Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -397,3 +397,5 @@
 {"step": 7940, "tokens": 4162846720, "loss": 2.8642766952514647, "tokens_per_second": 973947.4194234799, "lr": 0.0015, "gradient_norm": 0.07104819267988205, "benchmark": false}
 {"step": 7960, "tokens": 4173332480, "loss": 2.85324649810791, "tokens_per_second": 978055.0444066898, "lr": 0.0015, "gradient_norm": 0.07692700624465942, "benchmark": false}
 {"step": 7980, "tokens": 4183818240, "loss": 2.8597362518310545, "tokens_per_second": 976117.0131650119, "lr": 0.0015, "gradient_norm": 0.07403507828712463, "benchmark": false}
+{"step": 8000, "tokens": 4194304000, "loss": 2.8754581451416015, "tokens_per_second": 992788.4029002639, "lr": 0.0015, "gradient_norm": 0.06957851350307465, "benchmark": false}
+{"step": 8020, "tokens": 4204789760, "loss": 2.8420690536499023, "tokens_per_second": 837867.9176074439, "lr": 0.0015, "gradient_norm": 0.07526803761720657, "benchmark": false}

reports/run/validation.jsonl

@@ -5,3 +5,4 @@
 {"step": 5000, "source_validation_loss": {"fineweb_edu": 3.097052812576294, "dclm_edu": 3.1254994869232178, "cosmopedia": 2.2071025371551514, "finemath": 2.3563427925109863}}
 {"step": 6000, "source_validation_loss": {"fineweb_edu": 3.0642285346984863, "dclm_edu": 3.0997276306152344, "cosmopedia": 2.1647276878356934, "finemath": 2.31894588470459}}
 {"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}}