Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -199,3 +199,8 @@
 {"step": 3980, "tokens": 2086666240, "loss": 2.9722618103027343, "tokens_per_second": 936205.2487208435, "lr": 0.0015, "gradient_norm": 0.13280029594898224, "benchmark": false}
 {"step": 4000, "tokens": 2097152000, "loss": 3.018246078491211, "tokens_per_second": 980606.9515894684, "lr": 0.0015, "gradient_norm": 0.1104731634259224, "benchmark": false}
 {"step": 4020, "tokens": 2107637760, "loss": 3.0328603744506837, "tokens_per_second": 817031.6501473963, "lr": 0.0015, "gradient_norm": 0.10828843712806702, "benchmark": false}
+{"step": 4040, "tokens": 2118123520, "loss": 2.989377975463867, "tokens_per_second": 980716.3058365815, "lr": 0.0015, "gradient_norm": 0.12038014829158783, "benchmark": false}
+{"step": 4060, "tokens": 2128609280, "loss": 2.9882110595703124, "tokens_per_second": 971402.7047296836, "lr": 0.0015, "gradient_norm": 0.1103377491235733, "benchmark": false}
+{"step": 4080, "tokens": 2139095040, "loss": 2.9773632049560548, "tokens_per_second": 972889.6590334383, "lr": 0.0015, "gradient_norm": 0.1051115095615387, "benchmark": false}
+{"step": 4100, "tokens": 2149580800, "loss": 2.9693840026855467, "tokens_per_second": 839565.0333332941, "lr": 0.0015, "gradient_norm": 0.12314511835575104, "benchmark": false}
+{"step": 4120, "tokens": 2160066560, "loss": 3.0354469299316404, "tokens_per_second": 988168.614991664, "lr": 0.0015, "gradient_norm": 0.11598093062639236, "benchmark": false}