Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1099 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1099 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.996699669966997,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1514,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.033003300330033,
|
| 13 |
+
"grad_norm": 6.397974014282227,
|
| 14 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 15 |
+
"loss": 1.174,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.066006600660066,
|
| 20 |
+
"grad_norm": 6.199707984924316,
|
| 21 |
+
"learning_rate": 5.2631578947368424e-05,
|
| 22 |
+
"loss": 0.4999,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09900990099009901,
|
| 27 |
+
"grad_norm": 4.184757709503174,
|
| 28 |
+
"learning_rate": 7.894736842105263e-05,
|
| 29 |
+
"loss": 0.4234,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.132013201320132,
|
| 34 |
+
"grad_norm": 2.2750799655914307,
|
| 35 |
+
"learning_rate": 0.00010526315789473685,
|
| 36 |
+
"loss": 0.3396,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16501650165016502,
|
| 41 |
+
"grad_norm": 1.9745649099349976,
|
| 42 |
+
"learning_rate": 0.00013157894736842108,
|
| 43 |
+
"loss": 0.2269,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19801980198019803,
|
| 48 |
+
"grad_norm": 5.893553733825684,
|
| 49 |
+
"learning_rate": 0.00015789473684210527,
|
| 50 |
+
"loss": 0.2303,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.23102310231023102,
|
| 55 |
+
"grad_norm": 1.9844739437103271,
|
| 56 |
+
"learning_rate": 0.00018421052631578948,
|
| 57 |
+
"loss": 0.1964,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.264026402640264,
|
| 62 |
+
"grad_norm": 1.8762494325637817,
|
| 63 |
+
"learning_rate": 0.00019999618170748185,
|
| 64 |
+
"loss": 0.1872,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.297029702970297,
|
| 69 |
+
"grad_norm": 1.8214250802993774,
|
| 70 |
+
"learning_rate": 0.0001999532292652782,
|
| 71 |
+
"loss": 0.1714,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.33003300330033003,
|
| 76 |
+
"grad_norm": 1.117044448852539,
|
| 77 |
+
"learning_rate": 0.00019986257208350297,
|
| 78 |
+
"loss": 0.1655,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.36303630363036304,
|
| 83 |
+
"grad_norm": 0.7621971964836121,
|
| 84 |
+
"learning_rate": 0.0001997242534301653,
|
| 85 |
+
"loss": 0.1597,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.39603960396039606,
|
| 90 |
+
"grad_norm": 1.5328593254089355,
|
| 91 |
+
"learning_rate": 0.0001995383393206945,
|
| 92 |
+
"loss": 0.1619,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.429042904290429,
|
| 97 |
+
"grad_norm": 0.9974318146705627,
|
| 98 |
+
"learning_rate": 0.0001993049184864326,
|
| 99 |
+
"loss": 0.1283,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.46204620462046203,
|
| 104 |
+
"grad_norm": 1.477512240409851,
|
| 105 |
+
"learning_rate": 0.0001990241023322856,
|
| 106 |
+
"loss": 0.1243,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.49504950495049505,
|
| 111 |
+
"grad_norm": 1.3623563051223755,
|
| 112 |
+
"learning_rate": 0.00019869602488355311,
|
| 113 |
+
"loss": 0.1286,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.528052805280528,
|
| 118 |
+
"grad_norm": 1.1071594953536987,
|
| 119 |
+
"learning_rate": 0.00019832084272196205,
|
| 120 |
+
"loss": 0.1143,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5610561056105611,
|
| 125 |
+
"grad_norm": 1.2614643573760986,
|
| 126 |
+
"learning_rate": 0.00019789873491093488,
|
| 127 |
+
"loss": 0.1135,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.594059405940594,
|
| 132 |
+
"grad_norm": 0.8214598298072815,
|
| 133 |
+
"learning_rate": 0.00019742990291012763,
|
| 134 |
+
"loss": 0.0976,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6270627062706271,
|
| 139 |
+
"grad_norm": 0.6513738036155701,
|
| 140 |
+
"learning_rate": 0.00019691457047927942,
|
| 141 |
+
"loss": 0.0996,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6600660066006601,
|
| 146 |
+
"grad_norm": 1.2949241399765015,
|
| 147 |
+
"learning_rate": 0.00019635298357141836,
|
| 148 |
+
"loss": 0.1249,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.693069306930693,
|
| 153 |
+
"grad_norm": 1.1320366859436035,
|
| 154 |
+
"learning_rate": 0.00019574541021547543,
|
| 155 |
+
"loss": 0.1023,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7260726072607261,
|
| 160 |
+
"grad_norm": 1.2242770195007324,
|
| 161 |
+
"learning_rate": 0.0001950921403883621,
|
| 162 |
+
"loss": 0.1021,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.759075907590759,
|
| 167 |
+
"grad_norm": 0.6682835817337036,
|
| 168 |
+
"learning_rate": 0.00019439348587657292,
|
| 169 |
+
"loss": 0.0939,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7920792079207921,
|
| 174 |
+
"grad_norm": 0.6505287885665894,
|
| 175 |
+
"learning_rate": 0.000193649780127379,
|
| 176 |
+
"loss": 0.0907,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8250825082508251,
|
| 181 |
+
"grad_norm": 1.006945013999939,
|
| 182 |
+
"learning_rate": 0.00019286137808968343,
|
| 183 |
+
"loss": 0.0857,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.858085808580858,
|
| 188 |
+
"grad_norm": 0.905126690864563,
|
| 189 |
+
"learning_rate": 0.00019202865604461442,
|
| 190 |
+
"loss": 0.0916,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8910891089108911,
|
| 195 |
+
"grad_norm": 0.9964673519134521,
|
| 196 |
+
"learning_rate": 0.00019115201142593766,
|
| 197 |
+
"loss": 0.1034,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9240924092409241,
|
| 202 |
+
"grad_norm": 0.8413060307502747,
|
| 203 |
+
"learning_rate": 0.00019023186263037272,
|
| 204 |
+
"loss": 0.1056,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9570957095709571,
|
| 209 |
+
"grad_norm": 0.8259803652763367,
|
| 210 |
+
"learning_rate": 0.00018926864881790446,
|
| 211 |
+
"loss": 0.0856,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9900990099009901,
|
| 216 |
+
"grad_norm": 0.9525623321533203,
|
| 217 |
+
"learning_rate": 0.00018826282970218518,
|
| 218 |
+
"loss": 0.098,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.023102310231023,
|
| 223 |
+
"grad_norm": 0.793368399143219,
|
| 224 |
+
"learning_rate": 0.0001872148853311265,
|
| 225 |
+
"loss": 0.0905,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.056105610561056,
|
| 230 |
+
"grad_norm": 0.6828791499137878,
|
| 231 |
+
"learning_rate": 0.0001861253158577867,
|
| 232 |
+
"loss": 0.0874,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0891089108910892,
|
| 237 |
+
"grad_norm": 0.9090235829353333,
|
| 238 |
+
"learning_rate": 0.00018499464130166247,
|
| 239 |
+
"loss": 0.0889,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.1221122112211221,
|
| 244 |
+
"grad_norm": 1.1500462293624878,
|
| 245 |
+
"learning_rate": 0.00018382340130049878,
|
| 246 |
+
"loss": 0.0678,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.155115511551155,
|
| 251 |
+
"grad_norm": 0.46532729268074036,
|
| 252 |
+
"learning_rate": 0.00018261215485273558,
|
| 253 |
+
"loss": 0.0867,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.188118811881188,
|
| 258 |
+
"grad_norm": 0.48733001947402954,
|
| 259 |
+
"learning_rate": 0.00018136148005071438,
|
| 260 |
+
"loss": 0.0815,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.221122112211221,
|
| 265 |
+
"grad_norm": 0.531043291091919,
|
| 266 |
+
"learning_rate": 0.00018007197380477173,
|
| 267 |
+
"loss": 0.0713,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.2541254125412542,
|
| 272 |
+
"grad_norm": 0.5529260635375977,
|
| 273 |
+
"learning_rate": 0.00017874425155835146,
|
| 274 |
+
"loss": 0.0644,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2871287128712872,
|
| 279 |
+
"grad_norm": 1.0257015228271484,
|
| 280 |
+
"learning_rate": 0.00017737894699427153,
|
| 281 |
+
"loss": 0.0809,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.3201320132013201,
|
| 286 |
+
"grad_norm": 0.5263957381248474,
|
| 287 |
+
"learning_rate": 0.00017597671173228595,
|
| 288 |
+
"loss": 0.0645,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.353135313531353,
|
| 293 |
+
"grad_norm": 0.3812405467033386,
|
| 294 |
+
"learning_rate": 0.00017453821501808582,
|
| 295 |
+
"loss": 0.0611,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.386138613861386,
|
| 300 |
+
"grad_norm": 0.8025540113449097,
|
| 301 |
+
"learning_rate": 0.0001730641434038878,
|
| 302 |
+
"loss": 0.0766,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.4191419141914192,
|
| 307 |
+
"grad_norm": 0.7956365346908569,
|
| 308 |
+
"learning_rate": 0.00017155520042076328,
|
| 309 |
+
"loss": 0.0801,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4521452145214522,
|
| 314 |
+
"grad_norm": 0.7980107665061951,
|
| 315 |
+
"learning_rate": 0.0001700121062428634,
|
| 316 |
+
"loss": 0.0765,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4851485148514851,
|
| 321 |
+
"grad_norm": 0.863133430480957,
|
| 322 |
+
"learning_rate": 0.00016843559734370107,
|
| 323 |
+
"loss": 0.072,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.5181518151815183,
|
| 328 |
+
"grad_norm": 0.7694889903068542,
|
| 329 |
+
"learning_rate": 0.00016682642614465398,
|
| 330 |
+
"loss": 0.0627,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.551155115511551,
|
| 335 |
+
"grad_norm": 0.8058327436447144,
|
| 336 |
+
"learning_rate": 0.0001651853606558559,
|
| 337 |
+
"loss": 0.0734,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.5841584158415842,
|
| 342 |
+
"grad_norm": 0.7808170914649963,
|
| 343 |
+
"learning_rate": 0.00016351318410964802,
|
| 344 |
+
"loss": 0.0683,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.6171617161716172,
|
| 349 |
+
"grad_norm": 0.8645703196525574,
|
| 350 |
+
"learning_rate": 0.00016181069458676505,
|
| 351 |
+
"loss": 0.065,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6501650165016502,
|
| 356 |
+
"grad_norm": 0.6377092003822327,
|
| 357 |
+
"learning_rate": 0.00016007870463543493,
|
| 358 |
+
"loss": 0.0561,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.6831683168316833,
|
| 363 |
+
"grad_norm": 0.8465257883071899,
|
| 364 |
+
"learning_rate": 0.00015831804088357326,
|
| 365 |
+
"loss": 0.0642,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.716171617161716,
|
| 370 |
+
"grad_norm": 0.33361324667930603,
|
| 371 |
+
"learning_rate": 0.00015652954364425822,
|
| 372 |
+
"loss": 0.0508,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7491749174917492,
|
| 377 |
+
"grad_norm": 0.740112841129303,
|
| 378 |
+
"learning_rate": 0.00015471406651467376,
|
| 379 |
+
"loss": 0.0628,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.7821782178217822,
|
| 384 |
+
"grad_norm": 0.5477822422981262,
|
| 385 |
+
"learning_rate": 0.00015287247596871274,
|
| 386 |
+
"loss": 0.0696,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.8151815181518152,
|
| 391 |
+
"grad_norm": 0.48374637961387634,
|
| 392 |
+
"learning_rate": 0.00015100565094343436,
|
| 393 |
+
"loss": 0.0601,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8481848184818483,
|
| 398 |
+
"grad_norm": 0.6817354559898376,
|
| 399 |
+
"learning_rate": 0.00014911448241957347,
|
| 400 |
+
"loss": 0.0684,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.881188118811881,
|
| 405 |
+
"grad_norm": 0.6513104438781738,
|
| 406 |
+
"learning_rate": 0.00014719987299630148,
|
| 407 |
+
"loss": 0.0712,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.9141914191419143,
|
| 412 |
+
"grad_norm": 0.6665838360786438,
|
| 413 |
+
"learning_rate": 0.00014526273646044239,
|
| 414 |
+
"loss": 0.0691,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.9471947194719472,
|
| 419 |
+
"grad_norm": 0.5764182209968567,
|
| 420 |
+
"learning_rate": 0.00014330399735034915,
|
| 421 |
+
"loss": 0.0531,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.9801980198019802,
|
| 426 |
+
"grad_norm": 0.5813863277435303,
|
| 427 |
+
"learning_rate": 0.00014132459051464873,
|
| 428 |
+
"loss": 0.058,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.0132013201320134,
|
| 433 |
+
"grad_norm": 0.4850797653198242,
|
| 434 |
+
"learning_rate": 0.0001393254606660662,
|
| 435 |
+
"loss": 0.0639,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.046204620462046,
|
| 440 |
+
"grad_norm": 0.59760582447052,
|
| 441 |
+
"learning_rate": 0.00013730756193054127,
|
| 442 |
+
"loss": 0.0636,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.0792079207920793,
|
| 447 |
+
"grad_norm": 0.4797247350215912,
|
| 448 |
+
"learning_rate": 0.00013527185739185187,
|
| 449 |
+
"loss": 0.0546,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.112211221122112,
|
| 454 |
+
"grad_norm": 0.5719553232192993,
|
| 455 |
+
"learning_rate": 0.00013321931863196268,
|
| 456 |
+
"loss": 0.0557,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.145214521452145,
|
| 461 |
+
"grad_norm": 0.5088202357292175,
|
| 462 |
+
"learning_rate": 0.00013115092526731748,
|
| 463 |
+
"loss": 0.0514,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.1782178217821784,
|
| 468 |
+
"grad_norm": 0.5406337380409241,
|
| 469 |
+
"learning_rate": 0.00012906766448129718,
|
| 470 |
+
"loss": 0.0518,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.211221122112211,
|
| 475 |
+
"grad_norm": 0.5986407399177551,
|
| 476 |
+
"learning_rate": 0.00012697053055306596,
|
| 477 |
+
"loss": 0.0443,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.2442244224422443,
|
| 482 |
+
"grad_norm": 0.5725324153900146,
|
| 483 |
+
"learning_rate": 0.00012486052438303124,
|
| 484 |
+
"loss": 0.0581,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.2772277227722775,
|
| 489 |
+
"grad_norm": 0.5339573621749878,
|
| 490 |
+
"learning_rate": 0.0001227386530151432,
|
| 491 |
+
"loss": 0.043,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.31023102310231,
|
| 496 |
+
"grad_norm": 0.37319841980934143,
|
| 497 |
+
"learning_rate": 0.00012060592915626252,
|
| 498 |
+
"loss": 0.0415,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.3432343234323434,
|
| 503 |
+
"grad_norm": 0.41099220514297485,
|
| 504 |
+
"learning_rate": 0.00011846337069282463,
|
| 505 |
+
"loss": 0.0513,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.376237623762376,
|
| 510 |
+
"grad_norm": 0.42775505781173706,
|
| 511 |
+
"learning_rate": 0.00011631200020503307,
|
| 512 |
+
"loss": 0.0461,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.4092409240924093,
|
| 517 |
+
"grad_norm": 0.37573903799057007,
|
| 518 |
+
"learning_rate": 0.00011415284447881159,
|
| 519 |
+
"loss": 0.0539,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.442244224422442,
|
| 524 |
+
"grad_norm": 0.2992411255836487,
|
| 525 |
+
"learning_rate": 0.00011198693401574945,
|
| 526 |
+
"loss": 0.0547,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.4752475247524752,
|
| 531 |
+
"grad_norm": 0.47858911752700806,
|
| 532 |
+
"learning_rate": 0.00010981530254127341,
|
| 533 |
+
"loss": 0.047,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.5082508250825084,
|
| 538 |
+
"grad_norm": 0.4435286521911621,
|
| 539 |
+
"learning_rate": 0.00010763898651128084,
|
| 540 |
+
"loss": 0.0456,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.541254125412541,
|
| 545 |
+
"grad_norm": 0.6310757994651794,
|
| 546 |
+
"learning_rate": 0.00010545902461746984,
|
| 547 |
+
"loss": 0.0462,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.5742574257425743,
|
| 552 |
+
"grad_norm": 0.4391770362854004,
|
| 553 |
+
"learning_rate": 0.00010327645729160223,
|
| 554 |
+
"loss": 0.0524,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.6072607260726075,
|
| 559 |
+
"grad_norm": 0.6589362621307373,
|
| 560 |
+
"learning_rate": 0.00010109232620893592,
|
| 561 |
+
"loss": 0.0374,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.6402640264026402,
|
| 566 |
+
"grad_norm": 0.7139827609062195,
|
| 567 |
+
"learning_rate": 9.89076737910641e-05,
|
| 568 |
+
"loss": 0.0449,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.6732673267326734,
|
| 573 |
+
"grad_norm": 0.4805195927619934,
|
| 574 |
+
"learning_rate": 9.672354270839778e-05,
|
| 575 |
+
"loss": 0.0424,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.706270627062706,
|
| 580 |
+
"grad_norm": 0.5566986203193665,
|
| 581 |
+
"learning_rate": 9.45409753825302e-05,
|
| 582 |
+
"loss": 0.0506,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.7392739273927393,
|
| 587 |
+
"grad_norm": 0.47570672631263733,
|
| 588 |
+
"learning_rate": 9.236101348871921e-05,
|
| 589 |
+
"loss": 0.0521,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.772277227722772,
|
| 594 |
+
"grad_norm": 0.5024830102920532,
|
| 595 |
+
"learning_rate": 9.018469745872661e-05,
|
| 596 |
+
"loss": 0.0372,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.8052805280528053,
|
| 601 |
+
"grad_norm": 0.30473190546035767,
|
| 602 |
+
"learning_rate": 8.801306598425059e-05,
|
| 603 |
+
"loss": 0.0407,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.8382838283828384,
|
| 608 |
+
"grad_norm": 0.4293649196624756,
|
| 609 |
+
"learning_rate": 8.584715552118842e-05,
|
| 610 |
+
"loss": 0.0487,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.871287128712871,
|
| 615 |
+
"grad_norm": 0.3333033621311188,
|
| 616 |
+
"learning_rate": 8.368799979496694e-05,
|
| 617 |
+
"loss": 0.0432,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.9042904290429044,
|
| 622 |
+
"grad_norm": 0.4249730706214905,
|
| 623 |
+
"learning_rate": 8.153662930717538e-05,
|
| 624 |
+
"loss": 0.0434,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.9372937293729375,
|
| 629 |
+
"grad_norm": 0.36155596375465393,
|
| 630 |
+
"learning_rate": 7.939407084373751e-05,
|
| 631 |
+
"loss": 0.0447,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.9702970297029703,
|
| 636 |
+
"grad_norm": 0.5141147971153259,
|
| 637 |
+
"learning_rate": 7.726134698485679e-05,
|
| 638 |
+
"loss": 0.0448,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.0033003300330035,
|
| 643 |
+
"grad_norm": 0.28401315212249756,
|
| 644 |
+
"learning_rate": 7.513947561696877e-05,
|
| 645 |
+
"loss": 0.0437,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.036303630363036,
|
| 650 |
+
"grad_norm": 0.5779618620872498,
|
| 651 |
+
"learning_rate": 7.302946944693408e-05,
|
| 652 |
+
"loss": 0.045,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.0693069306930694,
|
| 657 |
+
"grad_norm": 0.3839390277862549,
|
| 658 |
+
"learning_rate": 7.093233551870287e-05,
|
| 659 |
+
"loss": 0.0505,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.102310231023102,
|
| 664 |
+
"grad_norm": 0.28532204031944275,
|
| 665 |
+
"learning_rate": 6.884907473268254e-05,
|
| 666 |
+
"loss": 0.0299,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.1353135313531353,
|
| 671 |
+
"grad_norm": 0.32001256942749023,
|
| 672 |
+
"learning_rate": 6.678068136803736e-05,
|
| 673 |
+
"loss": 0.0375,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.1683168316831685,
|
| 678 |
+
"grad_norm": 0.6117786765098572,
|
| 679 |
+
"learning_rate": 6.472814260814814e-05,
|
| 680 |
+
"loss": 0.0446,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.201320132013201,
|
| 685 |
+
"grad_norm": 0.3738768398761749,
|
| 686 |
+
"learning_rate": 6.269243806945878e-05,
|
| 687 |
+
"loss": 0.0307,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.2343234323432344,
|
| 692 |
+
"grad_norm": 0.5009806752204895,
|
| 693 |
+
"learning_rate": 6.067453933393383e-05,
|
| 694 |
+
"loss": 0.0389,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.2673267326732676,
|
| 699 |
+
"grad_norm": 0.2887628972530365,
|
| 700 |
+
"learning_rate": 5.86754094853513e-05,
|
| 701 |
+
"loss": 0.0326,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.3003300330033003,
|
| 706 |
+
"grad_norm": 0.38890546560287476,
|
| 707 |
+
"learning_rate": 5.669600264965088e-05,
|
| 708 |
+
"loss": 0.0393,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.3333333333333335,
|
| 713 |
+
"grad_norm": 0.3082311153411865,
|
| 714 |
+
"learning_rate": 5.473726353955764e-05,
|
| 715 |
+
"loss": 0.0433,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.366336633663366,
|
| 720 |
+
"grad_norm": 0.3528409004211426,
|
| 721 |
+
"learning_rate": 5.280012700369856e-05,
|
| 722 |
+
"loss": 0.0338,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.3993399339933994,
|
| 727 |
+
"grad_norm": 0.33807092905044556,
|
| 728 |
+
"learning_rate": 5.088551758042655e-05,
|
| 729 |
+
"loss": 0.0295,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.432343234323432,
|
| 734 |
+
"grad_norm": 0.26223328709602356,
|
| 735 |
+
"learning_rate": 4.899434905656563e-05,
|
| 736 |
+
"loss": 0.0309,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.4653465346534653,
|
| 741 |
+
"grad_norm": 0.3232835829257965,
|
| 742 |
+
"learning_rate": 4.712752403128729e-05,
|
| 743 |
+
"loss": 0.0382,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.4983498349834985,
|
| 748 |
+
"grad_norm": 0.2711802124977112,
|
| 749 |
+
"learning_rate": 4.5285933485326226e-05,
|
| 750 |
+
"loss": 0.0396,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.5313531353135312,
|
| 755 |
+
"grad_norm": 0.2992013096809387,
|
| 756 |
+
"learning_rate": 4.3470456355741786e-05,
|
| 757 |
+
"loss": 0.0296,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.5643564356435644,
|
| 762 |
+
"grad_norm": 0.5572153925895691,
|
| 763 |
+
"learning_rate": 4.168195911642677e-05,
|
| 764 |
+
"loss": 0.0411,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.5973597359735976,
|
| 769 |
+
"grad_norm": 0.3480948507785797,
|
| 770 |
+
"learning_rate": 3.9921295364565094e-05,
|
| 771 |
+
"loss": 0.032,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.6303630363036303,
|
| 776 |
+
"grad_norm": 0.2131304293870926,
|
| 777 |
+
"learning_rate": 3.8189305413234975e-05,
|
| 778 |
+
"loss": 0.0294,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.6633663366336635,
|
| 783 |
+
"grad_norm": 0.21037542819976807,
|
| 784 |
+
"learning_rate": 3.6486815890352e-05,
|
| 785 |
+
"loss": 0.0309,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6963696369636962,
|
| 790 |
+
"grad_norm": 0.34865525364875793,
|
| 791 |
+
"learning_rate": 3.481463934414412e-05,
|
| 792 |
+
"loss": 0.0283,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.7293729372937294,
|
| 797 |
+
"grad_norm": 0.39114323258399963,
|
| 798 |
+
"learning_rate": 3.3173573855346043e-05,
|
| 799 |
+
"loss": 0.0327,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.762376237623762,
|
| 804 |
+
"grad_norm": 0.3465011715888977,
|
| 805 |
+
"learning_rate": 3.156440265629899e-05,
|
| 806 |
+
"loss": 0.0239,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.7953795379537953,
|
| 811 |
+
"grad_norm": 0.2945576608181,
|
| 812 |
+
"learning_rate": 2.998789375713663e-05,
|
| 813 |
+
"loss": 0.0404,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.8283828382838285,
|
| 818 |
+
"grad_norm": 0.6535521149635315,
|
| 819 |
+
"learning_rate": 2.8444799579236704e-05,
|
| 820 |
+
"loss": 0.0341,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.8613861386138613,
|
| 825 |
+
"grad_norm": 0.40984803438186646,
|
| 826 |
+
"learning_rate": 2.6935856596112197e-05,
|
| 827 |
+
"loss": 0.0341,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.8943894389438944,
|
| 832 |
+
"grad_norm": 0.3325531780719757,
|
| 833 |
+
"learning_rate": 2.5461784981914193e-05,
|
| 834 |
+
"loss": 0.03,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.9273927392739276,
|
| 839 |
+
"grad_norm": 0.43570443987846375,
|
| 840 |
+
"learning_rate": 2.4023288267714038e-05,
|
| 841 |
+
"loss": 0.0305,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.9603960396039604,
|
| 846 |
+
"grad_norm": 0.33430153131484985,
|
| 847 |
+
"learning_rate": 2.2621053005728498e-05,
|
| 848 |
+
"loss": 0.0354,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.9933993399339935,
|
| 853 |
+
"grad_norm": 0.35682499408721924,
|
| 854 |
+
"learning_rate": 2.125574844164856e-05,
|
| 855 |
+
"loss": 0.0382,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.026402640264027,
|
| 860 |
+
"grad_norm": 0.25153276324272156,
|
| 861 |
+
"learning_rate": 1.9928026195228288e-05,
|
| 862 |
+
"loss": 0.0303,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.0594059405940595,
|
| 867 |
+
"grad_norm": 0.3027781546115875,
|
| 868 |
+
"learning_rate": 1.8638519949285636e-05,
|
| 869 |
+
"loss": 0.0295,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.092409240924092,
|
| 874 |
+
"grad_norm": 0.3476581275463104,
|
| 875 |
+
"learning_rate": 1.7387845147264446e-05,
|
| 876 |
+
"loss": 0.0311,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.125412541254126,
|
| 881 |
+
"grad_norm": 0.28510499000549316,
|
| 882 |
+
"learning_rate": 1.6176598699501256e-05,
|
| 883 |
+
"loss": 0.0295,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.158415841584159,
|
| 888 |
+
"grad_norm": 0.4659450054168701,
|
| 889 |
+
"learning_rate": 1.5005358698337535e-05,
|
| 890 |
+
"loss": 0.0236,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.191419141914191,
|
| 895 |
+
"grad_norm": 0.2653041183948517,
|
| 896 |
+
"learning_rate": 1.3874684142213312e-05,
|
| 897 |
+
"loss": 0.0275,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.224422442244224,
|
| 902 |
+
"grad_norm": 0.34631630778312683,
|
| 903 |
+
"learning_rate": 1.278511466887351e-05,
|
| 904 |
+
"loss": 0.0317,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.257425742574258,
|
| 909 |
+
"grad_norm": 0.356421560049057,
|
| 910 |
+
"learning_rate": 1.1737170297814826e-05,
|
| 911 |
+
"loss": 0.031,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.29042904290429,
|
| 916 |
+
"grad_norm": 0.44884151220321655,
|
| 917 |
+
"learning_rate": 1.0731351182095539e-05,
|
| 918 |
+
"loss": 0.0279,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.323432343234323,
|
| 923 |
+
"grad_norm": 0.28433260321617126,
|
| 924 |
+
"learning_rate": 9.768137369627296e-06,
|
| 925 |
+
"loss": 0.0249,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.356435643564357,
|
| 930 |
+
"grad_norm": 0.29540589451789856,
|
| 931 |
+
"learning_rate": 8.84798857406235e-06,
|
| 932 |
+
"loss": 0.0294,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.3894389438943895,
|
| 937 |
+
"grad_norm": 0.21095190942287445,
|
| 938 |
+
"learning_rate": 7.97134395538559e-06,
|
| 939 |
+
"loss": 0.0314,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.422442244224422,
|
| 944 |
+
"grad_norm": 0.30297479033470154,
|
| 945 |
+
"learning_rate": 7.138621910316589e-06,
|
| 946 |
+
"loss": 0.0391,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.455445544554456,
|
| 951 |
+
"grad_norm": 0.28455203771591187,
|
| 952 |
+
"learning_rate": 6.3502198726209885e-06,
|
| 953 |
+
"loss": 0.0333,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.488448844884489,
|
| 958 |
+
"grad_norm": 0.3449389338493347,
|
| 959 |
+
"learning_rate": 5.606514123427076e-06,
|
| 960 |
+
"loss": 0.0274,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.521452145214521,
|
| 965 |
+
"grad_norm": 0.21530762314796448,
|
| 966 |
+
"learning_rate": 4.907859611637911e-06,
|
| 967 |
+
"loss": 0.0317,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.554455445544555,
|
| 972 |
+
"grad_norm": 0.24574680626392365,
|
| 973 |
+
"learning_rate": 4.254589784524576e-06,
|
| 974 |
+
"loss": 0.0272,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.587458745874588,
|
| 979 |
+
"grad_norm": 0.22646954655647278,
|
| 980 |
+
"learning_rate": 3.64701642858164e-06,
|
| 981 |
+
"loss": 0.0278,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.62046204620462,
|
| 986 |
+
"grad_norm": 0.22831633687019348,
|
| 987 |
+
"learning_rate": 3.085429520720595e-06,
|
| 988 |
+
"loss": 0.0333,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.653465346534653,
|
| 993 |
+
"grad_norm": 0.3778948485851288,
|
| 994 |
+
"learning_rate": 2.570097089872392e-06,
|
| 995 |
+
"loss": 0.0378,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.686468646864687,
|
| 1000 |
+
"grad_norm": 0.3244766294956207,
|
| 1001 |
+
"learning_rate": 2.101265089065152e-06,
|
| 1002 |
+
"loss": 0.0342,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.7194719471947195,
|
| 1007 |
+
"grad_norm": 0.2885381579399109,
|
| 1008 |
+
"learning_rate": 1.679157278037946e-06,
|
| 1009 |
+
"loss": 0.0339,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.752475247524752,
|
| 1014 |
+
"grad_norm": 0.2981841266155243,
|
| 1015 |
+
"learning_rate": 1.3039751164469095e-06,
|
| 1016 |
+
"loss": 0.0252,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.785478547854785,
|
| 1021 |
+
"grad_norm": 0.3433663249015808,
|
| 1022 |
+
"learning_rate": 9.758976677144206e-07,
|
| 1023 |
+
"loss": 0.032,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.818481848184819,
|
| 1028 |
+
"grad_norm": 0.37242624163627625,
|
| 1029 |
+
"learning_rate": 6.950815135674083e-07,
|
| 1030 |
+
"loss": 0.0257,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.851485148514851,
|
| 1035 |
+
"grad_norm": 0.18495431542396545,
|
| 1036 |
+
"learning_rate": 4.6166067930550895e-07,
|
| 1037 |
+
"loss": 0.0169,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.884488448844884,
|
| 1042 |
+
"grad_norm": 0.31304922699928284,
|
| 1043 |
+
"learning_rate": 2.757465698347117e-07,
|
| 1044 |
+
"loss": 0.026,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.917491749174918,
|
| 1049 |
+
"grad_norm": 0.29433882236480713,
|
| 1050 |
+
"learning_rate": 1.374279164970571e-07,
|
| 1051 |
+
"loss": 0.024,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.9504950495049505,
|
| 1056 |
+
"grad_norm": 0.20669183135032654,
|
| 1057 |
+
"learning_rate": 4.67707347218016e-08,
|
| 1058 |
+
"loss": 0.0332,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.983498349834983,
|
| 1063 |
+
"grad_norm": 0.456485778093338,
|
| 1064 |
+
"learning_rate": 3.818292518176225e-09,
|
| 1065 |
+
"loss": 0.0263,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.996699669966997,
|
| 1070 |
+
"step": 1514,
|
| 1071 |
+
"total_flos": 5.36485853572416e+16,
|
| 1072 |
+
"train_loss": 0.07512911051373973,
|
| 1073 |
+
"train_runtime": 690.98,
|
| 1074 |
+
"train_samples_per_second": 35.057,
|
| 1075 |
+
"train_steps_per_second": 2.191
|
| 1076 |
+
}
|
| 1077 |
+
],
|
| 1078 |
+
"logging_steps": 10,
|
| 1079 |
+
"max_steps": 1514,
|
| 1080 |
+
"num_input_tokens_seen": 0,
|
| 1081 |
+
"num_train_epochs": 5,
|
| 1082 |
+
"save_steps": 500,
|
| 1083 |
+
"stateful_callbacks": {
|
| 1084 |
+
"TrainerControl": {
|
| 1085 |
+
"args": {
|
| 1086 |
+
"should_epoch_stop": false,
|
| 1087 |
+
"should_evaluate": false,
|
| 1088 |
+
"should_log": false,
|
| 1089 |
+
"should_save": true,
|
| 1090 |
+
"should_training_stop": true
|
| 1091 |
+
},
|
| 1092 |
+
"attributes": {}
|
| 1093 |
+
}
|
| 1094 |
+
},
|
| 1095 |
+
"total_flos": 5.36485853572416e+16,
|
| 1096 |
+
"train_batch_size": 16,
|
| 1097 |
+
"trial_name": null,
|
| 1098 |
+
"trial_params": null
|
| 1099 |
+
}
|