Update tokenizer, configuration and training reports

reports/run/metrics.jsonl

@@ -847,3 +847,5 @@
 {"step": 16940, "tokens": 8881438720, "loss": 2.787833404541016, "tokens_per_second": 988446.4059700933, "lr": 0.0015, "gradient_norm": 0.13221263885498047, "benchmark": false}
 {"step": 16960, "tokens": 8891924480, "loss": 2.739687919616699, "tokens_per_second": 946188.3591020354, "lr": 0.0015, "gradient_norm": 0.15024445950984955, "benchmark": false}
 {"step": 16980, "tokens": 8902410240, "loss": 2.764535331726074, "tokens_per_second": 975047.8654111625, "lr": 0.0015, "gradient_norm": 0.1423749327659607, "benchmark": false}
+{"step": 17000, "tokens": 8912896000, "loss": 2.7730123519897463, "tokens_per_second": 981977.874173138, "lr": 0.0015, "gradient_norm": 0.13165350258350372, "benchmark": false}
+{"step": 17020, "tokens": 8923381760, "loss": 2.743798828125, "tokens_per_second": 818942.7578559162, "lr": 0.0015, "gradient_norm": 0.13705229759216309, "benchmark": false}

reports/run/validation.jsonl

@@ -14,3 +14,4 @@
 {"step": 14000, "source_validation_loss": {"fineweb_edu": 2.948702096939087, "dclm_edu": 2.9865922927856445, "cosmopedia": 2.0160582065582275, "finemath": 2.18440580368042}}
 {"step": 15000, "source_validation_loss": {"fineweb_edu": 2.9435317516326904, "dclm_edu": 2.9802112579345703, "cosmopedia": 2.0042612552642822, "finemath": 2.18125581741333}}
 {"step": 16000, "source_validation_loss": {"fineweb_edu": 2.9361798763275146, "dclm_edu": 2.9739556312561035, "cosmopedia": 1.9942559003829956, "finemath": 2.171142578125}}
+{"step": 17000, "source_validation_loss": {"fineweb_edu": 2.931060791015625, "dclm_edu": 2.966684341430664, "cosmopedia": 1.9891211986541748, "finemath": 2.1641457080841064}}