Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1113 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1113 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.990259740259741,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1537,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.032467532467532464,
|
| 13 |
+
"grad_norm": 11.862669944763184,
|
| 14 |
+
"learning_rate": 2.5974025974025972e-05,
|
| 15 |
+
"loss": 2.2746,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06493506493506493,
|
| 20 |
+
"grad_norm": 14.773530006408691,
|
| 21 |
+
"learning_rate": 5.1948051948051944e-05,
|
| 22 |
+
"loss": 0.6306,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09740259740259741,
|
| 27 |
+
"grad_norm": 2.444714069366455,
|
| 28 |
+
"learning_rate": 7.792207792207793e-05,
|
| 29 |
+
"loss": 0.2859,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12987012987012986,
|
| 34 |
+
"grad_norm": 1.5860096216201782,
|
| 35 |
+
"learning_rate": 0.00010389610389610389,
|
| 36 |
+
"loss": 0.1714,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16233766233766234,
|
| 41 |
+
"grad_norm": 1.3271065950393677,
|
| 42 |
+
"learning_rate": 0.00012987012987012987,
|
| 43 |
+
"loss": 0.1621,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19480519480519481,
|
| 48 |
+
"grad_norm": 1.9922699928283691,
|
| 49 |
+
"learning_rate": 0.00015584415584415587,
|
| 50 |
+
"loss": 0.164,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.22727272727272727,
|
| 55 |
+
"grad_norm": 1.0023343563079834,
|
| 56 |
+
"learning_rate": 0.00018181818181818183,
|
| 57 |
+
"loss": 0.1308,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2597402597402597,
|
| 62 |
+
"grad_norm": 0.9265744686126709,
|
| 63 |
+
"learning_rate": 0.00019999791644466247,
|
| 64 |
+
"loss": 0.1217,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2922077922077922,
|
| 69 |
+
"grad_norm": 1.0755594968795776,
|
| 70 |
+
"learning_rate": 0.00019996087787618624,
|
| 71 |
+
"loss": 0.1126,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3246753246753247,
|
| 76 |
+
"grad_norm": 1.6246726512908936,
|
| 77 |
+
"learning_rate": 0.00019987755781707668,
|
| 78 |
+
"loss": 0.1066,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.35714285714285715,
|
| 83 |
+
"grad_norm": 1.3438688516616821,
|
| 84 |
+
"learning_rate": 0.00019974799484419106,
|
| 85 |
+
"loss": 0.0931,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.38961038961038963,
|
| 90 |
+
"grad_norm": 1.0927611589431763,
|
| 91 |
+
"learning_rate": 0.0001995722489446729,
|
| 92 |
+
"loss": 0.0955,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.42207792207792205,
|
| 97 |
+
"grad_norm": 0.5384888648986816,
|
| 98 |
+
"learning_rate": 0.0001993504014881777,
|
| 99 |
+
"loss": 0.0543,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.45454545454545453,
|
| 104 |
+
"grad_norm": 1.163609504699707,
|
| 105 |
+
"learning_rate": 0.0001990825551891994,
|
| 106 |
+
"loss": 0.0649,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.487012987012987,
|
| 111 |
+
"grad_norm": 1.2434951066970825,
|
| 112 |
+
"learning_rate": 0.00019876883405951377,
|
| 113 |
+
"loss": 0.0798,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5194805194805194,
|
| 118 |
+
"grad_norm": 0.865424633026123,
|
| 119 |
+
"learning_rate": 0.00019840938335076162,
|
| 120 |
+
"loss": 0.082,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.551948051948052,
|
| 125 |
+
"grad_norm": 0.8742974400520325,
|
| 126 |
+
"learning_rate": 0.00019800436948719775,
|
| 127 |
+
"loss": 0.0725,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5844155844155844,
|
| 132 |
+
"grad_norm": 0.5820501446723938,
|
| 133 |
+
"learning_rate": 0.0001975539799886372,
|
| 134 |
+
"loss": 0.0602,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6168831168831169,
|
| 139 |
+
"grad_norm": 0.6646164655685425,
|
| 140 |
+
"learning_rate": 0.00019705842338363434,
|
| 141 |
+
"loss": 0.0614,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6493506493506493,
|
| 146 |
+
"grad_norm": 1.1510143280029297,
|
| 147 |
+
"learning_rate": 0.00019651792911293508,
|
| 148 |
+
"loss": 0.066,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6818181818181818,
|
| 153 |
+
"grad_norm": 0.9399957060813904,
|
| 154 |
+
"learning_rate": 0.0001959327474232464,
|
| 155 |
+
"loss": 0.0627,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7142857142857143,
|
| 160 |
+
"grad_norm": 0.9072800874710083,
|
| 161 |
+
"learning_rate": 0.00019530314925137355,
|
| 162 |
+
"loss": 0.0666,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7467532467532467,
|
| 167 |
+
"grad_norm": 0.8320746421813965,
|
| 168 |
+
"learning_rate": 0.00019462942609877696,
|
| 169 |
+
"loss": 0.0518,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7792207792207793,
|
| 174 |
+
"grad_norm": 0.820354163646698,
|
| 175 |
+
"learning_rate": 0.00019391188989660848,
|
| 176 |
+
"loss": 0.0857,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8116883116883117,
|
| 181 |
+
"grad_norm": 0.9190356731414795,
|
| 182 |
+
"learning_rate": 0.0001931508728612885,
|
| 183 |
+
"loss": 0.0651,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8441558441558441,
|
| 188 |
+
"grad_norm": 0.6117440462112427,
|
| 189 |
+
"learning_rate": 0.0001923467273406908,
|
| 190 |
+
"loss": 0.0526,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8766233766233766,
|
| 195 |
+
"grad_norm": 0.8513985276222229,
|
| 196 |
+
"learning_rate": 0.00019149982565100728,
|
| 197 |
+
"loss": 0.0549,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9090909090909091,
|
| 202 |
+
"grad_norm": 0.5364561080932617,
|
| 203 |
+
"learning_rate": 0.00019061055990436665,
|
| 204 |
+
"loss": 0.0612,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9415584415584416,
|
| 209 |
+
"grad_norm": 0.8399089574813843,
|
| 210 |
+
"learning_rate": 0.00018967934182728837,
|
| 211 |
+
"loss": 0.0417,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.974025974025974,
|
| 216 |
+
"grad_norm": 0.7521604299545288,
|
| 217 |
+
"learning_rate": 0.00018870660257005466,
|
| 218 |
+
"loss": 0.0535,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.0064935064935066,
|
| 223 |
+
"grad_norm": 0.5698351263999939,
|
| 224 |
+
"learning_rate": 0.00018769279250708976,
|
| 225 |
+
"loss": 0.0489,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0389610389610389,
|
| 230 |
+
"grad_norm": 0.6740209460258484,
|
| 231 |
+
"learning_rate": 0.0001866383810284384,
|
| 232 |
+
"loss": 0.0569,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0714285714285714,
|
| 237 |
+
"grad_norm": 0.49634265899658203,
|
| 238 |
+
"learning_rate": 0.00018554385632244,
|
| 239 |
+
"loss": 0.0561,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.103896103896104,
|
| 244 |
+
"grad_norm": 0.6321184039115906,
|
| 245 |
+
"learning_rate": 0.00018440972514969926,
|
| 246 |
+
"loss": 0.0413,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.1363636363636362,
|
| 251 |
+
"grad_norm": 0.41812852025032043,
|
| 252 |
+
"learning_rate": 0.00018323651260845832,
|
| 253 |
+
"loss": 0.0411,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1688311688311688,
|
| 258 |
+
"grad_norm": 0.3897482454776764,
|
| 259 |
+
"learning_rate": 0.00018202476189147803,
|
| 260 |
+
"loss": 0.0505,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.2012987012987013,
|
| 265 |
+
"grad_norm": 0.6317518353462219,
|
| 266 |
+
"learning_rate": 0.00018077503403454216,
|
| 267 |
+
"loss": 0.0387,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.2337662337662338,
|
| 272 |
+
"grad_norm": 0.49315449595451355,
|
| 273 |
+
"learning_rate": 0.0001794879076566998,
|
| 274 |
+
"loss": 0.0458,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2662337662337662,
|
| 279 |
+
"grad_norm": 0.522301971912384,
|
| 280 |
+
"learning_rate": 0.00017816397869236717,
|
| 281 |
+
"loss": 0.0509,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2987012987012987,
|
| 286 |
+
"grad_norm": 0.6012665629386902,
|
| 287 |
+
"learning_rate": 0.00017680386011541222,
|
| 288 |
+
"loss": 0.0379,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.3311688311688312,
|
| 293 |
+
"grad_norm": 0.783263087272644,
|
| 294 |
+
"learning_rate": 0.0001754081816553504,
|
| 295 |
+
"loss": 0.047,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3636363636363638,
|
| 300 |
+
"grad_norm": 0.7606521248817444,
|
| 301 |
+
"learning_rate": 0.00017397758950578207,
|
| 302 |
+
"loss": 0.0733,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.396103896103896,
|
| 307 |
+
"grad_norm": 0.4706479012966156,
|
| 308 |
+
"learning_rate": 0.00017251274602520766,
|
| 309 |
+
"loss": 0.0556,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4285714285714286,
|
| 314 |
+
"grad_norm": 0.5615910887718201,
|
| 315 |
+
"learning_rate": 0.00017101432943035825,
|
| 316 |
+
"loss": 0.0491,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4610389610389611,
|
| 321 |
+
"grad_norm": 0.6203674674034119,
|
| 322 |
+
"learning_rate": 0.0001694830334821838,
|
| 323 |
+
"loss": 0.0414,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4935064935064934,
|
| 328 |
+
"grad_norm": 0.6230823397636414,
|
| 329 |
+
"learning_rate": 0.00016791956716464472,
|
| 330 |
+
"loss": 0.0484,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.525974025974026,
|
| 335 |
+
"grad_norm": 0.49440866708755493,
|
| 336 |
+
"learning_rate": 0.0001663246543564551,
|
| 337 |
+
"loss": 0.056,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.5584415584415585,
|
| 342 |
+
"grad_norm": 0.5235504508018494,
|
| 343 |
+
"learning_rate": 0.00016469903349592956,
|
| 344 |
+
"loss": 0.036,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5909090909090908,
|
| 349 |
+
"grad_norm": 0.7212300300598145,
|
| 350 |
+
"learning_rate": 0.0001630434572390895,
|
| 351 |
+
"loss": 0.0464,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6233766233766234,
|
| 356 |
+
"grad_norm": 0.45477214455604553,
|
| 357 |
+
"learning_rate": 0.00016135869211118603,
|
| 358 |
+
"loss": 0.04,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.655844155844156,
|
| 363 |
+
"grad_norm": 0.6098753809928894,
|
| 364 |
+
"learning_rate": 0.00015964551815180213,
|
| 365 |
+
"loss": 0.0396,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6883116883116882,
|
| 370 |
+
"grad_norm": 1.0758476257324219,
|
| 371 |
+
"learning_rate": 0.00015790472855369716,
|
| 372 |
+
"loss": 0.0567,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7207792207792207,
|
| 377 |
+
"grad_norm": 0.8209078311920166,
|
| 378 |
+
"learning_rate": 0.00015613712929556193,
|
| 379 |
+
"loss": 0.0506,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.7532467532467533,
|
| 384 |
+
"grad_norm": 0.7815719246864319,
|
| 385 |
+
"learning_rate": 0.00015434353876885362,
|
| 386 |
+
"loss": 0.0474,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7857142857142856,
|
| 391 |
+
"grad_norm": 0.3328838050365448,
|
| 392 |
+
"learning_rate": 0.00015252478739888385,
|
| 393 |
+
"loss": 0.0333,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8181818181818183,
|
| 398 |
+
"grad_norm": 0.6883902549743652,
|
| 399 |
+
"learning_rate": 0.0001506817172603351,
|
| 400 |
+
"loss": 0.0342,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.8506493506493507,
|
| 405 |
+
"grad_norm": 0.4302406311035156,
|
| 406 |
+
"learning_rate": 0.0001488151816873834,
|
| 407 |
+
"loss": 0.0415,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.883116883116883,
|
| 412 |
+
"grad_norm": 0.5647683143615723,
|
| 413 |
+
"learning_rate": 0.00014692604487860785,
|
| 414 |
+
"loss": 0.0492,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.9155844155844157,
|
| 419 |
+
"grad_norm": 0.46459880471229553,
|
| 420 |
+
"learning_rate": 0.00014501518149687042,
|
| 421 |
+
"loss": 0.035,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.948051948051948,
|
| 426 |
+
"grad_norm": 0.49570703506469727,
|
| 427 |
+
"learning_rate": 0.0001430834762643502,
|
| 428 |
+
"loss": 0.0402,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9805194805194806,
|
| 433 |
+
"grad_norm": 0.44045427441596985,
|
| 434 |
+
"learning_rate": 0.00014113182355292078,
|
| 435 |
+
"loss": 0.045,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.012987012987013,
|
| 440 |
+
"grad_norm": 0.22453299164772034,
|
| 441 |
+
"learning_rate": 0.0001391611269700594,
|
| 442 |
+
"loss": 0.0419,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.0454545454545454,
|
| 447 |
+
"grad_norm": 0.4719075560569763,
|
| 448 |
+
"learning_rate": 0.00013717229894048038,
|
| 449 |
+
"loss": 0.037,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.0779220779220777,
|
| 454 |
+
"grad_norm": 0.4710325598716736,
|
| 455 |
+
"learning_rate": 0.0001351662602836861,
|
| 456 |
+
"loss": 0.0368,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.1103896103896105,
|
| 461 |
+
"grad_norm": 0.5058510303497314,
|
| 462 |
+
"learning_rate": 0.0001331439397876312,
|
| 463 |
+
"loss": 0.0385,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.142857142857143,
|
| 468 |
+
"grad_norm": 0.5033309459686279,
|
| 469 |
+
"learning_rate": 0.0001311062737786974,
|
| 470 |
+
"loss": 0.0357,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.175324675324675,
|
| 475 |
+
"grad_norm": 0.36022046208381653,
|
| 476 |
+
"learning_rate": 0.0001290542056881781,
|
| 477 |
+
"loss": 0.0243,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.207792207792208,
|
| 482 |
+
"grad_norm": 0.3985787034034729,
|
| 483 |
+
"learning_rate": 0.0001269886856154735,
|
| 484 |
+
"loss": 0.0471,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.24025974025974,
|
| 489 |
+
"grad_norm": 0.47278791666030884,
|
| 490 |
+
"learning_rate": 0.0001249106698881982,
|
| 491 |
+
"loss": 0.0353,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.2727272727272725,
|
| 496 |
+
"grad_norm": 0.5603681802749634,
|
| 497 |
+
"learning_rate": 0.0001228211206194055,
|
| 498 |
+
"loss": 0.0324,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.3051948051948052,
|
| 503 |
+
"grad_norm": 0.28177714347839355,
|
| 504 |
+
"learning_rate": 0.0001207210052621327,
|
| 505 |
+
"loss": 0.0432,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.3376623376623376,
|
| 510 |
+
"grad_norm": 0.5803484320640564,
|
| 511 |
+
"learning_rate": 0.00011861129616147418,
|
| 512 |
+
"loss": 0.0322,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.3701298701298703,
|
| 517 |
+
"grad_norm": 0.298414409160614,
|
| 518 |
+
"learning_rate": 0.00011649297010438956,
|
| 519 |
+
"loss": 0.0428,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.4025974025974026,
|
| 524 |
+
"grad_norm": 0.21691694855690002,
|
| 525 |
+
"learning_rate": 0.00011436700786745515,
|
| 526 |
+
"loss": 0.0243,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.435064935064935,
|
| 531 |
+
"grad_norm": 0.5338732600212097,
|
| 532 |
+
"learning_rate": 0.00011223439376276835,
|
| 533 |
+
"loss": 0.0325,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.4675324675324677,
|
| 538 |
+
"grad_norm": 0.5570558905601501,
|
| 539 |
+
"learning_rate": 0.00011009611518221489,
|
| 540 |
+
"loss": 0.0368,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.5,
|
| 545 |
+
"grad_norm": 0.4559939503669739,
|
| 546 |
+
"learning_rate": 0.00010795316214031048,
|
| 547 |
+
"loss": 0.0268,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.5324675324675323,
|
| 552 |
+
"grad_norm": 0.33397603034973145,
|
| 553 |
+
"learning_rate": 0.0001058065268158279,
|
| 554 |
+
"loss": 0.0359,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.564935064935065,
|
| 559 |
+
"grad_norm": 0.22799064218997955,
|
| 560 |
+
"learning_rate": 0.00010365720309242217,
|
| 561 |
+
"loss": 0.0264,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.5974025974025974,
|
| 566 |
+
"grad_norm": 0.38165828585624695,
|
| 567 |
+
"learning_rate": 0.00010150618609846638,
|
| 568 |
+
"loss": 0.0301,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.62987012987013,
|
| 573 |
+
"grad_norm": 0.5442892909049988,
|
| 574 |
+
"learning_rate": 9.935447174631119e-05,
|
| 575 |
+
"loss": 0.0238,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.6623376623376624,
|
| 580 |
+
"grad_norm": 0.22587241232395172,
|
| 581 |
+
"learning_rate": 9.720305627118126e-05,
|
| 582 |
+
"loss": 0.0337,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.6948051948051948,
|
| 587 |
+
"grad_norm": 0.2878405451774597,
|
| 588 |
+
"learning_rate": 9.505293576992251e-05,
|
| 589 |
+
"loss": 0.0253,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.7272727272727275,
|
| 594 |
+
"grad_norm": 0.5154989957809448,
|
| 595 |
+
"learning_rate": 9.290510573981298e-05,
|
| 596 |
+
"loss": 0.0279,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.75974025974026,
|
| 601 |
+
"grad_norm": 0.42716383934020996,
|
| 602 |
+
"learning_rate": 9.076056061765173e-05,
|
| 603 |
+
"loss": 0.0358,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.792207792207792,
|
| 608 |
+
"grad_norm": 0.6430768370628357,
|
| 609 |
+
"learning_rate": 8.862029331933828e-05,
|
| 610 |
+
"loss": 0.0315,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.824675324675325,
|
| 615 |
+
"grad_norm": 0.6324707269668579,
|
| 616 |
+
"learning_rate": 8.648529478015685e-05,
|
| 617 |
+
"loss": 0.0305,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.857142857142857,
|
| 622 |
+
"grad_norm": 0.27602899074554443,
|
| 623 |
+
"learning_rate": 8.435655349597689e-05,
|
| 624 |
+
"loss": 0.026,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.8896103896103895,
|
| 629 |
+
"grad_norm": 0.5011958479881287,
|
| 630 |
+
"learning_rate": 8.223505506558375e-05,
|
| 631 |
+
"loss": 0.0274,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.9220779220779223,
|
| 636 |
+
"grad_norm": 0.3036370277404785,
|
| 637 |
+
"learning_rate": 8.012178173434986e-05,
|
| 638 |
+
"loss": 0.0306,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.9545454545454546,
|
| 643 |
+
"grad_norm": 0.3664824366569519,
|
| 644 |
+
"learning_rate": 7.80177119394596e-05,
|
| 645 |
+
"loss": 0.0241,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.987012987012987,
|
| 650 |
+
"grad_norm": 0.43233174085617065,
|
| 651 |
+
"learning_rate": 7.592381985689628e-05,
|
| 652 |
+
"loss": 0.0309,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.0194805194805197,
|
| 657 |
+
"grad_norm": 0.3537318706512451,
|
| 658 |
+
"learning_rate": 7.384107495040284e-05,
|
| 659 |
+
"loss": 0.0321,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.051948051948052,
|
| 664 |
+
"grad_norm": 0.2594110071659088,
|
| 665 |
+
"learning_rate": 7.177044152262369e-05,
|
| 666 |
+
"loss": 0.0245,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.0844155844155843,
|
| 671 |
+
"grad_norm": 0.3139798939228058,
|
| 672 |
+
"learning_rate": 6.97128782686367e-05,
|
| 673 |
+
"loss": 0.0218,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.116883116883117,
|
| 678 |
+
"grad_norm": 0.2899845540523529,
|
| 679 |
+
"learning_rate": 6.766933783208092e-05,
|
| 680 |
+
"loss": 0.029,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.1493506493506493,
|
| 685 |
+
"grad_norm": 0.320487916469574,
|
| 686 |
+
"learning_rate": 6.564076636408656e-05,
|
| 687 |
+
"loss": 0.0226,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.1818181818181817,
|
| 692 |
+
"grad_norm": 0.4478653371334076,
|
| 693 |
+
"learning_rate": 6.362810308521054e-05,
|
| 694 |
+
"loss": 0.0224,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.2142857142857144,
|
| 699 |
+
"grad_norm": 0.20480212569236755,
|
| 700 |
+
"learning_rate": 6.16322798505813e-05,
|
| 701 |
+
"loss": 0.0224,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.2467532467532467,
|
| 706 |
+
"grad_norm": 0.4546569287776947,
|
| 707 |
+
"learning_rate": 5.9654220718453424e-05,
|
| 708 |
+
"loss": 0.0311,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.279220779220779,
|
| 713 |
+
"grad_norm": 0.2200387865304947,
|
| 714 |
+
"learning_rate": 5.769484152237249e-05,
|
| 715 |
+
"loss": 0.026,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.311688311688312,
|
| 720 |
+
"grad_norm": 0.40406161546707153,
|
| 721 |
+
"learning_rate": 5.5755049447147444e-05,
|
| 722 |
+
"loss": 0.028,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.344155844155844,
|
| 727 |
+
"grad_norm": 0.41747692227363586,
|
| 728 |
+
"learning_rate": 5.383574260882802e-05,
|
| 729 |
+
"loss": 0.0255,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.3766233766233764,
|
| 734 |
+
"grad_norm": 0.4012701213359833,
|
| 735 |
+
"learning_rate": 5.1937809638880374e-05,
|
| 736 |
+
"loss": 0.0289,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.409090909090909,
|
| 741 |
+
"grad_norm": 0.49661171436309814,
|
| 742 |
+
"learning_rate": 5.0062129272754146e-05,
|
| 743 |
+
"loss": 0.0249,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.4415584415584415,
|
| 748 |
+
"grad_norm": 0.225033700466156,
|
| 749 |
+
"learning_rate": 4.8209569943031516e-05,
|
| 750 |
+
"loss": 0.0274,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.474025974025974,
|
| 755 |
+
"grad_norm": 0.2046009600162506,
|
| 756 |
+
"learning_rate": 4.638098937734648e-05,
|
| 757 |
+
"loss": 0.0232,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.5064935064935066,
|
| 762 |
+
"grad_norm": 0.4959232211112976,
|
| 763 |
+
"learning_rate": 4.4577234201260196e-05,
|
| 764 |
+
"loss": 0.0233,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.538961038961039,
|
| 769 |
+
"grad_norm": 0.42567142844200134,
|
| 770 |
+
"learning_rate": 4.279913954627667e-05,
|
| 771 |
+
"loss": 0.0188,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.571428571428571,
|
| 776 |
+
"grad_norm": 0.22492988407611847,
|
| 777 |
+
"learning_rate": 4.104752866318026e-05,
|
| 778 |
+
"loss": 0.0178,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.603896103896104,
|
| 783 |
+
"grad_norm": 0.2631309926509857,
|
| 784 |
+
"learning_rate": 3.932321254087389e-05,
|
| 785 |
+
"loss": 0.0187,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6363636363636362,
|
| 790 |
+
"grad_norm": 0.45389196276664734,
|
| 791 |
+
"learning_rate": 3.7626989530894285e-05,
|
| 792 |
+
"loss": 0.0211,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6688311688311686,
|
| 797 |
+
"grad_norm": 0.20797108113765717,
|
| 798 |
+
"learning_rate": 3.5959644977778386e-05,
|
| 799 |
+
"loss": 0.024,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.7012987012987013,
|
| 804 |
+
"grad_norm": 0.22460906207561493,
|
| 805 |
+
"learning_rate": 3.432195085545191e-05,
|
| 806 |
+
"loss": 0.0161,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.7337662337662336,
|
| 811 |
+
"grad_norm": 0.3012770116329193,
|
| 812 |
+
"learning_rate": 3.271466540980862e-05,
|
| 813 |
+
"loss": 0.0232,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.7662337662337664,
|
| 818 |
+
"grad_norm": 0.38262003660202026,
|
| 819 |
+
"learning_rate": 3.1138532807645394e-05,
|
| 820 |
+
"loss": 0.0174,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.7987012987012987,
|
| 825 |
+
"grad_norm": 0.22977425158023834,
|
| 826 |
+
"learning_rate": 2.9594282792115857e-05,
|
| 827 |
+
"loss": 0.0227,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.8311688311688314,
|
| 832 |
+
"grad_norm": 0.16646578907966614,
|
| 833 |
+
"learning_rate": 2.808263034486226e-05,
|
| 834 |
+
"loss": 0.0155,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8636363636363638,
|
| 839 |
+
"grad_norm": 0.257495254278183,
|
| 840 |
+
"learning_rate": 2.660427535498191e-05,
|
| 841 |
+
"loss": 0.0218,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.896103896103896,
|
| 846 |
+
"grad_norm": 0.3283085227012634,
|
| 847 |
+
"learning_rate": 2.5159902294981197e-05,
|
| 848 |
+
"loss": 0.0224,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.928571428571429,
|
| 853 |
+
"grad_norm": 0.36105039715766907,
|
| 854 |
+
"learning_rate": 2.3750179903867443e-05,
|
| 855 |
+
"loss": 0.0278,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.961038961038961,
|
| 860 |
+
"grad_norm": 0.3740982115268707,
|
| 861 |
+
"learning_rate": 2.237576087752554e-05,
|
| 862 |
+
"loss": 0.0241,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.9935064935064934,
|
| 867 |
+
"grad_norm": 0.19882728159427643,
|
| 868 |
+
"learning_rate": 2.1037281566522304e-05,
|
| 869 |
+
"loss": 0.0174,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.025974025974026,
|
| 874 |
+
"grad_norm": 0.12800416350364685,
|
| 875 |
+
"learning_rate": 1.973536168147867e-05,
|
| 876 |
+
"loss": 0.0137,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.058441558441558,
|
| 881 |
+
"grad_norm": 0.352993905544281,
|
| 882 |
+
"learning_rate": 1.8470604006146064e-05,
|
| 883 |
+
"loss": 0.0221,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.090909090909091,
|
| 888 |
+
"grad_norm": 0.3981715738773346,
|
| 889 |
+
"learning_rate": 1.7243594118319985e-05,
|
| 890 |
+
"loss": 0.0162,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.123376623376624,
|
| 895 |
+
"grad_norm": 0.23238126933574677,
|
| 896 |
+
"learning_rate": 1.6054900118719807e-05,
|
| 897 |
+
"loss": 0.0201,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.1558441558441555,
|
| 902 |
+
"grad_norm": 0.3487814962863922,
|
| 903 |
+
"learning_rate": 1.4905072367960437e-05,
|
| 904 |
+
"loss": 0.0182,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.188311688311688,
|
| 909 |
+
"grad_norm": 0.2413235455751419,
|
| 910 |
+
"learning_rate": 1.3794643231737348e-05,
|
| 911 |
+
"loss": 0.0161,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.220779220779221,
|
| 916 |
+
"grad_norm": 0.3678736984729767,
|
| 917 |
+
"learning_rate": 1.2724126834343564e-05,
|
| 918 |
+
"loss": 0.02,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.253246753246753,
|
| 923 |
+
"grad_norm": 0.0916026160120964,
|
| 924 |
+
"learning_rate": 1.1694018820632068e-05,
|
| 925 |
+
"loss": 0.0179,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.285714285714286,
|
| 930 |
+
"grad_norm": 0.16847121715545654,
|
| 931 |
+
"learning_rate": 1.0704796126534234e-05,
|
| 932 |
+
"loss": 0.0182,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.318181818181818,
|
| 937 |
+
"grad_norm": 0.10810457170009613,
|
| 938 |
+
"learning_rate": 9.756916758240286e-06,
|
| 939 |
+
"loss": 0.0219,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.35064935064935,
|
| 944 |
+
"grad_norm": 0.31443580985069275,
|
| 945 |
+
"learning_rate": 8.850819580144387e-06,
|
| 946 |
+
"loss": 0.0254,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.383116883116883,
|
| 951 |
+
"grad_norm": 0.2282458394765854,
|
| 952 |
+
"learning_rate": 7.986924111652006e-06,
|
| 953 |
+
"loss": 0.0177,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.415584415584416,
|
| 958 |
+
"grad_norm": 0.34387949109077454,
|
| 959 |
+
"learning_rate": 7.16563033294424e-06,
|
| 960 |
+
"loss": 0.0167,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.448051948051948,
|
| 965 |
+
"grad_norm": 0.12599095702171326,
|
| 966 |
+
"learning_rate": 6.387318499788497e-06,
|
| 967 |
+
"loss": 0.0205,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.48051948051948,
|
| 972 |
+
"grad_norm": 0.14966098964214325,
|
| 973 |
+
"learning_rate": 5.652348967481569e-06,
|
| 974 |
+
"loss": 0.0201,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.512987012987013,
|
| 979 |
+
"grad_norm": 0.28549131751060486,
|
| 980 |
+
"learning_rate": 4.961062024006591e-06,
|
| 981 |
+
"loss": 0.019,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.545454545454545,
|
| 986 |
+
"grad_norm": 0.22848688066005707,
|
| 987 |
+
"learning_rate": 4.313777732481039e-06,
|
| 988 |
+
"loss": 0.0182,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.577922077922078,
|
| 993 |
+
"grad_norm": 0.20382894575595856,
|
| 994 |
+
"learning_rate": 3.7107957829688234e-06,
|
| 995 |
+
"loss": 0.0228,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.6103896103896105,
|
| 1000 |
+
"grad_norm": 0.1720917969942093,
|
| 1001 |
+
"learning_rate": 3.1523953537248684e-06,
|
| 1002 |
+
"loss": 0.0189,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.642857142857143,
|
| 1007 |
+
"grad_norm": 0.10501822084188461,
|
| 1008 |
+
"learning_rate": 2.638834981936744e-06,
|
| 1009 |
+
"loss": 0.015,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.675324675324675,
|
| 1014 |
+
"grad_norm": 0.13112539052963257,
|
| 1015 |
+
"learning_rate": 2.1703524440230383e-06,
|
| 1016 |
+
"loss": 0.0118,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.707792207792208,
|
| 1021 |
+
"grad_norm": 0.17099051177501678,
|
| 1022 |
+
"learning_rate": 1.7471646455437085e-06,
|
| 1023 |
+
"loss": 0.0151,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.740259740259741,
|
| 1028 |
+
"grad_norm": 0.3271525204181671,
|
| 1029 |
+
"learning_rate": 1.3694675207737151e-06,
|
| 1030 |
+
"loss": 0.0223,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.7727272727272725,
|
| 1035 |
+
"grad_norm": 0.35389673709869385,
|
| 1036 |
+
"learning_rate": 1.03743594198622e-06,
|
| 1037 |
+
"loss": 0.0133,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.805194805194805,
|
| 1042 |
+
"grad_norm": 0.25904619693756104,
|
| 1043 |
+
"learning_rate": 7.512236384874305e-07,
|
| 1044 |
+
"loss": 0.0166,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.837662337662338,
|
| 1049 |
+
"grad_norm": 0.14534501731395721,
|
| 1050 |
+
"learning_rate": 5.109631254405445e-07,
|
| 1051 |
+
"loss": 0.0133,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.87012987012987,
|
| 1056 |
+
"grad_norm": 0.17412607371807098,
|
| 1057 |
+
"learning_rate": 3.1676564251171824e-07,
|
| 1058 |
+
"loss": 0.0188,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.902597402597403,
|
| 1063 |
+
"grad_norm": 0.20664988458156586,
|
| 1064 |
+
"learning_rate": 1.687211023665647e-07,
|
| 1065 |
+
"loss": 0.0114,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.935064935064935,
|
| 1070 |
+
"grad_norm": 0.19708259403705597,
|
| 1071 |
+
"learning_rate": 6.68980490409421e-08,
|
| 1072 |
+
"loss": 0.0179,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.967532467532467,
|
| 1077 |
+
"grad_norm": 0.5721464157104492,
|
| 1078 |
+
"learning_rate": 1.134362620534013e-08,
|
| 1079 |
+
"loss": 0.0192,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.990259740259741,
|
| 1084 |
+
"step": 1537,
|
| 1085 |
+
"total_flos": 5.134548117135398e+16,
|
| 1086 |
+
"train_loss": 0.05982626907687413,
|
| 1087 |
+
"train_runtime": 715.8651,
|
| 1088 |
+
"train_samples_per_second": 34.353,
|
| 1089 |
+
"train_steps_per_second": 2.147
|
| 1090 |
+
}
|
| 1091 |
+
],
|
| 1092 |
+
"logging_steps": 10,
|
| 1093 |
+
"max_steps": 1537,
|
| 1094 |
+
"num_input_tokens_seen": 0,
|
| 1095 |
+
"num_train_epochs": 5,
|
| 1096 |
+
"save_steps": 500,
|
| 1097 |
+
"stateful_callbacks": {
|
| 1098 |
+
"TrainerControl": {
|
| 1099 |
+
"args": {
|
| 1100 |
+
"should_epoch_stop": false,
|
| 1101 |
+
"should_evaluate": false,
|
| 1102 |
+
"should_log": false,
|
| 1103 |
+
"should_save": true,
|
| 1104 |
+
"should_training_stop": true
|
| 1105 |
+
},
|
| 1106 |
+
"attributes": {}
|
| 1107 |
+
}
|
| 1108 |
+
},
|
| 1109 |
+
"total_flos": 5.134548117135398e+16,
|
| 1110 |
+
"train_batch_size": 16,
|
| 1111 |
+
"trial_name": null,
|
| 1112 |
+
"trial_params": null
|
| 1113 |
+
}
|