Uploading fine-tuned LLaMA-EatFit-2-7b model
Browse files- trainer_state.json +1017 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1017 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3092,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008085381630012937,
|
| 13 |
+
"grad_norm": 0.3009789288043976,
|
| 14 |
+
"learning_rate": 0.00013440860215053763,
|
| 15 |
+
"loss": 1.7048,
|
| 16 |
+
"mean_token_accuracy": 0.661682380437851,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.016170763260025874,
|
| 21 |
+
"grad_norm": 0.4460005462169647,
|
| 22 |
+
"learning_rate": 0.00026881720430107527,
|
| 23 |
+
"loss": 1.2718,
|
| 24 |
+
"mean_token_accuracy": 0.713244981765747,
|
| 25 |
+
"step": 50
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.02425614489003881,
|
| 29 |
+
"grad_norm": 0.25601527094841003,
|
| 30 |
+
"learning_rate": 0.0004032258064516129,
|
| 31 |
+
"loss": 0.9995,
|
| 32 |
+
"mean_token_accuracy": 0.7553433787822723,
|
| 33 |
+
"step": 75
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.03234152652005175,
|
| 37 |
+
"grad_norm": 0.2627596855163574,
|
| 38 |
+
"learning_rate": 0.0004999932787358948,
|
| 39 |
+
"loss": 0.9036,
|
| 40 |
+
"mean_token_accuracy": 0.7764883422851563,
|
| 41 |
+
"step": 100
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.04042690815006468,
|
| 45 |
+
"grad_norm": 0.4143337905406952,
|
| 46 |
+
"learning_rate": 0.0004998595518201121,
|
| 47 |
+
"loss": 0.8985,
|
| 48 |
+
"mean_token_accuracy": 0.7727250349521637,
|
| 49 |
+
"step": 125
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.04851228978007762,
|
| 53 |
+
"grad_norm": 0.23922297358512878,
|
| 54 |
+
"learning_rate": 0.00049955446943686,
|
| 55 |
+
"loss": 0.7909,
|
| 56 |
+
"mean_token_accuracy": 0.7968144822120666,
|
| 57 |
+
"step": 150
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.056597671410090554,
|
| 61 |
+
"grad_norm": 0.2834464907646179,
|
| 62 |
+
"learning_rate": 0.0004990782408138185,
|
| 63 |
+
"loss": 0.822,
|
| 64 |
+
"mean_token_accuracy": 0.7941457486152649,
|
| 65 |
+
"step": 175
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0646830530401035,
|
| 69 |
+
"grad_norm": 0.29430675506591797,
|
| 70 |
+
"learning_rate": 0.000498431192551983,
|
| 71 |
+
"loss": 0.8546,
|
| 72 |
+
"mean_token_accuracy": 0.7836022305488587,
|
| 73 |
+
"step": 200
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07276843467011643,
|
| 77 |
+
"grad_norm": 0.23746679723262787,
|
| 78 |
+
"learning_rate": 0.0004976137684016788,
|
| 79 |
+
"loss": 0.7788,
|
| 80 |
+
"mean_token_accuracy": 0.8025749707221985,
|
| 81 |
+
"step": 225
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.08085381630012936,
|
| 85 |
+
"grad_norm": 0.24229322373867035,
|
| 86 |
+
"learning_rate": 0.0004966265289582338,
|
| 87 |
+
"loss": 0.8583,
|
| 88 |
+
"mean_token_accuracy": 0.7800547671318054,
|
| 89 |
+
"step": 250
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.08893919793014231,
|
| 93 |
+
"grad_norm": 0.24889522790908813,
|
| 94 |
+
"learning_rate": 0.0004954701512775184,
|
| 95 |
+
"loss": 0.685,
|
| 96 |
+
"mean_token_accuracy": 0.8196275687217712,
|
| 97 |
+
"step": 275
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.09702457956015524,
|
| 101 |
+
"grad_norm": 0.22737586498260498,
|
| 102 |
+
"learning_rate": 0.0004941454284116157,
|
| 103 |
+
"loss": 0.8068,
|
| 104 |
+
"mean_token_accuracy": 0.7950991153717041,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.10510996119016817,
|
| 109 |
+
"grad_norm": 0.2224387228488922,
|
| 110 |
+
"learning_rate": 0.0004926532688649407,
|
| 111 |
+
"loss": 0.8054,
|
| 112 |
+
"mean_token_accuracy": 0.7932506108283996,
|
| 113 |
+
"step": 325
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.11319534282018111,
|
| 117 |
+
"grad_norm": 0.2293468415737152,
|
| 118 |
+
"learning_rate": 0.0004909946959711816,
|
| 119 |
+
"loss": 0.7816,
|
| 120 |
+
"mean_token_accuracy": 0.7965369141101837,
|
| 121 |
+
"step": 350
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.12128072445019406,
|
| 125 |
+
"grad_norm": 0.2561907470226288,
|
| 126 |
+
"learning_rate": 0.0004891708471914897,
|
| 127 |
+
"loss": 0.7487,
|
| 128 |
+
"mean_token_accuracy": 0.8107043826580047,
|
| 129 |
+
"step": 375
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.129366106080207,
|
| 133 |
+
"grad_norm": 0.2583022117614746,
|
| 134 |
+
"learning_rate": 0.0004871829733344012,
|
| 135 |
+
"loss": 0.7786,
|
| 136 |
+
"mean_token_accuracy": 0.7999367988109589,
|
| 137 |
+
"step": 400
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.13745148771021992,
|
| 141 |
+
"grad_norm": 0.2550680339336395,
|
| 142 |
+
"learning_rate": 0.00048503243769802327,
|
| 143 |
+
"loss": 0.7576,
|
| 144 |
+
"mean_token_accuracy": 0.8000009500980377,
|
| 145 |
+
"step": 425
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.14553686934023286,
|
| 149 |
+
"grad_norm": 0.24982689321041107,
|
| 150 |
+
"learning_rate": 0.0004827207151350745,
|
| 151 |
+
"loss": 0.7897,
|
| 152 |
+
"mean_token_accuracy": 0.7945795893669129,
|
| 153 |
+
"step": 450
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.1536222509702458,
|
| 157 |
+
"grad_norm": 0.2784912884235382,
|
| 158 |
+
"learning_rate": 0.0004802493910414205,
|
| 159 |
+
"loss": 0.8013,
|
| 160 |
+
"mean_token_accuracy": 0.792938197851181,
|
| 161 |
+
"step": 475
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.16170763260025872,
|
| 165 |
+
"grad_norm": 0.2220473736524582,
|
| 166 |
+
"learning_rate": 0.00047762016026879807,
|
| 167 |
+
"loss": 0.711,
|
| 168 |
+
"mean_token_accuracy": 0.8124627935886383,
|
| 169 |
+
"step": 500
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.16979301423027165,
|
| 173 |
+
"grad_norm": 0.2841341197490692,
|
| 174 |
+
"learning_rate": 0.00047483482596247353,
|
| 175 |
+
"loss": 0.7382,
|
| 176 |
+
"mean_token_accuracy": 0.8129073297977447,
|
| 177 |
+
"step": 525
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17787839586028462,
|
| 181 |
+
"grad_norm": 0.23232664167881012,
|
| 182 |
+
"learning_rate": 0.00047189529832463296,
|
| 183 |
+
"loss": 0.7588,
|
| 184 |
+
"mean_token_accuracy": 0.8059421420097351,
|
| 185 |
+
"step": 550
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.18596377749029755,
|
| 189 |
+
"grad_norm": 0.22178468108177185,
|
| 190 |
+
"learning_rate": 0.00046880359330435216,
|
| 191 |
+
"loss": 0.8146,
|
| 192 |
+
"mean_token_accuracy": 0.7915572142601013,
|
| 193 |
+
"step": 575
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.19404915912031048,
|
| 197 |
+
"grad_norm": 0.23006677627563477,
|
| 198 |
+
"learning_rate": 0.0004655618312150437,
|
| 199 |
+
"loss": 0.7822,
|
| 200 |
+
"mean_token_accuracy": 0.7942324769496918,
|
| 201 |
+
"step": 600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.20213454075032342,
|
| 205 |
+
"grad_norm": 0.32772719860076904,
|
| 206 |
+
"learning_rate": 0.00046217223528033146,
|
| 207 |
+
"loss": 0.7632,
|
| 208 |
+
"mean_token_accuracy": 0.8043809008598327,
|
| 209 |
+
"step": 625
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.21021992238033635,
|
| 213 |
+
"grad_norm": 0.18335266411304474,
|
| 214 |
+
"learning_rate": 0.0004586371301093476,
|
| 215 |
+
"loss": 0.7218,
|
| 216 |
+
"mean_token_accuracy": 0.816128705739975,
|
| 217 |
+
"step": 650
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.21830530401034928,
|
| 221 |
+
"grad_norm": 0.22064125537872314,
|
| 222 |
+
"learning_rate": 0.00045495894010249915,
|
| 223 |
+
"loss": 0.7364,
|
| 224 |
+
"mean_token_accuracy": 0.8087958478927613,
|
| 225 |
+
"step": 675
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.22639068564036222,
|
| 229 |
+
"grad_norm": 0.21018439531326294,
|
| 230 |
+
"learning_rate": 0.0004511401877887967,
|
| 231 |
+
"loss": 0.738,
|
| 232 |
+
"mean_token_accuracy": 0.8069515633583069,
|
| 233 |
+
"step": 700
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.23447606727037515,
|
| 237 |
+
"grad_norm": 0.21225683391094208,
|
| 238 |
+
"learning_rate": 0.0004471834920958864,
|
| 239 |
+
"loss": 0.7397,
|
| 240 |
+
"mean_token_accuracy": 0.8051351046562195,
|
| 241 |
+
"step": 725
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2425614489003881,
|
| 245 |
+
"grad_norm": 0.22577480971813202,
|
| 246 |
+
"learning_rate": 0.00044309156655397003,
|
| 247 |
+
"loss": 0.7872,
|
| 248 |
+
"mean_token_accuracy": 0.7951467871665955,
|
| 249 |
+
"step": 750
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.25064683053040104,
|
| 253 |
+
"grad_norm": 0.19751884043216705,
|
| 254 |
+
"learning_rate": 0.000438867217434847,
|
| 255 |
+
"loss": 0.7147,
|
| 256 |
+
"mean_token_accuracy": 0.8134795188903808,
|
| 257 |
+
"step": 775
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.258732212160414,
|
| 261 |
+
"grad_norm": 0.2011658251285553,
|
| 262 |
+
"learning_rate": 0.0004345133418273529,
|
| 263 |
+
"loss": 0.7923,
|
| 264 |
+
"mean_token_accuracy": 0.7959077858924866,
|
| 265 |
+
"step": 800
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.2668175937904269,
|
| 269 |
+
"grad_norm": 0.2174764722585678,
|
| 270 |
+
"learning_rate": 0.00043003292565051544,
|
| 271 |
+
"loss": 0.7576,
|
| 272 |
+
"mean_token_accuracy": 0.8044932758808137,
|
| 273 |
+
"step": 825
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.27490297542043984,
|
| 277 |
+
"grad_norm": 0.20990514755249023,
|
| 278 |
+
"learning_rate": 0.0004254290416057898,
|
| 279 |
+
"loss": 0.739,
|
| 280 |
+
"mean_token_accuracy": 0.8073045027256012,
|
| 281 |
+
"step": 850
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.2829883570504528,
|
| 285 |
+
"grad_norm": 0.1976221799850464,
|
| 286 |
+
"learning_rate": 0.0004207048470697777,
|
| 287 |
+
"loss": 0.6717,
|
| 288 |
+
"mean_token_accuracy": 0.824974125623703,
|
| 289 |
+
"step": 875
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2910737386804657,
|
| 293 |
+
"grad_norm": 0.2552309036254883,
|
| 294 |
+
"learning_rate": 0.0004158635819288762,
|
| 295 |
+
"loss": 0.7311,
|
| 296 |
+
"mean_token_accuracy": 0.8078971183300019,
|
| 297 |
+
"step": 900
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.29915912031047864,
|
| 301 |
+
"grad_norm": 0.26244303584098816,
|
| 302 |
+
"learning_rate": 0.00041090856635734067,
|
| 303 |
+
"loss": 0.7264,
|
| 304 |
+
"mean_token_accuracy": 0.8127052938938141,
|
| 305 |
+
"step": 925
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.3072445019404916,
|
| 309 |
+
"grad_norm": 0.24333028495311737,
|
| 310 |
+
"learning_rate": 0.000405843198540285,
|
| 311 |
+
"loss": 0.7184,
|
| 312 |
+
"mean_token_accuracy": 0.8114526355266571,
|
| 313 |
+
"step": 950
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.3153298835705045,
|
| 317 |
+
"grad_norm": 0.19133129715919495,
|
| 318 |
+
"learning_rate": 0.0004006709523431822,
|
| 319 |
+
"loss": 0.7538,
|
| 320 |
+
"mean_token_accuracy": 0.8016650295257568,
|
| 321 |
+
"step": 975
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.32341526520051744,
|
| 325 |
+
"grad_norm": 0.25047338008880615,
|
| 326 |
+
"learning_rate": 0.00039539537492946285,
|
| 327 |
+
"loss": 0.8019,
|
| 328 |
+
"mean_token_accuracy": 0.7935136258602142,
|
| 329 |
+
"step": 1000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.3315006468305304,
|
| 333 |
+
"grad_norm": 0.19724124670028687,
|
| 334 |
+
"learning_rate": 0.0003900200843278449,
|
| 335 |
+
"loss": 0.6892,
|
| 336 |
+
"mean_token_accuracy": 0.8166925406455994,
|
| 337 |
+
"step": 1025
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.3395860284605433,
|
| 341 |
+
"grad_norm": 0.21111617982387543,
|
| 342 |
+
"learning_rate": 0.0003845487669510631,
|
| 343 |
+
"loss": 0.7281,
|
| 344 |
+
"mean_token_accuracy": 0.8110716784000397,
|
| 345 |
+
"step": 1050
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3476714100905563,
|
| 349 |
+
"grad_norm": 0.2487187534570694,
|
| 350 |
+
"learning_rate": 0.00037898517506770196,
|
| 351 |
+
"loss": 0.7962,
|
| 352 |
+
"mean_token_accuracy": 0.7921491277217865,
|
| 353 |
+
"step": 1075
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.35575679172056923,
|
| 357 |
+
"grad_norm": 0.2757723927497864,
|
| 358 |
+
"learning_rate": 0.0003733331242288622,
|
| 359 |
+
"loss": 0.7533,
|
| 360 |
+
"mean_token_accuracy": 0.8056223785877228,
|
| 361 |
+
"step": 1100
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.36384217335058217,
|
| 365 |
+
"grad_norm": 0.27630847692489624,
|
| 366 |
+
"learning_rate": 0.0003675964906514289,
|
| 367 |
+
"loss": 0.7885,
|
| 368 |
+
"mean_token_accuracy": 0.7973137283325196,
|
| 369 |
+
"step": 1125
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.3719275549805951,
|
| 373 |
+
"grad_norm": 0.2365068644285202,
|
| 374 |
+
"learning_rate": 0.00036177920855973405,
|
| 375 |
+
"loss": 0.7275,
|
| 376 |
+
"mean_token_accuracy": 0.8077067303657531,
|
| 377 |
+
"step": 1150
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.38001293661060803,
|
| 381 |
+
"grad_norm": 0.2511255443096161,
|
| 382 |
+
"learning_rate": 0.00035588526748743754,
|
| 383 |
+
"loss": 0.8052,
|
| 384 |
+
"mean_token_accuracy": 0.7931141972541809,
|
| 385 |
+
"step": 1175
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.38809831824062097,
|
| 389 |
+
"grad_norm": 0.2489156723022461,
|
| 390 |
+
"learning_rate": 0.0003499187095414763,
|
| 391 |
+
"loss": 0.7369,
|
| 392 |
+
"mean_token_accuracy": 0.8076127851009369,
|
| 393 |
+
"step": 1200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.3961836998706339,
|
| 397 |
+
"grad_norm": 0.23067978024482727,
|
| 398 |
+
"learning_rate": 0.00034388362662995855,
|
| 399 |
+
"loss": 0.7393,
|
| 400 |
+
"mean_token_accuracy": 0.8086310243606567,
|
| 401 |
+
"step": 1225
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.40426908150064683,
|
| 405 |
+
"grad_norm": 0.2202920764684677,
|
| 406 |
+
"learning_rate": 0.000337784157655904,
|
| 407 |
+
"loss": 0.7578,
|
| 408 |
+
"mean_token_accuracy": 0.8064273929595948,
|
| 409 |
+
"step": 1250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.41235446313065977,
|
| 413 |
+
"grad_norm": 0.21269731223583221,
|
| 414 |
+
"learning_rate": 0.0003316244856787544,
|
| 415 |
+
"loss": 0.7937,
|
| 416 |
+
"mean_token_accuracy": 0.795601452589035,
|
| 417 |
+
"step": 1275
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.4204398447606727,
|
| 421 |
+
"grad_norm": 0.24461720883846283,
|
| 422 |
+
"learning_rate": 0.0003254088350456017,
|
| 423 |
+
"loss": 0.6656,
|
| 424 |
+
"mean_token_accuracy": 0.8226878666877746,
|
| 425 |
+
"step": 1300
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.42852522639068563,
|
| 429 |
+
"grad_norm": 0.24032443761825562,
|
| 430 |
+
"learning_rate": 0.0003191414684941003,
|
| 431 |
+
"loss": 0.7778,
|
| 432 |
+
"mean_token_accuracy": 0.7960509729385375,
|
| 433 |
+
"step": 1325
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.43661060802069857,
|
| 437 |
+
"grad_norm": 0.25362804532051086,
|
| 438 |
+
"learning_rate": 0.0003128266842290513,
|
| 439 |
+
"loss": 0.6967,
|
| 440 |
+
"mean_token_accuracy": 0.8132575881481171,
|
| 441 |
+
"step": 1350
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4446959896507115,
|
| 445 |
+
"grad_norm": 0.2388894110918045,
|
| 446 |
+
"learning_rate": 0.0003064688129746629,
|
| 447 |
+
"loss": 0.716,
|
| 448 |
+
"mean_token_accuracy": 0.8112483811378479,
|
| 449 |
+
"step": 1375
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.45278137128072443,
|
| 453 |
+
"grad_norm": 0.2121650129556656,
|
| 454 |
+
"learning_rate": 0.0003000722150045085,
|
| 455 |
+
"loss": 0.6942,
|
| 456 |
+
"mean_token_accuracy": 0.8156200110912323,
|
| 457 |
+
"step": 1400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.46086675291073737,
|
| 461 |
+
"grad_norm": 0.2317550778388977,
|
| 462 |
+
"learning_rate": 0.0002936412771512206,
|
| 463 |
+
"loss": 0.7493,
|
| 464 |
+
"mean_token_accuracy": 0.8051575112342835,
|
| 465 |
+
"step": 1425
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.4689521345407503,
|
| 469 |
+
"grad_norm": 0.23476050794124603,
|
| 470 |
+
"learning_rate": 0.0002871804097979687,
|
| 471 |
+
"loss": 0.7136,
|
| 472 |
+
"mean_token_accuracy": 0.8104170382022857,
|
| 473 |
+
"step": 1450
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4770375161707633,
|
| 477 |
+
"grad_norm": 0.2009560465812683,
|
| 478 |
+
"learning_rate": 0.00028069404385378736,
|
| 479 |
+
"loss": 0.7117,
|
| 480 |
+
"mean_token_accuracy": 0.8178813803195953,
|
| 481 |
+
"step": 1475
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.4851228978007762,
|
| 485 |
+
"grad_norm": 0.21633079648017883,
|
| 486 |
+
"learning_rate": 0.0002741866277148276,
|
| 487 |
+
"loss": 0.7392,
|
| 488 |
+
"mean_token_accuracy": 0.8080459308624267,
|
| 489 |
+
"step": 1500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.49320827943078915,
|
| 493 |
+
"grad_norm": 0.2863864004611969,
|
| 494 |
+
"learning_rate": 0.00026766262421361407,
|
| 495 |
+
"loss": 0.7429,
|
| 496 |
+
"mean_token_accuracy": 0.8051086151599884,
|
| 497 |
+
"step": 1525
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.5012936610608021,
|
| 501 |
+
"grad_norm": 0.2746196389198303,
|
| 502 |
+
"learning_rate": 0.0002611265075584034,
|
| 503 |
+
"loss": 0.7378,
|
| 504 |
+
"mean_token_accuracy": 0.8071331679821014,
|
| 505 |
+
"step": 1550
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.509379042690815,
|
| 509 |
+
"grad_norm": 0.1745605319738388,
|
| 510 |
+
"learning_rate": 0.0002545827602647397,
|
| 511 |
+
"loss": 0.7329,
|
| 512 |
+
"mean_token_accuracy": 0.8071370398998261,
|
| 513 |
+
"step": 1575
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.517464424320828,
|
| 517 |
+
"grad_norm": 0.2692930996417999,
|
| 518 |
+
"learning_rate": 0.0002480358700813135,
|
| 519 |
+
"loss": 0.6469,
|
| 520 |
+
"mean_token_accuracy": 0.8244569575786591,
|
| 521 |
+
"step": 1600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5255498059508409,
|
| 525 |
+
"grad_norm": 0.2551010549068451,
|
| 526 |
+
"learning_rate": 0.00024149032691223173,
|
| 527 |
+
"loss": 0.7093,
|
| 528 |
+
"mean_token_accuracy": 0.8107299065589905,
|
| 529 |
+
"step": 1625
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.5336351875808538,
|
| 533 |
+
"grad_norm": 0.19366180896759033,
|
| 534 |
+
"learning_rate": 0.0002349506197378092,
|
| 535 |
+
"loss": 0.7518,
|
| 536 |
+
"mean_token_accuracy": 0.8024619662761688,
|
| 537 |
+
"step": 1650
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.5417205692108668,
|
| 541 |
+
"grad_norm": 0.22391283512115479,
|
| 542 |
+
"learning_rate": 0.00022842123353599369,
|
| 543 |
+
"loss": 0.6933,
|
| 544 |
+
"mean_token_accuracy": 0.8174584257602692,
|
| 545 |
+
"step": 1675
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.5498059508408797,
|
| 549 |
+
"grad_norm": 0.24723011255264282,
|
| 550 |
+
"learning_rate": 0.0002219066462065364,
|
| 551 |
+
"loss": 0.669,
|
| 552 |
+
"mean_token_accuracy": 0.822363510131836,
|
| 553 |
+
"step": 1700
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.5578913324708926,
|
| 557 |
+
"grad_norm": 0.20229819416999817,
|
| 558 |
+
"learning_rate": 0.00021541132550001584,
|
| 559 |
+
"loss": 0.6508,
|
| 560 |
+
"mean_token_accuracy": 0.8255820453166962,
|
| 561 |
+
"step": 1725
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5659767141009056,
|
| 565 |
+
"grad_norm": 0.2818906605243683,
|
| 566 |
+
"learning_rate": 0.00020893972595382274,
|
| 567 |
+
"loss": 0.6417,
|
| 568 |
+
"mean_token_accuracy": 0.830688863992691,
|
| 569 |
+
"step": 1750
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5740620957309185,
|
| 573 |
+
"grad_norm": 0.24256223440170288,
|
| 574 |
+
"learning_rate": 0.00020249628583720672,
|
| 575 |
+
"loss": 0.7353,
|
| 576 |
+
"mean_token_accuracy": 0.8104202616214752,
|
| 577 |
+
"step": 1775
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.5821474773609314,
|
| 581 |
+
"grad_norm": 0.24385611712932587,
|
| 582 |
+
"learning_rate": 0.00019608542410747888,
|
| 583 |
+
"loss": 0.6876,
|
| 584 |
+
"mean_token_accuracy": 0.819042581319809,
|
| 585 |
+
"step": 1800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.5902328589909444,
|
| 589 |
+
"grad_norm": 0.15913242101669312,
|
| 590 |
+
"learning_rate": 0.00018971153737945968,
|
| 591 |
+
"loss": 0.646,
|
| 592 |
+
"mean_token_accuracy": 0.8284247839450836,
|
| 593 |
+
"step": 1825
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5983182406209573,
|
| 597 |
+
"grad_norm": 0.17542122304439545,
|
| 598 |
+
"learning_rate": 0.00018337899691024914,
|
| 599 |
+
"loss": 0.6216,
|
| 600 |
+
"mean_token_accuracy": 0.83616614818573,
|
| 601 |
+
"step": 1850
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.6064036222509702,
|
| 605 |
+
"grad_norm": 0.22214658558368683,
|
| 606 |
+
"learning_rate": 0.0001770921456013872,
|
| 607 |
+
"loss": 0.6947,
|
| 608 |
+
"mean_token_accuracy": 0.8170740747451782,
|
| 609 |
+
"step": 1875
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.6144890038809832,
|
| 613 |
+
"grad_norm": 0.22132755815982819,
|
| 614 |
+
"learning_rate": 0.00017085529502046073,
|
| 615 |
+
"loss": 0.6788,
|
| 616 |
+
"mean_token_accuracy": 0.820680763721466,
|
| 617 |
+
"step": 1900
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.6225743855109961,
|
| 621 |
+
"grad_norm": 0.26897531747817993,
|
| 622 |
+
"learning_rate": 0.00016467272244420029,
|
| 623 |
+
"loss": 0.6833,
|
| 624 |
+
"mean_token_accuracy": 0.822064242362976,
|
| 625 |
+
"step": 1925
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.630659767141009,
|
| 629 |
+
"grad_norm": 0.25688934326171875,
|
| 630 |
+
"learning_rate": 0.0001585486679250922,
|
| 631 |
+
"loss": 0.6945,
|
| 632 |
+
"mean_token_accuracy": 0.8143122482299805,
|
| 633 |
+
"step": 1950
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.638745148771022,
|
| 637 |
+
"grad_norm": 0.21207553148269653,
|
| 638 |
+
"learning_rate": 0.0001524873313835208,
|
| 639 |
+
"loss": 0.6596,
|
| 640 |
+
"mean_token_accuracy": 0.8273860597610474,
|
| 641 |
+
"step": 1975
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.6468305304010349,
|
| 645 |
+
"grad_norm": 0.281393438577652,
|
| 646 |
+
"learning_rate": 0.00014649286972743319,
|
| 647 |
+
"loss": 0.6767,
|
| 648 |
+
"mean_token_accuracy": 0.8178416419029236,
|
| 649 |
+
"step": 2000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.6549159120310478,
|
| 653 |
+
"grad_norm": 0.27408191561698914,
|
| 654 |
+
"learning_rate": 0.00014056939400150143,
|
| 655 |
+
"loss": 0.6974,
|
| 656 |
+
"mean_token_accuracy": 0.8189209842681885,
|
| 657 |
+
"step": 2025
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.6630012936610608,
|
| 661 |
+
"grad_norm": 0.26886942982673645,
|
| 662 |
+
"learning_rate": 0.00013472096656773913,
|
| 663 |
+
"loss": 0.6497,
|
| 664 |
+
"mean_token_accuracy": 0.8288757252693176,
|
| 665 |
+
"step": 2050
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.6710866752910737,
|
| 669 |
+
"grad_norm": 0.21919454634189606,
|
| 670 |
+
"learning_rate": 0.00012895159831950462,
|
| 671 |
+
"loss": 0.693,
|
| 672 |
+
"mean_token_accuracy": 0.8163833570480347,
|
| 673 |
+
"step": 2075
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.6791720569210866,
|
| 677 |
+
"grad_norm": 0.21283280849456787,
|
| 678 |
+
"learning_rate": 0.0001232652459308012,
|
| 679 |
+
"loss": 0.7117,
|
| 680 |
+
"mean_token_accuracy": 0.808628898859024,
|
| 681 |
+
"step": 2100
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6872574385510997,
|
| 685 |
+
"grad_norm": 0.229765385389328,
|
| 686 |
+
"learning_rate": 0.00011766580914276209,
|
| 687 |
+
"loss": 0.7317,
|
| 688 |
+
"mean_token_accuracy": 0.8035627353191376,
|
| 689 |
+
"step": 2125
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6953428201811126,
|
| 693 |
+
"grad_norm": 0.21411098539829254,
|
| 694 |
+
"learning_rate": 0.00011215712808918003,
|
| 695 |
+
"loss": 0.6469,
|
| 696 |
+
"mean_token_accuracy": 0.8277445828914642,
|
| 697 |
+
"step": 2150
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.7034282018111255,
|
| 701 |
+
"grad_norm": 0.2254790961742401,
|
| 702 |
+
"learning_rate": 0.00010674298066291601,
|
| 703 |
+
"loss": 0.6976,
|
| 704 |
+
"mean_token_accuracy": 0.8171502375602722,
|
| 705 |
+
"step": 2175
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.7115135834411385,
|
| 709 |
+
"grad_norm": 0.27148380875587463,
|
| 710 |
+
"learning_rate": 0.0001014270799249933,
|
| 711 |
+
"loss": 0.717,
|
| 712 |
+
"mean_token_accuracy": 0.8086051964759826,
|
| 713 |
+
"step": 2200
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.7195989650711514,
|
| 717 |
+
"grad_norm": 0.2047407031059265,
|
| 718 |
+
"learning_rate": 9.621307155815398e-05,
|
| 719 |
+
"loss": 0.718,
|
| 720 |
+
"mean_token_accuracy": 0.8121638822555542,
|
| 721 |
+
"step": 2225
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.7276843467011643,
|
| 725 |
+
"grad_norm": 0.22144050896167755,
|
| 726 |
+
"learning_rate": 9.11045313666231e-05,
|
| 727 |
+
"loss": 0.6623,
|
| 728 |
+
"mean_token_accuracy": 0.8254709720611573,
|
| 729 |
+
"step": 2250
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.7357697283311773,
|
| 733 |
+
"grad_norm": 0.27873218059539795,
|
| 734 |
+
"learning_rate": 8.610496282379687e-05,
|
| 735 |
+
"loss": 0.7073,
|
| 736 |
+
"mean_token_accuracy": 0.8168034076690673,
|
| 737 |
+
"step": 2275
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7438551099611902,
|
| 741 |
+
"grad_norm": 0.25058674812316895,
|
| 742 |
+
"learning_rate": 8.121779466953572e-05,
|
| 743 |
+
"loss": 0.7961,
|
| 744 |
+
"mean_token_accuracy": 0.7905523943901062,
|
| 745 |
+
"step": 2300
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.7519404915912031,
|
| 749 |
+
"grad_norm": 0.2540716826915741,
|
| 750 |
+
"learning_rate": 7.644637855870959e-05,
|
| 751 |
+
"loss": 0.7561,
|
| 752 |
+
"mean_token_accuracy": 0.8036962306499481,
|
| 753 |
+
"step": 2325
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.7600258732212161,
|
| 757 |
+
"grad_norm": 0.2069474756717682,
|
| 758 |
+
"learning_rate": 7.179398676260923e-05,
|
| 759 |
+
"loss": 0.7163,
|
| 760 |
+
"mean_token_accuracy": 0.8117474913597107,
|
| 761 |
+
"step": 2350
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.768111254851229,
|
| 765 |
+
"grad_norm": 0.23127734661102295,
|
| 766 |
+
"learning_rate": 6.726380992479941e-05,
|
| 767 |
+
"loss": 0.6983,
|
| 768 |
+
"mean_token_accuracy": 0.8151715826988221,
|
| 769 |
+
"step": 2375
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.7761966364812419,
|
| 773 |
+
"grad_norm": 0.19075877964496613,
|
| 774 |
+
"learning_rate": 6.285895487295229e-05,
|
| 775 |
+
"loss": 0.6644,
|
| 776 |
+
"mean_token_accuracy": 0.8226857626438141,
|
| 777 |
+
"step": 2400
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.7842820181112549,
|
| 781 |
+
"grad_norm": 0.26920101046562195,
|
| 782 |
+
"learning_rate": 5.858244248816302e-05,
|
| 783 |
+
"loss": 0.678,
|
| 784 |
+
"mean_token_accuracy": 0.8184169673919678,
|
| 785 |
+
"step": 2425
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.7923673997412678,
|
| 789 |
+
"grad_norm": 0.22957506775856018,
|
| 790 |
+
"learning_rate": 5.443720563320792e-05,
|
| 791 |
+
"loss": 0.7125,
|
| 792 |
+
"mean_token_accuracy": 0.8165527045726776,
|
| 793 |
+
"step": 2450
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.8004527813712807,
|
| 797 |
+
"grad_norm": 0.23255349695682526,
|
| 798 |
+
"learning_rate": 5.042608714116612e-05,
|
| 799 |
+
"loss": 0.6648,
|
| 800 |
+
"mean_token_accuracy": 0.8205063927173615,
|
| 801 |
+
"step": 2475
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.8085381630012937,
|
| 805 |
+
"grad_norm": 0.21010981500148773,
|
| 806 |
+
"learning_rate": 4.655183786578426e-05,
|
| 807 |
+
"loss": 0.6833,
|
| 808 |
+
"mean_token_accuracy": 0.8196286606788635,
|
| 809 |
+
"step": 2500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.8166235446313066,
|
| 813 |
+
"grad_norm": 0.21943055093288422,
|
| 814 |
+
"learning_rate": 4.2817114794921677e-05,
|
| 815 |
+
"loss": 0.6897,
|
| 816 |
+
"mean_token_accuracy": 0.8164256310462952,
|
| 817 |
+
"step": 2525
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.8247089262613195,
|
| 821 |
+
"grad_norm": 0.21263104677200317,
|
| 822 |
+
"learning_rate": 3.92244792283685e-05,
|
| 823 |
+
"loss": 0.6553,
|
| 824 |
+
"mean_token_accuracy": 0.8244921159744263,
|
| 825 |
+
"step": 2550
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.8327943078913325,
|
| 829 |
+
"grad_norm": 0.30405542254447937,
|
| 830 |
+
"learning_rate": 3.577639502128843e-05,
|
| 831 |
+
"loss": 0.6771,
|
| 832 |
+
"mean_token_accuracy": 0.8220798122882843,
|
| 833 |
+
"step": 2575
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.8408796895213454,
|
| 837 |
+
"grad_norm": 0.2544702887535095,
|
| 838 |
+
"learning_rate": 3.247522689448923e-05,
|
| 839 |
+
"loss": 0.6825,
|
| 840 |
+
"mean_token_accuracy": 0.8175348448753357,
|
| 841 |
+
"step": 2600
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.8489650711513583,
|
| 845 |
+
"grad_norm": 0.2168809473514557,
|
| 846 |
+
"learning_rate": 2.9323238812679982e-05,
|
| 847 |
+
"loss": 0.7437,
|
| 848 |
+
"mean_token_accuracy": 0.8062794303894043,
|
| 849 |
+
"step": 2625
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8570504527813713,
|
| 853 |
+
"grad_norm": 0.32856041193008423,
|
| 854 |
+
"learning_rate": 2.6322592431828136e-05,
|
| 855 |
+
"loss": 0.6968,
|
| 856 |
+
"mean_token_accuracy": 0.8184567129611969,
|
| 857 |
+
"step": 2650
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.8651358344113842,
|
| 861 |
+
"grad_norm": 0.2495715171098709,
|
| 862 |
+
"learning_rate": 2.3475345616680327e-05,
|
| 863 |
+
"loss": 0.6864,
|
| 864 |
+
"mean_token_accuracy": 0.817262338399887,
|
| 865 |
+
"step": 2675
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.8732212160413971,
|
| 869 |
+
"grad_norm": 0.21743454039096832,
|
| 870 |
+
"learning_rate": 2.0783451029463995e-05,
|
| 871 |
+
"loss": 0.6554,
|
| 872 |
+
"mean_token_accuracy": 0.8261248970031738,
|
| 873 |
+
"step": 2700
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.8813065976714101,
|
| 877 |
+
"grad_norm": 0.19413378834724426,
|
| 878 |
+
"learning_rate": 1.8248754790737733e-05,
|
| 879 |
+
"loss": 0.691,
|
| 880 |
+
"mean_token_accuracy": 0.8169219958782196,
|
| 881 |
+
"step": 2725
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.889391979301423,
|
| 885 |
+
"grad_norm": 0.2755376100540161,
|
| 886 |
+
"learning_rate": 1.5872995213308566e-05,
|
| 887 |
+
"loss": 0.6868,
|
| 888 |
+
"mean_token_accuracy": 0.8156666767597198,
|
| 889 |
+
"step": 2750
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.8974773609314359,
|
| 893 |
+
"grad_norm": 0.23579077422618866,
|
| 894 |
+
"learning_rate": 1.3657801610084563e-05,
|
| 895 |
+
"loss": 0.669,
|
| 896 |
+
"mean_token_accuracy": 0.8254081463813782,
|
| 897 |
+
"step": 2775
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.9055627425614489,
|
| 901 |
+
"grad_norm": 0.21246632933616638,
|
| 902 |
+
"learning_rate": 1.1604693176680392e-05,
|
| 903 |
+
"loss": 0.6519,
|
| 904 |
+
"mean_token_accuracy": 0.8244655966758728,
|
| 905 |
+
"step": 2800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.9136481241914618,
|
| 909 |
+
"grad_norm": 0.29115888476371765,
|
| 910 |
+
"learning_rate": 9.715077949542184e-06,
|
| 911 |
+
"loss": 0.6673,
|
| 912 |
+
"mean_token_accuracy": 0.825350991487503,
|
| 913 |
+
"step": 2825
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.9217335058214747,
|
| 917 |
+
"grad_norm": 0.24020685255527496,
|
| 918 |
+
"learning_rate": 7.990251840305996e-06,
|
| 919 |
+
"loss": 0.6349,
|
| 920 |
+
"mean_token_accuracy": 0.8334643471240998,
|
| 921 |
+
"step": 2850
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.9298188874514877,
|
| 925 |
+
"grad_norm": 0.2604895532131195,
|
| 926 |
+
"learning_rate": 6.431397747052342e-06,
|
| 927 |
+
"loss": 0.6659,
|
| 928 |
+
"mean_token_accuracy": 0.8195065236091614,
|
| 929 |
+
"step": 2875
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.9379042690815006,
|
| 933 |
+
"grad_norm": 0.19995881617069244,
|
| 934 |
+
"learning_rate": 5.039584743066344e-06,
|
| 935 |
+
"loss": 0.706,
|
| 936 |
+
"mean_token_accuracy": 0.8151924252510071,
|
| 937 |
+
"step": 2900
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.9459896507115135,
|
| 941 |
+
"grad_norm": 0.22976571321487427,
|
| 942 |
+
"learning_rate": 3.815767343659377e-06,
|
| 943 |
+
"loss": 0.6477,
|
| 944 |
+
"mean_token_accuracy": 0.8256394731998443,
|
| 945 |
+
"step": 2925
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.9540750323415266,
|
| 949 |
+
"grad_norm": 0.22688381373882294,
|
| 950 |
+
"learning_rate": 2.760784851555953e-06,
|
| 951 |
+
"loss": 0.672,
|
| 952 |
+
"mean_token_accuracy": 0.8214276111125947,
|
| 953 |
+
"step": 2950
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.9621604139715395,
|
| 957 |
+
"grad_norm": 0.23771828413009644,
|
| 958 |
+
"learning_rate": 1.875360781293689e-06,
|
| 959 |
+
"loss": 0.6642,
|
| 960 |
+
"mean_token_accuracy": 0.8211515319347381,
|
| 961 |
+
"step": 2975
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9702457956015524,
|
| 965 |
+
"grad_norm": 0.22044353187084198,
|
| 966 |
+
"learning_rate": 1.1601023630319064e-06,
|
| 967 |
+
"loss": 0.6148,
|
| 968 |
+
"mean_token_accuracy": 0.838649377822876,
|
| 969 |
+
"step": 3000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.9783311772315654,
|
| 973 |
+
"grad_norm": 0.21916313469409943,
|
| 974 |
+
"learning_rate": 6.155001261089477e-07,
|
| 975 |
+
"loss": 0.7242,
|
| 976 |
+
"mean_token_accuracy": 0.8094534778594971,
|
| 977 |
+
"step": 3025
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.9864165588615783,
|
| 981 |
+
"grad_norm": 0.20908524096012115,
|
| 982 |
+
"learning_rate": 2.4192756263349826e-07,
|
| 983 |
+
"loss": 0.7293,
|
| 984 |
+
"mean_token_accuracy": 0.8055576062202454,
|
| 985 |
+
"step": 3050
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.9945019404915912,
|
| 989 |
+
"grad_norm": 0.2526475489139557,
|
| 990 |
+
"learning_rate": 3.9640871341173336e-08,
|
| 991 |
+
"loss": 0.6961,
|
| 992 |
+
"mean_token_accuracy": 0.8134376859664917,
|
| 993 |
+
"step": 3075
|
| 994 |
+
}
|
| 995 |
+
],
|
| 996 |
+
"logging_steps": 25,
|
| 997 |
+
"max_steps": 3092,
|
| 998 |
+
"num_input_tokens_seen": 0,
|
| 999 |
+
"num_train_epochs": 1,
|
| 1000 |
+
"save_steps": 0,
|
| 1001 |
+
"stateful_callbacks": {
|
| 1002 |
+
"TrainerControl": {
|
| 1003 |
+
"args": {
|
| 1004 |
+
"should_epoch_stop": false,
|
| 1005 |
+
"should_evaluate": false,
|
| 1006 |
+
"should_log": false,
|
| 1007 |
+
"should_save": true,
|
| 1008 |
+
"should_training_stop": true
|
| 1009 |
+
},
|
| 1010 |
+
"attributes": {}
|
| 1011 |
+
}
|
| 1012 |
+
},
|
| 1013 |
+
"total_flos": 1.2584006622747034e+17,
|
| 1014 |
+
"train_batch_size": 1,
|
| 1015 |
+
"trial_name": null,
|
| 1016 |
+
"trial_params": null
|
| 1017 |
+
}
|