Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -496,3 +496,5 @@
 {"step": 9920, "tokens": 5200936960, "loss": 2.8068157196044923, "tokens_per_second": 995466.1046283207, "lr": 0.0015, "gradient_norm": 0.09151347726583481, "benchmark": false}
 {"step": 9940, "tokens": 5211422720, "loss": 2.83557071685791, "tokens_per_second": 988697.6764334616, "lr": 0.0015, "gradient_norm": 0.08406423777341843, "benchmark": false}
 {"step": 9960, "tokens": 5221908480, "loss": 2.83525390625, "tokens_per_second": 985306.7973998113, "lr": 0.0015, "gradient_norm": 0.07898715138435364, "benchmark": false}
+{"step": 9980, "tokens": 5232394240, "loss": 2.809004211425781, "tokens_per_second": 990607.4548488081, "lr": 0.0015, "gradient_norm": 0.07946421951055527, "benchmark": false}
+{"step": 10000, "tokens": 5242880000, "loss": 2.8216739654541017, "tokens_per_second": 986454.10697014, "lr": 0.0015, "gradient_norm": 0.08943367004394531, "benchmark": false}

reports/run/validation.jsonl

@@ -7,3 +7,4 @@
 {"step": 7000, "source_validation_loss": {"fineweb_edu": 3.0427091121673584, "dclm_edu": 3.072791576385498, "cosmopedia": 2.1341612339019775, "finemath": 2.284400701522827}}
 {"step": 8000, "source_validation_loss": {"fineweb_edu": 3.022184133529663, "dclm_edu": 3.0521011352539062, "cosmopedia": 2.1067497730255127, "finemath": 2.2722864151000977}}
 {"step": 9000, "source_validation_loss": {"fineweb_edu": 3.0031962394714355, "dclm_edu": 3.0357918739318848, "cosmopedia": 2.0862884521484375, "finemath": 2.2452752590179443}}
+{"step": 10000, "source_validation_loss": {"fineweb_edu": 2.9892542362213135, "dclm_edu": 3.0245914459228516, "cosmopedia": 2.067286968231201, "finemath": 2.238250970840454}}