Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1057 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1057 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.5,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1455,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01718213058419244,
|
| 13 |
+
"grad_norm": 4.919781684875488,
|
| 14 |
+
"learning_rate": 2.7397260273972603e-05,
|
| 15 |
+
"loss": 0.9081,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03436426116838488,
|
| 20 |
+
"grad_norm": 5.393226623535156,
|
| 21 |
+
"learning_rate": 5.479452054794521e-05,
|
| 22 |
+
"loss": 0.4511,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05154639175257732,
|
| 27 |
+
"grad_norm": 2.135725259780884,
|
| 28 |
+
"learning_rate": 8.219178082191781e-05,
|
| 29 |
+
"loss": 0.289,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06872852233676977,
|
| 34 |
+
"grad_norm": 2.005138397216797,
|
| 35 |
+
"learning_rate": 0.00010958904109589041,
|
| 36 |
+
"loss": 0.2236,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0859106529209622,
|
| 41 |
+
"grad_norm": 2.274747371673584,
|
| 42 |
+
"learning_rate": 0.000136986301369863,
|
| 43 |
+
"loss": 0.2057,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10309278350515463,
|
| 48 |
+
"grad_norm": 2.1100614070892334,
|
| 49 |
+
"learning_rate": 0.00016438356164383562,
|
| 50 |
+
"loss": 0.1831,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12027491408934708,
|
| 55 |
+
"grad_norm": 0.9432219862937927,
|
| 56 |
+
"learning_rate": 0.0001917808219178082,
|
| 57 |
+
"loss": 0.1452,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13745704467353953,
|
| 62 |
+
"grad_norm": 1.4754092693328857,
|
| 63 |
+
"learning_rate": 0.00019998733979961563,
|
| 64 |
+
"loss": 0.1496,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.15463917525773196,
|
| 69 |
+
"grad_norm": 1.504097580909729,
|
| 70 |
+
"learning_rate": 0.0001999253383717226,
|
| 71 |
+
"loss": 0.1542,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1718213058419244,
|
| 76 |
+
"grad_norm": 1.2584880590438843,
|
| 77 |
+
"learning_rate": 0.00019981170237143067,
|
| 78 |
+
"loss": 0.1341,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.18900343642611683,
|
| 83 |
+
"grad_norm": 1.030260443687439,
|
| 84 |
+
"learning_rate": 0.00019964649051804355,
|
| 85 |
+
"loss": 0.1153,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.20618556701030927,
|
| 90 |
+
"grad_norm": 0.6816794276237488,
|
| 91 |
+
"learning_rate": 0.000199429788181734,
|
| 92 |
+
"loss": 0.1458,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.22336769759450173,
|
| 97 |
+
"grad_norm": 1.3301407098770142,
|
| 98 |
+
"learning_rate": 0.0001991617073394306,
|
| 99 |
+
"loss": 0.0968,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24054982817869416,
|
| 104 |
+
"grad_norm": 1.0439300537109375,
|
| 105 |
+
"learning_rate": 0.00019884238651695556,
|
| 106 |
+
"loss": 0.1089,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.25773195876288657,
|
| 111 |
+
"grad_norm": 1.2270495891571045,
|
| 112 |
+
"learning_rate": 0.00019847199071744415,
|
| 113 |
+
"loss": 0.1115,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.27491408934707906,
|
| 118 |
+
"grad_norm": 0.5990088582038879,
|
| 119 |
+
"learning_rate": 0.00019805071133608242,
|
| 120 |
+
"loss": 0.101,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2920962199312715,
|
| 125 |
+
"grad_norm": 0.9263975024223328,
|
| 126 |
+
"learning_rate": 0.0001975787660612072,
|
| 127 |
+
"loss": 0.0987,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.30927835051546393,
|
| 132 |
+
"grad_norm": 0.6488857865333557,
|
| 133 |
+
"learning_rate": 0.00019705639876181969,
|
| 134 |
+
"loss": 0.0761,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.32646048109965636,
|
| 139 |
+
"grad_norm": 1.0950772762298584,
|
| 140 |
+
"learning_rate": 0.00019648387936157068,
|
| 141 |
+
"loss": 0.1128,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3436426116838488,
|
| 146 |
+
"grad_norm": 1.5179388523101807,
|
| 147 |
+
"learning_rate": 0.00019586150369928245,
|
| 148 |
+
"loss": 0.1069,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.36082474226804123,
|
| 153 |
+
"grad_norm": 0.7192593216896057,
|
| 154 |
+
"learning_rate": 0.00019518959337607957,
|
| 155 |
+
"loss": 0.0902,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.37800687285223367,
|
| 160 |
+
"grad_norm": 0.8463940620422363,
|
| 161 |
+
"learning_rate": 0.0001944684955892075,
|
| 162 |
+
"loss": 0.0883,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3951890034364261,
|
| 167 |
+
"grad_norm": 0.8777188658714294,
|
| 168 |
+
"learning_rate": 0.0001936985829526247,
|
| 169 |
+
"loss": 0.0783,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.41237113402061853,
|
| 174 |
+
"grad_norm": 0.8422673940658569,
|
| 175 |
+
"learning_rate": 0.00019288025330446126,
|
| 176 |
+
"loss": 0.0842,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.42955326460481097,
|
| 181 |
+
"grad_norm": 0.7569606900215149,
|
| 182 |
+
"learning_rate": 0.00019201392950144363,
|
| 183 |
+
"loss": 0.0765,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.44673539518900346,
|
| 188 |
+
"grad_norm": 0.7297337055206299,
|
| 189 |
+
"learning_rate": 0.0001911000592003909,
|
| 190 |
+
"loss": 0.0763,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4639175257731959,
|
| 195 |
+
"grad_norm": 0.570925772190094,
|
| 196 |
+
"learning_rate": 0.00019013911462689668,
|
| 197 |
+
"loss": 0.0687,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.48109965635738833,
|
| 202 |
+
"grad_norm": 0.7235879302024841,
|
| 203 |
+
"learning_rate": 0.000189131592331315,
|
| 204 |
+
"loss": 0.0884,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.49828178694158076,
|
| 209 |
+
"grad_norm": 0.3659440875053406,
|
| 210 |
+
"learning_rate": 0.00018807801293217735,
|
| 211 |
+
"loss": 0.0613,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5154639175257731,
|
| 216 |
+
"grad_norm": 0.5860876441001892,
|
| 217 |
+
"learning_rate": 0.00018697892084717238,
|
| 218 |
+
"loss": 0.0739,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5326460481099656,
|
| 223 |
+
"grad_norm": 0.7031033039093018,
|
| 224 |
+
"learning_rate": 0.00018583488401182843,
|
| 225 |
+
"loss": 0.0637,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5498281786941581,
|
| 230 |
+
"grad_norm": 0.6644700765609741,
|
| 231 |
+
"learning_rate": 0.0001846464935860431,
|
| 232 |
+
"loss": 0.0682,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5670103092783505,
|
| 237 |
+
"grad_norm": 0.7531244158744812,
|
| 238 |
+
"learning_rate": 0.0001834143636486124,
|
| 239 |
+
"loss": 0.0723,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.584192439862543,
|
| 244 |
+
"grad_norm": 0.7708925008773804,
|
| 245 |
+
"learning_rate": 0.00018213913087991685,
|
| 246 |
+
"loss": 0.0619,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6013745704467354,
|
| 251 |
+
"grad_norm": 0.716632604598999,
|
| 252 |
+
"learning_rate": 0.00018082145423292868,
|
| 253 |
+
"loss": 0.0613,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6185567010309279,
|
| 258 |
+
"grad_norm": 0.4382629096508026,
|
| 259 |
+
"learning_rate": 0.0001794620145927101,
|
| 260 |
+
"loss": 0.0614,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6357388316151202,
|
| 265 |
+
"grad_norm": 0.6322966814041138,
|
| 266 |
+
"learning_rate": 0.00017806151442457827,
|
| 267 |
+
"loss": 0.0581,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6529209621993127,
|
| 272 |
+
"grad_norm": 0.5152290463447571,
|
| 273 |
+
"learning_rate": 0.00017662067741111974,
|
| 274 |
+
"loss": 0.07,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6701030927835051,
|
| 279 |
+
"grad_norm": 0.6691680550575256,
|
| 280 |
+
"learning_rate": 0.00017514024807824055,
|
| 281 |
+
"loss": 0.0839,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6872852233676976,
|
| 286 |
+
"grad_norm": 0.5262071490287781,
|
| 287 |
+
"learning_rate": 0.00017362099141044626,
|
| 288 |
+
"loss": 0.052,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7044673539518901,
|
| 293 |
+
"grad_norm": 0.6715118288993835,
|
| 294 |
+
"learning_rate": 0.00017206369245555036,
|
| 295 |
+
"loss": 0.0682,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7216494845360825,
|
| 300 |
+
"grad_norm": 0.6439153552055359,
|
| 301 |
+
"learning_rate": 0.0001704691559190155,
|
| 302 |
+
"loss": 0.0733,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.738831615120275,
|
| 307 |
+
"grad_norm": 0.37341853976249695,
|
| 308 |
+
"learning_rate": 0.0001688382057481364,
|
| 309 |
+
"loss": 0.0653,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7560137457044673,
|
| 314 |
+
"grad_norm": 1.1194217205047607,
|
| 315 |
+
"learning_rate": 0.00016717168470628077,
|
| 316 |
+
"loss": 0.0735,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7731958762886598,
|
| 321 |
+
"grad_norm": 0.6904529333114624,
|
| 322 |
+
"learning_rate": 0.0001654704539374066,
|
| 323 |
+
"loss": 0.0652,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7903780068728522,
|
| 328 |
+
"grad_norm": 0.8207924962043762,
|
| 329 |
+
"learning_rate": 0.00016373539252108202,
|
| 330 |
+
"loss": 0.0479,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8075601374570447,
|
| 335 |
+
"grad_norm": 0.5141025185585022,
|
| 336 |
+
"learning_rate": 0.00016196739701823716,
|
| 337 |
+
"loss": 0.0626,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8247422680412371,
|
| 342 |
+
"grad_norm": 0.46161457896232605,
|
| 343 |
+
"learning_rate": 0.00016016738100788297,
|
| 344 |
+
"loss": 0.0469,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8419243986254296,
|
| 349 |
+
"grad_norm": 0.3261314034461975,
|
| 350 |
+
"learning_rate": 0.00015833627461503595,
|
| 351 |
+
"loss": 0.0489,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8591065292096219,
|
| 356 |
+
"grad_norm": 0.564329981803894,
|
| 357 |
+
"learning_rate": 0.0001564750240300934,
|
| 358 |
+
"loss": 0.0445,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8762886597938144,
|
| 363 |
+
"grad_norm": 0.5079665184020996,
|
| 364 |
+
"learning_rate": 0.00015458459101990693,
|
| 365 |
+
"loss": 0.0557,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8934707903780069,
|
| 370 |
+
"grad_norm": 0.41103053092956543,
|
| 371 |
+
"learning_rate": 0.00015266595243080714,
|
| 372 |
+
"loss": 0.042,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9106529209621993,
|
| 377 |
+
"grad_norm": 0.5678098201751709,
|
| 378 |
+
"learning_rate": 0.00015072009968383656,
|
| 379 |
+
"loss": 0.0526,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9278350515463918,
|
| 384 |
+
"grad_norm": 0.6878070831298828,
|
| 385 |
+
"learning_rate": 0.00014874803826245089,
|
| 386 |
+
"loss": 0.0605,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9450171821305842,
|
| 391 |
+
"grad_norm": 0.586892306804657,
|
| 392 |
+
"learning_rate": 0.00014675078719295415,
|
| 393 |
+
"loss": 0.0494,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9621993127147767,
|
| 398 |
+
"grad_norm": 0.4587668776512146,
|
| 399 |
+
"learning_rate": 0.00014472937851793557,
|
| 400 |
+
"loss": 0.0515,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.979381443298969,
|
| 405 |
+
"grad_norm": 0.5171328186988831,
|
| 406 |
+
"learning_rate": 0.00014268485676298078,
|
| 407 |
+
"loss": 0.0576,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.9965635738831615,
|
| 412 |
+
"grad_norm": 0.6215698719024658,
|
| 413 |
+
"learning_rate": 0.0001406182783969324,
|
| 414 |
+
"loss": 0.0537,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.013745704467354,
|
| 419 |
+
"grad_norm": 0.40490490198135376,
|
| 420 |
+
"learning_rate": 0.00013853071128597924,
|
| 421 |
+
"loss": 0.0467,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0309278350515463,
|
| 426 |
+
"grad_norm": 0.7031866312026978,
|
| 427 |
+
"learning_rate": 0.0001364232341418564,
|
| 428 |
+
"loss": 0.049,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.0481099656357389,
|
| 433 |
+
"grad_norm": 0.3372012972831726,
|
| 434 |
+
"learning_rate": 0.00013429693596444067,
|
| 435 |
+
"loss": 0.0598,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.0652920962199313,
|
| 440 |
+
"grad_norm": 0.4208463728427887,
|
| 441 |
+
"learning_rate": 0.00013215291547903006,
|
| 442 |
+
"loss": 0.0549,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.0824742268041236,
|
| 447 |
+
"grad_norm": 0.5574418902397156,
|
| 448 |
+
"learning_rate": 0.00012999228056859784,
|
| 449 |
+
"loss": 0.0469,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.0996563573883162,
|
| 454 |
+
"grad_norm": 0.5417460203170776,
|
| 455 |
+
"learning_rate": 0.00012781614770131442,
|
| 456 |
+
"loss": 0.0397,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.1168384879725086,
|
| 461 |
+
"grad_norm": 0.38392218947410583,
|
| 462 |
+
"learning_rate": 0.00012562564135363313,
|
| 463 |
+
"loss": 0.0409,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.134020618556701,
|
| 468 |
+
"grad_norm": 0.3199153244495392,
|
| 469 |
+
"learning_rate": 0.0001234218934292376,
|
| 470 |
+
"loss": 0.0402,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.1512027491408934,
|
| 475 |
+
"grad_norm": 0.33709654211997986,
|
| 476 |
+
"learning_rate": 0.00012120604267415172,
|
| 477 |
+
"loss": 0.0338,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.168384879725086,
|
| 482 |
+
"grad_norm": 0.685562789440155,
|
| 483 |
+
"learning_rate": 0.00011897923408831346,
|
| 484 |
+
"loss": 0.0514,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.1855670103092784,
|
| 489 |
+
"grad_norm": 0.6378128528594971,
|
| 490 |
+
"learning_rate": 0.0001167426183339174,
|
| 491 |
+
"loss": 0.0385,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.2027491408934707,
|
| 496 |
+
"grad_norm": 0.39844685792922974,
|
| 497 |
+
"learning_rate": 0.00011449735114083127,
|
| 498 |
+
"loss": 0.0355,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.2199312714776633,
|
| 503 |
+
"grad_norm": 0.5079712271690369,
|
| 504 |
+
"learning_rate": 0.00011224459270939384,
|
| 505 |
+
"loss": 0.0433,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.2371134020618557,
|
| 510 |
+
"grad_norm": 0.3737178444862366,
|
| 511 |
+
"learning_rate": 0.000109985507110903,
|
| 512 |
+
"loss": 0.0418,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.254295532646048,
|
| 517 |
+
"grad_norm": 0.5172721147537231,
|
| 518 |
+
"learning_rate": 0.00010772126168610325,
|
| 519 |
+
"loss": 0.0484,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.2714776632302405,
|
| 524 |
+
"grad_norm": 0.3860677480697632,
|
| 525 |
+
"learning_rate": 0.00010545302644198405,
|
| 526 |
+
"loss": 0.0421,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.2886597938144329,
|
| 531 |
+
"grad_norm": 0.37599268555641174,
|
| 532 |
+
"learning_rate": 0.00010318197344720018,
|
| 533 |
+
"loss": 0.0428,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.3058419243986255,
|
| 538 |
+
"grad_norm": 0.26397764682769775,
|
| 539 |
+
"learning_rate": 0.0001009092762264271,
|
| 540 |
+
"loss": 0.0402,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.3230240549828178,
|
| 545 |
+
"grad_norm": 0.35941213369369507,
|
| 546 |
+
"learning_rate": 9.863610915396365e-05,
|
| 547 |
+
"loss": 0.0404,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.3402061855670104,
|
| 552 |
+
"grad_norm": 0.47325387597084045,
|
| 553 |
+
"learning_rate": 9.63636468468959e-05,
|
| 554 |
+
"loss": 0.0446,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.3573883161512028,
|
| 559 |
+
"grad_norm": 0.4716935455799103,
|
| 560 |
+
"learning_rate": 9.409306355813529e-05,
|
| 561 |
+
"loss": 0.0342,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.3745704467353952,
|
| 566 |
+
"grad_norm": 0.2812814712524414,
|
| 567 |
+
"learning_rate": 9.18255325696454e-05,
|
| 568 |
+
"loss": 0.0343,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.3917525773195876,
|
| 573 |
+
"grad_norm": 0.29519158601760864,
|
| 574 |
+
"learning_rate": 8.956222558616998e-05,
|
| 575 |
+
"loss": 0.0372,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.40893470790378,
|
| 580 |
+
"grad_norm": 0.41123560070991516,
|
| 581 |
+
"learning_rate": 8.730431212977625e-05,
|
| 582 |
+
"loss": 0.0567,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.4261168384879725,
|
| 587 |
+
"grad_norm": 0.3533785343170166,
|
| 588 |
+
"learning_rate": 8.505295893552594e-05,
|
| 589 |
+
"loss": 0.0385,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.443298969072165,
|
| 594 |
+
"grad_norm": 0.4819997251033783,
|
| 595 |
+
"learning_rate": 8.280932934858652e-05,
|
| 596 |
+
"loss": 0.0392,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.4604810996563573,
|
| 601 |
+
"grad_norm": 0.285748690366745,
|
| 602 |
+
"learning_rate": 8.05745827230941e-05,
|
| 603 |
+
"loss": 0.0362,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.47766323024055,
|
| 608 |
+
"grad_norm": 0.39609745144844055,
|
| 609 |
+
"learning_rate": 7.834987382307861e-05,
|
| 610 |
+
"loss": 0.0432,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.4948453608247423,
|
| 615 |
+
"grad_norm": 0.33522850275039673,
|
| 616 |
+
"learning_rate": 7.613635222576072e-05,
|
| 617 |
+
"loss": 0.04,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.5120274914089347,
|
| 622 |
+
"grad_norm": 0.47676974534988403,
|
| 623 |
+
"learning_rate": 7.393516172752919e-05,
|
| 624 |
+
"loss": 0.0347,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.529209621993127,
|
| 629 |
+
"grad_norm": 0.5136341452598572,
|
| 630 |
+
"learning_rate": 7.174743975290513e-05,
|
| 631 |
+
"loss": 0.0374,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.5463917525773194,
|
| 636 |
+
"grad_norm": 0.361741840839386,
|
| 637 |
+
"learning_rate": 6.957431676679896e-05,
|
| 638 |
+
"loss": 0.0358,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.563573883161512,
|
| 643 |
+
"grad_norm": 0.49855756759643555,
|
| 644 |
+
"learning_rate": 6.741691569036338e-05,
|
| 645 |
+
"loss": 0.0352,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.5807560137457046,
|
| 650 |
+
"grad_norm": 0.26860520243644714,
|
| 651 |
+
"learning_rate": 6.527635132074493e-05,
|
| 652 |
+
"loss": 0.0397,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.597938144329897,
|
| 657 |
+
"grad_norm": 0.4026907682418823,
|
| 658 |
+
"learning_rate": 6.315372975503285e-05,
|
| 659 |
+
"loss": 0.0495,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.6151202749140894,
|
| 664 |
+
"grad_norm": 0.377718448638916,
|
| 665 |
+
"learning_rate": 6.1050147818704e-05,
|
| 666 |
+
"loss": 0.0328,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.6323024054982818,
|
| 671 |
+
"grad_norm": 0.41204559803009033,
|
| 672 |
+
"learning_rate": 5.896669249885851e-05,
|
| 673 |
+
"loss": 0.0336,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.6494845360824741,
|
| 678 |
+
"grad_norm": 0.41056016087532043,
|
| 679 |
+
"learning_rate": 5.690444038253935e-05,
|
| 680 |
+
"loss": 0.0406,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.6666666666666665,
|
| 685 |
+
"grad_norm": 0.4092923402786255,
|
| 686 |
+
"learning_rate": 5.4864457100425783e-05,
|
| 687 |
+
"loss": 0.033,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.6838487972508591,
|
| 692 |
+
"grad_norm": 0.35451173782348633,
|
| 693 |
+
"learning_rate": 5.284779677618841e-05,
|
| 694 |
+
"loss": 0.0351,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.7010309278350515,
|
| 699 |
+
"grad_norm": 0.5240267515182495,
|
| 700 |
+
"learning_rate": 5.0855501481790305e-05,
|
| 701 |
+
"loss": 0.0326,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.718213058419244,
|
| 706 |
+
"grad_norm": 0.6117306351661682,
|
| 707 |
+
"learning_rate": 4.8888600699015496e-05,
|
| 708 |
+
"loss": 0.0327,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.7353951890034365,
|
| 713 |
+
"grad_norm": 0.27679693698883057,
|
| 714 |
+
"learning_rate": 4.694811078750338e-05,
|
| 715 |
+
"loss": 0.0421,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.7525773195876289,
|
| 720 |
+
"grad_norm": 0.44769006967544556,
|
| 721 |
+
"learning_rate": 4.50350344595635e-05,
|
| 722 |
+
"loss": 0.0313,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.7697594501718212,
|
| 727 |
+
"grad_norm": 0.35016706585884094,
|
| 728 |
+
"learning_rate": 4.315036026204262e-05,
|
| 729 |
+
"loss": 0.0388,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.7869415807560136,
|
| 734 |
+
"grad_norm": 0.2692789137363434,
|
| 735 |
+
"learning_rate": 4.129506206551138e-05,
|
| 736 |
+
"loss": 0.0295,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.8041237113402062,
|
| 741 |
+
"grad_norm": 0.45466527342796326,
|
| 742 |
+
"learning_rate": 3.947009856103465e-05,
|
| 743 |
+
"loss": 0.0367,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.8213058419243986,
|
| 748 |
+
"grad_norm": 0.2106892317533493,
|
| 749 |
+
"learning_rate": 3.767641276478563e-05,
|
| 750 |
+
"loss": 0.0369,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.8384879725085912,
|
| 755 |
+
"grad_norm": 0.2619366943836212,
|
| 756 |
+
"learning_rate": 3.591493153075966e-05,
|
| 757 |
+
"loss": 0.0278,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.8556701030927836,
|
| 762 |
+
"grad_norm": 0.35229456424713135,
|
| 763 |
+
"learning_rate": 3.41865650718396e-05,
|
| 764 |
+
"loss": 0.034,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.872852233676976,
|
| 769 |
+
"grad_norm": 0.325093537569046,
|
| 770 |
+
"learning_rate": 3.24922064894601e-05,
|
| 771 |
+
"loss": 0.034,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.8900343642611683,
|
| 776 |
+
"grad_norm": 0.13483747839927673,
|
| 777 |
+
"learning_rate": 3.083273131211382e-05,
|
| 778 |
+
"loss": 0.0254,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.9072164948453607,
|
| 783 |
+
"grad_norm": 0.37040096521377563,
|
| 784 |
+
"learning_rate": 2.920899704293849e-05,
|
| 785 |
+
"loss": 0.027,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.9243986254295533,
|
| 790 |
+
"grad_norm": 0.2273675948381424,
|
| 791 |
+
"learning_rate": 2.762184271661785e-05,
|
| 792 |
+
"loss": 0.0265,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.9415807560137457,
|
| 797 |
+
"grad_norm": 0.12444432824850082,
|
| 798 |
+
"learning_rate": 2.6072088465826038e-05,
|
| 799 |
+
"loss": 0.0248,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.9587628865979383,
|
| 804 |
+
"grad_norm": 0.3245929479598999,
|
| 805 |
+
"learning_rate": 2.4560535097439108e-05,
|
| 806 |
+
"loss": 0.0296,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.9759450171821307,
|
| 811 |
+
"grad_norm": 0.28723689913749695,
|
| 812 |
+
"learning_rate": 2.308796367873296e-05,
|
| 813 |
+
"loss": 0.03,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.993127147766323,
|
| 818 |
+
"grad_norm": 0.2640382647514343,
|
| 819 |
+
"learning_rate": 2.165513513378121e-05,
|
| 820 |
+
"loss": 0.0237,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.0103092783505154,
|
| 825 |
+
"grad_norm": 0.2659141421318054,
|
| 826 |
+
"learning_rate": 2.0262789850261798e-05,
|
| 827 |
+
"loss": 0.034,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.027491408934708,
|
| 832 |
+
"grad_norm": 0.41219818592071533,
|
| 833 |
+
"learning_rate": 1.8911647296875147e-05,
|
| 834 |
+
"loss": 0.0282,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.0446735395189,
|
| 839 |
+
"grad_norm": 0.23828770220279694,
|
| 840 |
+
"learning_rate": 1.7602405651572275e-05,
|
| 841 |
+
"loss": 0.0288,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.0618556701030926,
|
| 846 |
+
"grad_norm": 0.18391673266887665,
|
| 847 |
+
"learning_rate": 1.6335741440784035e-05,
|
| 848 |
+
"loss": 0.0314,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.0790378006872854,
|
| 853 |
+
"grad_norm": 0.27285298705101013,
|
| 854 |
+
"learning_rate": 1.511230918983867e-05,
|
| 855 |
+
"loss": 0.0367,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.0962199312714778,
|
| 860 |
+
"grad_norm": 0.3976253271102905,
|
| 861 |
+
"learning_rate": 1.3932741084747913e-05,
|
| 862 |
+
"loss": 0.0322,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.11340206185567,
|
| 867 |
+
"grad_norm": 0.17294034361839294,
|
| 868 |
+
"learning_rate": 1.2797646645536566e-05,
|
| 869 |
+
"loss": 0.0221,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.1305841924398625,
|
| 874 |
+
"grad_norm": 0.39969944953918457,
|
| 875 |
+
"learning_rate": 1.1707612411284253e-05,
|
| 876 |
+
"loss": 0.0318,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.147766323024055,
|
| 881 |
+
"grad_norm": 0.2953219711780548,
|
| 882 |
+
"learning_rate": 1.0663201637042252e-05,
|
| 883 |
+
"loss": 0.0267,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.1649484536082473,
|
| 888 |
+
"grad_norm": 0.24471993744373322,
|
| 889 |
+
"learning_rate": 9.664954002781745e-06,
|
| 890 |
+
"loss": 0.0219,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.1821305841924397,
|
| 895 |
+
"grad_norm": 0.26958975195884705,
|
| 896 |
+
"learning_rate": 8.713385334524283e-06,
|
| 897 |
+
"loss": 0.0227,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.1993127147766325,
|
| 902 |
+
"grad_norm": 0.3202515244483948,
|
| 903 |
+
"learning_rate": 7.808987337798158e-06,
|
| 904 |
+
"loss": 0.0258,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.216494845360825,
|
| 909 |
+
"grad_norm": 0.23253372311592102,
|
| 910 |
+
"learning_rate": 6.952227343558671e-06,
|
| 911 |
+
"loss": 0.0254,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.2336769759450172,
|
| 916 |
+
"grad_norm": 0.2511325776576996,
|
| 917 |
+
"learning_rate": 6.143548066703475e-06,
|
| 918 |
+
"loss": 0.0282,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.2508591065292096,
|
| 923 |
+
"grad_norm": 0.21489648520946503,
|
| 924 |
+
"learning_rate": 5.383367377307857e-06,
|
| 925 |
+
"loss": 0.02,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.268041237113402,
|
| 930 |
+
"grad_norm": 0.3450476825237274,
|
| 931 |
+
"learning_rate": 4.672078084698095e-06,
|
| 932 |
+
"loss": 0.0279,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.2852233676975944,
|
| 937 |
+
"grad_norm": 0.36600273847579956,
|
| 938 |
+
"learning_rate": 4.010047734474454e-06,
|
| 939 |
+
"loss": 0.034,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.3024054982817868,
|
| 944 |
+
"grad_norm": 0.48743385076522827,
|
| 945 |
+
"learning_rate": 3.397618418588877e-06,
|
| 946 |
+
"loss": 0.0384,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.319587628865979,
|
| 951 |
+
"grad_norm": 0.30427098274230957,
|
| 952 |
+
"learning_rate": 2.8351065985751766e-06,
|
| 953 |
+
"loss": 0.0226,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.336769759450172,
|
| 958 |
+
"grad_norm": 0.36907103657722473,
|
| 959 |
+
"learning_rate": 2.322802942023461e-06,
|
| 960 |
+
"loss": 0.0334,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.3539518900343643,
|
| 965 |
+
"grad_norm": 0.184714213013649,
|
| 966 |
+
"learning_rate": 1.8609721723830132e-06,
|
| 967 |
+
"loss": 0.0223,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.3711340206185567,
|
| 972 |
+
"grad_norm": 0.43203970789909363,
|
| 973 |
+
"learning_rate": 1.4498529321713584e-06,
|
| 974 |
+
"loss": 0.0273,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.388316151202749,
|
| 979 |
+
"grad_norm": 0.26931440830230713,
|
| 980 |
+
"learning_rate": 1.0896576596600705e-06,
|
| 981 |
+
"loss": 0.0265,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.4054982817869415,
|
| 986 |
+
"grad_norm": 0.19834183156490326,
|
| 987 |
+
"learning_rate": 7.80572479101327e-07,
|
| 988 |
+
"loss": 0.0253,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.422680412371134,
|
| 993 |
+
"grad_norm": 0.22906753420829773,
|
| 994 |
+
"learning_rate": 5.227571045515633e-07,
|
| 995 |
+
"loss": 0.0331,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.4398625429553267,
|
| 1000 |
+
"grad_norm": 0.307211697101593,
|
| 1001 |
+
"learning_rate": 3.163447573422351e-07,
|
| 1002 |
+
"loss": 0.03,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.457044673539519,
|
| 1007 |
+
"grad_norm": 0.31240877509117126,
|
| 1008 |
+
"learning_rate": 1.614420972401165e-07,
|
| 1009 |
+
"loss": 0.0307,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.4742268041237114,
|
| 1014 |
+
"grad_norm": 0.28886109590530396,
|
| 1015 |
+
"learning_rate": 5.812916733284324e-08,
|
| 1016 |
+
"loss": 0.0214,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.491408934707904,
|
| 1021 |
+
"grad_norm": 0.29304051399230957,
|
| 1022 |
+
"learning_rate": 6.459352668164442e-09,
|
| 1023 |
+
"loss": 0.0285,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.5,
|
| 1028 |
+
"step": 1455,
|
| 1029 |
+
"total_flos": 4.98510260177783e+16,
|
| 1030 |
+
"train_loss": 0.0646941511594143,
|
| 1031 |
+
"train_runtime": 645.0485,
|
| 1032 |
+
"train_samples_per_second": 36.09,
|
| 1033 |
+
"train_steps_per_second": 2.256
|
| 1034 |
+
}
|
| 1035 |
+
],
|
| 1036 |
+
"logging_steps": 10,
|
| 1037 |
+
"max_steps": 1455,
|
| 1038 |
+
"num_input_tokens_seen": 0,
|
| 1039 |
+
"num_train_epochs": 3,
|
| 1040 |
+
"save_steps": 500,
|
| 1041 |
+
"stateful_callbacks": {
|
| 1042 |
+
"TrainerControl": {
|
| 1043 |
+
"args": {
|
| 1044 |
+
"should_epoch_stop": false,
|
| 1045 |
+
"should_evaluate": false,
|
| 1046 |
+
"should_log": false,
|
| 1047 |
+
"should_save": true,
|
| 1048 |
+
"should_training_stop": true
|
| 1049 |
+
},
|
| 1050 |
+
"attributes": {}
|
| 1051 |
+
}
|
| 1052 |
+
},
|
| 1053 |
+
"total_flos": 4.98510260177783e+16,
|
| 1054 |
+
"train_batch_size": 16,
|
| 1055 |
+
"trial_name": null,
|
| 1056 |
+
"trial_params": null
|
| 1057 |
+
}
|