Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -1023,3 +1023,7 @@
 {"step": 20460, "tokens": 10726932480, "loss": 2.7352483749389647, "tokens_per_second": 978425.8752704178, "lr": 0.0015, "gradient_norm": 0.15846577286720276, "benchmark": false}
 {"step": 20480, "tokens": 10737418240, "loss": 2.733202362060547, "tokens_per_second": 989329.4571691591, "lr": 0.0015, "gradient_norm": 0.14727097749710083, "benchmark": false}
 {"step": 20500, "tokens": 10747904000, "loss": 2.743163299560547, "tokens_per_second": 990398.8649932749, "lr": 0.0015, "gradient_norm": 0.14675293862819672, "benchmark": false}
+{"step": 20520, "tokens": 10758389760, "loss": 2.7502891540527346, "tokens_per_second": 778152.2190613195, "lr": 0.0015, "gradient_norm": 0.1588241308927536, "benchmark": false}
+{"step": 20540, "tokens": 10768875520, "loss": 2.7529842376708986, "tokens_per_second": 993998.8605152208, "lr": 0.0015, "gradient_norm": 0.13833782076835632, "benchmark": false}
+{"step": 20560, "tokens": 10779361280, "loss": 2.7573177337646486, "tokens_per_second": 984597.292859046, "lr": 0.0015, "gradient_norm": 0.14055199921131134, "benchmark": false}
+{"step": 20580, "tokens": 10789847040, "loss": 2.762322998046875, "tokens_per_second": 983528.5535243816, "lr": 0.0015, "gradient_norm": 0.17151477932929993, "benchmark": false}