Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -597,3 +597,5 @@
 {"step": 11940, "tokens": 6259998720, "loss": 2.7949031829833983, "tokens_per_second": 985018.5760938062, "lr": 0.0015, "gradient_norm": 0.09925438463687897, "benchmark": false}
 {"step": 11960, "tokens": 6270484480, "loss": 2.800166893005371, "tokens_per_second": 989799.1647813246, "lr": 0.0015, "gradient_norm": 0.1009145975112915, "benchmark": false}
 {"step": 11980, "tokens": 6280970240, "loss": 2.8234445571899416, "tokens_per_second": 847536.274677034, "lr": 0.0015, "gradient_norm": 0.10739849507808685, "benchmark": false}
+{"step": 12000, "tokens": 6291456000, "loss": 2.754237174987793, "tokens_per_second": 987605.9263153236, "lr": 0.0015, "gradient_norm": 0.10888316482305527, "benchmark": false}
+{"step": 12020, "tokens": 6301941760, "loss": 2.8089101791381834, "tokens_per_second": 822342.5921110236, "lr": 0.0015, "gradient_norm": 0.10958659648895264, "benchmark": false}

reports/run/validation.jsonl

@@ -9,3 +9,4 @@
 {"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}}
 {"step": 11000, "source_validation_loss": {"fineweb_edu": 2.9761719703674316, "dclm_edu": 3.0146288871765137, "cosmopedia": 2.05277943611145, "finemath": 2.2114815711975098}}
+{"step": 12000, "source_validation_loss": {"fineweb_edu": 2.9684195518493652, "dclm_edu": 3.0029680728912354, "cosmopedia": 2.0343093872070312, "finemath": 2.201632261276245}}