Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -620,3 +620,5 @@
 {"step": 12400, "tokens": 6501171200, "loss": 2.78804988861084, "tokens_per_second": 960203.670722355, "lr": 0.0015, "gradient_norm": 0.08776278793811798, "benchmark": false}
 {"step": 12420, "tokens": 6511656960, "loss": 2.8205455780029296, "tokens_per_second": 981290.9366722134, "lr": 0.0015, "gradient_norm": 0.10064391046762466, "benchmark": false}
 {"step": 12440, "tokens": 6522142720, "loss": 2.7763082504272463, "tokens_per_second": 950548.1511525006, "lr": 0.0015, "gradient_norm": 0.10321742296218872, "benchmark": false}
+{"step": 12460, "tokens": 6532628480, "loss": 2.7925830841064454, "tokens_per_second": 941279.0475299946, "lr": 0.0015, "gradient_norm": 0.09877070784568787, "benchmark": false}
+{"step": 12480, "tokens": 6543114240, "loss": 2.778891944885254, "tokens_per_second": 850421.2073633118, "lr": 0.0015, "gradient_norm": 0.11233720183372498, "benchmark": false}