Upload trainer_state.json
Browse files- trainer_state.json +1020 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1020 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 77340,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"learning_rate": 9.935350400827516e-05,
|
| 13 |
+
"loss": 8.0382,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.13,
|
| 18 |
+
"learning_rate": 9.87070080165503e-05,
|
| 19 |
+
"loss": 7.6429,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.19,
|
| 24 |
+
"learning_rate": 9.806051202482544e-05,
|
| 25 |
+
"loss": 7.1341,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.26,
|
| 30 |
+
"learning_rate": 9.74140160331006e-05,
|
| 31 |
+
"loss": 6.9301,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.32,
|
| 36 |
+
"learning_rate": 9.676752004137574e-05,
|
| 37 |
+
"loss": 6.8441,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.39,
|
| 42 |
+
"learning_rate": 9.61210240496509e-05,
|
| 43 |
+
"loss": 6.7735,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.45,
|
| 48 |
+
"learning_rate": 9.547452805792605e-05,
|
| 49 |
+
"loss": 6.3512,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.52,
|
| 54 |
+
"learning_rate": 9.482803206620118e-05,
|
| 55 |
+
"loss": 5.6406,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.58,
|
| 60 |
+
"learning_rate": 9.418153607447634e-05,
|
| 61 |
+
"loss": 5.2743,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.65,
|
| 66 |
+
"learning_rate": 9.35350400827515e-05,
|
| 67 |
+
"loss": 5.0399,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.71,
|
| 72 |
+
"learning_rate": 9.288854409102664e-05,
|
| 73 |
+
"loss": 4.8407,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.78,
|
| 78 |
+
"learning_rate": 9.224204809930179e-05,
|
| 79 |
+
"loss": 4.6694,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.84,
|
| 84 |
+
"learning_rate": 9.159555210757695e-05,
|
| 85 |
+
"loss": 4.5262,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.91,
|
| 90 |
+
"learning_rate": 9.094905611585208e-05,
|
| 91 |
+
"loss": 4.4021,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.97,
|
| 96 |
+
"learning_rate": 9.030256012412723e-05,
|
| 97 |
+
"loss": 4.2888,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.0,
|
| 102 |
+
"eval_loss": 4.0826334953308105,
|
| 103 |
+
"eval_runtime": 475.8214,
|
| 104 |
+
"eval_samples_per_second": 693.498,
|
| 105 |
+
"eval_steps_per_second": 1.807,
|
| 106 |
+
"step": 7734
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.03,
|
| 110 |
+
"learning_rate": 8.965606413240239e-05,
|
| 111 |
+
"loss": 4.1903,
|
| 112 |
+
"step": 8000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.1,
|
| 116 |
+
"learning_rate": 8.900956814067753e-05,
|
| 117 |
+
"loss": 4.1066,
|
| 118 |
+
"step": 8500
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 1.16,
|
| 122 |
+
"learning_rate": 8.836307214895269e-05,
|
| 123 |
+
"loss": 4.0331,
|
| 124 |
+
"step": 9000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.23,
|
| 128 |
+
"learning_rate": 8.771657615722783e-05,
|
| 129 |
+
"loss": 3.9644,
|
| 130 |
+
"step": 9500
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.29,
|
| 134 |
+
"learning_rate": 8.707008016550297e-05,
|
| 135 |
+
"loss": 3.9044,
|
| 136 |
+
"step": 10000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.36,
|
| 140 |
+
"learning_rate": 8.642358417377813e-05,
|
| 141 |
+
"loss": 3.851,
|
| 142 |
+
"step": 10500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.42,
|
| 146 |
+
"learning_rate": 8.577708818205328e-05,
|
| 147 |
+
"loss": 3.8044,
|
| 148 |
+
"step": 11000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 1.49,
|
| 152 |
+
"learning_rate": 8.513059219032842e-05,
|
| 153 |
+
"loss": 3.761,
|
| 154 |
+
"step": 11500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.55,
|
| 158 |
+
"learning_rate": 8.448409619860357e-05,
|
| 159 |
+
"loss": 3.7197,
|
| 160 |
+
"step": 12000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.62,
|
| 164 |
+
"learning_rate": 8.383760020687872e-05,
|
| 165 |
+
"loss": 3.6803,
|
| 166 |
+
"step": 12500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.68,
|
| 170 |
+
"learning_rate": 8.319110421515386e-05,
|
| 171 |
+
"loss": 3.6381,
|
| 172 |
+
"step": 13000
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.75,
|
| 176 |
+
"learning_rate": 8.254460822342902e-05,
|
| 177 |
+
"loss": 3.6119,
|
| 178 |
+
"step": 13500
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.81,
|
| 182 |
+
"learning_rate": 8.189811223170418e-05,
|
| 183 |
+
"loss": 3.577,
|
| 184 |
+
"step": 14000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.87,
|
| 188 |
+
"learning_rate": 8.12516162399793e-05,
|
| 189 |
+
"loss": 3.5515,
|
| 190 |
+
"step": 14500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 1.94,
|
| 194 |
+
"learning_rate": 8.060512024825446e-05,
|
| 195 |
+
"loss": 3.5228,
|
| 196 |
+
"step": 15000
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 2.0,
|
| 200 |
+
"eval_loss": 3.361564874649048,
|
| 201 |
+
"eval_runtime": 476.9803,
|
| 202 |
+
"eval_samples_per_second": 691.813,
|
| 203 |
+
"eval_steps_per_second": 1.803,
|
| 204 |
+
"step": 15468
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.0,
|
| 208 |
+
"learning_rate": 7.995862425652962e-05,
|
| 209 |
+
"loss": 3.4922,
|
| 210 |
+
"step": 15500
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 2.07,
|
| 214 |
+
"learning_rate": 7.931212826480476e-05,
|
| 215 |
+
"loss": 3.4583,
|
| 216 |
+
"step": 16000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.13,
|
| 220 |
+
"learning_rate": 7.866563227307991e-05,
|
| 221 |
+
"loss": 3.4363,
|
| 222 |
+
"step": 16500
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 2.2,
|
| 226 |
+
"learning_rate": 7.801913628135506e-05,
|
| 227 |
+
"loss": 3.4124,
|
| 228 |
+
"step": 17000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.26,
|
| 232 |
+
"learning_rate": 7.73726402896302e-05,
|
| 233 |
+
"loss": 3.3911,
|
| 234 |
+
"step": 17500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.33,
|
| 238 |
+
"learning_rate": 7.672614429790535e-05,
|
| 239 |
+
"loss": 3.3731,
|
| 240 |
+
"step": 18000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.39,
|
| 244 |
+
"learning_rate": 7.607964830618051e-05,
|
| 245 |
+
"loss": 3.3487,
|
| 246 |
+
"step": 18500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 2.46,
|
| 250 |
+
"learning_rate": 7.543315231445565e-05,
|
| 251 |
+
"loss": 3.333,
|
| 252 |
+
"step": 19000
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 2.52,
|
| 256 |
+
"learning_rate": 7.478665632273081e-05,
|
| 257 |
+
"loss": 3.3162,
|
| 258 |
+
"step": 19500
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.59,
|
| 262 |
+
"learning_rate": 7.414016033100595e-05,
|
| 263 |
+
"loss": 3.2947,
|
| 264 |
+
"step": 20000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 2.65,
|
| 268 |
+
"learning_rate": 7.349366433928109e-05,
|
| 269 |
+
"loss": 3.2783,
|
| 270 |
+
"step": 20500
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.72,
|
| 274 |
+
"learning_rate": 7.284716834755625e-05,
|
| 275 |
+
"loss": 3.2641,
|
| 276 |
+
"step": 21000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.78,
|
| 280 |
+
"learning_rate": 7.22006723558314e-05,
|
| 281 |
+
"loss": 3.2446,
|
| 282 |
+
"step": 21500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.84,
|
| 286 |
+
"learning_rate": 7.155417636410655e-05,
|
| 287 |
+
"loss": 3.2326,
|
| 288 |
+
"step": 22000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.91,
|
| 292 |
+
"learning_rate": 7.090768037238169e-05,
|
| 293 |
+
"loss": 3.2149,
|
| 294 |
+
"step": 22500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.97,
|
| 298 |
+
"learning_rate": 7.026118438065684e-05,
|
| 299 |
+
"loss": 3.2034,
|
| 300 |
+
"step": 23000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 3.0,
|
| 304 |
+
"eval_loss": 3.070152997970581,
|
| 305 |
+
"eval_runtime": 476.4651,
|
| 306 |
+
"eval_samples_per_second": 692.561,
|
| 307 |
+
"eval_steps_per_second": 1.805,
|
| 308 |
+
"step": 23202
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.04,
|
| 312 |
+
"learning_rate": 6.961468838893199e-05,
|
| 313 |
+
"loss": 3.1826,
|
| 314 |
+
"step": 23500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.1,
|
| 318 |
+
"learning_rate": 6.896819239720714e-05,
|
| 319 |
+
"loss": 3.1653,
|
| 320 |
+
"step": 24000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.17,
|
| 324 |
+
"learning_rate": 6.83216964054823e-05,
|
| 325 |
+
"loss": 3.1519,
|
| 326 |
+
"step": 24500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.23,
|
| 330 |
+
"learning_rate": 6.767520041375744e-05,
|
| 331 |
+
"loss": 3.1387,
|
| 332 |
+
"step": 25000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.3,
|
| 336 |
+
"learning_rate": 6.702870442203258e-05,
|
| 337 |
+
"loss": 3.1317,
|
| 338 |
+
"step": 25500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.36,
|
| 342 |
+
"learning_rate": 6.638220843030774e-05,
|
| 343 |
+
"loss": 3.1206,
|
| 344 |
+
"step": 26000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.43,
|
| 348 |
+
"learning_rate": 6.573571243858288e-05,
|
| 349 |
+
"loss": 3.1074,
|
| 350 |
+
"step": 26500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.49,
|
| 354 |
+
"learning_rate": 6.508921644685804e-05,
|
| 355 |
+
"loss": 3.0959,
|
| 356 |
+
"step": 27000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 3.56,
|
| 360 |
+
"learning_rate": 6.444272045513318e-05,
|
| 361 |
+
"loss": 3.0846,
|
| 362 |
+
"step": 27500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 3.62,
|
| 366 |
+
"learning_rate": 6.379622446340833e-05,
|
| 367 |
+
"loss": 3.0691,
|
| 368 |
+
"step": 28000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.69,
|
| 372 |
+
"learning_rate": 6.314972847168348e-05,
|
| 373 |
+
"loss": 3.0621,
|
| 374 |
+
"step": 28500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.75,
|
| 378 |
+
"learning_rate": 6.250323247995863e-05,
|
| 379 |
+
"loss": 3.0496,
|
| 380 |
+
"step": 29000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.81,
|
| 384 |
+
"learning_rate": 6.185673648823377e-05,
|
| 385 |
+
"loss": 3.0406,
|
| 386 |
+
"step": 29500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.88,
|
| 390 |
+
"learning_rate": 6.121024049650892e-05,
|
| 391 |
+
"loss": 3.0332,
|
| 392 |
+
"step": 30000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.94,
|
| 396 |
+
"learning_rate": 6.056374450478407e-05,
|
| 397 |
+
"loss": 3.0235,
|
| 398 |
+
"step": 30500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 4.0,
|
| 402 |
+
"eval_loss": 2.890948534011841,
|
| 403 |
+
"eval_runtime": 475.9375,
|
| 404 |
+
"eval_samples_per_second": 693.328,
|
| 405 |
+
"eval_steps_per_second": 1.807,
|
| 406 |
+
"step": 30936
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 4.01,
|
| 410 |
+
"learning_rate": 5.991724851305922e-05,
|
| 411 |
+
"loss": 3.0144,
|
| 412 |
+
"step": 31000
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 4.07,
|
| 416 |
+
"learning_rate": 5.927075252133437e-05,
|
| 417 |
+
"loss": 2.9923,
|
| 418 |
+
"step": 31500
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 4.14,
|
| 422 |
+
"learning_rate": 5.8624256529609526e-05,
|
| 423 |
+
"loss": 2.9902,
|
| 424 |
+
"step": 32000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 4.2,
|
| 428 |
+
"learning_rate": 5.7977760537884675e-05,
|
| 429 |
+
"loss": 2.9746,
|
| 430 |
+
"step": 32500
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.27,
|
| 434 |
+
"learning_rate": 5.733126454615981e-05,
|
| 435 |
+
"loss": 2.9708,
|
| 436 |
+
"step": 33000
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.33,
|
| 440 |
+
"learning_rate": 5.6684768554434966e-05,
|
| 441 |
+
"loss": 2.9586,
|
| 442 |
+
"step": 33500
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.4,
|
| 446 |
+
"learning_rate": 5.6038272562710115e-05,
|
| 447 |
+
"loss": 2.9556,
|
| 448 |
+
"step": 34000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.46,
|
| 452 |
+
"learning_rate": 5.5391776570985264e-05,
|
| 453 |
+
"loss": 2.9471,
|
| 454 |
+
"step": 34500
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 4.53,
|
| 458 |
+
"learning_rate": 5.474528057926042e-05,
|
| 459 |
+
"loss": 2.941,
|
| 460 |
+
"step": 35000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.59,
|
| 464 |
+
"learning_rate": 5.4098784587535555e-05,
|
| 465 |
+
"loss": 2.9353,
|
| 466 |
+
"step": 35500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.65,
|
| 470 |
+
"learning_rate": 5.3452288595810704e-05,
|
| 471 |
+
"loss": 2.9263,
|
| 472 |
+
"step": 36000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.72,
|
| 476 |
+
"learning_rate": 5.280579260408586e-05,
|
| 477 |
+
"loss": 2.9193,
|
| 478 |
+
"step": 36500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 4.78,
|
| 482 |
+
"learning_rate": 5.215929661236101e-05,
|
| 483 |
+
"loss": 2.9101,
|
| 484 |
+
"step": 37000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 4.85,
|
| 488 |
+
"learning_rate": 5.151280062063616e-05,
|
| 489 |
+
"loss": 2.9071,
|
| 490 |
+
"step": 37500
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 4.91,
|
| 494 |
+
"learning_rate": 5.08663046289113e-05,
|
| 495 |
+
"loss": 2.9019,
|
| 496 |
+
"step": 38000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.98,
|
| 500 |
+
"learning_rate": 5.021980863718645e-05,
|
| 501 |
+
"loss": 2.89,
|
| 502 |
+
"step": 38500
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.0,
|
| 506 |
+
"eval_loss": 2.768686294555664,
|
| 507 |
+
"eval_runtime": 485.6271,
|
| 508 |
+
"eval_samples_per_second": 679.495,
|
| 509 |
+
"eval_steps_per_second": 1.771,
|
| 510 |
+
"step": 38670
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 5.04,
|
| 514 |
+
"learning_rate": 4.95733126454616e-05,
|
| 515 |
+
"loss": 2.8791,
|
| 516 |
+
"step": 39000
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 5.11,
|
| 520 |
+
"learning_rate": 4.892681665373675e-05,
|
| 521 |
+
"loss": 2.87,
|
| 522 |
+
"step": 39500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 5.17,
|
| 526 |
+
"learning_rate": 4.8280320662011896e-05,
|
| 527 |
+
"loss": 2.8664,
|
| 528 |
+
"step": 40000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 5.24,
|
| 532 |
+
"learning_rate": 4.7633824670287045e-05,
|
| 533 |
+
"loss": 2.859,
|
| 534 |
+
"step": 40500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 5.3,
|
| 538 |
+
"learning_rate": 4.6987328678562193e-05,
|
| 539 |
+
"loss": 2.8562,
|
| 540 |
+
"step": 41000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 5.37,
|
| 544 |
+
"learning_rate": 4.634083268683734e-05,
|
| 545 |
+
"loss": 2.8489,
|
| 546 |
+
"step": 41500
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 5.43,
|
| 550 |
+
"learning_rate": 4.569433669511249e-05,
|
| 551 |
+
"loss": 2.846,
|
| 552 |
+
"step": 42000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 5.5,
|
| 556 |
+
"learning_rate": 4.504784070338764e-05,
|
| 557 |
+
"loss": 2.8342,
|
| 558 |
+
"step": 42500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 5.56,
|
| 562 |
+
"learning_rate": 4.440134471166279e-05,
|
| 563 |
+
"loss": 2.8313,
|
| 564 |
+
"step": 43000
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 5.62,
|
| 568 |
+
"learning_rate": 4.375484871993794e-05,
|
| 569 |
+
"loss": 2.8217,
|
| 570 |
+
"step": 43500
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 5.69,
|
| 574 |
+
"learning_rate": 4.310835272821309e-05,
|
| 575 |
+
"loss": 2.8197,
|
| 576 |
+
"step": 44000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.75,
|
| 580 |
+
"learning_rate": 4.2461856736488236e-05,
|
| 581 |
+
"loss": 2.8158,
|
| 582 |
+
"step": 44500
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 5.82,
|
| 586 |
+
"learning_rate": 4.1815360744763385e-05,
|
| 587 |
+
"loss": 2.8094,
|
| 588 |
+
"step": 45000
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 5.88,
|
| 592 |
+
"learning_rate": 4.1168864753038534e-05,
|
| 593 |
+
"loss": 2.8061,
|
| 594 |
+
"step": 45500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 5.95,
|
| 598 |
+
"learning_rate": 4.052236876131368e-05,
|
| 599 |
+
"loss": 2.8032,
|
| 600 |
+
"step": 46000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 6.0,
|
| 604 |
+
"eval_loss": 2.681367874145508,
|
| 605 |
+
"eval_runtime": 476.3627,
|
| 606 |
+
"eval_samples_per_second": 692.71,
|
| 607 |
+
"eval_steps_per_second": 1.805,
|
| 608 |
+
"step": 46404
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 6.01,
|
| 612 |
+
"learning_rate": 3.987587276958883e-05,
|
| 613 |
+
"loss": 2.794,
|
| 614 |
+
"step": 46500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 6.08,
|
| 618 |
+
"learning_rate": 3.922937677786398e-05,
|
| 619 |
+
"loss": 2.7858,
|
| 620 |
+
"step": 47000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 6.14,
|
| 624 |
+
"learning_rate": 3.858288078613912e-05,
|
| 625 |
+
"loss": 2.7765,
|
| 626 |
+
"step": 47500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 6.21,
|
| 630 |
+
"learning_rate": 3.793638479441428e-05,
|
| 631 |
+
"loss": 2.7752,
|
| 632 |
+
"step": 48000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 6.27,
|
| 636 |
+
"learning_rate": 3.728988880268943e-05,
|
| 637 |
+
"loss": 2.7725,
|
| 638 |
+
"step": 48500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 6.34,
|
| 642 |
+
"learning_rate": 3.664339281096457e-05,
|
| 643 |
+
"loss": 2.7696,
|
| 644 |
+
"step": 49000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 6.4,
|
| 648 |
+
"learning_rate": 3.5996896819239726e-05,
|
| 649 |
+
"loss": 2.7665,
|
| 650 |
+
"step": 49500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 6.46,
|
| 654 |
+
"learning_rate": 3.5350400827514875e-05,
|
| 655 |
+
"loss": 2.7622,
|
| 656 |
+
"step": 50000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 6.53,
|
| 660 |
+
"learning_rate": 3.470390483579002e-05,
|
| 661 |
+
"loss": 2.7566,
|
| 662 |
+
"step": 50500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 6.59,
|
| 666 |
+
"learning_rate": 3.4057408844065166e-05,
|
| 667 |
+
"loss": 2.7543,
|
| 668 |
+
"step": 51000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 6.66,
|
| 672 |
+
"learning_rate": 3.3410912852340315e-05,
|
| 673 |
+
"loss": 2.7451,
|
| 674 |
+
"step": 51500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.72,
|
| 678 |
+
"learning_rate": 3.2764416860615464e-05,
|
| 679 |
+
"loss": 2.7392,
|
| 680 |
+
"step": 52000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 6.79,
|
| 684 |
+
"learning_rate": 3.211792086889061e-05,
|
| 685 |
+
"loss": 2.7384,
|
| 686 |
+
"step": 52500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 6.85,
|
| 690 |
+
"learning_rate": 3.147142487716576e-05,
|
| 691 |
+
"loss": 2.7341,
|
| 692 |
+
"step": 53000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.92,
|
| 696 |
+
"learning_rate": 3.082492888544091e-05,
|
| 697 |
+
"loss": 2.7317,
|
| 698 |
+
"step": 53500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 6.98,
|
| 702 |
+
"learning_rate": 3.0178432893716057e-05,
|
| 703 |
+
"loss": 2.7308,
|
| 704 |
+
"step": 54000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 7.0,
|
| 708 |
+
"eval_loss": 2.6155083179473877,
|
| 709 |
+
"eval_runtime": 476.4568,
|
| 710 |
+
"eval_samples_per_second": 692.573,
|
| 711 |
+
"eval_steps_per_second": 1.805,
|
| 712 |
+
"step": 54138
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 7.05,
|
| 716 |
+
"learning_rate": 2.953193690199121e-05,
|
| 717 |
+
"loss": 2.7178,
|
| 718 |
+
"step": 54500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 7.11,
|
| 722 |
+
"learning_rate": 2.8885440910266358e-05,
|
| 723 |
+
"loss": 2.712,
|
| 724 |
+
"step": 55000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 7.18,
|
| 728 |
+
"learning_rate": 2.8238944918541504e-05,
|
| 729 |
+
"loss": 2.7152,
|
| 730 |
+
"step": 55500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 7.24,
|
| 734 |
+
"learning_rate": 2.7592448926816656e-05,
|
| 735 |
+
"loss": 2.7076,
|
| 736 |
+
"step": 56000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 7.31,
|
| 740 |
+
"learning_rate": 2.6945952935091805e-05,
|
| 741 |
+
"loss": 2.7048,
|
| 742 |
+
"step": 56500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 7.37,
|
| 746 |
+
"learning_rate": 2.629945694336695e-05,
|
| 747 |
+
"loss": 2.7038,
|
| 748 |
+
"step": 57000
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 7.43,
|
| 752 |
+
"learning_rate": 2.5652960951642103e-05,
|
| 753 |
+
"loss": 2.6975,
|
| 754 |
+
"step": 57500
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 7.5,
|
| 758 |
+
"learning_rate": 2.500646495991725e-05,
|
| 759 |
+
"loss": 2.696,
|
| 760 |
+
"step": 58000
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 7.56,
|
| 764 |
+
"learning_rate": 2.4359968968192397e-05,
|
| 765 |
+
"loss": 2.6941,
|
| 766 |
+
"step": 58500
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 7.63,
|
| 770 |
+
"learning_rate": 2.3713472976467546e-05,
|
| 771 |
+
"loss": 2.6884,
|
| 772 |
+
"step": 59000
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 7.69,
|
| 776 |
+
"learning_rate": 2.3066976984742695e-05,
|
| 777 |
+
"loss": 2.6844,
|
| 778 |
+
"step": 59500
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 7.76,
|
| 782 |
+
"learning_rate": 2.2420480993017844e-05,
|
| 783 |
+
"loss": 2.6791,
|
| 784 |
+
"step": 60000
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 7.82,
|
| 788 |
+
"learning_rate": 2.1773985001292993e-05,
|
| 789 |
+
"loss": 2.6759,
|
| 790 |
+
"step": 60500
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 7.89,
|
| 794 |
+
"learning_rate": 2.1127489009568142e-05,
|
| 795 |
+
"loss": 2.6779,
|
| 796 |
+
"step": 61000
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 7.95,
|
| 800 |
+
"learning_rate": 2.0480993017843288e-05,
|
| 801 |
+
"loss": 2.6749,
|
| 802 |
+
"step": 61500
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 8.0,
|
| 806 |
+
"eval_loss": 2.561079502105713,
|
| 807 |
+
"eval_runtime": 476.3937,
|
| 808 |
+
"eval_samples_per_second": 692.665,
|
| 809 |
+
"eval_steps_per_second": 1.805,
|
| 810 |
+
"step": 61872
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 8.02,
|
| 814 |
+
"learning_rate": 1.983449702611844e-05,
|
| 815 |
+
"loss": 2.6705,
|
| 816 |
+
"step": 62000
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 8.08,
|
| 820 |
+
"learning_rate": 1.918800103439359e-05,
|
| 821 |
+
"loss": 2.6613,
|
| 822 |
+
"step": 62500
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 8.15,
|
| 826 |
+
"learning_rate": 1.8541505042668735e-05,
|
| 827 |
+
"loss": 2.6649,
|
| 828 |
+
"step": 63000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 8.21,
|
| 832 |
+
"learning_rate": 1.7895009050943884e-05,
|
| 833 |
+
"loss": 2.6597,
|
| 834 |
+
"step": 63500
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 8.28,
|
| 838 |
+
"learning_rate": 1.7248513059219036e-05,
|
| 839 |
+
"loss": 2.6568,
|
| 840 |
+
"step": 64000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 8.34,
|
| 844 |
+
"learning_rate": 1.660201706749418e-05,
|
| 845 |
+
"loss": 2.6551,
|
| 846 |
+
"step": 64500
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 8.4,
|
| 850 |
+
"learning_rate": 1.595552107576933e-05,
|
| 851 |
+
"loss": 2.6518,
|
| 852 |
+
"step": 65000
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 8.47,
|
| 856 |
+
"learning_rate": 1.530902508404448e-05,
|
| 857 |
+
"loss": 2.6501,
|
| 858 |
+
"step": 65500
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 8.53,
|
| 862 |
+
"learning_rate": 1.466252909231963e-05,
|
| 863 |
+
"loss": 2.6469,
|
| 864 |
+
"step": 66000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 8.6,
|
| 868 |
+
"learning_rate": 1.4016033100594778e-05,
|
| 869 |
+
"loss": 2.6465,
|
| 870 |
+
"step": 66500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 8.66,
|
| 874 |
+
"learning_rate": 1.3369537108869925e-05,
|
| 875 |
+
"loss": 2.6455,
|
| 876 |
+
"step": 67000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 8.73,
|
| 880 |
+
"learning_rate": 1.2723041117145074e-05,
|
| 881 |
+
"loss": 2.6398,
|
| 882 |
+
"step": 67500
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 8.79,
|
| 886 |
+
"learning_rate": 1.2076545125420223e-05,
|
| 887 |
+
"loss": 2.6436,
|
| 888 |
+
"step": 68000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 8.86,
|
| 892 |
+
"learning_rate": 1.1430049133695372e-05,
|
| 893 |
+
"loss": 2.6395,
|
| 894 |
+
"step": 68500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 8.92,
|
| 898 |
+
"learning_rate": 1.078355314197052e-05,
|
| 899 |
+
"loss": 2.634,
|
| 900 |
+
"step": 69000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 8.99,
|
| 904 |
+
"learning_rate": 1.0137057150245668e-05,
|
| 905 |
+
"loss": 2.6359,
|
| 906 |
+
"step": 69500
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 9.0,
|
| 910 |
+
"eval_loss": 2.5289456844329834,
|
| 911 |
+
"eval_runtime": 476.5662,
|
| 912 |
+
"eval_samples_per_second": 692.414,
|
| 913 |
+
"eval_steps_per_second": 1.805,
|
| 914 |
+
"step": 69606
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 9.05,
|
| 918 |
+
"learning_rate": 9.490561158520819e-06,
|
| 919 |
+
"loss": 2.6261,
|
| 920 |
+
"step": 70000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 9.12,
|
| 924 |
+
"learning_rate": 8.844065166795966e-06,
|
| 925 |
+
"loss": 2.6228,
|
| 926 |
+
"step": 70500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 9.18,
|
| 930 |
+
"learning_rate": 8.197569175071115e-06,
|
| 931 |
+
"loss": 2.6303,
|
| 932 |
+
"step": 71000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 9.24,
|
| 936 |
+
"learning_rate": 7.551073183346263e-06,
|
| 937 |
+
"loss": 2.622,
|
| 938 |
+
"step": 71500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 9.31,
|
| 942 |
+
"learning_rate": 6.904577191621413e-06,
|
| 943 |
+
"loss": 2.6257,
|
| 944 |
+
"step": 72000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 9.37,
|
| 948 |
+
"learning_rate": 6.258081199896561e-06,
|
| 949 |
+
"loss": 2.6244,
|
| 950 |
+
"step": 72500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 9.44,
|
| 954 |
+
"learning_rate": 5.611585208171709e-06,
|
| 955 |
+
"loss": 2.6238,
|
| 956 |
+
"step": 73000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 9.5,
|
| 960 |
+
"learning_rate": 4.965089216446858e-06,
|
| 961 |
+
"loss": 2.6207,
|
| 962 |
+
"step": 73500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 9.57,
|
| 966 |
+
"learning_rate": 4.318593224722007e-06,
|
| 967 |
+
"loss": 2.6168,
|
| 968 |
+
"step": 74000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 9.63,
|
| 972 |
+
"learning_rate": 3.6720972329971556e-06,
|
| 973 |
+
"loss": 2.619,
|
| 974 |
+
"step": 74500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 9.7,
|
| 978 |
+
"learning_rate": 3.025601241272304e-06,
|
| 979 |
+
"loss": 2.6185,
|
| 980 |
+
"step": 75000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 9.76,
|
| 984 |
+
"learning_rate": 2.379105249547453e-06,
|
| 985 |
+
"loss": 2.6173,
|
| 986 |
+
"step": 75500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 9.83,
|
| 990 |
+
"learning_rate": 1.7326092578226015e-06,
|
| 991 |
+
"loss": 2.6114,
|
| 992 |
+
"step": 76000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 9.89,
|
| 996 |
+
"learning_rate": 1.0861132660977503e-06,
|
| 997 |
+
"loss": 2.6091,
|
| 998 |
+
"step": 76500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 9.96,
|
| 1002 |
+
"learning_rate": 4.396172743728989e-07,
|
| 1003 |
+
"loss": 2.6133,
|
| 1004 |
+
"step": 77000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 10.0,
|
| 1008 |
+
"eval_loss": 2.5088417530059814,
|
| 1009 |
+
"eval_runtime": 484.8466,
|
| 1010 |
+
"eval_samples_per_second": 680.588,
|
| 1011 |
+
"eval_steps_per_second": 1.774,
|
| 1012 |
+
"step": 77340
|
| 1013 |
+
}
|
| 1014 |
+
],
|
| 1015 |
+
"max_steps": 77340,
|
| 1016 |
+
"num_train_epochs": 10,
|
| 1017 |
+
"total_flos": 1.9542406291981978e+18,
|
| 1018 |
+
"trial_name": null,
|
| 1019 |
+
"trial_params": null
|
| 1020 |
+
}
|