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": 4.993485342019544,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1533,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03257328990228013,
|
| 13 |
+
"grad_norm": 10.128602027893066,
|
| 14 |
+
"learning_rate": 2.5974025974025972e-05,
|
| 15 |
+
"loss": 1.7762,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06514657980456026,
|
| 20 |
+
"grad_norm": 11.557284355163574,
|
| 21 |
+
"learning_rate": 5.1948051948051944e-05,
|
| 22 |
+
"loss": 0.5874,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09771986970684039,
|
| 27 |
+
"grad_norm": 1.7021989822387695,
|
| 28 |
+
"learning_rate": 7.792207792207793e-05,
|
| 29 |
+
"loss": 0.2361,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.13029315960912052,
|
| 34 |
+
"grad_norm": 2.088670015335083,
|
| 35 |
+
"learning_rate": 0.00010389610389610389,
|
| 36 |
+
"loss": 0.176,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16286644951140064,
|
| 41 |
+
"grad_norm": 2.070906400680542,
|
| 42 |
+
"learning_rate": 0.00012987012987012987,
|
| 43 |
+
"loss": 0.1546,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19543973941368079,
|
| 48 |
+
"grad_norm": 1.9704210758209229,
|
| 49 |
+
"learning_rate": 0.00015584415584415587,
|
| 50 |
+
"loss": 0.1463,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2280130293159609,
|
| 55 |
+
"grad_norm": 1.2319754362106323,
|
| 56 |
+
"learning_rate": 0.00018181818181818183,
|
| 57 |
+
"loss": 0.1306,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.26058631921824105,
|
| 62 |
+
"grad_norm": 1.4384595155715942,
|
| 63 |
+
"learning_rate": 0.0001999979049808708,
|
| 64 |
+
"loss": 0.1218,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2931596091205212,
|
| 69 |
+
"grad_norm": 1.3861066102981567,
|
| 70 |
+
"learning_rate": 0.00019996066263830531,
|
| 71 |
+
"loss": 0.1031,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3257328990228013,
|
| 76 |
+
"grad_norm": 1.3069312572479248,
|
| 77 |
+
"learning_rate": 0.00019987688427197897,
|
| 78 |
+
"loss": 0.1125,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3583061889250814,
|
| 83 |
+
"grad_norm": 1.1740531921386719,
|
| 84 |
+
"learning_rate": 0.0001997466088843548,
|
| 85 |
+
"loss": 0.0984,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.39087947882736157,
|
| 90 |
+
"grad_norm": 0.831253170967102,
|
| 91 |
+
"learning_rate": 0.00019956989712427577,
|
| 92 |
+
"loss": 0.1013,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4234527687296417,
|
| 97 |
+
"grad_norm": 1.1512802839279175,
|
| 98 |
+
"learning_rate": 0.0001993468312587303,
|
| 99 |
+
"loss": 0.0827,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4560260586319218,
|
| 104 |
+
"grad_norm": 0.7890095710754395,
|
| 105 |
+
"learning_rate": 0.00019907751513455302,
|
| 106 |
+
"loss": 0.0747,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.48859934853420195,
|
| 111 |
+
"grad_norm": 0.8635206818580627,
|
| 112 |
+
"learning_rate": 0.00019876207413008015,
|
| 113 |
+
"loss": 0.0911,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5211726384364821,
|
| 118 |
+
"grad_norm": 0.984118640422821,
|
| 119 |
+
"learning_rate": 0.00019840065509677988,
|
| 120 |
+
"loss": 0.0708,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5537459283387622,
|
| 125 |
+
"grad_norm": 1.1899775266647339,
|
| 126 |
+
"learning_rate": 0.00019799342629088702,
|
| 127 |
+
"loss": 0.0759,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5863192182410424,
|
| 132 |
+
"grad_norm": 0.8736202120780945,
|
| 133 |
+
"learning_rate": 0.00019754057729507227,
|
| 134 |
+
"loss": 0.0747,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6188925081433225,
|
| 139 |
+
"grad_norm": 0.6165608763694763,
|
| 140 |
+
"learning_rate": 0.0001970423189301833,
|
| 141 |
+
"loss": 0.0697,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6514657980456026,
|
| 146 |
+
"grad_norm": 1.2145845890045166,
|
| 147 |
+
"learning_rate": 0.00019649888315709843,
|
| 148 |
+
"loss": 0.0768,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6840390879478827,
|
| 153 |
+
"grad_norm": 0.7146447896957397,
|
| 154 |
+
"learning_rate": 0.00019591052296873888,
|
| 155 |
+
"loss": 0.0575,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7166123778501629,
|
| 160 |
+
"grad_norm": 1.0241068601608276,
|
| 161 |
+
"learning_rate": 0.00019527751227228963,
|
| 162 |
+
"loss": 0.0688,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.749185667752443,
|
| 167 |
+
"grad_norm": 0.8579204678535461,
|
| 168 |
+
"learning_rate": 0.00019460014576168358,
|
| 169 |
+
"loss": 0.0544,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7817589576547231,
|
| 174 |
+
"grad_norm": 0.5284867286682129,
|
| 175 |
+
"learning_rate": 0.0001938787387804088,
|
| 176 |
+
"loss": 0.0516,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8143322475570033,
|
| 181 |
+
"grad_norm": 0.6213068962097168,
|
| 182 |
+
"learning_rate": 0.00019311362717470268,
|
| 183 |
+
"loss": 0.0501,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8469055374592834,
|
| 188 |
+
"grad_norm": 0.7320097088813782,
|
| 189 |
+
"learning_rate": 0.00019230516713720052,
|
| 190 |
+
"loss": 0.0581,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8794788273615635,
|
| 195 |
+
"grad_norm": 0.36288219690322876,
|
| 196 |
+
"learning_rate": 0.00019145373504111279,
|
| 197 |
+
"loss": 0.0478,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9120521172638436,
|
| 202 |
+
"grad_norm": 0.6096987724304199,
|
| 203 |
+
"learning_rate": 0.00019055972726500695,
|
| 204 |
+
"loss": 0.0548,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9446254071661238,
|
| 209 |
+
"grad_norm": 0.7391396760940552,
|
| 210 |
+
"learning_rate": 0.0001896235600082759,
|
| 211 |
+
"loss": 0.0444,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9771986970684039,
|
| 216 |
+
"grad_norm": 0.7820376753807068,
|
| 217 |
+
"learning_rate": 0.00018864566909737937,
|
| 218 |
+
"loss": 0.0566,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.009771986970684,
|
| 223 |
+
"grad_norm": 0.3828499913215637,
|
| 224 |
+
"learning_rate": 0.00018762650978294758,
|
| 225 |
+
"loss": 0.0533,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0423452768729642,
|
| 230 |
+
"grad_norm": 0.49720415472984314,
|
| 231 |
+
"learning_rate": 0.0001865665565278424,
|
| 232 |
+
"loss": 0.0582,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0749185667752443,
|
| 237 |
+
"grad_norm": 0.4968441128730774,
|
| 238 |
+
"learning_rate": 0.00018546630278627437,
|
| 239 |
+
"loss": 0.0473,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.1074918566775245,
|
| 244 |
+
"grad_norm": 0.812972366809845,
|
| 245 |
+
"learning_rate": 0.00018432626077407829,
|
| 246 |
+
"loss": 0.0382,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.1400651465798046,
|
| 251 |
+
"grad_norm": 0.4055006802082062,
|
| 252 |
+
"learning_rate": 0.00018314696123025454,
|
| 253 |
+
"loss": 0.043,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1726384364820848,
|
| 258 |
+
"grad_norm": 0.6709649562835693,
|
| 259 |
+
"learning_rate": 0.0001819289531698871,
|
| 260 |
+
"loss": 0.0343,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.205211726384365,
|
| 265 |
+
"grad_norm": 0.6343613862991333,
|
| 266 |
+
"learning_rate": 0.0001806728036285532,
|
| 267 |
+
"loss": 0.0388,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.237785016286645,
|
| 272 |
+
"grad_norm": 0.44080787897109985,
|
| 273 |
+
"learning_rate": 0.00017937909739834367,
|
| 274 |
+
"loss": 0.0419,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2703583061889252,
|
| 279 |
+
"grad_norm": 0.714772641658783,
|
| 280 |
+
"learning_rate": 0.00017804843675561677,
|
| 281 |
+
"loss": 0.0476,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.3029315960912053,
|
| 286 |
+
"grad_norm": 0.4127221703529358,
|
| 287 |
+
"learning_rate": 0.00017668144118061262,
|
| 288 |
+
"loss": 0.0354,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.3355048859934853,
|
| 293 |
+
"grad_norm": 0.44271162152290344,
|
| 294 |
+
"learning_rate": 0.00017527874706905805,
|
| 295 |
+
"loss": 0.0385,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3680781758957654,
|
| 300 |
+
"grad_norm": 0.6225533485412598,
|
| 301 |
+
"learning_rate": 0.00017384100743589697,
|
| 302 |
+
"loss": 0.0665,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.4006514657980456,
|
| 307 |
+
"grad_norm": 0.4418353736400604,
|
| 308 |
+
"learning_rate": 0.0001723688916112835,
|
| 309 |
+
"loss": 0.0554,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4332247557003257,
|
| 314 |
+
"grad_norm": 0.3992595672607422,
|
| 315 |
+
"learning_rate": 0.00017086308492897983,
|
| 316 |
+
"loss": 0.0522,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4657980456026058,
|
| 321 |
+
"grad_norm": 0.45478880405426025,
|
| 322 |
+
"learning_rate": 0.0001693242884073035,
|
| 323 |
+
"loss": 0.0398,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.498371335504886,
|
| 328 |
+
"grad_norm": 0.5688516497612,
|
| 329 |
+
"learning_rate": 0.00016775321842277312,
|
| 330 |
+
"loss": 0.032,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.5309446254071661,
|
| 335 |
+
"grad_norm": 0.4958943724632263,
|
| 336 |
+
"learning_rate": 0.00016615060637660388,
|
| 337 |
+
"loss": 0.0422,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.5635179153094463,
|
| 342 |
+
"grad_norm": 0.4975818991661072,
|
| 343 |
+
"learning_rate": 0.00016451719835420877,
|
| 344 |
+
"loss": 0.0317,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5960912052117264,
|
| 349 |
+
"grad_norm": 0.5288175344467163,
|
| 350 |
+
"learning_rate": 0.00016285375477786322,
|
| 351 |
+
"loss": 0.0412,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6286644951140063,
|
| 356 |
+
"grad_norm": 0.6416569352149963,
|
| 357 |
+
"learning_rate": 0.0001611610500526957,
|
| 358 |
+
"loss": 0.0275,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.6612377850162865,
|
| 363 |
+
"grad_norm": 0.42349156737327576,
|
| 364 |
+
"learning_rate": 0.00015943987220616855,
|
| 365 |
+
"loss": 0.029,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6938110749185666,
|
| 370 |
+
"grad_norm": 0.2572161853313446,
|
| 371 |
+
"learning_rate": 0.00015769102252121702,
|
| 372 |
+
"loss": 0.0308,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7263843648208468,
|
| 377 |
+
"grad_norm": 0.5944684147834778,
|
| 378 |
+
"learning_rate": 0.0001559153151632171,
|
| 379 |
+
"loss": 0.0354,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.758957654723127,
|
| 384 |
+
"grad_norm": 0.5154451727867126,
|
| 385 |
+
"learning_rate": 0.0001541135768009566,
|
| 386 |
+
"loss": 0.0424,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.791530944625407,
|
| 391 |
+
"grad_norm": 0.5298171043395996,
|
| 392 |
+
"learning_rate": 0.00015228664622178467,
|
| 393 |
+
"loss": 0.0263,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8241042345276872,
|
| 398 |
+
"grad_norm": 0.3582160472869873,
|
| 399 |
+
"learning_rate": 0.00015043537394112007,
|
| 400 |
+
"loss": 0.0386,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.8566775244299674,
|
| 405 |
+
"grad_norm": 0.6350813508033752,
|
| 406 |
+
"learning_rate": 0.0001485606218064993,
|
| 407 |
+
"loss": 0.0444,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8892508143322475,
|
| 412 |
+
"grad_norm": 0.6265957355499268,
|
| 413 |
+
"learning_rate": 0.00014666326259634918,
|
| 414 |
+
"loss": 0.0417,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.9218241042345277,
|
| 419 |
+
"grad_norm": 0.41339483857154846,
|
| 420 |
+
"learning_rate": 0.00014474417961367065,
|
| 421 |
+
"loss": 0.0309,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.9543973941368078,
|
| 426 |
+
"grad_norm": 0.41287094354629517,
|
| 427 |
+
"learning_rate": 0.000142804266274823,
|
| 428 |
+
"loss": 0.0308,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.986970684039088,
|
| 433 |
+
"grad_norm": 0.5833337903022766,
|
| 434 |
+
"learning_rate": 0.00014084442569359964,
|
| 435 |
+
"loss": 0.0346,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.019543973941368,
|
| 440 |
+
"grad_norm": 0.3714321255683899,
|
| 441 |
+
"learning_rate": 0.00013886557026078955,
|
| 442 |
+
"loss": 0.0337,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.0521172638436482,
|
| 447 |
+
"grad_norm": 0.2712138295173645,
|
| 448 |
+
"learning_rate": 0.0001368686212194199,
|
| 449 |
+
"loss": 0.0196,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.0846905537459284,
|
| 454 |
+
"grad_norm": 0.6208952069282532,
|
| 455 |
+
"learning_rate": 0.00013485450823587725,
|
| 456 |
+
"loss": 0.0288,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.1172638436482085,
|
| 461 |
+
"grad_norm": 0.34229573607444763,
|
| 462 |
+
"learning_rate": 0.00013282416896710778,
|
| 463 |
+
"loss": 0.0246,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.1498371335504887,
|
| 468 |
+
"grad_norm": 0.4573960304260254,
|
| 469 |
+
"learning_rate": 0.00013077854862409696,
|
| 470 |
+
"loss": 0.0249,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.182410423452769,
|
| 475 |
+
"grad_norm": 0.3675532042980194,
|
| 476 |
+
"learning_rate": 0.0001287185995318333,
|
| 477 |
+
"loss": 0.022,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.214983713355049,
|
| 482 |
+
"grad_norm": 0.5372172594070435,
|
| 483 |
+
"learning_rate": 0.0001266452806859594,
|
| 484 |
+
"loss": 0.0343,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.247557003257329,
|
| 489 |
+
"grad_norm": 0.46904900670051575,
|
| 490 |
+
"learning_rate": 0.00012455955730631804,
|
| 491 |
+
"loss": 0.0233,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.2801302931596092,
|
| 496 |
+
"grad_norm": 0.4394093155860901,
|
| 497 |
+
"learning_rate": 0.00012246240038760043,
|
| 498 |
+
"loss": 0.0209,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.3127035830618894,
|
| 503 |
+
"grad_norm": 0.3127492368221283,
|
| 504 |
+
"learning_rate": 0.00012035478624730608,
|
| 505 |
+
"loss": 0.0321,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.3452768729641695,
|
| 510 |
+
"grad_norm": 0.601370096206665,
|
| 511 |
+
"learning_rate": 0.00011823769607122479,
|
| 512 |
+
"loss": 0.0243,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.3778501628664497,
|
| 517 |
+
"grad_norm": 0.5871070623397827,
|
| 518 |
+
"learning_rate": 0.00011611211545665184,
|
| 519 |
+
"loss": 0.0337,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.41042345276873,
|
| 524 |
+
"grad_norm": 0.3546801805496216,
|
| 525 |
+
"learning_rate": 0.00011397903395354996,
|
| 526 |
+
"loss": 0.0288,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.44299674267101,
|
| 531 |
+
"grad_norm": 0.8319407105445862,
|
| 532 |
+
"learning_rate": 0.0001118394446038708,
|
| 533 |
+
"loss": 0.0337,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.47557003257329,
|
| 538 |
+
"grad_norm": 0.5210663080215454,
|
| 539 |
+
"learning_rate": 0.00010969434347925076,
|
| 540 |
+
"loss": 0.026,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.5081433224755703,
|
| 545 |
+
"grad_norm": 0.5834184288978577,
|
| 546 |
+
"learning_rate": 0.00010754472921729661,
|
| 547 |
+
"loss": 0.0282,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.5407166123778504,
|
| 552 |
+
"grad_norm": 0.42890864610671997,
|
| 553 |
+
"learning_rate": 0.00010539160255667623,
|
| 554 |
+
"loss": 0.028,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.5732899022801305,
|
| 559 |
+
"grad_norm": 0.4473400413990021,
|
| 560 |
+
"learning_rate": 0.00010323596587123145,
|
| 561 |
+
"loss": 0.025,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.6058631921824107,
|
| 566 |
+
"grad_norm": 0.5189303159713745,
|
| 567 |
+
"learning_rate": 0.00010107882270332952,
|
| 568 |
+
"loss": 0.0293,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.6384364820846904,
|
| 573 |
+
"grad_norm": 0.43365001678466797,
|
| 574 |
+
"learning_rate": 9.892117729667052e-05,
|
| 575 |
+
"loss": 0.0175,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.6710097719869705,
|
| 580 |
+
"grad_norm": 0.28346696496009827,
|
| 581 |
+
"learning_rate": 9.676403412876856e-05,
|
| 582 |
+
"loss": 0.0334,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.7035830618892507,
|
| 587 |
+
"grad_norm": 0.3956477642059326,
|
| 588 |
+
"learning_rate": 9.460839744332378e-05,
|
| 589 |
+
"loss": 0.0271,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.736156351791531,
|
| 594 |
+
"grad_norm": 0.30705949664115906,
|
| 595 |
+
"learning_rate": 9.245527078270341e-05,
|
| 596 |
+
"loss": 0.0217,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.768729641693811,
|
| 601 |
+
"grad_norm": 0.40188854932785034,
|
| 602 |
+
"learning_rate": 9.030565652074926e-05,
|
| 603 |
+
"loss": 0.019,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.801302931596091,
|
| 608 |
+
"grad_norm": 0.3447129428386688,
|
| 609 |
+
"learning_rate": 8.816055539612924e-05,
|
| 610 |
+
"loss": 0.028,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.8338762214983713,
|
| 615 |
+
"grad_norm": 0.38768622279167175,
|
| 616 |
+
"learning_rate": 8.602096604645009e-05,
|
| 617 |
+
"loss": 0.0218,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.8664495114006514,
|
| 622 |
+
"grad_norm": 0.26912721991539,
|
| 623 |
+
"learning_rate": 8.388788454334817e-05,
|
| 624 |
+
"loss": 0.0173,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.8990228013029316,
|
| 629 |
+
"grad_norm": 0.33078861236572266,
|
| 630 |
+
"learning_rate": 8.176230392877523e-05,
|
| 631 |
+
"loss": 0.0233,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.9315960912052117,
|
| 636 |
+
"grad_norm": 0.24832488596439362,
|
| 637 |
+
"learning_rate": 7.964521375269396e-05,
|
| 638 |
+
"loss": 0.0171,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.964169381107492,
|
| 643 |
+
"grad_norm": 0.6595136523246765,
|
| 644 |
+
"learning_rate": 7.753759961239964e-05,
|
| 645 |
+
"loss": 0.0272,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.996742671009772,
|
| 650 |
+
"grad_norm": 0.2780207097530365,
|
| 651 |
+
"learning_rate": 7.544044269368197e-05,
|
| 652 |
+
"loss": 0.0338,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.029315960912052,
|
| 657 |
+
"grad_norm": 0.7173179388046265,
|
| 658 |
+
"learning_rate": 7.335471931404063e-05,
|
| 659 |
+
"loss": 0.0365,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.0618892508143323,
|
| 664 |
+
"grad_norm": 0.33753442764282227,
|
| 665 |
+
"learning_rate": 7.128140046816671e-05,
|
| 666 |
+
"loss": 0.0195,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.0944625407166124,
|
| 671 |
+
"grad_norm": 0.35064950585365295,
|
| 672 |
+
"learning_rate": 6.922145137590306e-05,
|
| 673 |
+
"loss": 0.02,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.1270358306188926,
|
| 678 |
+
"grad_norm": 0.39598166942596436,
|
| 679 |
+
"learning_rate": 6.717583103289229e-05,
|
| 680 |
+
"loss": 0.0203,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.1596091205211727,
|
| 685 |
+
"grad_norm": 0.18257524073123932,
|
| 686 |
+
"learning_rate": 6.514549176412275e-05,
|
| 687 |
+
"loss": 0.0134,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.192182410423453,
|
| 692 |
+
"grad_norm": 0.4458347260951996,
|
| 693 |
+
"learning_rate": 6.313137878058013e-05,
|
| 694 |
+
"loss": 0.0236,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.224755700325733,
|
| 699 |
+
"grad_norm": 0.22742605209350586,
|
| 700 |
+
"learning_rate": 6.113442973921046e-05,
|
| 701 |
+
"loss": 0.0208,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.257328990228013,
|
| 706 |
+
"grad_norm": 0.1858537793159485,
|
| 707 |
+
"learning_rate": 5.9155574306400395e-05,
|
| 708 |
+
"loss": 0.0218,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.2899022801302933,
|
| 713 |
+
"grad_norm": 0.24626286327838898,
|
| 714 |
+
"learning_rate": 5.7195733725176994e-05,
|
| 715 |
+
"loss": 0.0232,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.3224755700325734,
|
| 720 |
+
"grad_norm": 0.2719153165817261,
|
| 721 |
+
"learning_rate": 5.525582038632934e-05,
|
| 722 |
+
"loss": 0.0148,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.3550488599348536,
|
| 727 |
+
"grad_norm": 0.218730166554451,
|
| 728 |
+
"learning_rate": 5.333673740365083e-05,
|
| 729 |
+
"loss": 0.0157,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.3876221498371337,
|
| 734 |
+
"grad_norm": 0.20292945206165314,
|
| 735 |
+
"learning_rate": 5.1439378193500707e-05,
|
| 736 |
+
"loss": 0.0143,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.420195439739414,
|
| 741 |
+
"grad_norm": 0.2846449017524719,
|
| 742 |
+
"learning_rate": 4.956462605887994e-05,
|
| 743 |
+
"loss": 0.0177,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.4527687296416936,
|
| 748 |
+
"grad_norm": 0.322721391916275,
|
| 749 |
+
"learning_rate": 4.771335377821535e-05,
|
| 750 |
+
"loss": 0.0224,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.4853420195439737,
|
| 755 |
+
"grad_norm": 0.1719449758529663,
|
| 756 |
+
"learning_rate": 4.588642319904343e-05,
|
| 757 |
+
"loss": 0.0234,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.517915309446254,
|
| 762 |
+
"grad_norm": 0.44704851508140564,
|
| 763 |
+
"learning_rate": 4.408468483678293e-05,
|
| 764 |
+
"loss": 0.019,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.550488599348534,
|
| 769 |
+
"grad_norm": 0.4159814417362213,
|
| 770 |
+
"learning_rate": 4.230897747878303e-05,
|
| 771 |
+
"loss": 0.0156,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.583061889250814,
|
| 776 |
+
"grad_norm": 0.19604472815990448,
|
| 777 |
+
"learning_rate": 4.056012779383145e-05,
|
| 778 |
+
"loss": 0.0158,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.6156351791530943,
|
| 783 |
+
"grad_norm": 0.19116809964179993,
|
| 784 |
+
"learning_rate": 3.883894994730428e-05,
|
| 785 |
+
"loss": 0.0174,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6482084690553744,
|
| 790 |
+
"grad_norm": 0.3637801706790924,
|
| 791 |
+
"learning_rate": 3.714624522213681e-05,
|
| 792 |
+
"loss": 0.0162,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6807817589576546,
|
| 797 |
+
"grad_norm": 0.1877295821905136,
|
| 798 |
+
"learning_rate": 3.548280164579126e-05,
|
| 799 |
+
"loss": 0.0142,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.7133550488599347,
|
| 804 |
+
"grad_norm": 0.1830226182937622,
|
| 805 |
+
"learning_rate": 3.384939362339614e-05,
|
| 806 |
+
"loss": 0.0119,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.745928338762215,
|
| 811 |
+
"grad_norm": 0.15163740515708923,
|
| 812 |
+
"learning_rate": 3.224678157722689e-05,
|
| 813 |
+
"loss": 0.0181,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.778501628664495,
|
| 818 |
+
"grad_norm": 0.2479788213968277,
|
| 819 |
+
"learning_rate": 3.067571159269651e-05,
|
| 820 |
+
"loss": 0.0138,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.811074918566775,
|
| 825 |
+
"grad_norm": 0.6171669960021973,
|
| 826 |
+
"learning_rate": 2.913691507102019e-05,
|
| 827 |
+
"loss": 0.0197,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.8436482084690553,
|
| 832 |
+
"grad_norm": 0.18519634008407593,
|
| 833 |
+
"learning_rate": 2.763110838871651e-05,
|
| 834 |
+
"loss": 0.0137,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8762214983713354,
|
| 839 |
+
"grad_norm": 0.26303982734680176,
|
| 840 |
+
"learning_rate": 2.6158992564103058e-05,
|
| 841 |
+
"loss": 0.0172,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.9087947882736156,
|
| 846 |
+
"grad_norm": 0.28331807255744934,
|
| 847 |
+
"learning_rate": 2.4721252930941974e-05,
|
| 848 |
+
"loss": 0.0168,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.9413680781758957,
|
| 853 |
+
"grad_norm": 0.20530906319618225,
|
| 854 |
+
"learning_rate": 2.3318558819387404e-05,
|
| 855 |
+
"loss": 0.0199,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.973941368078176,
|
| 860 |
+
"grad_norm": 0.16924133896827698,
|
| 861 |
+
"learning_rate": 2.1951563244383233e-05,
|
| 862 |
+
"loss": 0.0146,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.006514657980456,
|
| 867 |
+
"grad_norm": 0.13186028599739075,
|
| 868 |
+
"learning_rate": 2.0620902601656345e-05,
|
| 869 |
+
"loss": 0.0124,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.039087947882736,
|
| 874 |
+
"grad_norm": 0.24360792338848114,
|
| 875 |
+
"learning_rate": 1.9327196371446776e-05,
|
| 876 |
+
"loss": 0.0119,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.071661237785016,
|
| 881 |
+
"grad_norm": 0.09876150637865067,
|
| 882 |
+
"learning_rate": 1.807104683011289e-05,
|
| 883 |
+
"loss": 0.012,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.1042345276872965,
|
| 888 |
+
"grad_norm": 0.2283184826374054,
|
| 889 |
+
"learning_rate": 1.6853038769745467e-05,
|
| 890 |
+
"loss": 0.0142,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.136807817589577,
|
| 895 |
+
"grad_norm": 0.32383596897125244,
|
| 896 |
+
"learning_rate": 1.5673739225921758e-05,
|
| 897 |
+
"loss": 0.012,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.169381107491857,
|
| 902 |
+
"grad_norm": 0.2783248722553253,
|
| 903 |
+
"learning_rate": 1.4533697213725662e-05,
|
| 904 |
+
"loss": 0.0163,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.201954397394137,
|
| 909 |
+
"grad_norm": 0.17678265273571014,
|
| 910 |
+
"learning_rate": 1.3433443472157613e-05,
|
| 911 |
+
"loss": 0.012,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.234527687296417,
|
| 916 |
+
"grad_norm": 0.25102487206459045,
|
| 917 |
+
"learning_rate": 1.237349021705243e-05,
|
| 918 |
+
"loss": 0.0158,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.267100977198697,
|
| 923 |
+
"grad_norm": 0.15461167693138123,
|
| 924 |
+
"learning_rate": 1.1354330902620636e-05,
|
| 925 |
+
"loss": 0.0126,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.299674267100977,
|
| 930 |
+
"grad_norm": 0.24122057855129242,
|
| 931 |
+
"learning_rate": 1.0376439991724096e-05,
|
| 932 |
+
"loss": 0.0168,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.3322475570032575,
|
| 937 |
+
"grad_norm": 0.14669205248355865,
|
| 938 |
+
"learning_rate": 9.440272734993072e-06,
|
| 939 |
+
"loss": 0.0179,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.364820846905538,
|
| 944 |
+
"grad_norm": 0.32440969347953796,
|
| 945 |
+
"learning_rate": 8.546264958887219e-06,
|
| 946 |
+
"loss": 0.0197,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.397394136807818,
|
| 951 |
+
"grad_norm": 0.14456795156002045,
|
| 952 |
+
"learning_rate": 7.694832862799505e-06,
|
| 953 |
+
"loss": 0.0111,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.429967426710098,
|
| 958 |
+
"grad_norm": 0.17956456542015076,
|
| 959 |
+
"learning_rate": 6.886372825297349e-06,
|
| 960 |
+
"loss": 0.0085,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.462540716612378,
|
| 965 |
+
"grad_norm": 0.30424752831459045,
|
| 966 |
+
"learning_rate": 6.12126121959119e-06,
|
| 967 |
+
"loss": 0.0207,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.495114006514658,
|
| 972 |
+
"grad_norm": 0.18671758472919464,
|
| 973 |
+
"learning_rate": 5.399854238316437e-06,
|
| 974 |
+
"loss": 0.013,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.527687296416938,
|
| 979 |
+
"grad_norm": 0.3565406799316406,
|
| 980 |
+
"learning_rate": 4.722487727710368e-06,
|
| 981 |
+
"loss": 0.0165,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.5602605863192185,
|
| 986 |
+
"grad_norm": 0.26344749331474304,
|
| 987 |
+
"learning_rate": 4.089477031261113e-06,
|
| 988 |
+
"loss": 0.0148,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.592833876221499,
|
| 993 |
+
"grad_norm": 0.18339155614376068,
|
| 994 |
+
"learning_rate": 3.5011168429016083e-06,
|
| 995 |
+
"loss": 0.0195,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.625407166123779,
|
| 1000 |
+
"grad_norm": 0.2598022222518921,
|
| 1001 |
+
"learning_rate": 2.95768106981672e-06,
|
| 1002 |
+
"loss": 0.0135,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.657980456026059,
|
| 1007 |
+
"grad_norm": 0.3853515684604645,
|
| 1008 |
+
"learning_rate": 2.4594227049277386e-06,
|
| 1009 |
+
"loss": 0.0177,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.690553745928339,
|
| 1014 |
+
"grad_norm": 0.13664180040359497,
|
| 1015 |
+
"learning_rate": 2.006573709112991e-06,
|
| 1016 |
+
"loss": 0.0086,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.723127035830619,
|
| 1021 |
+
"grad_norm": 0.1015399917960167,
|
| 1022 |
+
"learning_rate": 1.5993449032201458e-06,
|
| 1023 |
+
"loss": 0.0116,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.755700325732899,
|
| 1028 |
+
"grad_norm": 0.18885648250579834,
|
| 1029 |
+
"learning_rate": 1.237925869919887e-06,
|
| 1030 |
+
"loss": 0.0175,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.7882736156351795,
|
| 1035 |
+
"grad_norm": 0.18131224811077118,
|
| 1036 |
+
"learning_rate": 9.224848654469931e-07,
|
| 1037 |
+
"loss": 0.0088,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.82084690553746,
|
| 1042 |
+
"grad_norm": 0.194551482796669,
|
| 1043 |
+
"learning_rate": 6.531687412697496e-07,
|
| 1044 |
+
"loss": 0.014,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.85342019543974,
|
| 1049 |
+
"grad_norm": 0.23798178136348724,
|
| 1050 |
+
"learning_rate": 4.3010287572422537e-07,
|
| 1051 |
+
"loss": 0.0097,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.88599348534202,
|
| 1056 |
+
"grad_norm": 0.141094371676445,
|
| 1057 |
+
"learning_rate": 2.5339111564521844e-07,
|
| 1058 |
+
"loss": 0.0151,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.918566775244299,
|
| 1063 |
+
"grad_norm": 0.17839759588241577,
|
| 1064 |
+
"learning_rate": 1.2311572802105043e-07,
|
| 1065 |
+
"loss": 0.0097,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.95114006514658,
|
| 1070 |
+
"grad_norm": 0.3124100863933563,
|
| 1071 |
+
"learning_rate": 3.933736169471347e-08,
|
| 1072 |
+
"loss": 0.0098,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.9837133550488595,
|
| 1077 |
+
"grad_norm": 0.17664480209350586,
|
| 1078 |
+
"learning_rate": 2.0950191292112842e-09,
|
| 1079 |
+
"loss": 0.0142,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.993485342019544,
|
| 1084 |
+
"step": 1533,
|
| 1085 |
+
"total_flos": 5.185032946443418e+16,
|
| 1086 |
+
"train_loss": 0.05182663513868756,
|
| 1087 |
+
"train_runtime": 722.6992,
|
| 1088 |
+
"train_samples_per_second": 33.939,
|
| 1089 |
+
"train_steps_per_second": 2.121
|
| 1090 |
+
}
|
| 1091 |
+
],
|
| 1092 |
+
"logging_steps": 10,
|
| 1093 |
+
"max_steps": 1533,
|
| 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": 5.185032946443418e+16,
|
| 1110 |
+
"train_batch_size": 16,
|
| 1111 |
+
"trial_name": null,
|
| 1112 |
+
"trial_params": null
|
| 1113 |
+
}
|