Upload checkpoint-140/trainer_state.json with huggingface_hub
Browse files- checkpoint-140/trainer_state.json +1032 -0
checkpoint-140/trainer_state.json
ADDED
|
@@ -0,0 +1,1032 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8296296296296296,
|
| 6 |
+
"eval_steps": 70,
|
| 7 |
+
"global_step": 140,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.005925925925925926,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 0.0,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.011851851851851851,
|
| 21 |
+
"grad_norm": NaN,
|
| 22 |
+
"learning_rate": 5e-05,
|
| 23 |
+
"loss": 0.0,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.017777777777777778,
|
| 28 |
+
"grad_norm": NaN,
|
| 29 |
+
"learning_rate": 5e-05,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.023703703703703703,
|
| 35 |
+
"grad_norm": NaN,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 0.0,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02962962962962963,
|
| 42 |
+
"grad_norm": NaN,
|
| 43 |
+
"learning_rate": 5e-05,
|
| 44 |
+
"loss": 0.0,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.035555555555555556,
|
| 49 |
+
"grad_norm": NaN,
|
| 50 |
+
"learning_rate": 5e-05,
|
| 51 |
+
"loss": 0.0,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.04148148148148148,
|
| 56 |
+
"grad_norm": NaN,
|
| 57 |
+
"learning_rate": 5e-05,
|
| 58 |
+
"loss": 0.0,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.047407407407407405,
|
| 63 |
+
"grad_norm": NaN,
|
| 64 |
+
"learning_rate": 5e-05,
|
| 65 |
+
"loss": 0.0,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.05333333333333334,
|
| 70 |
+
"grad_norm": NaN,
|
| 71 |
+
"learning_rate": 5e-05,
|
| 72 |
+
"loss": 0.0,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.05925925925925926,
|
| 77 |
+
"grad_norm": NaN,
|
| 78 |
+
"learning_rate": 5e-05,
|
| 79 |
+
"loss": 0.0,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06518518518518518,
|
| 84 |
+
"grad_norm": NaN,
|
| 85 |
+
"learning_rate": 5e-05,
|
| 86 |
+
"loss": 0.0,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.07111111111111111,
|
| 91 |
+
"grad_norm": NaN,
|
| 92 |
+
"learning_rate": 5e-05,
|
| 93 |
+
"loss": 0.0,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07703703703703704,
|
| 98 |
+
"grad_norm": NaN,
|
| 99 |
+
"learning_rate": 5e-05,
|
| 100 |
+
"loss": 0.0,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.08296296296296296,
|
| 105 |
+
"grad_norm": NaN,
|
| 106 |
+
"learning_rate": 5e-05,
|
| 107 |
+
"loss": 0.0,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.08888888888888889,
|
| 112 |
+
"grad_norm": NaN,
|
| 113 |
+
"learning_rate": 5e-05,
|
| 114 |
+
"loss": 0.0,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.09481481481481481,
|
| 119 |
+
"grad_norm": NaN,
|
| 120 |
+
"learning_rate": 5e-05,
|
| 121 |
+
"loss": 0.0,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.10074074074074074,
|
| 126 |
+
"grad_norm": NaN,
|
| 127 |
+
"learning_rate": 5e-05,
|
| 128 |
+
"loss": 0.0,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10666666666666667,
|
| 133 |
+
"grad_norm": NaN,
|
| 134 |
+
"learning_rate": 5e-05,
|
| 135 |
+
"loss": 0.0,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.11259259259259259,
|
| 140 |
+
"grad_norm": NaN,
|
| 141 |
+
"learning_rate": 5e-05,
|
| 142 |
+
"loss": 0.0,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.11851851851851852,
|
| 147 |
+
"grad_norm": NaN,
|
| 148 |
+
"learning_rate": 5e-05,
|
| 149 |
+
"loss": 0.0,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12444444444444444,
|
| 154 |
+
"grad_norm": NaN,
|
| 155 |
+
"learning_rate": 5e-05,
|
| 156 |
+
"loss": 0.0,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.13037037037037036,
|
| 161 |
+
"grad_norm": NaN,
|
| 162 |
+
"learning_rate": 5e-05,
|
| 163 |
+
"loss": 0.0,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.1362962962962963,
|
| 168 |
+
"grad_norm": NaN,
|
| 169 |
+
"learning_rate": 5e-05,
|
| 170 |
+
"loss": 0.0,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.14222222222222222,
|
| 175 |
+
"grad_norm": NaN,
|
| 176 |
+
"learning_rate": 5e-05,
|
| 177 |
+
"loss": 0.0,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.14814814814814814,
|
| 182 |
+
"grad_norm": NaN,
|
| 183 |
+
"learning_rate": 5e-05,
|
| 184 |
+
"loss": 0.0,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.15407407407407409,
|
| 189 |
+
"grad_norm": NaN,
|
| 190 |
+
"learning_rate": 5e-05,
|
| 191 |
+
"loss": 0.0,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.16,
|
| 196 |
+
"grad_norm": NaN,
|
| 197 |
+
"learning_rate": 5e-05,
|
| 198 |
+
"loss": 0.0,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.16592592592592592,
|
| 203 |
+
"grad_norm": NaN,
|
| 204 |
+
"learning_rate": 5e-05,
|
| 205 |
+
"loss": 0.0,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.17185185185185184,
|
| 210 |
+
"grad_norm": NaN,
|
| 211 |
+
"learning_rate": 5e-05,
|
| 212 |
+
"loss": 0.0,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.17777777777777778,
|
| 217 |
+
"grad_norm": NaN,
|
| 218 |
+
"learning_rate": 5e-05,
|
| 219 |
+
"loss": 0.0,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.1837037037037037,
|
| 224 |
+
"grad_norm": NaN,
|
| 225 |
+
"learning_rate": 5e-05,
|
| 226 |
+
"loss": 0.0,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.18962962962962962,
|
| 231 |
+
"grad_norm": NaN,
|
| 232 |
+
"learning_rate": 5e-05,
|
| 233 |
+
"loss": 0.0,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.19555555555555557,
|
| 238 |
+
"grad_norm": NaN,
|
| 239 |
+
"learning_rate": 5e-05,
|
| 240 |
+
"loss": 0.0,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.20148148148148148,
|
| 245 |
+
"grad_norm": NaN,
|
| 246 |
+
"learning_rate": 5e-05,
|
| 247 |
+
"loss": 0.0,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2074074074074074,
|
| 252 |
+
"grad_norm": NaN,
|
| 253 |
+
"learning_rate": 5e-05,
|
| 254 |
+
"loss": 0.0,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.21333333333333335,
|
| 259 |
+
"grad_norm": NaN,
|
| 260 |
+
"learning_rate": 5e-05,
|
| 261 |
+
"loss": 0.0,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.21925925925925926,
|
| 266 |
+
"grad_norm": NaN,
|
| 267 |
+
"learning_rate": 5e-05,
|
| 268 |
+
"loss": 0.0,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.22518518518518518,
|
| 273 |
+
"grad_norm": NaN,
|
| 274 |
+
"learning_rate": 5e-05,
|
| 275 |
+
"loss": 0.0,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2311111111111111,
|
| 280 |
+
"grad_norm": NaN,
|
| 281 |
+
"learning_rate": 5e-05,
|
| 282 |
+
"loss": 0.0,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.23703703703703705,
|
| 287 |
+
"grad_norm": NaN,
|
| 288 |
+
"learning_rate": 5e-05,
|
| 289 |
+
"loss": 0.0,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.24296296296296296,
|
| 294 |
+
"grad_norm": NaN,
|
| 295 |
+
"learning_rate": 5e-05,
|
| 296 |
+
"loss": 0.0,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.24888888888888888,
|
| 301 |
+
"grad_norm": NaN,
|
| 302 |
+
"learning_rate": 5e-05,
|
| 303 |
+
"loss": 0.0,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2548148148148148,
|
| 308 |
+
"grad_norm": NaN,
|
| 309 |
+
"learning_rate": 5e-05,
|
| 310 |
+
"loss": 0.0,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2607407407407407,
|
| 315 |
+
"grad_norm": NaN,
|
| 316 |
+
"learning_rate": 5e-05,
|
| 317 |
+
"loss": 0.0,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.26666666666666666,
|
| 322 |
+
"grad_norm": NaN,
|
| 323 |
+
"learning_rate": 5e-05,
|
| 324 |
+
"loss": 0.0,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2725925925925926,
|
| 329 |
+
"grad_norm": NaN,
|
| 330 |
+
"learning_rate": 5e-05,
|
| 331 |
+
"loss": 0.0,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2785185185185185,
|
| 336 |
+
"grad_norm": NaN,
|
| 337 |
+
"learning_rate": 5e-05,
|
| 338 |
+
"loss": 0.0,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.28444444444444444,
|
| 343 |
+
"grad_norm": NaN,
|
| 344 |
+
"learning_rate": 5e-05,
|
| 345 |
+
"loss": 0.0,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2903703703703704,
|
| 350 |
+
"grad_norm": NaN,
|
| 351 |
+
"learning_rate": 5e-05,
|
| 352 |
+
"loss": 0.0,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2962962962962963,
|
| 357 |
+
"grad_norm": NaN,
|
| 358 |
+
"learning_rate": 5e-05,
|
| 359 |
+
"loss": 0.0,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3022222222222222,
|
| 364 |
+
"grad_norm": NaN,
|
| 365 |
+
"learning_rate": 5e-05,
|
| 366 |
+
"loss": 0.0,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.30814814814814817,
|
| 371 |
+
"grad_norm": NaN,
|
| 372 |
+
"learning_rate": 5e-05,
|
| 373 |
+
"loss": 0.0,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.31407407407407406,
|
| 378 |
+
"grad_norm": NaN,
|
| 379 |
+
"learning_rate": 5e-05,
|
| 380 |
+
"loss": 0.0,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.32,
|
| 385 |
+
"grad_norm": NaN,
|
| 386 |
+
"learning_rate": 5e-05,
|
| 387 |
+
"loss": 0.0,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.32592592592592595,
|
| 392 |
+
"grad_norm": NaN,
|
| 393 |
+
"learning_rate": 5e-05,
|
| 394 |
+
"loss": 0.0,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.33185185185185184,
|
| 399 |
+
"grad_norm": NaN,
|
| 400 |
+
"learning_rate": 5e-05,
|
| 401 |
+
"loss": 0.0,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.3377777777777778,
|
| 406 |
+
"grad_norm": NaN,
|
| 407 |
+
"learning_rate": 5e-05,
|
| 408 |
+
"loss": 0.0,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3437037037037037,
|
| 413 |
+
"grad_norm": NaN,
|
| 414 |
+
"learning_rate": 5e-05,
|
| 415 |
+
"loss": 0.0,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3496296296296296,
|
| 420 |
+
"grad_norm": NaN,
|
| 421 |
+
"learning_rate": 5e-05,
|
| 422 |
+
"loss": 0.0,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.35555555555555557,
|
| 427 |
+
"grad_norm": NaN,
|
| 428 |
+
"learning_rate": 5e-05,
|
| 429 |
+
"loss": 0.0,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.36148148148148146,
|
| 434 |
+
"grad_norm": NaN,
|
| 435 |
+
"learning_rate": 5e-05,
|
| 436 |
+
"loss": 0.0,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3674074074074074,
|
| 441 |
+
"grad_norm": NaN,
|
| 442 |
+
"learning_rate": 5e-05,
|
| 443 |
+
"loss": 0.0,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.37333333333333335,
|
| 448 |
+
"grad_norm": NaN,
|
| 449 |
+
"learning_rate": 5e-05,
|
| 450 |
+
"loss": 0.0,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.37925925925925924,
|
| 455 |
+
"grad_norm": NaN,
|
| 456 |
+
"learning_rate": 5e-05,
|
| 457 |
+
"loss": 0.0,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3851851851851852,
|
| 462 |
+
"grad_norm": NaN,
|
| 463 |
+
"learning_rate": 5e-05,
|
| 464 |
+
"loss": 0.0,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.39111111111111113,
|
| 469 |
+
"grad_norm": NaN,
|
| 470 |
+
"learning_rate": 5e-05,
|
| 471 |
+
"loss": 0.0,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.397037037037037,
|
| 476 |
+
"grad_norm": NaN,
|
| 477 |
+
"learning_rate": 5e-05,
|
| 478 |
+
"loss": 0.0,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.40296296296296297,
|
| 483 |
+
"grad_norm": NaN,
|
| 484 |
+
"learning_rate": 5e-05,
|
| 485 |
+
"loss": 0.0,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.4088888888888889,
|
| 490 |
+
"grad_norm": NaN,
|
| 491 |
+
"learning_rate": 5e-05,
|
| 492 |
+
"loss": 0.0,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.4148148148148148,
|
| 497 |
+
"grad_norm": NaN,
|
| 498 |
+
"learning_rate": 5e-05,
|
| 499 |
+
"loss": 0.0,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.4148148148148148,
|
| 504 |
+
"eval_loss": 40297.44140625,
|
| 505 |
+
"eval_runtime": 63.992,
|
| 506 |
+
"eval_samples_per_second": 7.032,
|
| 507 |
+
"eval_steps_per_second": 1.172,
|
| 508 |
+
"eval_wer": 1.0381992969974168,
|
| 509 |
+
"step": 70
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.42074074074074075,
|
| 513 |
+
"grad_norm": NaN,
|
| 514 |
+
"learning_rate": 5e-05,
|
| 515 |
+
"loss": 0.0,
|
| 516 |
+
"step": 71
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.4266666666666667,
|
| 520 |
+
"grad_norm": NaN,
|
| 521 |
+
"learning_rate": 5e-05,
|
| 522 |
+
"loss": 0.0,
|
| 523 |
+
"step": 72
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.4325925925925926,
|
| 527 |
+
"grad_norm": NaN,
|
| 528 |
+
"learning_rate": 5e-05,
|
| 529 |
+
"loss": 0.0,
|
| 530 |
+
"step": 73
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.43851851851851853,
|
| 534 |
+
"grad_norm": NaN,
|
| 535 |
+
"learning_rate": 5e-05,
|
| 536 |
+
"loss": 0.0,
|
| 537 |
+
"step": 74
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.4444444444444444,
|
| 541 |
+
"grad_norm": NaN,
|
| 542 |
+
"learning_rate": 5e-05,
|
| 543 |
+
"loss": 0.0,
|
| 544 |
+
"step": 75
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.45037037037037037,
|
| 548 |
+
"grad_norm": NaN,
|
| 549 |
+
"learning_rate": 5e-05,
|
| 550 |
+
"loss": 0.0,
|
| 551 |
+
"step": 76
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.4562962962962963,
|
| 555 |
+
"grad_norm": NaN,
|
| 556 |
+
"learning_rate": 5e-05,
|
| 557 |
+
"loss": 0.0,
|
| 558 |
+
"step": 77
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.4622222222222222,
|
| 562 |
+
"grad_norm": NaN,
|
| 563 |
+
"learning_rate": 5e-05,
|
| 564 |
+
"loss": 0.0,
|
| 565 |
+
"step": 78
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.46814814814814815,
|
| 569 |
+
"grad_norm": NaN,
|
| 570 |
+
"learning_rate": 5e-05,
|
| 571 |
+
"loss": 0.0,
|
| 572 |
+
"step": 79
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.4740740740740741,
|
| 576 |
+
"grad_norm": NaN,
|
| 577 |
+
"learning_rate": 5e-05,
|
| 578 |
+
"loss": 0.0,
|
| 579 |
+
"step": 80
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.48,
|
| 583 |
+
"grad_norm": NaN,
|
| 584 |
+
"learning_rate": 5e-05,
|
| 585 |
+
"loss": 0.0,
|
| 586 |
+
"step": 81
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.48592592592592593,
|
| 590 |
+
"grad_norm": NaN,
|
| 591 |
+
"learning_rate": 5e-05,
|
| 592 |
+
"loss": 0.0,
|
| 593 |
+
"step": 82
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.4918518518518519,
|
| 597 |
+
"grad_norm": NaN,
|
| 598 |
+
"learning_rate": 5e-05,
|
| 599 |
+
"loss": 0.0,
|
| 600 |
+
"step": 83
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.49777777777777776,
|
| 604 |
+
"grad_norm": NaN,
|
| 605 |
+
"learning_rate": 5e-05,
|
| 606 |
+
"loss": 0.0,
|
| 607 |
+
"step": 84
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5037037037037037,
|
| 611 |
+
"grad_norm": NaN,
|
| 612 |
+
"learning_rate": 5e-05,
|
| 613 |
+
"loss": 0.0,
|
| 614 |
+
"step": 85
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5096296296296297,
|
| 618 |
+
"grad_norm": NaN,
|
| 619 |
+
"learning_rate": 5e-05,
|
| 620 |
+
"loss": 0.0,
|
| 621 |
+
"step": 86
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5155555555555555,
|
| 625 |
+
"grad_norm": NaN,
|
| 626 |
+
"learning_rate": 5e-05,
|
| 627 |
+
"loss": 0.0,
|
| 628 |
+
"step": 87
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.5214814814814814,
|
| 632 |
+
"grad_norm": NaN,
|
| 633 |
+
"learning_rate": 5e-05,
|
| 634 |
+
"loss": 0.0,
|
| 635 |
+
"step": 88
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.5274074074074074,
|
| 639 |
+
"grad_norm": NaN,
|
| 640 |
+
"learning_rate": 5e-05,
|
| 641 |
+
"loss": 0.0,
|
| 642 |
+
"step": 89
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.5333333333333333,
|
| 646 |
+
"grad_norm": NaN,
|
| 647 |
+
"learning_rate": 5e-05,
|
| 648 |
+
"loss": 0.0,
|
| 649 |
+
"step": 90
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.5392592592592592,
|
| 653 |
+
"grad_norm": NaN,
|
| 654 |
+
"learning_rate": 5e-05,
|
| 655 |
+
"loss": 0.0,
|
| 656 |
+
"step": 91
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.5451851851851852,
|
| 660 |
+
"grad_norm": NaN,
|
| 661 |
+
"learning_rate": 5e-05,
|
| 662 |
+
"loss": 0.0,
|
| 663 |
+
"step": 92
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.5511111111111111,
|
| 667 |
+
"grad_norm": NaN,
|
| 668 |
+
"learning_rate": 5e-05,
|
| 669 |
+
"loss": 0.0,
|
| 670 |
+
"step": 93
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.557037037037037,
|
| 674 |
+
"grad_norm": NaN,
|
| 675 |
+
"learning_rate": 5e-05,
|
| 676 |
+
"loss": 0.0,
|
| 677 |
+
"step": 94
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.562962962962963,
|
| 681 |
+
"grad_norm": NaN,
|
| 682 |
+
"learning_rate": 5e-05,
|
| 683 |
+
"loss": 0.0,
|
| 684 |
+
"step": 95
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.5688888888888889,
|
| 688 |
+
"grad_norm": NaN,
|
| 689 |
+
"learning_rate": 5e-05,
|
| 690 |
+
"loss": 0.0,
|
| 691 |
+
"step": 96
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5748148148148148,
|
| 695 |
+
"grad_norm": NaN,
|
| 696 |
+
"learning_rate": 5e-05,
|
| 697 |
+
"loss": 0.0,
|
| 698 |
+
"step": 97
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5807407407407408,
|
| 702 |
+
"grad_norm": NaN,
|
| 703 |
+
"learning_rate": 5e-05,
|
| 704 |
+
"loss": 0.0,
|
| 705 |
+
"step": 98
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5866666666666667,
|
| 709 |
+
"grad_norm": NaN,
|
| 710 |
+
"learning_rate": 5e-05,
|
| 711 |
+
"loss": 0.0,
|
| 712 |
+
"step": 99
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 0.5925925925925926,
|
| 716 |
+
"grad_norm": NaN,
|
| 717 |
+
"learning_rate": 5e-05,
|
| 718 |
+
"loss": 0.0,
|
| 719 |
+
"step": 100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.5985185185185186,
|
| 723 |
+
"grad_norm": NaN,
|
| 724 |
+
"learning_rate": 5e-05,
|
| 725 |
+
"loss": 0.0,
|
| 726 |
+
"step": 101
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.6044444444444445,
|
| 730 |
+
"grad_norm": NaN,
|
| 731 |
+
"learning_rate": 5e-05,
|
| 732 |
+
"loss": 0.0,
|
| 733 |
+
"step": 102
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.6103703703703703,
|
| 737 |
+
"grad_norm": NaN,
|
| 738 |
+
"learning_rate": 5e-05,
|
| 739 |
+
"loss": 0.0,
|
| 740 |
+
"step": 103
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.6162962962962963,
|
| 744 |
+
"grad_norm": NaN,
|
| 745 |
+
"learning_rate": 5e-05,
|
| 746 |
+
"loss": 0.0,
|
| 747 |
+
"step": 104
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.6222222222222222,
|
| 751 |
+
"grad_norm": NaN,
|
| 752 |
+
"learning_rate": 5e-05,
|
| 753 |
+
"loss": 0.0,
|
| 754 |
+
"step": 105
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.6281481481481481,
|
| 758 |
+
"grad_norm": NaN,
|
| 759 |
+
"learning_rate": 5e-05,
|
| 760 |
+
"loss": 0.0,
|
| 761 |
+
"step": 106
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.6340740740740741,
|
| 765 |
+
"grad_norm": NaN,
|
| 766 |
+
"learning_rate": 5e-05,
|
| 767 |
+
"loss": 0.0,
|
| 768 |
+
"step": 107
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.64,
|
| 772 |
+
"grad_norm": NaN,
|
| 773 |
+
"learning_rate": 5e-05,
|
| 774 |
+
"loss": 0.0,
|
| 775 |
+
"step": 108
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.6459259259259259,
|
| 779 |
+
"grad_norm": NaN,
|
| 780 |
+
"learning_rate": 5e-05,
|
| 781 |
+
"loss": 0.0,
|
| 782 |
+
"step": 109
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.6518518518518519,
|
| 786 |
+
"grad_norm": NaN,
|
| 787 |
+
"learning_rate": 5e-05,
|
| 788 |
+
"loss": 0.0,
|
| 789 |
+
"step": 110
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.6577777777777778,
|
| 793 |
+
"grad_norm": NaN,
|
| 794 |
+
"learning_rate": 5e-05,
|
| 795 |
+
"loss": 0.0,
|
| 796 |
+
"step": 111
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.6637037037037037,
|
| 800 |
+
"grad_norm": NaN,
|
| 801 |
+
"learning_rate": 5e-05,
|
| 802 |
+
"loss": 0.0,
|
| 803 |
+
"step": 112
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.6696296296296296,
|
| 807 |
+
"grad_norm": NaN,
|
| 808 |
+
"learning_rate": 5e-05,
|
| 809 |
+
"loss": 0.0,
|
| 810 |
+
"step": 113
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.6755555555555556,
|
| 814 |
+
"grad_norm": NaN,
|
| 815 |
+
"learning_rate": 5e-05,
|
| 816 |
+
"loss": 0.0,
|
| 817 |
+
"step": 114
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.6814814814814815,
|
| 821 |
+
"grad_norm": NaN,
|
| 822 |
+
"learning_rate": 5e-05,
|
| 823 |
+
"loss": 0.0,
|
| 824 |
+
"step": 115
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.6874074074074074,
|
| 828 |
+
"grad_norm": NaN,
|
| 829 |
+
"learning_rate": 5e-05,
|
| 830 |
+
"loss": 0.0,
|
| 831 |
+
"step": 116
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.6933333333333334,
|
| 835 |
+
"grad_norm": NaN,
|
| 836 |
+
"learning_rate": 5e-05,
|
| 837 |
+
"loss": 0.0,
|
| 838 |
+
"step": 117
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.6992592592592592,
|
| 842 |
+
"grad_norm": NaN,
|
| 843 |
+
"learning_rate": 5e-05,
|
| 844 |
+
"loss": 0.0,
|
| 845 |
+
"step": 118
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.7051851851851851,
|
| 849 |
+
"grad_norm": NaN,
|
| 850 |
+
"learning_rate": 5e-05,
|
| 851 |
+
"loss": 0.0,
|
| 852 |
+
"step": 119
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.7111111111111111,
|
| 856 |
+
"grad_norm": NaN,
|
| 857 |
+
"learning_rate": 5e-05,
|
| 858 |
+
"loss": 0.0,
|
| 859 |
+
"step": 120
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.717037037037037,
|
| 863 |
+
"grad_norm": NaN,
|
| 864 |
+
"learning_rate": 5e-05,
|
| 865 |
+
"loss": 0.0,
|
| 866 |
+
"step": 121
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.7229629629629629,
|
| 870 |
+
"grad_norm": NaN,
|
| 871 |
+
"learning_rate": 5e-05,
|
| 872 |
+
"loss": 0.0,
|
| 873 |
+
"step": 122
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.7288888888888889,
|
| 877 |
+
"grad_norm": NaN,
|
| 878 |
+
"learning_rate": 5e-05,
|
| 879 |
+
"loss": 0.0,
|
| 880 |
+
"step": 123
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.7348148148148148,
|
| 884 |
+
"grad_norm": NaN,
|
| 885 |
+
"learning_rate": 5e-05,
|
| 886 |
+
"loss": 0.0,
|
| 887 |
+
"step": 124
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.7407407407407407,
|
| 891 |
+
"grad_norm": NaN,
|
| 892 |
+
"learning_rate": 5e-05,
|
| 893 |
+
"loss": 0.0,
|
| 894 |
+
"step": 125
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.7466666666666667,
|
| 898 |
+
"grad_norm": NaN,
|
| 899 |
+
"learning_rate": 5e-05,
|
| 900 |
+
"loss": 0.0,
|
| 901 |
+
"step": 126
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.7525925925925926,
|
| 905 |
+
"grad_norm": NaN,
|
| 906 |
+
"learning_rate": 5e-05,
|
| 907 |
+
"loss": 0.0,
|
| 908 |
+
"step": 127
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.7585185185185185,
|
| 912 |
+
"grad_norm": NaN,
|
| 913 |
+
"learning_rate": 5e-05,
|
| 914 |
+
"loss": 0.0,
|
| 915 |
+
"step": 128
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.7644444444444445,
|
| 919 |
+
"grad_norm": NaN,
|
| 920 |
+
"learning_rate": 5e-05,
|
| 921 |
+
"loss": 0.0,
|
| 922 |
+
"step": 129
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.7703703703703704,
|
| 926 |
+
"grad_norm": NaN,
|
| 927 |
+
"learning_rate": 5e-05,
|
| 928 |
+
"loss": 0.0,
|
| 929 |
+
"step": 130
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.7762962962962963,
|
| 933 |
+
"grad_norm": NaN,
|
| 934 |
+
"learning_rate": 5e-05,
|
| 935 |
+
"loss": 0.0,
|
| 936 |
+
"step": 131
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.7822222222222223,
|
| 940 |
+
"grad_norm": NaN,
|
| 941 |
+
"learning_rate": 5e-05,
|
| 942 |
+
"loss": 0.0,
|
| 943 |
+
"step": 132
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.7881481481481482,
|
| 947 |
+
"grad_norm": NaN,
|
| 948 |
+
"learning_rate": 5e-05,
|
| 949 |
+
"loss": 0.0,
|
| 950 |
+
"step": 133
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.794074074074074,
|
| 954 |
+
"grad_norm": NaN,
|
| 955 |
+
"learning_rate": 5e-05,
|
| 956 |
+
"loss": 0.0,
|
| 957 |
+
"step": 134
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.8,
|
| 961 |
+
"grad_norm": NaN,
|
| 962 |
+
"learning_rate": 5e-05,
|
| 963 |
+
"loss": 0.0,
|
| 964 |
+
"step": 135
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8059259259259259,
|
| 968 |
+
"grad_norm": NaN,
|
| 969 |
+
"learning_rate": 5e-05,
|
| 970 |
+
"loss": 0.0,
|
| 971 |
+
"step": 136
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8118518518518518,
|
| 975 |
+
"grad_norm": NaN,
|
| 976 |
+
"learning_rate": 5e-05,
|
| 977 |
+
"loss": 0.0,
|
| 978 |
+
"step": 137
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8177777777777778,
|
| 982 |
+
"grad_norm": NaN,
|
| 983 |
+
"learning_rate": 5e-05,
|
| 984 |
+
"loss": 0.0,
|
| 985 |
+
"step": 138
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8237037037037037,
|
| 989 |
+
"grad_norm": NaN,
|
| 990 |
+
"learning_rate": 5e-05,
|
| 991 |
+
"loss": 0.0,
|
| 992 |
+
"step": 139
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8296296296296296,
|
| 996 |
+
"grad_norm": NaN,
|
| 997 |
+
"learning_rate": 5e-05,
|
| 998 |
+
"loss": 0.0,
|
| 999 |
+
"step": 140
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.8296296296296296,
|
| 1003 |
+
"eval_loss": 40297.44140625,
|
| 1004 |
+
"eval_runtime": 63.7847,
|
| 1005 |
+
"eval_samples_per_second": 7.055,
|
| 1006 |
+
"eval_steps_per_second": 1.176,
|
| 1007 |
+
"eval_wer": 1.0381992969974168,
|
| 1008 |
+
"step": 140
|
| 1009 |
+
}
|
| 1010 |
+
],
|
| 1011 |
+
"logging_steps": 1,
|
| 1012 |
+
"max_steps": 1008,
|
| 1013 |
+
"num_input_tokens_seen": 0,
|
| 1014 |
+
"num_train_epochs": 6,
|
| 1015 |
+
"save_steps": 70,
|
| 1016 |
+
"stateful_callbacks": {
|
| 1017 |
+
"TrainerControl": {
|
| 1018 |
+
"args": {
|
| 1019 |
+
"should_epoch_stop": false,
|
| 1020 |
+
"should_evaluate": false,
|
| 1021 |
+
"should_log": false,
|
| 1022 |
+
"should_save": true,
|
| 1023 |
+
"should_training_stop": false
|
| 1024 |
+
},
|
| 1025 |
+
"attributes": {}
|
| 1026 |
+
}
|
| 1027 |
+
},
|
| 1028 |
+
"total_flos": 1.149038015806368e+18,
|
| 1029 |
+
"train_batch_size": 6,
|
| 1030 |
+
"trial_name": null,
|
| 1031 |
+
"trial_params": null
|
| 1032 |
+
}
|