Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +874 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,874 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.087527352297593,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 120,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0007293946024799417,
|
| 14 |
+
"grad_norm": 3.4206197261810303,
|
| 15 |
+
"learning_rate": 4e-05,
|
| 16 |
+
"loss": 1.0407,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0014587892049598833,
|
| 21 |
+
"grad_norm": 2.839146614074707,
|
| 22 |
+
"learning_rate": 8e-05,
|
| 23 |
+
"loss": 1.0393,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.002188183807439825,
|
| 28 |
+
"grad_norm": 2.998342275619507,
|
| 29 |
+
"learning_rate": 0.00012,
|
| 30 |
+
"loss": 1.0121,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0029175784099197666,
|
| 35 |
+
"grad_norm": 2.9921820163726807,
|
| 36 |
+
"learning_rate": 0.00016,
|
| 37 |
+
"loss": 0.8346,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0036469730123997084,
|
| 42 |
+
"grad_norm": 3.6432197093963623,
|
| 43 |
+
"learning_rate": 0.0002,
|
| 44 |
+
"loss": 0.9709,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.00437636761487965,
|
| 49 |
+
"grad_norm": 3.543360948562622,
|
| 50 |
+
"learning_rate": 0.0001982608695652174,
|
| 51 |
+
"loss": 0.888,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.005105762217359592,
|
| 56 |
+
"grad_norm": 3.7492313385009766,
|
| 57 |
+
"learning_rate": 0.0001965217391304348,
|
| 58 |
+
"loss": 0.8022,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.005835156819839533,
|
| 63 |
+
"grad_norm": 3.5265772342681885,
|
| 64 |
+
"learning_rate": 0.00019478260869565218,
|
| 65 |
+
"loss": 0.9769,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.006564551422319475,
|
| 70 |
+
"grad_norm": 4.798442363739014,
|
| 71 |
+
"learning_rate": 0.00019304347826086958,
|
| 72 |
+
"loss": 1.1666,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.007293946024799417,
|
| 77 |
+
"grad_norm": 4.2874436378479,
|
| 78 |
+
"learning_rate": 0.00019130434782608697,
|
| 79 |
+
"loss": 1.2155,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.008023340627279359,
|
| 84 |
+
"grad_norm": 4.175326824188232,
|
| 85 |
+
"learning_rate": 0.00018956521739130436,
|
| 86 |
+
"loss": 1.1292,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0087527352297593,
|
| 91 |
+
"grad_norm": 3.6792125701904297,
|
| 92 |
+
"learning_rate": 0.00018782608695652175,
|
| 93 |
+
"loss": 1.0746,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.009482129832239242,
|
| 98 |
+
"grad_norm": 3.5794081687927246,
|
| 99 |
+
"learning_rate": 0.00018608695652173914,
|
| 100 |
+
"loss": 1.0253,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.010211524434719184,
|
| 105 |
+
"grad_norm": 3.7175612449645996,
|
| 106 |
+
"learning_rate": 0.00018434782608695653,
|
| 107 |
+
"loss": 1.2297,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.010940919037199124,
|
| 112 |
+
"grad_norm": 2.981461763381958,
|
| 113 |
+
"learning_rate": 0.00018260869565217392,
|
| 114 |
+
"loss": 1.0819,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.011670313639679067,
|
| 119 |
+
"grad_norm": 3.286484479904175,
|
| 120 |
+
"learning_rate": 0.00018086956521739132,
|
| 121 |
+
"loss": 0.9201,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.012399708242159009,
|
| 126 |
+
"grad_norm": 3.14170241355896,
|
| 127 |
+
"learning_rate": 0.0001791304347826087,
|
| 128 |
+
"loss": 0.9966,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01312910284463895,
|
| 133 |
+
"grad_norm": 3.2921481132507324,
|
| 134 |
+
"learning_rate": 0.0001773913043478261,
|
| 135 |
+
"loss": 1.0784,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.013858497447118891,
|
| 140 |
+
"grad_norm": 4.169837474822998,
|
| 141 |
+
"learning_rate": 0.0001756521739130435,
|
| 142 |
+
"loss": 1.3222,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.014587892049598834,
|
| 147 |
+
"grad_norm": 3.0340821743011475,
|
| 148 |
+
"learning_rate": 0.00017391304347826088,
|
| 149 |
+
"loss": 1.2287,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.015317286652078774,
|
| 154 |
+
"grad_norm": 3.216778516769409,
|
| 155 |
+
"learning_rate": 0.00017217391304347827,
|
| 156 |
+
"loss": 1.0829,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.016046681254558718,
|
| 161 |
+
"grad_norm": 3.5301125049591064,
|
| 162 |
+
"learning_rate": 0.00017043478260869566,
|
| 163 |
+
"loss": 1.2479,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.016776075857038657,
|
| 168 |
+
"grad_norm": 3.6915533542633057,
|
| 169 |
+
"learning_rate": 0.00016869565217391306,
|
| 170 |
+
"loss": 1.2559,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0175054704595186,
|
| 175 |
+
"grad_norm": 4.012074947357178,
|
| 176 |
+
"learning_rate": 0.00016695652173913042,
|
| 177 |
+
"loss": 1.4941,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.01823486506199854,
|
| 182 |
+
"grad_norm": 3.469731092453003,
|
| 183 |
+
"learning_rate": 0.00016521739130434784,
|
| 184 |
+
"loss": 1.3116,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.018964259664478483,
|
| 189 |
+
"grad_norm": 3.1409828662872314,
|
| 190 |
+
"learning_rate": 0.00016347826086956523,
|
| 191 |
+
"loss": 1.2485,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.019693654266958426,
|
| 196 |
+
"grad_norm": 3.4184839725494385,
|
| 197 |
+
"learning_rate": 0.00016173913043478262,
|
| 198 |
+
"loss": 1.5548,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.020423048869438368,
|
| 203 |
+
"grad_norm": 3.288313388824463,
|
| 204 |
+
"learning_rate": 0.00016,
|
| 205 |
+
"loss": 1.3852,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.021152443471918306,
|
| 210 |
+
"grad_norm": 3.2971737384796143,
|
| 211 |
+
"learning_rate": 0.0001582608695652174,
|
| 212 |
+
"loss": 1.4702,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.02188183807439825,
|
| 217 |
+
"grad_norm": 3.3972525596618652,
|
| 218 |
+
"learning_rate": 0.0001565217391304348,
|
| 219 |
+
"loss": 1.6733,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.02261123267687819,
|
| 224 |
+
"grad_norm": 3.280540704727173,
|
| 225 |
+
"learning_rate": 0.0001547826086956522,
|
| 226 |
+
"loss": 1.6348,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.023340627279358133,
|
| 231 |
+
"grad_norm": 3.1876862049102783,
|
| 232 |
+
"learning_rate": 0.00015304347826086958,
|
| 233 |
+
"loss": 1.8272,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.024070021881838075,
|
| 238 |
+
"grad_norm": 3.2754197120666504,
|
| 239 |
+
"learning_rate": 0.00015130434782608694,
|
| 240 |
+
"loss": 1.6901,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.024799416484318017,
|
| 245 |
+
"grad_norm": 3.196545124053955,
|
| 246 |
+
"learning_rate": 0.00014956521739130436,
|
| 247 |
+
"loss": 1.7045,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.025528811086797956,
|
| 252 |
+
"grad_norm": 2.9313583374023438,
|
| 253 |
+
"learning_rate": 0.00014782608695652173,
|
| 254 |
+
"loss": 1.3504,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0262582056892779,
|
| 259 |
+
"grad_norm": 3.0764999389648438,
|
| 260 |
+
"learning_rate": 0.00014608695652173914,
|
| 261 |
+
"loss": 1.6135,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.02698760029175784,
|
| 266 |
+
"grad_norm": 3.048112392425537,
|
| 267 |
+
"learning_rate": 0.00014434782608695654,
|
| 268 |
+
"loss": 1.7037,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.027716994894237783,
|
| 273 |
+
"grad_norm": 2.651155948638916,
|
| 274 |
+
"learning_rate": 0.00014260869565217393,
|
| 275 |
+
"loss": 1.2778,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.028446389496717725,
|
| 280 |
+
"grad_norm": 2.9630019664764404,
|
| 281 |
+
"learning_rate": 0.00014086956521739132,
|
| 282 |
+
"loss": 1.4965,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.029175784099197667,
|
| 287 |
+
"grad_norm": 3.275195598602295,
|
| 288 |
+
"learning_rate": 0.0001391304347826087,
|
| 289 |
+
"loss": 1.8892,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.029905178701677606,
|
| 294 |
+
"grad_norm": 2.9564414024353027,
|
| 295 |
+
"learning_rate": 0.0001373913043478261,
|
| 296 |
+
"loss": 1.5416,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.030634573304157548,
|
| 301 |
+
"grad_norm": 2.5216736793518066,
|
| 302 |
+
"learning_rate": 0.00013565217391304347,
|
| 303 |
+
"loss": 1.3729,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.031363967906637494,
|
| 308 |
+
"grad_norm": 4.906222343444824,
|
| 309 |
+
"learning_rate": 0.00013391304347826088,
|
| 310 |
+
"loss": 1.6411,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.032093362509117436,
|
| 315 |
+
"grad_norm": 2.5856339931488037,
|
| 316 |
+
"learning_rate": 0.00013217391304347825,
|
| 317 |
+
"loss": 1.482,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.03282275711159737,
|
| 322 |
+
"grad_norm": 2.53202223777771,
|
| 323 |
+
"learning_rate": 0.00013043478260869567,
|
| 324 |
+
"loss": 1.6198,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.03355215171407731,
|
| 329 |
+
"grad_norm": 2.823518753051758,
|
| 330 |
+
"learning_rate": 0.00012869565217391303,
|
| 331 |
+
"loss": 1.5032,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.034281546316557256,
|
| 336 |
+
"grad_norm": 5.919870376586914,
|
| 337 |
+
"learning_rate": 0.00012695652173913045,
|
| 338 |
+
"loss": 1.6127,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.0350109409190372,
|
| 343 |
+
"grad_norm": 3.135148525238037,
|
| 344 |
+
"learning_rate": 0.00012521739130434784,
|
| 345 |
+
"loss": 1.7123,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.03574033552151714,
|
| 350 |
+
"grad_norm": 2.7107229232788086,
|
| 351 |
+
"learning_rate": 0.00012347826086956523,
|
| 352 |
+
"loss": 1.3955,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.03646973012399708,
|
| 357 |
+
"grad_norm": 2.641472578048706,
|
| 358 |
+
"learning_rate": 0.00012173913043478263,
|
| 359 |
+
"loss": 1.3128,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.037199124726477024,
|
| 364 |
+
"grad_norm": 2.926663398742676,
|
| 365 |
+
"learning_rate": 0.00012,
|
| 366 |
+
"loss": 1.5215,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.03792851932895697,
|
| 371 |
+
"grad_norm": 2.6990513801574707,
|
| 372 |
+
"learning_rate": 0.00011826086956521741,
|
| 373 |
+
"loss": 1.4037,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.03865791393143691,
|
| 378 |
+
"grad_norm": 3.1344690322875977,
|
| 379 |
+
"learning_rate": 0.00011652173913043479,
|
| 380 |
+
"loss": 1.7596,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.03938730853391685,
|
| 385 |
+
"grad_norm": 2.6223137378692627,
|
| 386 |
+
"learning_rate": 0.00011478260869565218,
|
| 387 |
+
"loss": 1.243,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.04011670313639679,
|
| 392 |
+
"grad_norm": 2.5806047916412354,
|
| 393 |
+
"learning_rate": 0.00011304347826086956,
|
| 394 |
+
"loss": 1.2454,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.040846097738876735,
|
| 399 |
+
"grad_norm": 4.001445293426514,
|
| 400 |
+
"learning_rate": 0.00011130434782608696,
|
| 401 |
+
"loss": 1.9131,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.04157549234135667,
|
| 406 |
+
"grad_norm": 2.714655637741089,
|
| 407 |
+
"learning_rate": 0.00010956521739130434,
|
| 408 |
+
"loss": 1.3318,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.04230488694383661,
|
| 413 |
+
"grad_norm": 2.9351234436035156,
|
| 414 |
+
"learning_rate": 0.00010782608695652174,
|
| 415 |
+
"loss": 1.3198,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.043034281546316555,
|
| 420 |
+
"grad_norm": 3.144646644592285,
|
| 421 |
+
"learning_rate": 0.00010608695652173915,
|
| 422 |
+
"loss": 1.5481,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.0437636761487965,
|
| 427 |
+
"grad_norm": 2.698659658432007,
|
| 428 |
+
"learning_rate": 0.00010434782608695653,
|
| 429 |
+
"loss": 1.5793,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.04449307075127644,
|
| 434 |
+
"grad_norm": 2.60088849067688,
|
| 435 |
+
"learning_rate": 0.00010260869565217393,
|
| 436 |
+
"loss": 1.5449,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.04522246535375638,
|
| 441 |
+
"grad_norm": 2.440284490585327,
|
| 442 |
+
"learning_rate": 0.00010086956521739131,
|
| 443 |
+
"loss": 1.4398,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.045951859956236324,
|
| 448 |
+
"grad_norm": 3.003382444381714,
|
| 449 |
+
"learning_rate": 9.91304347826087e-05,
|
| 450 |
+
"loss": 1.7254,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.046681254558716266,
|
| 455 |
+
"grad_norm": 2.8187639713287354,
|
| 456 |
+
"learning_rate": 9.739130434782609e-05,
|
| 457 |
+
"loss": 1.4895,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.04741064916119621,
|
| 462 |
+
"grad_norm": 2.6653711795806885,
|
| 463 |
+
"learning_rate": 9.565217391304348e-05,
|
| 464 |
+
"loss": 1.3805,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.04814004376367615,
|
| 469 |
+
"grad_norm": 2.566655158996582,
|
| 470 |
+
"learning_rate": 9.391304347826087e-05,
|
| 471 |
+
"loss": 1.4169,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.04886943836615609,
|
| 476 |
+
"grad_norm": 2.4838967323303223,
|
| 477 |
+
"learning_rate": 9.217391304347827e-05,
|
| 478 |
+
"loss": 1.3721,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.049598832968636035,
|
| 483 |
+
"grad_norm": 2.3984270095825195,
|
| 484 |
+
"learning_rate": 9.043478260869566e-05,
|
| 485 |
+
"loss": 1.3303,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.05032822757111598,
|
| 490 |
+
"grad_norm": 2.7791898250579834,
|
| 491 |
+
"learning_rate": 8.869565217391305e-05,
|
| 492 |
+
"loss": 1.435,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.05105762217359591,
|
| 497 |
+
"grad_norm": 2.599867105484009,
|
| 498 |
+
"learning_rate": 8.695652173913044e-05,
|
| 499 |
+
"loss": 1.4253,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.051787016776075855,
|
| 504 |
+
"grad_norm": 2.8419313430786133,
|
| 505 |
+
"learning_rate": 8.521739130434783e-05,
|
| 506 |
+
"loss": 1.4364,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.0525164113785558,
|
| 511 |
+
"grad_norm": 2.829371452331543,
|
| 512 |
+
"learning_rate": 8.347826086956521e-05,
|
| 513 |
+
"loss": 1.4784,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.05324580598103574,
|
| 518 |
+
"grad_norm": 2.3601653575897217,
|
| 519 |
+
"learning_rate": 8.173913043478262e-05,
|
| 520 |
+
"loss": 1.3688,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.05397520058351568,
|
| 525 |
+
"grad_norm": 3.19826340675354,
|
| 526 |
+
"learning_rate": 8e-05,
|
| 527 |
+
"loss": 1.5037,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.05470459518599562,
|
| 532 |
+
"grad_norm": 2.285161256790161,
|
| 533 |
+
"learning_rate": 7.82608695652174e-05,
|
| 534 |
+
"loss": 1.0791,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.055433989788475566,
|
| 539 |
+
"grad_norm": 2.3437016010284424,
|
| 540 |
+
"learning_rate": 7.652173913043479e-05,
|
| 541 |
+
"loss": 1.2071,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.05616338439095551,
|
| 546 |
+
"grad_norm": 2.8213608264923096,
|
| 547 |
+
"learning_rate": 7.478260869565218e-05,
|
| 548 |
+
"loss": 1.4101,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.05689277899343545,
|
| 553 |
+
"grad_norm": 2.7381675243377686,
|
| 554 |
+
"learning_rate": 7.304347826086957e-05,
|
| 555 |
+
"loss": 1.4432,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.05762217359591539,
|
| 560 |
+
"grad_norm": 2.711702823638916,
|
| 561 |
+
"learning_rate": 7.130434782608696e-05,
|
| 562 |
+
"loss": 1.6227,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.058351568198395334,
|
| 567 |
+
"grad_norm": 2.6544787883758545,
|
| 568 |
+
"learning_rate": 6.956521739130436e-05,
|
| 569 |
+
"loss": 1.6657,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.05908096280087528,
|
| 574 |
+
"grad_norm": 2.7479665279388428,
|
| 575 |
+
"learning_rate": 6.782608695652173e-05,
|
| 576 |
+
"loss": 1.4854,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.05981035740335521,
|
| 581 |
+
"grad_norm": 2.199681282043457,
|
| 582 |
+
"learning_rate": 6.608695652173912e-05,
|
| 583 |
+
"loss": 1.1914,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.060539752005835154,
|
| 588 |
+
"grad_norm": 3.368340253829956,
|
| 589 |
+
"learning_rate": 6.434782608695652e-05,
|
| 590 |
+
"loss": 1.1718,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.061269146608315096,
|
| 595 |
+
"grad_norm": 2.5759427547454834,
|
| 596 |
+
"learning_rate": 6.260869565217392e-05,
|
| 597 |
+
"loss": 1.2188,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.06199854121079504,
|
| 602 |
+
"grad_norm": 3.044017791748047,
|
| 603 |
+
"learning_rate": 6.086956521739131e-05,
|
| 604 |
+
"loss": 1.466,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.06272793581327499,
|
| 609 |
+
"grad_norm": 2.7837345600128174,
|
| 610 |
+
"learning_rate": 5.9130434782608704e-05,
|
| 611 |
+
"loss": 1.3557,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.06345733041575492,
|
| 616 |
+
"grad_norm": 2.68403697013855,
|
| 617 |
+
"learning_rate": 5.739130434782609e-05,
|
| 618 |
+
"loss": 1.3008,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.06418672501823487,
|
| 623 |
+
"grad_norm": 2.668792486190796,
|
| 624 |
+
"learning_rate": 5.565217391304348e-05,
|
| 625 |
+
"loss": 1.6496,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.06491611962071481,
|
| 630 |
+
"grad_norm": 2.3745946884155273,
|
| 631 |
+
"learning_rate": 5.391304347826087e-05,
|
| 632 |
+
"loss": 1.2317,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.06564551422319474,
|
| 637 |
+
"grad_norm": 3.019322156906128,
|
| 638 |
+
"learning_rate": 5.217391304347826e-05,
|
| 639 |
+
"loss": 1.4779,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.06637490882567469,
|
| 644 |
+
"grad_norm": 2.546433210372925,
|
| 645 |
+
"learning_rate": 5.0434782608695655e-05,
|
| 646 |
+
"loss": 1.3536,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.06710430342815463,
|
| 651 |
+
"grad_norm": 2.510876178741455,
|
| 652 |
+
"learning_rate": 4.8695652173913046e-05,
|
| 653 |
+
"loss": 1.3767,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.06783369803063458,
|
| 658 |
+
"grad_norm": 2.549628973007202,
|
| 659 |
+
"learning_rate": 4.695652173913044e-05,
|
| 660 |
+
"loss": 1.3804,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.06856309263311451,
|
| 665 |
+
"grad_norm": 2.5327932834625244,
|
| 666 |
+
"learning_rate": 4.521739130434783e-05,
|
| 667 |
+
"loss": 1.4853,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.06929248723559446,
|
| 672 |
+
"grad_norm": 2.2544209957122803,
|
| 673 |
+
"learning_rate": 4.347826086956522e-05,
|
| 674 |
+
"loss": 1.2218,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.0700218818380744,
|
| 679 |
+
"grad_norm": 2.487122058868408,
|
| 680 |
+
"learning_rate": 4.1739130434782605e-05,
|
| 681 |
+
"loss": 1.1685,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.07075127644055434,
|
| 686 |
+
"grad_norm": 2.6893062591552734,
|
| 687 |
+
"learning_rate": 4e-05,
|
| 688 |
+
"loss": 1.4211,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.07148067104303428,
|
| 693 |
+
"grad_norm": 2.7432773113250732,
|
| 694 |
+
"learning_rate": 3.8260869565217395e-05,
|
| 695 |
+
"loss": 1.3966,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.07221006564551423,
|
| 700 |
+
"grad_norm": 2.1431596279144287,
|
| 701 |
+
"learning_rate": 3.6521739130434786e-05,
|
| 702 |
+
"loss": 1.3541,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.07293946024799416,
|
| 707 |
+
"grad_norm": 2.632312774658203,
|
| 708 |
+
"learning_rate": 3.478260869565218e-05,
|
| 709 |
+
"loss": 1.3412,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.0736688548504741,
|
| 714 |
+
"grad_norm": 2.4512953758239746,
|
| 715 |
+
"learning_rate": 3.304347826086956e-05,
|
| 716 |
+
"loss": 1.2732,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.07439824945295405,
|
| 721 |
+
"grad_norm": 2.49603009223938,
|
| 722 |
+
"learning_rate": 3.130434782608696e-05,
|
| 723 |
+
"loss": 1.2898,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.07512764405543398,
|
| 728 |
+
"grad_norm": 2.4943108558654785,
|
| 729 |
+
"learning_rate": 2.9565217391304352e-05,
|
| 730 |
+
"loss": 1.5888,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.07585703865791393,
|
| 735 |
+
"grad_norm": 2.4405064582824707,
|
| 736 |
+
"learning_rate": 2.782608695652174e-05,
|
| 737 |
+
"loss": 1.5069,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.07658643326039387,
|
| 742 |
+
"grad_norm": 2.517979145050049,
|
| 743 |
+
"learning_rate": 2.608695652173913e-05,
|
| 744 |
+
"loss": 1.3677,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.07731582786287382,
|
| 749 |
+
"grad_norm": 2.156628370285034,
|
| 750 |
+
"learning_rate": 2.4347826086956523e-05,
|
| 751 |
+
"loss": 1.212,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.07804522246535375,
|
| 756 |
+
"grad_norm": 2.101703405380249,
|
| 757 |
+
"learning_rate": 2.2608695652173914e-05,
|
| 758 |
+
"loss": 1.0052,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.0787746170678337,
|
| 763 |
+
"grad_norm": 2.3143198490142822,
|
| 764 |
+
"learning_rate": 2.0869565217391303e-05,
|
| 765 |
+
"loss": 1.4326,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.07950401167031364,
|
| 770 |
+
"grad_norm": 2.255425214767456,
|
| 771 |
+
"learning_rate": 1.9130434782608697e-05,
|
| 772 |
+
"loss": 1.2735,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.08023340627279359,
|
| 777 |
+
"grad_norm": 2.6278059482574463,
|
| 778 |
+
"learning_rate": 1.739130434782609e-05,
|
| 779 |
+
"loss": 1.3857,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.08096280087527352,
|
| 784 |
+
"grad_norm": 2.3596839904785156,
|
| 785 |
+
"learning_rate": 1.565217391304348e-05,
|
| 786 |
+
"loss": 1.2475,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.08169219547775347,
|
| 791 |
+
"grad_norm": 2.384459972381592,
|
| 792 |
+
"learning_rate": 1.391304347826087e-05,
|
| 793 |
+
"loss": 1.2721,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.0824215900802334,
|
| 798 |
+
"grad_norm": 2.311274528503418,
|
| 799 |
+
"learning_rate": 1.2173913043478261e-05,
|
| 800 |
+
"loss": 1.2485,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.08315098468271334,
|
| 805 |
+
"grad_norm": 2.1393139362335205,
|
| 806 |
+
"learning_rate": 1.0434782608695651e-05,
|
| 807 |
+
"loss": 0.9907,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.08388037928519329,
|
| 812 |
+
"grad_norm": 2.6496288776397705,
|
| 813 |
+
"learning_rate": 8.695652173913044e-06,
|
| 814 |
+
"loss": 1.5222,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.08460977388767323,
|
| 819 |
+
"grad_norm": 2.1133055686950684,
|
| 820 |
+
"learning_rate": 6.956521739130435e-06,
|
| 821 |
+
"loss": 1.1745,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.08533916849015317,
|
| 826 |
+
"grad_norm": 2.4991602897644043,
|
| 827 |
+
"learning_rate": 5.217391304347826e-06,
|
| 828 |
+
"loss": 1.4379,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.08606856309263311,
|
| 833 |
+
"grad_norm": 2.292524814605713,
|
| 834 |
+
"learning_rate": 3.4782608695652175e-06,
|
| 835 |
+
"loss": 1.358,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.08679795769511306,
|
| 840 |
+
"grad_norm": 2.232152223587036,
|
| 841 |
+
"learning_rate": 1.7391304347826088e-06,
|
| 842 |
+
"loss": 1.1723,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.087527352297593,
|
| 847 |
+
"grad_norm": 2.235368013381958,
|
| 848 |
+
"learning_rate": 0.0,
|
| 849 |
+
"loss": 1.2209,
|
| 850 |
+
"step": 120
|
| 851 |
+
}
|
| 852 |
+
],
|
| 853 |
+
"logging_steps": 1,
|
| 854 |
+
"max_steps": 120,
|
| 855 |
+
"num_input_tokens_seen": 0,
|
| 856 |
+
"num_train_epochs": 1,
|
| 857 |
+
"save_steps": 500,
|
| 858 |
+
"stateful_callbacks": {
|
| 859 |
+
"TrainerControl": {
|
| 860 |
+
"args": {
|
| 861 |
+
"should_epoch_stop": false,
|
| 862 |
+
"should_evaluate": false,
|
| 863 |
+
"should_log": false,
|
| 864 |
+
"should_save": true,
|
| 865 |
+
"should_training_stop": true
|
| 866 |
+
},
|
| 867 |
+
"attributes": {}
|
| 868 |
+
}
|
| 869 |
+
},
|
| 870 |
+
"total_flos": 4012563229948416.0,
|
| 871 |
+
"train_batch_size": 2,
|
| 872 |
+
"trial_name": null,
|
| 873 |
+
"trial_params": null
|
| 874 |
+
}
|