Update training logs at step 4000

training_logs.jsonl

@@ -84,3 +84,5 @@
 {"step": 3975, "epoch": 2.5078864353312302, "time": "2026-07-09 00:21:05", "logs": {"loss": 0.048152894973754884, "grad_norm": 0.04799109697341919, "learning_rate": 7.350702929022739e-06, "epoch": 2.5078864353312302}}
 {"step": 4000, "epoch": 2.5236593059936907, "time": "2026-07-09 00:21:29", "logs": {"loss": 0.03499542236328125, "grad_norm": 0.6452091336250305, "learning_rate": 7.2182321627378e-06, "epoch": 2.5236593059936907}}
 {"step": 4000, "epoch": 2.5236593059936907, "time": "2026-07-09 00:22:28", "logs": {"eval_loss": 0.09398876130580902, "eval_accuracy": 0.9817034700315458, "eval_precision": 0.9698461538461538, "eval_recall": 0.9943217665615142, "eval_f1": 0.9819314641744549, "eval_runtime": 59.2692, "eval_samples_per_second": 53.485, "eval_steps_per_second": 3.358, "epoch": 2.5236593059936907}}
+{"step": 4000, "epoch": 2.5236593059936907, "time": "2026-07-09 00:22:49", "logs": {"train_runtime": 1116.7016, "train_samples_per_second": 90.817, "train_steps_per_second": 5.677, "total_flos": 1.67654364475392e+16, "train_loss": 0.012189212433993816, "epoch": 2.5236593059936907}}
+{"step": 4000, "epoch": 2.5236593059936907, "time": "2026-07-09 00:23:50", "logs": {"eval_loss": 0.09398876130580902, "eval_accuracy": 0.9817034700315458, "eval_precision": 0.9698461538461538, "eval_recall": 0.9943217665615142, "eval_f1": 0.9819314641744549, "eval_runtime": 61.1784, "eval_samples_per_second": 51.816, "eval_steps_per_second": 3.253, "epoch": 2.5236593059936907}}