Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -146,3 +146,5 @@
 {"step": 2920, "tokens": 1530920960, "loss": 3.062306022644043, "tokens_per_second": 970441.6601681139, "lr": 0.0015, "gradient_norm": 0.1730983555316925, "benchmark": false}
 {"step": 2940, "tokens": 1541406720, "loss": 3.0921581268310545, "tokens_per_second": 964402.069190114, "lr": 0.0015, "gradient_norm": 0.175042524933815, "benchmark": false}
 {"step": 2960, "tokens": 1551892480, "loss": 3.0917444229125977, "tokens_per_second": 976263.4119847118, "lr": 0.0015, "gradient_norm": 0.16529427468776703, "benchmark": false}
+{"step": 2980, "tokens": 1562378240, "loss": 3.107547378540039, "tokens_per_second": 974664.1426940245, "lr": 0.0015, "gradient_norm": 0.1620856076478958, "benchmark": false}
+{"step": 3000, "tokens": 1572864000, "loss": 3.053942108154297, "tokens_per_second": 961120.6402985074, "lr": 0.0015, "gradient_norm": 0.17983736097812653, "benchmark": false}

reports/run/validation.jsonl

@@ -1,2 +1,3 @@
 {"step": 1000, "source_validation_loss": {"fineweb_edu": 3.6357946395874023, "dclm_edu": 3.6318132877349854, "cosmopedia": 2.9174532890319824, "finemath": 2.940213680267334}}
 {"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}}