Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -1049,3 +1049,7 @@
 {"step": 20980, "tokens": 10999562240, "loss": 2.7450157165527345, "tokens_per_second": 987344.8799541212, "lr": 0.0015, "gradient_norm": 0.1675918698310852, "benchmark": false}
 {"step": 21000, "tokens": 11010048000, "loss": 2.7119800567626955, "tokens_per_second": 992182.775109319, "lr": 0.0015, "gradient_norm": 0.14855486154556274, "benchmark": false}
 {"step": 21020, "tokens": 11020533760, "loss": 2.7412237167358398, "tokens_per_second": 823605.7211595208, "lr": 0.0015, "gradient_norm": 0.1713978499174118, "benchmark": false}
+{"step": 21040, "tokens": 11031019520, "loss": 2.71264591217041, "tokens_per_second": 957452.7862116075, "lr": 0.0015, "gradient_norm": 0.16591550409793854, "benchmark": false}
+{"step": 21060, "tokens": 11041505280, "loss": 2.764055633544922, "tokens_per_second": 996040.605801838, "lr": 0.0015, "gradient_norm": 0.17413797974586487, "benchmark": false}
+{"step": 21080, "tokens": 11051991040, "loss": 2.7366050720214843, "tokens_per_second": 984438.3881484806, "lr": 0.0015, "gradient_norm": 0.16170263290405273, "benchmark": false}
+{"step": 21100, "tokens": 11062476800, "loss": 2.7537540435791015, "tokens_per_second": 974425.6193885907, "lr": 0.0015, "gradient_norm": 0.17343269288539886, "benchmark": false}