Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -247,3 +247,5 @@
 {"step": 4940, "tokens": 2589982720, "loss": 2.9586591720581055, "tokens_per_second": 964416.464912028, "lr": 0.0015, "gradient_norm": 0.09584059566259384, "benchmark": false}
 {"step": 4960, "tokens": 2600468480, "loss": 2.9198270797729493, "tokens_per_second": 967224.2776713663, "lr": 0.0015, "gradient_norm": 0.09711223840713501, "benchmark": false}
 {"step": 4980, "tokens": 2610954240, "loss": 2.9437145233154296, "tokens_per_second": 980325.3360152998, "lr": 0.0015, "gradient_norm": 0.08578862994909286, "benchmark": false}
+{"step": 5000, "tokens": 2621440000, "loss": 2.933154296875, "tokens_per_second": 969211.9684460608, "lr": 0.0015, "gradient_norm": 0.09016783535480499, "benchmark": false}
+{"step": 5020, "tokens": 2631925760, "loss": 2.91949462890625, "tokens_per_second": 807039.9908925168, "lr": 0.0015, "gradient_norm": 0.0832776427268982, "benchmark": false}

reports/run/validation.jsonl

@@ -2,3 +2,4 @@
 {"step": 2000, "source_validation_loss": {"fineweb_edu": 3.3628950119018555, "dclm_edu": 3.3812646865844727, "cosmopedia": 2.551811456680298, "finemath": 2.6517066955566406}}
 {"step": 3000, "source_validation_loss": {"fineweb_edu": 3.2158007621765137, "dclm_edu": 3.24115252494812, "cosmopedia": 2.362274169921875, "finemath": 2.494218349456787}}
 {"step": 4000, "source_validation_loss": {"fineweb_edu": 3.1462671756744385, "dclm_edu": 3.168818950653076, "cosmopedia": 2.2698538303375244, "finemath": 2.4097580909729004}}
+{"step": 5000, "source_validation_loss": {"fineweb_edu": 3.097052812576294, "dclm_edu": 3.1254994869232178, "cosmopedia": 2.2071025371551514, "finemath": 2.3563427925109863}}