Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -195,3 +195,5 @@
 {"step": 3900, "tokens": 2044723200, "loss": 2.9944433212280273, "tokens_per_second": 979193.8035194282, "lr": 0.0015, "gradient_norm": 0.10861857980489731, "benchmark": false}
 {"step": 3920, "tokens": 2055208960, "loss": 2.9812715530395506, "tokens_per_second": 981608.6602078674, "lr": 0.0015, "gradient_norm": 0.11419326812028885, "benchmark": false}
 {"step": 3940, "tokens": 2065694720, "loss": 3.01178035736084, "tokens_per_second": 980920.9083517338, "lr": 0.0015, "gradient_norm": 0.09572461247444153, "benchmark": false}
+{"step": 3960, "tokens": 2076180480, "loss": 3.002391815185547, "tokens_per_second": 982260.303099604, "lr": 0.0015, "gradient_norm": 0.11502530425786972, "benchmark": false}
+{"step": 3980, "tokens": 2086666240, "loss": 2.9722618103027343, "tokens_per_second": 936205.2487208435, "lr": 0.0015, "gradient_norm": 0.13280029594898224, "benchmark": false}