Upload trainer_state.json
Browse files- trainer_state.json +813 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,813 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.6711751222610474,
|
| 3 |
+
"best_model_checkpoint": "./checkpoints/KuLLM3_lora_clm_with_added_tokens_no_shorts-2batch-2epoch_1109/checkpoint-80000",
|
| 4 |
+
"epoch": 1.9412292834437408,
|
| 5 |
+
"eval_steps": 8000,
|
| 6 |
+
"global_step": 80000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01941229283443741,
|
| 13 |
+
"grad_norm": 179.34385681152344,
|
| 14 |
+
"learning_rate": 1.940805434255216e-06,
|
| 15 |
+
"loss": 1.3039,
|
| 16 |
+
"step": 800
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03882458566887482,
|
| 20 |
+
"grad_norm": 72.6114730834961,
|
| 21 |
+
"learning_rate": 3.881610868510432e-06,
|
| 22 |
+
"loss": 0.9632,
|
| 23 |
+
"step": 1600
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.058236878503312224,
|
| 27 |
+
"grad_norm": 76.58552551269531,
|
| 28 |
+
"learning_rate": 5.822416302765648e-06,
|
| 29 |
+
"loss": 0.8919,
|
| 30 |
+
"step": 2400
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07764917133774964,
|
| 34 |
+
"grad_norm": 121.52155303955078,
|
| 35 |
+
"learning_rate": 7.763221737020863e-06,
|
| 36 |
+
"loss": 0.868,
|
| 37 |
+
"step": 3200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09706146417218704,
|
| 41 |
+
"grad_norm": 61.98054885864258,
|
| 42 |
+
"learning_rate": 9.70402717127608e-06,
|
| 43 |
+
"loss": 0.8474,
|
| 44 |
+
"step": 4000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11647375700662445,
|
| 48 |
+
"grad_norm": 48.987022399902344,
|
| 49 |
+
"learning_rate": 9.913409961685825e-06,
|
| 50 |
+
"loss": 0.8259,
|
| 51 |
+
"step": 4800
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.13588604984106184,
|
| 55 |
+
"grad_norm": 52.976566314697266,
|
| 56 |
+
"learning_rate": 9.81123882503193e-06,
|
| 57 |
+
"loss": 0.8224,
|
| 58 |
+
"step": 5600
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15529834267549927,
|
| 62 |
+
"grad_norm": 74.86931610107422,
|
| 63 |
+
"learning_rate": 9.709067688378035e-06,
|
| 64 |
+
"loss": 0.8002,
|
| 65 |
+
"step": 6400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.17471063550993668,
|
| 69 |
+
"grad_norm": 58.481258392333984,
|
| 70 |
+
"learning_rate": 9.606896551724138e-06,
|
| 71 |
+
"loss": 0.8021,
|
| 72 |
+
"step": 7200
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.19412292834437408,
|
| 76 |
+
"grad_norm": 82.62439727783203,
|
| 77 |
+
"learning_rate": 9.504725415070245e-06,
|
| 78 |
+
"loss": 0.8004,
|
| 79 |
+
"step": 8000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19412292834437408,
|
| 83 |
+
"eval_loss": 0.7944459915161133,
|
| 84 |
+
"eval_runtime": 1619.7602,
|
| 85 |
+
"eval_samples_per_second": 6.337,
|
| 86 |
+
"eval_steps_per_second": 6.337,
|
| 87 |
+
"step": 8000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2135352211788115,
|
| 91 |
+
"grad_norm": 58.08147430419922,
|
| 92 |
+
"learning_rate": 9.402554278416348e-06,
|
| 93 |
+
"loss": 0.7754,
|
| 94 |
+
"step": 8800
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.2329475140132489,
|
| 98 |
+
"grad_norm": 85.25300598144531,
|
| 99 |
+
"learning_rate": 9.300383141762453e-06,
|
| 100 |
+
"loss": 0.7728,
|
| 101 |
+
"step": 9600
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2523598068476863,
|
| 105 |
+
"grad_norm": 139.68646240234375,
|
| 106 |
+
"learning_rate": 9.198212005108558e-06,
|
| 107 |
+
"loss": 0.768,
|
| 108 |
+
"step": 10400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.2717720996821237,
|
| 112 |
+
"grad_norm": 62.37124252319336,
|
| 113 |
+
"learning_rate": 9.096040868454661e-06,
|
| 114 |
+
"loss": 0.7773,
|
| 115 |
+
"step": 11200
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2911843925165611,
|
| 119 |
+
"grad_norm": 32.99665069580078,
|
| 120 |
+
"learning_rate": 8.993869731800768e-06,
|
| 121 |
+
"loss": 0.7657,
|
| 122 |
+
"step": 12000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.31059668535099855,
|
| 126 |
+
"grad_norm": 117.61981964111328,
|
| 127 |
+
"learning_rate": 8.891698595146871e-06,
|
| 128 |
+
"loss": 0.7613,
|
| 129 |
+
"step": 12800
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33000897818543595,
|
| 133 |
+
"grad_norm": 43.9254035949707,
|
| 134 |
+
"learning_rate": 8.789527458492976e-06,
|
| 135 |
+
"loss": 0.7573,
|
| 136 |
+
"step": 13600
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.34942127101987336,
|
| 140 |
+
"grad_norm": 31.05263328552246,
|
| 141 |
+
"learning_rate": 8.687356321839081e-06,
|
| 142 |
+
"loss": 0.7453,
|
| 143 |
+
"step": 14400
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.36883356385431076,
|
| 147 |
+
"grad_norm": 57.3399543762207,
|
| 148 |
+
"learning_rate": 8.585185185185186e-06,
|
| 149 |
+
"loss": 0.743,
|
| 150 |
+
"step": 15200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.38824585668874817,
|
| 154 |
+
"grad_norm": 29.631826400756836,
|
| 155 |
+
"learning_rate": 8.483014048531291e-06,
|
| 156 |
+
"loss": 0.7445,
|
| 157 |
+
"step": 16000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.38824585668874817,
|
| 161 |
+
"eval_loss": 0.750423789024353,
|
| 162 |
+
"eval_runtime": 1620.0402,
|
| 163 |
+
"eval_samples_per_second": 6.336,
|
| 164 |
+
"eval_steps_per_second": 6.336,
|
| 165 |
+
"step": 16000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.4076581495231856,
|
| 169 |
+
"grad_norm": 29.647502899169922,
|
| 170 |
+
"learning_rate": 8.380842911877395e-06,
|
| 171 |
+
"loss": 0.7331,
|
| 172 |
+
"step": 16800
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.427070442357623,
|
| 176 |
+
"grad_norm": 19.794334411621094,
|
| 177 |
+
"learning_rate": 8.2786717752235e-06,
|
| 178 |
+
"loss": 0.7329,
|
| 179 |
+
"step": 17600
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.4464827351920604,
|
| 183 |
+
"grad_norm": 114.88660430908203,
|
| 184 |
+
"learning_rate": 8.176500638569605e-06,
|
| 185 |
+
"loss": 0.7334,
|
| 186 |
+
"step": 18400
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.4658950280264978,
|
| 190 |
+
"grad_norm": 29.109107971191406,
|
| 191 |
+
"learning_rate": 8.07432950191571e-06,
|
| 192 |
+
"loss": 0.7294,
|
| 193 |
+
"step": 19200
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.4853073208609352,
|
| 197 |
+
"grad_norm": 25.62792205810547,
|
| 198 |
+
"learning_rate": 7.972158365261815e-06,
|
| 199 |
+
"loss": 0.7287,
|
| 200 |
+
"step": 20000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.5047196136953725,
|
| 204 |
+
"grad_norm": 40.84642028808594,
|
| 205 |
+
"learning_rate": 7.86998722860792e-06,
|
| 206 |
+
"loss": 0.7157,
|
| 207 |
+
"step": 20800
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.52413190652981,
|
| 211 |
+
"grad_norm": 46.767059326171875,
|
| 212 |
+
"learning_rate": 7.767816091954023e-06,
|
| 213 |
+
"loss": 0.7284,
|
| 214 |
+
"step": 21600
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.5435441993642474,
|
| 218 |
+
"grad_norm": 24.68915557861328,
|
| 219 |
+
"learning_rate": 7.66564495530013e-06,
|
| 220 |
+
"loss": 0.7314,
|
| 221 |
+
"step": 22400
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.5629564921986848,
|
| 225 |
+
"grad_norm": 31.020458221435547,
|
| 226 |
+
"learning_rate": 7.563473818646233e-06,
|
| 227 |
+
"loss": 0.7157,
|
| 228 |
+
"step": 23200
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.5823687850331222,
|
| 232 |
+
"grad_norm": 29.391735076904297,
|
| 233 |
+
"learning_rate": 7.461302681992337e-06,
|
| 234 |
+
"loss": 0.7058,
|
| 235 |
+
"step": 24000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.5823687850331222,
|
| 239 |
+
"eval_loss": 0.7259429693222046,
|
| 240 |
+
"eval_runtime": 1620.5504,
|
| 241 |
+
"eval_samples_per_second": 6.334,
|
| 242 |
+
"eval_steps_per_second": 6.334,
|
| 243 |
+
"step": 24000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.6017810778675596,
|
| 247 |
+
"grad_norm": 45.48763656616211,
|
| 248 |
+
"learning_rate": 7.359131545338443e-06,
|
| 249 |
+
"loss": 0.7117,
|
| 250 |
+
"step": 24800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.6211933707019971,
|
| 254 |
+
"grad_norm": 24.26145362854004,
|
| 255 |
+
"learning_rate": 7.256960408684547e-06,
|
| 256 |
+
"loss": 0.701,
|
| 257 |
+
"step": 25600
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.6406056635364344,
|
| 261 |
+
"grad_norm": 20.341548919677734,
|
| 262 |
+
"learning_rate": 7.154789272030652e-06,
|
| 263 |
+
"loss": 0.7227,
|
| 264 |
+
"step": 26400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.6600179563708719,
|
| 268 |
+
"grad_norm": 55.4034423828125,
|
| 269 |
+
"learning_rate": 7.052618135376756e-06,
|
| 270 |
+
"loss": 0.7125,
|
| 271 |
+
"step": 27200
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.6794302492053093,
|
| 275 |
+
"grad_norm": 27.05145263671875,
|
| 276 |
+
"learning_rate": 6.950446998722862e-06,
|
| 277 |
+
"loss": 0.7053,
|
| 278 |
+
"step": 28000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.6988425420397467,
|
| 282 |
+
"grad_norm": 30.649011611938477,
|
| 283 |
+
"learning_rate": 6.848275862068966e-06,
|
| 284 |
+
"loss": 0.6974,
|
| 285 |
+
"step": 28800
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.7182548348741841,
|
| 289 |
+
"grad_norm": 54.12739181518555,
|
| 290 |
+
"learning_rate": 6.746104725415071e-06,
|
| 291 |
+
"loss": 0.7018,
|
| 292 |
+
"step": 29600
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.7376671277086215,
|
| 296 |
+
"grad_norm": 22.7417049407959,
|
| 297 |
+
"learning_rate": 6.6439335887611755e-06,
|
| 298 |
+
"loss": 0.6966,
|
| 299 |
+
"step": 30400
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.7570794205430589,
|
| 303 |
+
"grad_norm": 50.62528991699219,
|
| 304 |
+
"learning_rate": 6.54176245210728e-06,
|
| 305 |
+
"loss": 0.6963,
|
| 306 |
+
"step": 31200
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.7764917133774963,
|
| 310 |
+
"grad_norm": 107.59382629394531,
|
| 311 |
+
"learning_rate": 6.4395913154533855e-06,
|
| 312 |
+
"loss": 0.7039,
|
| 313 |
+
"step": 32000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.7764917133774963,
|
| 317 |
+
"eval_loss": 0.7099697589874268,
|
| 318 |
+
"eval_runtime": 1619.6536,
|
| 319 |
+
"eval_samples_per_second": 6.338,
|
| 320 |
+
"eval_steps_per_second": 6.338,
|
| 321 |
+
"step": 32000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.7959040062119337,
|
| 325 |
+
"grad_norm": 24.90111541748047,
|
| 326 |
+
"learning_rate": 6.33742017879949e-06,
|
| 327 |
+
"loss": 0.7021,
|
| 328 |
+
"step": 32800
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.8153162990463712,
|
| 332 |
+
"grad_norm": 42.44001007080078,
|
| 333 |
+
"learning_rate": 6.235249042145595e-06,
|
| 334 |
+
"loss": 0.6929,
|
| 335 |
+
"step": 33600
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.8347285918808085,
|
| 339 |
+
"grad_norm": 29.183473587036133,
|
| 340 |
+
"learning_rate": 6.133077905491699e-06,
|
| 341 |
+
"loss": 0.6955,
|
| 342 |
+
"step": 34400
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.854140884715246,
|
| 346 |
+
"grad_norm": 38.95024490356445,
|
| 347 |
+
"learning_rate": 6.030906768837805e-06,
|
| 348 |
+
"loss": 0.6928,
|
| 349 |
+
"step": 35200
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.8735531775496833,
|
| 353 |
+
"grad_norm": 31.284547805786133,
|
| 354 |
+
"learning_rate": 5.928735632183909e-06,
|
| 355 |
+
"loss": 0.6919,
|
| 356 |
+
"step": 36000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.8929654703841208,
|
| 360 |
+
"grad_norm": 41.60378646850586,
|
| 361 |
+
"learning_rate": 5.826564495530014e-06,
|
| 362 |
+
"loss": 0.6955,
|
| 363 |
+
"step": 36800
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.9123777632185581,
|
| 367 |
+
"grad_norm": 47.39144515991211,
|
| 368 |
+
"learning_rate": 5.724393358876118e-06,
|
| 369 |
+
"loss": 0.7038,
|
| 370 |
+
"step": 37600
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.9317900560529956,
|
| 374 |
+
"grad_norm": 45.3167610168457,
|
| 375 |
+
"learning_rate": 5.622222222222222e-06,
|
| 376 |
+
"loss": 0.6838,
|
| 377 |
+
"step": 38400
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.9512023488874329,
|
| 381 |
+
"grad_norm": 17.134929656982422,
|
| 382 |
+
"learning_rate": 5.520051085568328e-06,
|
| 383 |
+
"loss": 0.6933,
|
| 384 |
+
"step": 39200
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.9706146417218704,
|
| 388 |
+
"grad_norm": 32.11737060546875,
|
| 389 |
+
"learning_rate": 5.417879948914432e-06,
|
| 390 |
+
"loss": 0.6861,
|
| 391 |
+
"step": 40000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.9706146417218704,
|
| 395 |
+
"eval_loss": 0.698592483997345,
|
| 396 |
+
"eval_runtime": 1620.335,
|
| 397 |
+
"eval_samples_per_second": 6.335,
|
| 398 |
+
"eval_steps_per_second": 6.335,
|
| 399 |
+
"step": 40000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.9900269345563077,
|
| 403 |
+
"grad_norm": 36.55326843261719,
|
| 404 |
+
"learning_rate": 5.315708812260537e-06,
|
| 405 |
+
"loss": 0.6859,
|
| 406 |
+
"step": 40800
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.009439227390745,
|
| 410 |
+
"grad_norm": 23.36037254333496,
|
| 411 |
+
"learning_rate": 5.213537675606641e-06,
|
| 412 |
+
"loss": 0.6842,
|
| 413 |
+
"step": 41600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.0288515202251827,
|
| 417 |
+
"grad_norm": 21.21329116821289,
|
| 418 |
+
"learning_rate": 5.111366538952746e-06,
|
| 419 |
+
"loss": 0.6687,
|
| 420 |
+
"step": 42400
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.04826381305962,
|
| 424 |
+
"grad_norm": 26.16172981262207,
|
| 425 |
+
"learning_rate": 5.009195402298851e-06,
|
| 426 |
+
"loss": 0.6859,
|
| 427 |
+
"step": 43200
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.0676761058940574,
|
| 431 |
+
"grad_norm": 34.661190032958984,
|
| 432 |
+
"learning_rate": 4.9070242656449555e-06,
|
| 433 |
+
"loss": 0.6667,
|
| 434 |
+
"step": 44000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.0870883987284947,
|
| 438 |
+
"grad_norm": 47.726898193359375,
|
| 439 |
+
"learning_rate": 4.8048531289910605e-06,
|
| 440 |
+
"loss": 0.6628,
|
| 441 |
+
"step": 44800
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.1065006915629323,
|
| 445 |
+
"grad_norm": 26.898998260498047,
|
| 446 |
+
"learning_rate": 4.7026819923371655e-06,
|
| 447 |
+
"loss": 0.6715,
|
| 448 |
+
"step": 45600
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.1259129843973696,
|
| 452 |
+
"grad_norm": 31.409574508666992,
|
| 453 |
+
"learning_rate": 4.60051085568327e-06,
|
| 454 |
+
"loss": 0.6725,
|
| 455 |
+
"step": 46400
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.145325277231807,
|
| 459 |
+
"grad_norm": 29.432598114013672,
|
| 460 |
+
"learning_rate": 4.498339719029375e-06,
|
| 461 |
+
"loss": 0.6589,
|
| 462 |
+
"step": 47200
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.1647375700662446,
|
| 466 |
+
"grad_norm": 33.80415725708008,
|
| 467 |
+
"learning_rate": 4.396168582375479e-06,
|
| 468 |
+
"loss": 0.667,
|
| 469 |
+
"step": 48000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.1647375700662446,
|
| 473 |
+
"eval_loss": 0.6902133226394653,
|
| 474 |
+
"eval_runtime": 1620.636,
|
| 475 |
+
"eval_samples_per_second": 6.334,
|
| 476 |
+
"eval_steps_per_second": 6.334,
|
| 477 |
+
"step": 48000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.184149862900682,
|
| 481 |
+
"grad_norm": 34.50639724731445,
|
| 482 |
+
"learning_rate": 4.293997445721584e-06,
|
| 483 |
+
"loss": 0.658,
|
| 484 |
+
"step": 48800
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 1.2035621557351193,
|
| 488 |
+
"grad_norm": 18.016963958740234,
|
| 489 |
+
"learning_rate": 4.191826309067689e-06,
|
| 490 |
+
"loss": 0.6753,
|
| 491 |
+
"step": 49600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.2229744485695566,
|
| 495 |
+
"grad_norm": 44.54015350341797,
|
| 496 |
+
"learning_rate": 4.089655172413794e-06,
|
| 497 |
+
"loss": 0.6644,
|
| 498 |
+
"step": 50400
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 1.242386741403994,
|
| 502 |
+
"grad_norm": 33.76991653442383,
|
| 503 |
+
"learning_rate": 3.987484035759898e-06,
|
| 504 |
+
"loss": 0.6583,
|
| 505 |
+
"step": 51200
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.2617990342384315,
|
| 509 |
+
"grad_norm": 36.015037536621094,
|
| 510 |
+
"learning_rate": 3.885312899106003e-06,
|
| 511 |
+
"loss": 0.6653,
|
| 512 |
+
"step": 52000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.2812113270728689,
|
| 516 |
+
"grad_norm": 31.974958419799805,
|
| 517 |
+
"learning_rate": 3.7831417624521076e-06,
|
| 518 |
+
"loss": 0.6666,
|
| 519 |
+
"step": 52800
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.3006236199073062,
|
| 523 |
+
"grad_norm": 23.773361206054688,
|
| 524 |
+
"learning_rate": 3.6809706257982126e-06,
|
| 525 |
+
"loss": 0.6696,
|
| 526 |
+
"step": 53600
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.3200359127417438,
|
| 530 |
+
"grad_norm": 23.204721450805664,
|
| 531 |
+
"learning_rate": 3.578799489144317e-06,
|
| 532 |
+
"loss": 0.6641,
|
| 533 |
+
"step": 54400
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.3394482055761812,
|
| 537 |
+
"grad_norm": 35.860713958740234,
|
| 538 |
+
"learning_rate": 3.4766283524904217e-06,
|
| 539 |
+
"loss": 0.6572,
|
| 540 |
+
"step": 55200
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.3588604984106185,
|
| 544 |
+
"grad_norm": 31.097198486328125,
|
| 545 |
+
"learning_rate": 3.3744572158365263e-06,
|
| 546 |
+
"loss": 0.6562,
|
| 547 |
+
"step": 56000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 1.3588604984106185,
|
| 551 |
+
"eval_loss": 0.6831924915313721,
|
| 552 |
+
"eval_runtime": 1619.9018,
|
| 553 |
+
"eval_samples_per_second": 6.337,
|
| 554 |
+
"eval_steps_per_second": 6.337,
|
| 555 |
+
"step": 56000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.3782727912450559,
|
| 559 |
+
"grad_norm": 24.997142791748047,
|
| 560 |
+
"learning_rate": 3.272286079182631e-06,
|
| 561 |
+
"loss": 0.6618,
|
| 562 |
+
"step": 56800
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.3976850840794932,
|
| 566 |
+
"grad_norm": 30.415740966796875,
|
| 567 |
+
"learning_rate": 3.170114942528736e-06,
|
| 568 |
+
"loss": 0.6653,
|
| 569 |
+
"step": 57600
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.4170973769139308,
|
| 573 |
+
"grad_norm": 49.423728942871094,
|
| 574 |
+
"learning_rate": 3.0679438058748405e-06,
|
| 575 |
+
"loss": 0.6456,
|
| 576 |
+
"step": 58400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.4365096697483681,
|
| 580 |
+
"grad_norm": 33.79461669921875,
|
| 581 |
+
"learning_rate": 2.9657726692209455e-06,
|
| 582 |
+
"loss": 0.6619,
|
| 583 |
+
"step": 59200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.4559219625828055,
|
| 587 |
+
"grad_norm": 41.93768310546875,
|
| 588 |
+
"learning_rate": 2.86360153256705e-06,
|
| 589 |
+
"loss": 0.6592,
|
| 590 |
+
"step": 60000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.475334255417243,
|
| 594 |
+
"grad_norm": 36.79194641113281,
|
| 595 |
+
"learning_rate": 2.761430395913155e-06,
|
| 596 |
+
"loss": 0.6457,
|
| 597 |
+
"step": 60800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.4947465482516804,
|
| 601 |
+
"grad_norm": 39.22713088989258,
|
| 602 |
+
"learning_rate": 2.6592592592592592e-06,
|
| 603 |
+
"loss": 0.6484,
|
| 604 |
+
"step": 61600
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5141588410861178,
|
| 608 |
+
"grad_norm": 57.07029724121094,
|
| 609 |
+
"learning_rate": 2.557088122605364e-06,
|
| 610 |
+
"loss": 0.6599,
|
| 611 |
+
"step": 62400
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.5335711339205553,
|
| 615 |
+
"grad_norm": 34.55560302734375,
|
| 616 |
+
"learning_rate": 2.4549169859514692e-06,
|
| 617 |
+
"loss": 0.6526,
|
| 618 |
+
"step": 63200
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.5529834267549925,
|
| 622 |
+
"grad_norm": 47.0441780090332,
|
| 623 |
+
"learning_rate": 2.3527458492975734e-06,
|
| 624 |
+
"loss": 0.6605,
|
| 625 |
+
"step": 64000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.5529834267549925,
|
| 629 |
+
"eval_loss": 0.6782345175743103,
|
| 630 |
+
"eval_runtime": 1619.8932,
|
| 631 |
+
"eval_samples_per_second": 6.337,
|
| 632 |
+
"eval_steps_per_second": 6.337,
|
| 633 |
+
"step": 64000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.57239571958943,
|
| 637 |
+
"grad_norm": 36.22556686401367,
|
| 638 |
+
"learning_rate": 2.2505747126436784e-06,
|
| 639 |
+
"loss": 0.6685,
|
| 640 |
+
"step": 64800
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5918080124238674,
|
| 644 |
+
"grad_norm": 51.00742721557617,
|
| 645 |
+
"learning_rate": 2.148403575989783e-06,
|
| 646 |
+
"loss": 0.6507,
|
| 647 |
+
"step": 65600
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6112203052583047,
|
| 651 |
+
"grad_norm": 36.90612030029297,
|
| 652 |
+
"learning_rate": 2.046232439335888e-06,
|
| 653 |
+
"loss": 0.6397,
|
| 654 |
+
"step": 66400
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.6306325980927423,
|
| 658 |
+
"grad_norm": 46.38152313232422,
|
| 659 |
+
"learning_rate": 1.9440613026819926e-06,
|
| 660 |
+
"loss": 0.65,
|
| 661 |
+
"step": 67200
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.6500448909271797,
|
| 665 |
+
"grad_norm": 41.90031051635742,
|
| 666 |
+
"learning_rate": 1.8418901660280971e-06,
|
| 667 |
+
"loss": 0.6554,
|
| 668 |
+
"step": 68000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.669457183761617,
|
| 672 |
+
"grad_norm": 30.752347946166992,
|
| 673 |
+
"learning_rate": 1.739719029374202e-06,
|
| 674 |
+
"loss": 0.6513,
|
| 675 |
+
"step": 68800
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.6888694765960546,
|
| 679 |
+
"grad_norm": 34.301795959472656,
|
| 680 |
+
"learning_rate": 1.6375478927203067e-06,
|
| 681 |
+
"loss": 0.6467,
|
| 682 |
+
"step": 69600
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.7082817694304917,
|
| 686 |
+
"grad_norm": 46.13114929199219,
|
| 687 |
+
"learning_rate": 1.5353767560664113e-06,
|
| 688 |
+
"loss": 0.6582,
|
| 689 |
+
"step": 70400
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.7276940622649293,
|
| 693 |
+
"grad_norm": 30.393938064575195,
|
| 694 |
+
"learning_rate": 1.4332056194125161e-06,
|
| 695 |
+
"loss": 0.6535,
|
| 696 |
+
"step": 71200
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.7471063550993666,
|
| 700 |
+
"grad_norm": 37.74268341064453,
|
| 701 |
+
"learning_rate": 1.331034482758621e-06,
|
| 702 |
+
"loss": 0.6474,
|
| 703 |
+
"step": 72000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.7471063550993666,
|
| 707 |
+
"eval_loss": 0.6739329695701599,
|
| 708 |
+
"eval_runtime": 1620.1308,
|
| 709 |
+
"eval_samples_per_second": 6.336,
|
| 710 |
+
"eval_steps_per_second": 6.336,
|
| 711 |
+
"step": 72000
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.766518647933804,
|
| 715 |
+
"grad_norm": 67.96341705322266,
|
| 716 |
+
"learning_rate": 1.2288633461047255e-06,
|
| 717 |
+
"loss": 0.6557,
|
| 718 |
+
"step": 72800
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.7859309407682415,
|
| 722 |
+
"grad_norm": 26.30314064025879,
|
| 723 |
+
"learning_rate": 1.1266922094508303e-06,
|
| 724 |
+
"loss": 0.6527,
|
| 725 |
+
"step": 73600
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.805343233602679,
|
| 729 |
+
"grad_norm": 46.09438705444336,
|
| 730 |
+
"learning_rate": 1.024521072796935e-06,
|
| 731 |
+
"loss": 0.6589,
|
| 732 |
+
"step": 74400
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.8247555264371162,
|
| 736 |
+
"grad_norm": 28.36471176147461,
|
| 737 |
+
"learning_rate": 9.223499361430396e-07,
|
| 738 |
+
"loss": 0.6506,
|
| 739 |
+
"step": 75200
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.8441678192715538,
|
| 743 |
+
"grad_norm": 50.00242614746094,
|
| 744 |
+
"learning_rate": 8.201787994891444e-07,
|
| 745 |
+
"loss": 0.6441,
|
| 746 |
+
"step": 76000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8635801121059912,
|
| 750 |
+
"grad_norm": 27.380474090576172,
|
| 751 |
+
"learning_rate": 7.18007662835249e-07,
|
| 752 |
+
"loss": 0.6662,
|
| 753 |
+
"step": 76800
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.8829924049404285,
|
| 757 |
+
"grad_norm": 43.15554428100586,
|
| 758 |
+
"learning_rate": 6.158365261813538e-07,
|
| 759 |
+
"loss": 0.6374,
|
| 760 |
+
"step": 77600
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.902404697774866,
|
| 764 |
+
"grad_norm": 41.739227294921875,
|
| 765 |
+
"learning_rate": 5.136653895274585e-07,
|
| 766 |
+
"loss": 0.6569,
|
| 767 |
+
"step": 78400
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.9218169906093032,
|
| 771 |
+
"grad_norm": 38.57143783569336,
|
| 772 |
+
"learning_rate": 4.1149425287356324e-07,
|
| 773 |
+
"loss": 0.6534,
|
| 774 |
+
"step": 79200
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9412292834437408,
|
| 778 |
+
"grad_norm": 52.96192932128906,
|
| 779 |
+
"learning_rate": 3.09323116219668e-07,
|
| 780 |
+
"loss": 0.6537,
|
| 781 |
+
"step": 80000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.9412292834437408,
|
| 785 |
+
"eval_loss": 0.6711751222610474,
|
| 786 |
+
"eval_runtime": 1619.9483,
|
| 787 |
+
"eval_samples_per_second": 6.337,
|
| 788 |
+
"eval_steps_per_second": 6.337,
|
| 789 |
+
"step": 80000
|
| 790 |
+
}
|
| 791 |
+
],
|
| 792 |
+
"logging_steps": 800,
|
| 793 |
+
"max_steps": 82422,
|
| 794 |
+
"num_input_tokens_seen": 0,
|
| 795 |
+
"num_train_epochs": 2,
|
| 796 |
+
"save_steps": 8000,
|
| 797 |
+
"stateful_callbacks": {
|
| 798 |
+
"TrainerControl": {
|
| 799 |
+
"args": {
|
| 800 |
+
"should_epoch_stop": false,
|
| 801 |
+
"should_evaluate": false,
|
| 802 |
+
"should_log": false,
|
| 803 |
+
"should_save": true,
|
| 804 |
+
"should_training_stop": false
|
| 805 |
+
},
|
| 806 |
+
"attributes": {}
|
| 807 |
+
}
|
| 808 |
+
},
|
| 809 |
+
"total_flos": 1.0083501150921034e+19,
|
| 810 |
+
"train_batch_size": 2,
|
| 811 |
+
"trial_name": null,
|
| 812 |
+
"trial_params": null
|
| 813 |
+
}
|