ckpt probe_log.jsonl

probe_log.jsonl

@@ -2780,3 +2780,12 @@
 {"type": "train", "step": 2779, "loss": 3.92891, "gnorm": 4.9688, "lr": 0.0001503359784158556, "tokens": 64445, "L": 4902, "bl2": 384473664, "ckpt_on": true, "samples_seen": 44464, "step_ms": 37533.9, "ts": 1790079120.5, "cache_emptied": true, "ema_loss": 1.26425}
 {"type": "train", "step": 2780, "loss": 2.12467, "gnorm": 2.9688, "lr": 0.00015018513630459375, "tokens": 60742, "L": 4823, "bl2": 372181264, "ckpt_on": true, "samples_seen": 44480, "step_ms": 39433.1, "ts": 1790079182.6, "mem_alloc_gb": 3.71, "mem_reserved_gb": 6.19, "cache_emptied": true, "ema_loss": 1.28145}
 {"type": "train", "step": 2781, "loss": 3.00747, "gnorm": 1.5938, "lr": 0.00015003437518434193, "tokens": 63680, "L": 4884, "bl2": 381655296, "ckpt_on": true, "samples_seen": 44496, "step_ms": 38366.6, "ts": 1790079247.8, "cache_emptied": true, "ema_loss": 1.31597}
+{"type": "train", "step": 2782, "loss": 1.74557, "gnorm": 1.8203, "lr": 0.0001498836951529272, "tokens": 65589, "L": 4823, "bl2": 372181264, "ckpt_on": true, "samples_seen": 44512, "step_ms": 37594.9, "ts": 1790079309.8, "cache_emptied": true, "ema_loss": 1.32457}
+{"type": "train", "step": 2783, "loss": 2.76571, "gnorm": 1.1562, "lr": 0.00014973309630812438, "tokens": 61957, "L": 4823, "bl2": 372181264, "ckpt_on": true, "samples_seen": 44528, "step_ms": 37801.2, "ts": 1790079371.8, "cache_emptied": true, "ema_loss": 1.35339}
+{"type": "train", "step": 2784, "loss": 2.91456, "gnorm": 1.5234, "lr": 0.00014958257874765534, "tokens": 58456, "L": 4902, "bl2": 384473664, "ckpt_on": true, "samples_seen": 44544, "step_ms": 38267.3, "ts": 1790079444.2, "cache_emptied": true, "ema_loss": 1.38461}
+{"type": "train", "step": 2785, "loss": 2.3933, "gnorm": 2.1406, "lr": 0.0001494321425691894, "tokens": 83147, "L": 5854, "bl2": 548309056, "ckpt_on": true, "samples_seen": 44560, "step_ms": 45125.5, "ts": 1790079516.0, "cache_emptied": true, "ema_loss": 1.40479}
+{"type": "train", "step": 2786, "loss": 2.88791, "gnorm": 3.7031, "lr": 0.0001492817878703428, "tokens": 83338, "L": 6168, "bl2": 608707584, "ckpt_on": true, "samples_seen": 44576, "step_ms": 50623.3, "ts": 1790079577.9, "cache_emptied": true, "ema_loss": 1.43445}
+{"type": "train", "step": 2787, "loss": 3.05543, "gnorm": 2.1094, "lr": 0.0001491315147486791, "tokens": 81581, "L": 5464, "bl2": 477684736, "ckpt_on": true, "samples_seen": 44592, "step_ms": 50357.4, "ts": 1790079637.8, "cache_emptied": true, "ema_loss": 1.46687}
+{"type": "train", "step": 2788, "loss": 2.86631, "gnorm": 2.9844, "lr": 0.00014898132330170907, "tokens": 87240, "L": 8744, "bl2": 1223320576, "ckpt_on": true, "samples_seen": 44608, "step_ms": 82033.3, "ts": 1790079731.5, "cache_emptied": true, "ema_loss": 1.49486}
+{"type": "train", "step": 2789, "loss": 2.46715, "gnorm": 2.5156, "lr": 0.0001488312136268901, "tokens": 85281, "L": 7548, "bl2": 911556864, "ckpt_on": true, "samples_seen": 44624, "step_ms": 66935.5, "ts": 1790079808.4, "cache_emptied": true, "ema_loss": 1.5143}
+{"type": "train", "step": 2790, "loss": 1.90468, "gnorm": 1.1797, "lr": 0.0001486811858216268, "tokens": 88131, "L": 9848, "bl2": 1551729664, "ckpt_on": true, "samples_seen": 44640, "step_ms": 101329.9, "ts": 1790079919.8, "cache_emptied": true, "ema_loss": 1.52211}