419532e-06,
|
| 610 |
+
"loss": 0.0346,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 10.815533980582524,
|
| 615 |
+
"grad_norm": 0.1295372098684311,
|
| 616 |
+
"learning_rate": 3.4794258383624115e-06,
|
| 617 |
+
"loss": 0.0347,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 10.932038834951456,
|
| 622 |
+
"grad_norm": 0.15434859693050385,
|
| 623 |
+
"learning_rate": 3.3572880904458267e-06,
|
| 624 |
+
"loss": 0.0402,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 11.116504854368932,
|
| 629 |
+
"grad_norm": 0.3346112370491028,
|
| 630 |
+
"learning_rate": 3.236239567254526e-06,
|
| 631 |
+
"loss": 0.0675,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 11.233009708737864,
|
| 636 |
+
"grad_norm": 0.19830213487148285,
|
| 637 |
+
"learning_rate": 3.116360531824074e-06,
|
| 638 |
+
"loss": 0.0298,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 11.349514563106796,
|
| 643 |
+
"grad_norm": 0.11921335756778717,
|
| 644 |
+
"learning_rate": 2.997730471743667e-06,
|
| 645 |
+
"loss": 0.0271,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 11.466019417475728,
|
| 650 |
+
"grad_norm": 0.16639156639575958,
|
| 651 |
+
"learning_rate": 2.880428046450697e-06,
|
| 652 |
+
"loss": 0.027,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 11.58252427184466,
|
| 657 |
+
"grad_norm": 0.20680218935012817,
|
| 658 |
+
"learning_rate": 2.7645310350744296e-06,
|
| 659 |
+
"loss": 0.0263,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 11.699029126213592,
|
| 664 |
+
"grad_norm": 0.23083102703094482,
|
| 665 |
+
"learning_rate": 2.6501162848634023e-06,
|
| 666 |
+
"loss": 0.0272,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 11.815533980582524,
|
| 671 |
+
"grad_norm": 0.1634109765291214,
|
| 672 |
+
"learning_rate": 2.537259660230679e-06,
|
| 673 |
+
"loss": 0.027,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 11.932038834951456,
|
| 678 |
+
"grad_norm": 0.14731772243976593,
|
| 679 |
+
"learning_rate": 2.426035992450848e-06,
|
| 680 |
+
"loss": 0.027,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 12.116504854368932,
|
| 685 |
+
"grad_norm": 0.18388910591602325,
|
| 686 |
+
"learning_rate": 2.316519030041998e-06,
|
| 687 |
+
"loss": 0.0456,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 12.233009708737864,
|
| 692 |
+
"grad_norm": 0.1441938877105713,
|
| 693 |
+
"learning_rate": 2.2087813898656775e-06,
|
| 694 |
+
"loss": 0.0193,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 12.349514563106796,
|
| 699 |
+
"grad_norm": 0.1402544379234314,
|
| 700 |
+
"learning_rate": 2.102894508977182e-06,
|
| 701 |
+
"loss": 0.0201,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 12.466019417475728,
|
| 706 |
+
"grad_norm": 0.13963298499584198,
|
| 707 |
+
"learning_rate": 1.9989285972581595e-06,
|
| 708 |
+
"loss": 0.0208,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 12.58252427184466,
|
| 713 |
+
"grad_norm": 0.15653668344020844,
|
| 714 |
+
"learning_rate": 1.896952590862886e-06,
|
| 715 |
+
"loss": 0.0186,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 12.699029126213592,
|
| 720 |
+
"grad_norm": 0.1681966930627823,
|
| 721 |
+
"learning_rate": 1.7970341065091246e-06,
|
| 722 |
+
"loss": 0.0185,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.815533980582524,
|
| 727 |
+
"grad_norm": 0.15047527849674225,
|
| 728 |
+
"learning_rate": 1.699239396643841e-06,
|
| 729 |
+
"loss": 0.0182,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 12.932038834951456,
|
| 734 |
+
"grad_norm": 0.10233399271965027,
|
| 735 |
+
"learning_rate": 1.6036333055135345e-06,
|
| 736 |
+
"loss": 0.0189,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 13.116504854368932,
|
| 741 |
+
"grad_norm": 0.21229885518550873,
|
| 742 |
+
"learning_rate": 1.5102792261682813e-06,
|
| 743 |
+
"loss": 0.0353,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 13.233009708737864,
|
| 748 |
+
"grad_norm": 0.12984932959079742,
|
| 749 |
+
"learning_rate": 1.4192390584280347e-06,
|
| 750 |
+
"loss": 0.0178,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 13.349514563106796,
|
| 755 |
+
"grad_norm": 0.10708415508270264,
|
| 756 |
+
"learning_rate": 1.330573167839005e-06,
|
| 757 |
+
"loss": 0.0147,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 13.466019417475728,
|
| 762 |
+
"grad_norm": 0.11098859459161758,
|
| 763 |
+
"learning_rate": 1.2443403456474017e-06,
|
| 764 |
+
"loss": 0.0139,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 13.58252427184466,
|
| 769 |
+
"grad_norm": 0.09491042047739029,
|
| 770 |
+
"learning_rate": 1.1605977698170001e-06,
|
| 771 |
+
"loss": 0.0142,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 13.699029126213592,
|
| 776 |
+
"grad_norm": 0.0828198492527008,
|
| 777 |
+
"learning_rate": 1.0794009671164484e-06,
|
| 778 |
+
"loss": 0.0141,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 13.815533980582524,
|
| 783 |
+
"grad_norm": 0.09377221018075943,
|
| 784 |
+
"learning_rate": 1.0008037763014033e-06,
|
| 785 |
+
"loss": 0.0156,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 13.932038834951456,
|
| 790 |
+
"grad_norm": 0.1043652817606926,
|
| 791 |
+
"learning_rate": 9.248583124159438e-07,
|
| 792 |
+
"loss": 0.0147,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 14.116504854368932,
|
| 797 |
+
"grad_norm": 0.1541670560836792,
|
| 798 |
+
"learning_rate": 8.516149322369055e-07,
|
| 799 |
+
"loss": 0.0246,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 14.233009708737864,
|
| 804 |
+
"grad_norm": 0.07124023884534836,
|
| 805 |
+
"learning_rate": 7.811222008840719e-07,
|
| 806 |
+
"loss": 0.0139,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 14.349514563106796,
|
| 811 |
+
"grad_norm": 0.07342305779457092,
|
| 812 |
+
"learning_rate": 7.13426859618338e-07,
|
| 813 |
+
"loss": 0.0126,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 14.466019417475728,
|
| 818 |
+
"grad_norm": 0.06642112135887146,
|
| 819 |
+
"learning_rate": 6.485737948492237e-07,
|
| 820 |
+
"loss": 0.0119,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 14.58252427184466,
|
| 825 |
+
"grad_norm": 0.06306291371583939,
|
| 826 |
+
"learning_rate": 5.866060083722624e-07,
|
| 827 |
+
"loss": 0.0112,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 14.699029126213592,
|
| 832 |
+
"grad_norm": 0.06630466133356094,
|
| 833 |
+
"learning_rate": 5.275645888560233e-07,
|
| 834 |
+
"loss": 0.0121,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 14.815533980582524,
|
| 839 |
+
"grad_norm": 0.06889720261096954,
|
| 840 |
+
"learning_rate": 4.71488684597643e-07,
|
| 841 |
+
"loss": 0.0116,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 14.932038834951456,
|
| 846 |
+
"grad_norm": 0.06075473129749298,
|
| 847 |
+
"learning_rate": 4.184154775649768e-07,
|
| 848 |
+
"loss": 0.014,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 15.116504854368932,
|
| 853 |
+
"grad_norm": 0.07335468381643295,
|
| 854 |
+
"learning_rate": 3.683801587425251e-07,
|
| 855 |
+
"loss": 0.0269,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 15.233009708737864,
|
| 860 |
+
"grad_norm": 0.13383059203624725,
|
| 861 |
+
"learning_rate": 3.214159047975324e-07,
|
| 862 |
+
"loss": 0.0105,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 15.349514563106796,
|
| 867 |
+
"grad_norm": 0.06708887219429016,
|
| 868 |
+
"learning_rate": 2.7755385608169374e-07,
|
| 869 |
+
"loss": 0.011,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 15.466019417475728,
|
| 874 |
+
"grad_norm": 0.061848320066928864,
|
| 875 |
+
"learning_rate": 2.368230959830875e-07,
|
| 876 |
+
"loss": 0.012,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 15.58252427184466,
|
| 881 |
+
"grad_norm": 0.0594082772731781,
|
| 882 |
+
"learning_rate": 1.992506316419912e-07,
|
| 883 |
+
"loss": 0.0113,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 15.699029126213592,
|
| 888 |
+
"grad_norm": 0.06530940532684326,
|
| 889 |
+
"learning_rate": 1.6486137604339813e-07,
|
| 890 |
+
"loss": 0.0117,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 15.815533980582524,
|
| 895 |
+
"grad_norm": 0.07374989986419678,
|
| 896 |
+
"learning_rate": 1.3367813149808728e-07,
|
| 897 |
+
"loss": 0.0106,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 15.932038834951456,
|
| 902 |
+
"grad_norm": 0.0554700568318367,
|
| 903 |
+
"learning_rate": 1.0572157452321097e-07,
|
| 904 |
+
"loss": 0.01,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 16.116504854368934,
|
| 909 |
+
"grad_norm": 0.11609751731157303,
|
| 910 |
+
"learning_rate": 8.101024213241826e-08,
|
| 911 |
+
"loss": 0.0216,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 16.233009708737864,
|
| 916 |
+
"grad_norm": 0.05223438888788223,
|
| 917 |
+
"learning_rate": 5.9560519544614725e-08,
|
| 918 |
+
"loss": 0.0108,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 16.349514563106798,
|
| 923 |
+
"grad_norm": 0.05844803899526596,
|
| 924 |
+
"learning_rate": 4.138662931949255e-08,
|
| 925 |
+
"loss": 0.0105,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 16.466019417475728,
|
| 930 |
+
"grad_norm": 0.052356619387865067,
|
| 931 |
+
"learning_rate": 2.6500621927054716e-08,
|
| 932 |
+
"loss": 0.0105,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 16.58252427184466,
|
| 937 |
+
"grad_norm": 0.05942930653691292,
|
| 938 |
+
"learning_rate": 1.4912367757366485e-08,
|
| 939 |
+
"loss": 0.0107,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 16.699029126213592,
|
| 944 |
+
"grad_norm": 0.05563804879784584,
|
| 945 |
+
"learning_rate": 6.629550575847355e-09,
|
| 946 |
+
"loss": 0.0115,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 16.815533980582526,
|
| 951 |
+
"grad_norm": 0.05824480950832367,
|
| 952 |
+
"learning_rate": 1.657662428434792e-09,
|
| 953 |
+
"loss": 0.0113,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 16.932038834951456,
|
| 958 |
+
"grad_norm": 0.05532587692141533,
|
| 959 |
+
"learning_rate": 0.0,
|
| 960 |
+
"loss": 0.0111,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 16.932038834951456,
|
| 965 |
+
"step": 136,
|
| 966 |
+
"total_flos": 493320970436608.0,
|
| 967 |
+
"train_loss": 0.15380282930838055,
|
| 968 |
+
"train_runtime": 33702.47,
|
| 969 |
+
"train_samples_per_second": 0.414,
|
| 970 |
+
"train_steps_per_second": 0.004
|
| 971 |
+
}
|
| 972 |
+
],
|
| 973 |
+
"logging_steps": 1,
|
| 974 |
+
"max_steps": 136,
|
| 975 |
+
"num_input_tokens_seen": 0,
|
| 976 |
+
"num_train_epochs": 17,
|
| 977 |
+
"save_steps": 10000,
|
| 978 |
+
"stateful_callbacks": {
|
| 979 |
+
"TrainerControl": {
|
| 980 |
+
"args": {
|
| 981 |
+
"should_epoch_stop": false,
|
| 982 |
+
"should_evaluate": false,
|
| 983 |
+
"should_log": false,
|
| 984 |
+
"should_save": true,
|
| 985 |
+
"should_training_stop": true
|
| 986 |
+
},
|
| 987 |
+
"attributes": {}
|
| 988 |
+
}
|
| 989 |
+
},
|
| 990 |
+
"total_flos": 493320970436608.0,
|
| 991 |
+
"train_batch_size": 1,
|
| 992 |
+
"trial_name": null,
|
| 993 |
+
"trial_params": null
|
| 994 |
+
}
|