Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -448,3 +448,7 @@
 {"step": 8960, "tokens": 4697620480, "loss": 2.8529546737670897, "tokens_per_second": 969905.4493514767, "lr": 0.0015, "gradient_norm": 0.08122095465660095, "benchmark": false}
 {"step": 8980, "tokens": 4708106240, "loss": 2.840422821044922, "tokens_per_second": 972420.6085403181, "lr": 0.0015, "gradient_norm": 0.0813227966427803, "benchmark": false}
 {"step": 9000, "tokens": 4718592000, "loss": 2.819308853149414, "tokens_per_second": 984628.671317153, "lr": 0.0015, "gradient_norm": 0.07416989654302597, "benchmark": false}
+{"step": 9020, "tokens": 4729077760, "loss": 2.8245450973510744, "tokens_per_second": 835198.1153051325, "lr": 0.0015, "gradient_norm": 0.06911784410476685, "benchmark": false}
+{"step": 9040, "tokens": 4739563520, "loss": 2.861843490600586, "tokens_per_second": 989974.9290623281, "lr": 0.0015, "gradient_norm": 0.0898256003856659, "benchmark": false}
+{"step": 9060, "tokens": 4750049280, "loss": 2.8607154846191407, "tokens_per_second": 991137.133589742, "lr": 0.0015, "gradient_norm": 0.08151320368051529, "benchmark": false}
+{"step": 9080, "tokens": 4760535040, "loss": 2.8201387405395506, "tokens_per_second": 994661.0410152077, "lr": 0.0015, "gradient_norm": 0.07083109766244888, "benchmark": false}