Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -210,3 +210,5 @@
 {"step": 4200, "tokens": 2202009600, "loss": 2.9926197052001955, "tokens_per_second": 984405.9331565329, "lr": 0.0015, "gradient_norm": 0.09584445506334305, "benchmark": false}
 {"step": 4220, "tokens": 2212495360, "loss": 2.98980655670166, "tokens_per_second": 849521.3373798685, "lr": 0.0015, "gradient_norm": 0.09558536857366562, "benchmark": false}
 {"step": 4240, "tokens": 2222981120, "loss": 2.976529121398926, "tokens_per_second": 983884.5041944423, "lr": 0.0015, "gradient_norm": 0.11071575433015823, "benchmark": false}
+{"step": 4260, "tokens": 2233466880, "loss": 2.9626989364624023, "tokens_per_second": 985666.2047822862, "lr": 0.0015, "gradient_norm": 0.09352339804172516, "benchmark": false}
+{"step": 4280, "tokens": 2243952640, "loss": 2.9729793548583983, "tokens_per_second": 983842.6010859351, "lr": 0.0015, "gradient_norm": 0.0999748557806015, "benchmark": false}