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": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1535,
|
| 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": 10.164939880371094,
|
| 14 |
+
"learning_rate": 2.5974025974025972e-05,
|
| 15 |
+
"loss": 3.0074,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06514657980456026,
|
| 20 |
+
"grad_norm": 11.759410858154297,
|
| 21 |
+
"learning_rate": 5.1948051948051944e-05,
|
| 22 |
+
"loss": 1.1608,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09771986970684039,
|
| 27 |
+
"grad_norm": 2.336061716079712,
|
| 28 |
+
"learning_rate": 7.792207792207793e-05,
|
| 29 |
+
"loss": 0.8385,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.13029315960912052,
|
| 34 |
+
"grad_norm": 1.3015629053115845,
|
| 35 |
+
"learning_rate": 0.00010389610389610389,
|
| 36 |
+
"loss": 0.2802,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16286644951140064,
|
| 41 |
+
"grad_norm": 1.0599477291107178,
|
| 42 |
+
"learning_rate": 0.00012987012987012987,
|
| 43 |
+
"loss": 0.1576,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19543973941368079,
|
| 48 |
+
"grad_norm": 3.056833267211914,
|
| 49 |
+
"learning_rate": 0.00015584415584415587,
|
| 50 |
+
"loss": 0.2403,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2280130293159609,
|
| 55 |
+
"grad_norm": 2.162693738937378,
|
| 56 |
+
"learning_rate": 0.00018181818181818183,
|
| 57 |
+
"loss": 0.251,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.26058631921824105,
|
| 62 |
+
"grad_norm": 3.423180341720581,
|
| 63 |
+
"learning_rate": 0.0001999979107245606,
|
| 64 |
+
"loss": 0.2418,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2931596091205212,
|
| 69 |
+
"grad_norm": 1.267954707145691,
|
| 70 |
+
"learning_rate": 0.0001999607704786645,
|
| 71 |
+
"loss": 0.157,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3257328990228013,
|
| 76 |
+
"grad_norm": 2.345654249191284,
|
| 77 |
+
"learning_rate": 0.00019987722173728587,
|
| 78 |
+
"loss": 0.2589,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3583061889250814,
|
| 83 |
+
"grad_norm": 3.0315651893615723,
|
| 84 |
+
"learning_rate": 0.00019974730328935534,
|
| 85 |
+
"loss": 0.2621,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.39087947882736157,
|
| 90 |
+
"grad_norm": 0.543853223323822,
|
| 91 |
+
"learning_rate": 0.00019957107545173192,
|
| 92 |
+
"loss": 0.1612,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4234527687296417,
|
| 97 |
+
"grad_norm": 0.9071061611175537,
|
| 98 |
+
"learning_rate": 0.00019934862004120004,
|
| 99 |
+
"loss": 0.2302,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4560260586319218,
|
| 104 |
+
"grad_norm": 2.0897481441497803,
|
| 105 |
+
"learning_rate": 0.00019908004033648453,
|
| 106 |
+
"loss": 0.1522,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.48859934853420195,
|
| 111 |
+
"grad_norm": 0.8175027370452881,
|
| 112 |
+
"learning_rate": 0.00019876546103030196,
|
| 113 |
+
"loss": 0.0943,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5211726384364821,
|
| 118 |
+
"grad_norm": 0.5068450570106506,
|
| 119 |
+
"learning_rate": 0.00019840502817146966,
|
| 120 |
+
"loss": 0.0865,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5537459283387622,
|
| 125 |
+
"grad_norm": 1.0016964673995972,
|
| 126 |
+
"learning_rate": 0.00019799890909710013,
|
| 127 |
+
"loss": 0.1229,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5863192182410424,
|
| 132 |
+
"grad_norm": 2.739112377166748,
|
| 133 |
+
"learning_rate": 0.00019754729235491207,
|
| 134 |
+
"loss": 0.1327,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6188925081433225,
|
| 139 |
+
"grad_norm": 1.9622081518173218,
|
| 140 |
+
"learning_rate": 0.0001970503876156937,
|
| 141 |
+
"loss": 0.1361,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6514657980456026,
|
| 146 |
+
"grad_norm": 3.4183578491210938,
|
| 147 |
+
"learning_rate": 0.00019650842557595967,
|
| 148 |
+
"loss": 0.1363,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6840390879478827,
|
| 153 |
+
"grad_norm": 0.8435428738594055,
|
| 154 |
+
"learning_rate": 0.00019592165785084603,
|
| 155 |
+
"loss": 0.1079,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7166123778501629,
|
| 160 |
+
"grad_norm": 2.0992093086242676,
|
| 161 |
+
"learning_rate": 0.00019529035685729391,
|
| 162 |
+
"loss": 0.2401,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.749185667752443,
|
| 167 |
+
"grad_norm": 2.749847650527954,
|
| 168 |
+
"learning_rate": 0.00019461481568757506,
|
| 169 |
+
"loss": 0.1147,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7817589576547231,
|
| 174 |
+
"grad_norm": 2.5013184547424316,
|
| 175 |
+
"learning_rate": 0.00019389534797321884,
|
| 176 |
+
"loss": 0.1661,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8143322475570033,
|
| 181 |
+
"grad_norm": 0.8295263648033142,
|
| 182 |
+
"learning_rate": 0.00019313228773940345,
|
| 183 |
+
"loss": 0.1267,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8469055374592834,
|
| 188 |
+
"grad_norm": 2.893522262573242,
|
| 189 |
+
"learning_rate": 0.00019232598924987903,
|
| 190 |
+
"loss": 0.0937,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8794788273615635,
|
| 195 |
+
"grad_norm": 3.0416808128356934,
|
| 196 |
+
"learning_rate": 0.0001914768268424946,
|
| 197 |
+
"loss": 0.1318,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9120521172638436,
|
| 202 |
+
"grad_norm": 0.6984137892723083,
|
| 203 |
+
"learning_rate": 0.00019058519475540538,
|
| 204 |
+
"loss": 0.1003,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9446254071661238,
|
| 209 |
+
"grad_norm": 1.0821751356124878,
|
| 210 |
+
"learning_rate": 0.00018965150694404094,
|
| 211 |
+
"loss": 0.0888,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9771986970684039,
|
| 216 |
+
"grad_norm": 0.9735985994338989,
|
| 217 |
+
"learning_rate": 0.00018867619688891937,
|
| 218 |
+
"loss": 0.207,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.009771986970684,
|
| 223 |
+
"grad_norm": 0.6221311688423157,
|
| 224 |
+
"learning_rate": 0.0001876597173943965,
|
| 225 |
+
"loss": 0.1472,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0423452768729642,
|
| 230 |
+
"grad_norm": 2.335994243621826,
|
| 231 |
+
"learning_rate": 0.00018660254037844388,
|
| 232 |
+
"loss": 0.071,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0749185667752443,
|
| 237 |
+
"grad_norm": 5.066413402557373,
|
| 238 |
+
"learning_rate": 0.00018550515665355247,
|
| 239 |
+
"loss": 0.0781,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.1074918566775245,
|
| 244 |
+
"grad_norm": 1.0136305093765259,
|
| 245 |
+
"learning_rate": 0.000184368075698865,
|
| 246 |
+
"loss": 0.0879,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.1400651465798046,
|
| 251 |
+
"grad_norm": 1.4384146928787231,
|
| 252 |
+
"learning_rate": 0.00018319182542364117,
|
| 253 |
+
"loss": 0.1173,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1726384364820848,
|
| 258 |
+
"grad_norm": 1.0204399824142456,
|
| 259 |
+
"learning_rate": 0.00018197695192216702,
|
| 260 |
+
"loss": 0.0738,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.205211726384365,
|
| 265 |
+
"grad_norm": 0.8423169255256653,
|
| 266 |
+
"learning_rate": 0.00018072401922022117,
|
| 267 |
+
"loss": 0.0999,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.237785016286645,
|
| 272 |
+
"grad_norm": 1.6605499982833862,
|
| 273 |
+
"learning_rate": 0.0001794336090132164,
|
| 274 |
+
"loss": 0.0888,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2703583061889252,
|
| 279 |
+
"grad_norm": 3.043694019317627,
|
| 280 |
+
"learning_rate": 0.00017810632039613736,
|
| 281 |
+
"loss": 0.1041,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.3029315960912053,
|
| 286 |
+
"grad_norm": 0.49020448327064514,
|
| 287 |
+
"learning_rate": 0.00017674276958540072,
|
| 288 |
+
"loss": 0.1021,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.3355048859934853,
|
| 293 |
+
"grad_norm": 2.167153835296631,
|
| 294 |
+
"learning_rate": 0.00017534358963276607,
|
| 295 |
+
"loss": 0.0971,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3680781758957654,
|
| 300 |
+
"grad_norm": 0.8787184953689575,
|
| 301 |
+
"learning_rate": 0.00017390943013143083,
|
| 302 |
+
"loss": 0.0941,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.4006514657980456,
|
| 307 |
+
"grad_norm": 1.053727626800537,
|
| 308 |
+
"learning_rate": 0.0001724409569144455,
|
| 309 |
+
"loss": 0.0756,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4332247557003257,
|
| 314 |
+
"grad_norm": 0.5299842357635498,
|
| 315 |
+
"learning_rate": 0.0001709388517455893,
|
| 316 |
+
"loss": 0.0927,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4657980456026058,
|
| 321 |
+
"grad_norm": 1.387248158454895,
|
| 322 |
+
"learning_rate": 0.00016940381200284972,
|
| 323 |
+
"loss": 0.0828,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.498371335504886,
|
| 328 |
+
"grad_norm": 0.581618070602417,
|
| 329 |
+
"learning_rate": 0.0001678365503546528,
|
| 330 |
+
"loss": 0.0697,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.5309446254071661,
|
| 335 |
+
"grad_norm": 0.41632965207099915,
|
| 336 |
+
"learning_rate": 0.0001662377944289948,
|
| 337 |
+
"loss": 0.1533,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.5635179153094463,
|
| 342 |
+
"grad_norm": 1.1524041891098022,
|
| 343 |
+
"learning_rate": 0.0001646082864756282,
|
| 344 |
+
"loss": 0.0879,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5960912052117264,
|
| 349 |
+
"grad_norm": 0.3662007451057434,
|
| 350 |
+
"learning_rate": 0.00016294878302145987,
|
| 351 |
+
"loss": 0.0421,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6286644951140063,
|
| 356 |
+
"grad_norm": 0.5205959677696228,
|
| 357 |
+
"learning_rate": 0.0001612600545193203,
|
| 358 |
+
"loss": 0.1127,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.6612377850162865,
|
| 363 |
+
"grad_norm": 0.5683305263519287,
|
| 364 |
+
"learning_rate": 0.00015954288499026782,
|
| 365 |
+
"loss": 0.0697,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6938110749185666,
|
| 370 |
+
"grad_norm": 1.1733250617980957,
|
| 371 |
+
"learning_rate": 0.0001577980716595934,
|
| 372 |
+
"loss": 0.0724,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7263843648208468,
|
| 377 |
+
"grad_norm": 3.809410810470581,
|
| 378 |
+
"learning_rate": 0.00015602642458669528,
|
| 379 |
+
"loss": 0.0993,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.758957654723127,
|
| 384 |
+
"grad_norm": 0.48794323205947876,
|
| 385 |
+
"learning_rate": 0.0001542287662889948,
|
| 386 |
+
"loss": 0.0576,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.791530944625407,
|
| 391 |
+
"grad_norm": 1.401726245880127,
|
| 392 |
+
"learning_rate": 0.00015240593136006897,
|
| 393 |
+
"loss": 0.1039,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8241042345276872,
|
| 398 |
+
"grad_norm": 0.33422133326530457,
|
| 399 |
+
"learning_rate": 0.0001505587660821759,
|
| 400 |
+
"loss": 0.0458,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.8566775244299674,
|
| 405 |
+
"grad_norm": 0.5557196140289307,
|
| 406 |
+
"learning_rate": 0.0001486881280333539,
|
| 407 |
+
"loss": 0.0489,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8892508143322475,
|
| 412 |
+
"grad_norm": 2.7507164478302,
|
| 413 |
+
"learning_rate": 0.00014679488568927616,
|
| 414 |
+
"loss": 0.0873,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.9218241042345277,
|
| 419 |
+
"grad_norm": 0.8317222595214844,
|
| 420 |
+
"learning_rate": 0.00014487991802004623,
|
| 421 |
+
"loss": 0.0402,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.9543973941368078,
|
| 426 |
+
"grad_norm": 2.6091113090515137,
|
| 427 |
+
"learning_rate": 0.0001429441140821209,
|
| 428 |
+
"loss": 0.0853,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.986970684039088,
|
| 433 |
+
"grad_norm": 0.29834985733032227,
|
| 434 |
+
"learning_rate": 0.00014098837260555084,
|
| 435 |
+
"loss": 0.0544,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.019543973941368,
|
| 440 |
+
"grad_norm": 0.7023916840553284,
|
| 441 |
+
"learning_rate": 0.0001390136015767295,
|
| 442 |
+
"loss": 0.0976,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.0521172638436482,
|
| 447 |
+
"grad_norm": 0.34739574790000916,
|
| 448 |
+
"learning_rate": 0.00013702071781684517,
|
| 449 |
+
"loss": 0.0612,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.0846905537459284,
|
| 454 |
+
"grad_norm": 0.43586787581443787,
|
| 455 |
+
"learning_rate": 0.00013501064655623094,
|
| 456 |
+
"loss": 0.0614,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.1172638436482085,
|
| 461 |
+
"grad_norm": 0.27919432520866394,
|
| 462 |
+
"learning_rate": 0.00013298432100481079,
|
| 463 |
+
"loss": 0.0539,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.1498371335504887,
|
| 468 |
+
"grad_norm": 0.5438246130943298,
|
| 469 |
+
"learning_rate": 0.0001309426819188409,
|
| 470 |
+
"loss": 0.0401,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.182410423452769,
|
| 475 |
+
"grad_norm": 0.6277980804443359,
|
| 476 |
+
"learning_rate": 0.0001288866771641474,
|
| 477 |
+
"loss": 0.0744,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.214983713355049,
|
| 482 |
+
"grad_norm": 0.9506388306617737,
|
| 483 |
+
"learning_rate": 0.00012681726127606376,
|
| 484 |
+
"loss": 0.073,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.247557003257329,
|
| 489 |
+
"grad_norm": 1.6029094457626343,
|
| 490 |
+
"learning_rate": 0.000124735395016271,
|
| 491 |
+
"loss": 0.0329,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.2801302931596092,
|
| 496 |
+
"grad_norm": 2.2171266078948975,
|
| 497 |
+
"learning_rate": 0.00012264204492674815,
|
| 498 |
+
"loss": 0.0885,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.3127035830618894,
|
| 503 |
+
"grad_norm": 1.0621176958084106,
|
| 504 |
+
"learning_rate": 0.0001205381828810382,
|
| 505 |
+
"loss": 0.0518,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.3452768729641695,
|
| 510 |
+
"grad_norm": 1.3911964893341064,
|
| 511 |
+
"learning_rate": 0.00011842478563303952,
|
| 512 |
+
"loss": 0.0694,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.3778501628664497,
|
| 517 |
+
"grad_norm": 0.586409330368042,
|
| 518 |
+
"learning_rate": 0.00011630283436353098,
|
| 519 |
+
"loss": 0.1295,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.41042345276873,
|
| 524 |
+
"grad_norm": 0.4753137230873108,
|
| 525 |
+
"learning_rate": 0.00011417331422464205,
|
| 526 |
+
"loss": 0.0438,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.44299674267101,
|
| 531 |
+
"grad_norm": 0.2599440813064575,
|
| 532 |
+
"learning_rate": 0.00011203721388247923,
|
| 533 |
+
"loss": 0.0446,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.47557003257329,
|
| 538 |
+
"grad_norm": 0.43588873744010925,
|
| 539 |
+
"learning_rate": 0.00010989552505812072,
|
| 540 |
+
"loss": 0.1583,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.5081433224755703,
|
| 545 |
+
"grad_norm": 0.36574748158454895,
|
| 546 |
+
"learning_rate": 0.0001077492420671931,
|
| 547 |
+
"loss": 0.0968,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.5407166123778504,
|
| 552 |
+
"grad_norm": 0.5202295780181885,
|
| 553 |
+
"learning_rate": 0.00010559936135824322,
|
| 554 |
+
"loss": 0.048,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.5732899022801305,
|
| 559 |
+
"grad_norm": 0.40310633182525635,
|
| 560 |
+
"learning_rate": 0.00010344688105012005,
|
| 561 |
+
"loss": 0.0422,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.6058631921824107,
|
| 566 |
+
"grad_norm": 0.2391999065876007,
|
| 567 |
+
"learning_rate": 0.00010129280046858086,
|
| 568 |
+
"loss": 0.092,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.6384364820846904,
|
| 573 |
+
"grad_norm": 0.390307754278183,
|
| 574 |
+
"learning_rate": 9.913811968233716e-05,
|
| 575 |
+
"loss": 0.0432,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.6710097719869705,
|
| 580 |
+
"grad_norm": 0.5276124477386475,
|
| 581 |
+
"learning_rate": 9.69838390387558e-05,
|
| 582 |
+
"loss": 0.0606,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.7035830618892507,
|
| 587 |
+
"grad_norm": 0.44748738408088684,
|
| 588 |
+
"learning_rate": 9.483095869943055e-05,
|
| 589 |
+
"loss": 0.0445,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.736156351791531,
|
| 594 |
+
"grad_norm": 1.4012900590896606,
|
| 595 |
+
"learning_rate": 9.268047817583998e-05,
|
| 596 |
+
"loss": 0.0372,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.768729641693811,
|
| 601 |
+
"grad_norm": 0.78389972448349,
|
| 602 |
+
"learning_rate": 9.053339586530723e-05,
|
| 603 |
+
"loss": 0.0623,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.801302931596091,
|
| 608 |
+
"grad_norm": 0.5246724486351013,
|
| 609 |
+
"learning_rate": 8.839070858747697e-05,
|
| 610 |
+
"loss": 0.0499,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.8338762214983713,
|
| 615 |
+
"grad_norm": 0.2494039684534073,
|
| 616 |
+
"learning_rate": 8.625341112152487e-05,
|
| 617 |
+
"loss": 0.0462,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.8664495114006514,
|
| 622 |
+
"grad_norm": 0.3887605667114258,
|
| 623 |
+
"learning_rate": 8.412249574431428e-05,
|
| 624 |
+
"loss": 0.0319,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.8990228013029316,
|
| 629 |
+
"grad_norm": 0.3090686500072479,
|
| 630 |
+
"learning_rate": 8.199895176971488e-05,
|
| 631 |
+
"loss": 0.0804,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.9315960912052117,
|
| 636 |
+
"grad_norm": 0.47201988101005554,
|
| 637 |
+
"learning_rate": 7.988376508929676e-05,
|
| 638 |
+
"loss": 0.0466,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.964169381107492,
|
| 643 |
+
"grad_norm": 0.4059586822986603,
|
| 644 |
+
"learning_rate": 7.777791771461332e-05,
|
| 645 |
+
"loss": 0.0337,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.996742671009772,
|
| 650 |
+
"grad_norm": 0.4746284782886505,
|
| 651 |
+
"learning_rate": 7.568238732128585e-05,
|
| 652 |
+
"loss": 0.0489,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.029315960912052,
|
| 657 |
+
"grad_norm": 0.24269038438796997,
|
| 658 |
+
"learning_rate": 7.359814679510065e-05,
|
| 659 |
+
"loss": 0.0408,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.0618892508143323,
|
| 664 |
+
"grad_norm": 0.6225023865699768,
|
| 665 |
+
"learning_rate": 7.152616378033042e-05,
|
| 666 |
+
"loss": 0.1245,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.0944625407166124,
|
| 671 |
+
"grad_norm": 0.3462587296962738,
|
| 672 |
+
"learning_rate": 6.94674002304887e-05,
|
| 673 |
+
"loss": 0.0369,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.1270358306188926,
|
| 678 |
+
"grad_norm": 1.3566789627075195,
|
| 679 |
+
"learning_rate": 6.742281196172663e-05,
|
| 680 |
+
"loss": 0.0499,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.1596091205211727,
|
| 685 |
+
"grad_norm": 0.45816901326179504,
|
| 686 |
+
"learning_rate": 6.539334820907888e-05,
|
| 687 |
+
"loss": 0.0368,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.192182410423453,
|
| 692 |
+
"grad_norm": 0.3949735760688782,
|
| 693 |
+
"learning_rate": 6.337995118576521e-05,
|
| 694 |
+
"loss": 0.0542,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.224755700325733,
|
| 699 |
+
"grad_norm": 0.7988837361335754,
|
| 700 |
+
"learning_rate": 6.138355564575169e-05,
|
| 701 |
+
"loss": 0.0415,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.257328990228013,
|
| 706 |
+
"grad_norm": 0.2154189944267273,
|
| 707 |
+
"learning_rate": 5.940508844977537e-05,
|
| 708 |
+
"loss": 0.0322,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.2899022801302933,
|
| 713 |
+
"grad_norm": 0.7597559690475464,
|
| 714 |
+
"learning_rate": 5.744546813503328e-05,
|
| 715 |
+
"loss": 0.0924,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.3224755700325734,
|
| 720 |
+
"grad_norm": 0.26576346158981323,
|
| 721 |
+
"learning_rate": 5.550560448873575e-05,
|
| 722 |
+
"loss": 0.0344,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.3550488599348536,
|
| 727 |
+
"grad_norm": 2.2215240001678467,
|
| 728 |
+
"learning_rate": 5.358639812572244e-05,
|
| 729 |
+
"loss": 0.0429,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.3876221498371337,
|
| 734 |
+
"grad_norm": 0.7066490650177002,
|
| 735 |
+
"learning_rate": 5.168874007033615e-05,
|
| 736 |
+
"loss": 0.0419,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.420195439739414,
|
| 741 |
+
"grad_norm": 0.23159299790859222,
|
| 742 |
+
"learning_rate": 4.9813511342749805e-05,
|
| 743 |
+
"loss": 0.0253,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.4527687296416936,
|
| 748 |
+
"grad_norm": 0.5614166259765625,
|
| 749 |
+
"learning_rate": 4.7961582549937675e-05,
|
| 750 |
+
"loss": 0.0264,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.4853420195439737,
|
| 755 |
+
"grad_norm": 0.4681215286254883,
|
| 756 |
+
"learning_rate": 4.6133813481481246e-05,
|
| 757 |
+
"loss": 0.0234,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.517915309446254,
|
| 762 |
+
"grad_norm": 0.3369932770729065,
|
| 763 |
+
"learning_rate": 4.433105271039721e-05,
|
| 764 |
+
"loss": 0.0441,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.550488599348534,
|
| 769 |
+
"grad_norm": 0.30148735642433167,
|
| 770 |
+
"learning_rate": 4.255413719917294e-05,
|
| 771 |
+
"loss": 0.0492,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.583061889250814,
|
| 776 |
+
"grad_norm": 0.3371269106864929,
|
| 777 |
+
"learning_rate": 4.080389191119241e-05,
|
| 778 |
+
"loss": 0.0325,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.6156351791530943,
|
| 783 |
+
"grad_norm": 0.8490313291549683,
|
| 784 |
+
"learning_rate": 3.9081129427732774e-05,
|
| 785 |
+
"loss": 0.0264,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6482084690553744,
|
| 790 |
+
"grad_norm": 0.30010277032852173,
|
| 791 |
+
"learning_rate": 3.7386649570709644e-05,
|
| 792 |
+
"loss": 0.0276,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6807817589576546,
|
| 797 |
+
"grad_norm": 0.2943625748157501,
|
| 798 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 799 |
+
"loss": 0.0945,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.7133550488599347,
|
| 804 |
+
"grad_norm": 0.19849584996700287,
|
| 805 |
+
"learning_rate": 3.408567100493787e-05,
|
| 806 |
+
"loss": 0.0543,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.745928338762215,
|
| 811 |
+
"grad_norm": 0.24954350292682648,
|
| 812 |
+
"learning_rate": 3.248070483188426e-05,
|
| 813 |
+
"loss": 0.0423,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.778501628664495,
|
| 818 |
+
"grad_norm": 0.425689697265625,
|
| 819 |
+
"learning_rate": 3.090708564515124e-05,
|
| 820 |
+
"loss": 0.0556,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.811074918566775,
|
| 825 |
+
"grad_norm": 0.18508583307266235,
|
| 826 |
+
"learning_rate": 2.936554402433087e-05,
|
| 827 |
+
"loss": 0.0217,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.8436482084690553,
|
| 832 |
+
"grad_norm": 4.1349196434021,
|
| 833 |
+
"learning_rate": 2.7856795656457257e-05,
|
| 834 |
+
"loss": 0.0729,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8762214983713354,
|
| 839 |
+
"grad_norm": 0.4305671155452728,
|
| 840 |
+
"learning_rate": 2.6381541003736486e-05,
|
| 841 |
+
"loss": 0.0327,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.9087947882736156,
|
| 846 |
+
"grad_norm": 0.16706599295139313,
|
| 847 |
+
"learning_rate": 2.494046497834518e-05,
|
| 848 |
+
"loss": 0.0202,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.9413680781758957,
|
| 853 |
+
"grad_norm": 0.40168455243110657,
|
| 854 |
+
"learning_rate": 2.3534236624448302e-05,
|
| 855 |
+
"loss": 0.0428,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.973941368078176,
|
| 860 |
+
"grad_norm": 0.33089303970336914,
|
| 861 |
+
"learning_rate": 2.2163508807583998e-05,
|
| 862 |
+
"loss": 0.0289,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.006514657980456,
|
| 867 |
+
"grad_norm": 0.1787663847208023,
|
| 868 |
+
"learning_rate": 2.082891791155954e-05,
|
| 869 |
+
"loss": 0.0152,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.039087947882736,
|
| 874 |
+
"grad_norm": 0.7322351932525635,
|
| 875 |
+
"learning_rate": 1.9531083542999317e-05,
|
| 876 |
+
"loss": 0.0303,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.071661237785016,
|
| 881 |
+
"grad_norm": 0.34002748131752014,
|
| 882 |
+
"learning_rate": 1.8270608243681953e-05,
|
| 883 |
+
"loss": 0.0144,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.1042345276872965,
|
| 888 |
+
"grad_norm": 0.2633480131626129,
|
| 889 |
+
"learning_rate": 1.7048077210799772e-05,
|
| 890 |
+
"loss": 0.0162,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.136807817589577,
|
| 895 |
+
"grad_norm": 0.32320234179496765,
|
| 896 |
+
"learning_rate": 1.5864058025271246e-05,
|
| 897 |
+
"loss": 0.0158,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.169381107491857,
|
| 902 |
+
"grad_norm": 0.2887153625488281,
|
| 903 |
+
"learning_rate": 1.47191003882317e-05,
|
| 904 |
+
"loss": 0.0275,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.201954397394137,
|
| 909 |
+
"grad_norm": 0.2007487416267395,
|
| 910 |
+
"learning_rate": 1.3613735865825305e-05,
|
| 911 |
+
"loss": 0.0192,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.234527687296417,
|
| 916 |
+
"grad_norm": 0.6144905090332031,
|
| 917 |
+
"learning_rate": 1.2548477642416256e-05,
|
| 918 |
+
"loss": 0.0225,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.267100977198697,
|
| 923 |
+
"grad_norm": 0.2809244692325592,
|
| 924 |
+
"learning_rate": 1.1523820282334219e-05,
|
| 925 |
+
"loss": 0.0201,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.299674267100977,
|
| 930 |
+
"grad_norm": 0.2203732132911682,
|
| 931 |
+
"learning_rate": 1.0540239500264516e-05,
|
| 932 |
+
"loss": 0.0213,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.3322475570032575,
|
| 937 |
+
"grad_norm": 0.17500793933868408,
|
| 938 |
+
"learning_rate": 9.598191940389256e-06,
|
| 939 |
+
"loss": 0.0193,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.364820846905538,
|
| 944 |
+
"grad_norm": 0.45830363035202026,
|
| 945 |
+
"learning_rate": 8.698114964382598e-06,
|
| 946 |
+
"loss": 0.0661,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.397394136807818,
|
| 951 |
+
"grad_norm": 0.34097039699554443,
|
| 952 |
+
"learning_rate": 7.840426448358085e-06,
|
| 953 |
+
"loss": 0.0155,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.429967426710098,
|
| 958 |
+
"grad_norm": 2.330317497253418,
|
| 959 |
+
"learning_rate": 7.025524588862542e-06,
|
| 960 |
+
"loss": 0.0291,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.462540716612378,
|
| 965 |
+
"grad_norm": 0.39467132091522217,
|
| 966 |
+
"learning_rate": 6.253787718006498e-06,
|
| 967 |
+
"loss": 0.0568,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.495114006514658,
|
| 972 |
+
"grad_norm": 0.28935298323631287,
|
| 973 |
+
"learning_rate": 5.525574127817046e-06,
|
| 974 |
+
"loss": 0.0302,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.527687296416938,
|
| 979 |
+
"grad_norm": 0.3630921542644501,
|
| 980 |
+
"learning_rate": 4.841221903894633e-06,
|
| 981 |
+
"loss": 0.027,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.5602605863192185,
|
| 986 |
+
"grad_norm": 0.1575985699892044,
|
| 987 |
+
"learning_rate": 4.20104876845111e-06,
|
| 988 |
+
"loss": 0.0208,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.592833876221499,
|
| 993 |
+
"grad_norm": 0.3565620481967926,
|
| 994 |
+
"learning_rate": 3.605351932801693e-06,
|
| 995 |
+
"loss": 0.0725,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.625407166123779,
|
| 1000 |
+
"grad_norm": 0.17438632249832153,
|
| 1001 |
+
"learning_rate": 3.0544079593795573e-06,
|
| 1002 |
+
"loss": 0.0167,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.657980456026059,
|
| 1007 |
+
"grad_norm": 0.15625396370887756,
|
| 1008 |
+
"learning_rate": 2.548472633337007e-06,
|
| 1009 |
+
"loss": 0.0162,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.690553745928339,
|
| 1014 |
+
"grad_norm": 0.21613460779190063,
|
| 1015 |
+
"learning_rate": 2.0877808437928637e-06,
|
| 1016 |
+
"loss": 0.02,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.723127035830619,
|
| 1021 |
+
"grad_norm": 0.25940829515457153,
|
| 1022 |
+
"learning_rate": 1.6725464747811447e-06,
|
| 1023 |
+
"loss": 0.0327,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.755700325732899,
|
| 1028 |
+
"grad_norm": 0.6820478439331055,
|
| 1029 |
+
"learning_rate": 1.3029623059517493e-06,
|
| 1030 |
+
"loss": 0.0182,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.7882736156351795,
|
| 1035 |
+
"grad_norm": 0.4143361747264862,
|
| 1036 |
+
"learning_rate": 9.791999230692629e-07,
|
| 1037 |
+
"loss": 0.0236,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.82084690553746,
|
| 1042 |
+
"grad_norm": 0.18389806151390076,
|
| 1043 |
+
"learning_rate": 7.014096383512802e-07,
|
| 1044 |
+
"loss": 0.0245,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.85342019543974,
|
| 1049 |
+
"grad_norm": 0.1316615343093872,
|
| 1050 |
+
"learning_rate": 4.6972042068341714e-07,
|
| 1051 |
+
"loss": 0.0162,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.88599348534202,
|
| 1056 |
+
"grad_norm": 0.7174280881881714,
|
| 1057 |
+
"learning_rate": 2.8423983574328295e-07,
|
| 1058 |
+
"loss": 0.0292,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.918566775244299,
|
| 1063 |
+
"grad_norm": 0.36342909932136536,
|
| 1064 |
+
"learning_rate": 1.4505399606130621e-07,
|
| 1065 |
+
"loss": 0.0269,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.95114006514658,
|
| 1070 |
+
"grad_norm": 0.12144973129034042,
|
| 1071 |
+
"learning_rate": 5.2227521041470216e-08,
|
| 1072 |
+
"loss": 0.0193,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.9837133550488595,
|
| 1077 |
+
"grad_norm": 0.2756155729293823,
|
| 1078 |
+
"learning_rate": 5.803506960722072e-09,
|
| 1079 |
+
"loss": 0.0375,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.0,
|
| 1084 |
+
"step": 1535,
|
| 1085 |
+
"total_flos": 5.197521391244208e+16,
|
| 1086 |
+
"train_loss": 0.10585675648980886,
|
| 1087 |
+
"train_runtime": 733.8736,
|
| 1088 |
+
"train_samples_per_second": 33.466,
|
| 1089 |
+
"train_steps_per_second": 2.092
|
| 1090 |
+
}
|
| 1091 |
+
],
|
| 1092 |
+
"logging_steps": 10,
|
| 1093 |
+
"max_steps": 1535,
|
| 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": 5.197521391244208e+16,
|
| 1110 |
+
"train_batch_size": 16,
|
| 1111 |
+
"trial_name": null,
|
| 1112 |
+
"trial_params": null
|
| 1113 |
+
}
|