Upload metrics_session_002.jsonl with huggingface_hub
Browse files- metrics_session_002.jsonl +56 -0
metrics_session_002.jsonl
CHANGED
|
@@ -563,3 +563,59 @@
|
|
| 563 |
{"timestamp": 1787528789.5252805, "elapsed_hours": 0.6015657703081767, "session_id": 2, "event": "train_step", "step": 5000, "epoch": 1, "loss": 2.727938725948334, "loss_raw": 2.1073598861694336, "perplexity": 15.301314288960437, "muon_lr": 0.009990884823293127, "adamw_lr": 0.00029972654469879377, "tokens_per_sec": 152098.13425009273, "vram_gb": 27.299458026885986, "grad_norm": 0.228515625, "tokens_processed": 327680000, "tokens_processed_billions": 0.32768}
|
| 564 |
{"timestamp": 1787528789.5309296, "elapsed_hours": 0.601567339764701, "session_id": 2, "event": "lambda_stats", "step": 5000, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2300347222222223, "lambda_std": 0.12358745677232225, "lambda_all": [1.6484375, 1.390625, 1.3125, 1.3125, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.140625, 1.234375], "local_layers": 9, "global_layers": 9}
|
| 565 |
{"timestamp": 1787528802.9331946, "elapsed_hours": 0.6052901911735534, "session_id": 2, "event": "validation", "step": 5000, "epoch": 1, "val_loss": 2.8289759135246277, "val_perplexity": 16.928116090061515, "is_best": true, "best_loss": 2.8289759135246277, "best_perplexity": 16.928116090061515, "train_loss": 2.1073598861694336, "train_loss_avg": 2.727938725948334}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 563 |
{"timestamp": 1787528789.5252805, "elapsed_hours": 0.6015657703081767, "session_id": 2, "event": "train_step", "step": 5000, "epoch": 1, "loss": 2.727938725948334, "loss_raw": 2.1073598861694336, "perplexity": 15.301314288960437, "muon_lr": 0.009990884823293127, "adamw_lr": 0.00029972654469879377, "tokens_per_sec": 152098.13425009273, "vram_gb": 27.299458026885986, "grad_norm": 0.228515625, "tokens_processed": 327680000, "tokens_processed_billions": 0.32768}
|
| 564 |
{"timestamp": 1787528789.5309296, "elapsed_hours": 0.601567339764701, "session_id": 2, "event": "lambda_stats", "step": 5000, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2300347222222223, "lambda_std": 0.12358745677232225, "lambda_all": [1.6484375, 1.390625, 1.3125, 1.3125, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.140625, 1.234375], "local_layers": 9, "global_layers": 9}
|
| 565 |
{"timestamp": 1787528802.9331946, "elapsed_hours": 0.6052901911735534, "session_id": 2, "event": "validation", "step": 5000, "epoch": 1, "val_loss": 2.8289759135246277, "val_perplexity": 16.928116090061515, "is_best": true, "best_loss": 2.8289759135246277, "best_perplexity": 16.928116090061515, "train_loss": 2.1073598861694336, "train_loss_avg": 2.727938725948334}
|
| 566 |
+
{"timestamp": 1787528807.2732413, "elapsed_hours": 0.6064957594209247, "session_id": 2, "event": "train_step", "step": 5010, "epoch": 1, "loss": 2.712250416278839, "loss_raw": 3.7293801307678223, "perplexity": 15.063135726649064, "muon_lr": 0.009990836144478946, "adamw_lr": 0.0002997250843343683, "tokens_per_sec": 151157.09923600766, "vram_gb": 27.710740089416504, "grad_norm": 0.2578125, "tokens_processed": 328335360, "tokens_processed_billions": 0.32833536}
|
| 567 |
+
{"timestamp": 1787528811.2351675, "elapsed_hours": 0.6075962945487764, "session_id": 2, "event": "train_step", "step": 5020, "epoch": 1, "loss": 2.7055386304855347, "loss_raw": 1.8473526239395142, "perplexity": 14.962373711069558, "muon_lr": 0.00999078733616067, "adamw_lr": 0.00029972362008482003, "tokens_per_sec": 151183.05983099822, "vram_gb": 27.710740089416504, "grad_norm": 0.251953125, "tokens_processed": 328990720, "tokens_processed_billions": 0.32899072}
|
| 568 |
+
{"timestamp": 1787528815.201365, "elapsed_hours": 0.6086980160739687, "session_id": 2, "event": "train_step", "step": 5030, "epoch": 1, "loss": 2.735182282924652, "loss_raw": 2.7709784507751465, "perplexity": 15.412552625056788, "muon_lr": 0.009990738398339707, "adamw_lr": 0.0002997221519501912, "tokens_per_sec": 151208.62991464953, "vram_gb": 27.710740089416504, "grad_norm": 0.2578125, "tokens_processed": 329646080, "tokens_processed_billions": 0.32964608}
|
| 569 |
+
{"timestamp": 1787528819.2981215, "elapsed_hours": 0.6098360039128198, "session_id": 2, "event": "train_step", "step": 5040, "epoch": 1, "loss": 2.7040195631980897, "loss_raw": 1.737249732017517, "perplexity": 14.939662113212595, "muon_lr": 0.009990640134195372, "adamw_lr": 0.00029971920402586114, "tokens_per_sec": 151225.05548014274, "vram_gb": 27.299458026885986, "grad_norm": 0.2255859375, "tokens_processed": 330301440, "tokens_processed_billions": 0.33030144}
|
| 570 |
+
{"timestamp": 1787528823.2603357, "elapsed_hours": 0.6109366189108955, "session_id": 2, "event": "train_step", "step": 5050, "epoch": 1, "loss": 2.792148549556732, "loss_raw": 3.146662712097168, "perplexity": 16.316037985346988, "muon_lr": 0.009990590807874837, "adamw_lr": 0.000299717724236245, "tokens_per_sec": 151250.7298966898, "vram_gb": 27.710740089416504, "grad_norm": 0.2431640625, "tokens_processed": 330956800, "tokens_processed_billions": 0.3309568}
|
| 571 |
+
{"timestamp": 1787528827.2185793, "elapsed_hours": 0.6120361311568154, "session_id": 2, "event": "train_step", "step": 5060, "epoch": 1, "loss": 2.6829313969612123, "loss_raw": 3.9356565475463867, "perplexity": 14.627910720820855, "muon_lr": 0.009990541352057278, "adamw_lr": 0.0002997162405617183, "tokens_per_sec": 151276.58061070173, "vram_gb": 27.710740089416504, "grad_norm": 0.236328125, "tokens_processed": 331612160, "tokens_processed_billions": 0.33161216}
|
| 572 |
+
{"timestamp": 1787528831.1749084, "elapsed_hours": 0.6131351113319397, "session_id": 2, "event": "train_step", "step": 5070, "epoch": 1, "loss": 2.632625732421875, "loss_raw": 2.380391836166382, "perplexity": 13.910246577613014, "muon_lr": 0.009990491766744127, "adamw_lr": 0.0002997147530023238, "tokens_per_sec": 151302.47440911288, "vram_gb": 27.710740089416504, "grad_norm": 0.2314453125, "tokens_processed": 332267520, "tokens_processed_billions": 0.33226752}
|
| 573 |
+
{"timestamp": 1787528835.2745762, "elapsed_hours": 0.6142739080720477, "session_id": 2, "event": "train_step", "step": 5080, "epoch": 1, "loss": 2.581952555179596, "loss_raw": 2.7909765243530273, "perplexity": 13.222931473367293, "muon_lr": 0.009990392207636777, "adamw_lr": 0.0002997117662291032, "tokens_per_sec": 151318.41018567473, "vram_gb": 27.299458026885986, "grad_norm": 0.2431640625, "tokens_processed": 332922880, "tokens_processed_billions": 0.33292288}
|
| 574 |
+
{"timestamp": 1787528839.2627404, "elapsed_hours": 0.6153817315896352, "session_id": 2, "event": "train_step", "step": 5090, "epoch": 1, "loss": 2.5685708785057066, "loss_raw": 3.9370381832122803, "perplexity": 13.047165126774619, "muon_lr": 0.009990342233845442, "adamw_lr": 0.00029971026701536323, "tokens_per_sec": 151341.95227362297, "vram_gb": 27.710740089416504, "grad_norm": 0.232421875, "tokens_processed": 333578240, "tokens_processed_billions": 0.33357824}
|
| 575 |
+
{"timestamp": 1787528843.2218785, "elapsed_hours": 0.6164814919895596, "session_id": 2, "event": "train_step", "step": 5100, "epoch": 1, "loss": 2.5933403038978575, "loss_raw": 2.924309492111206, "perplexity": 13.374371538439124, "muon_lr": 0.009990292130564259, "adamw_lr": 0.0002997087639169277, "tokens_per_sec": 151367.3926409371, "vram_gb": 27.710740089416504, "grad_norm": 0.279296875, "tokens_processed": 334233600, "tokens_processed_billions": 0.3342336}
|
| 576 |
+
{"timestamp": 1787528843.2254605, "elapsed_hours": 0.6164824872546726, "session_id": 2, "event": "lambda_stats", "step": 5100, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.65625, "lambda_mean": 1.2296006944444444, "lambda_std": 0.1250851188297103, "lambda_all": [1.65625, 1.390625, 1.3125, 1.3125, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1953125, 1.140625, 1.21875, 1.140625, 1.234375], "local_layers": 9, "global_layers": 9}
|
| 577 |
+
{"timestamp": 1787528847.192861, "elapsed_hours": 0.6175845427645578, "session_id": 2, "event": "train_step", "step": 5110, "epoch": 1, "loss": 2.5670921874046324, "loss_raw": 1.915006399154663, "perplexity": 13.027886656771171, "muon_lr": 0.009990241897794669, "adamw_lr": 0.00029970725693384, "tokens_per_sec": 151391.9314310033, "vram_gb": 27.710740089416504, "grad_norm": 0.232421875, "tokens_processed": 334888960, "tokens_processed_billions": 0.33488896}
|
| 578 |
+
{"timestamp": 1787528851.2719817, "elapsed_hours": 0.6187176317638821, "session_id": 2, "event": "train_step", "step": 5120, "epoch": 1, "loss": 2.558055188655853, "loss_raw": 2.918266534805298, "perplexity": 12.910684038944499, "muon_lr": 0.009990141043796068, "adamw_lr": 0.00029970423131388203, "tokens_per_sec": 151408.99405568096, "vram_gb": 27.299458026885986, "grad_norm": 0.2216796875, "tokens_processed": 335544320, "tokens_processed_billions": 0.33554432}
|
| 579 |
+
{"timestamp": 1787528855.2273033, "elapsed_hours": 0.6198163322607676, "session_id": 2, "event": "train_step", "step": 5130, "epoch": 1, "loss": 2.638927869796753, "loss_raw": 3.5663506984710693, "perplexity": 13.998187679854773, "muon_lr": 0.009990090422569969, "adamw_lr": 0.00029970271267709903, "tokens_per_sec": 151434.44618656332, "vram_gb": 27.710740089416504, "grad_norm": 0.283203125, "tokens_processed": 336199680, "tokens_processed_billions": 0.33619968}
|
| 580 |
+
{"timestamp": 1787528859.188793, "elapsed_hours": 0.6209167460600535, "session_id": 2, "event": "train_step", "step": 5140, "epoch": 1, "loss": 2.5815890765190126, "loss_raw": 3.846585988998413, "perplexity": 13.218126093325825, "muon_lr": 0.009990039671861281, "adamw_lr": 0.0002997011901558384, "tokens_per_sec": 151459.37374202002, "vram_gb": 27.710740089416504, "grad_norm": 0.2421875, "tokens_processed": 336855040, "tokens_processed_billions": 0.33685504}
|
| 581 |
+
{"timestamp": 1787528863.1529198, "elapsed_hours": 0.6220178923341964, "session_id": 2, "event": "train_step", "step": 5150, "epoch": 1, "loss": 2.571561210155487, "loss_raw": 2.679042100906372, "perplexity": 13.086238870202232, "muon_lr": 0.009989988791671466, "adamw_lr": 0.00029969966375014394, "tokens_per_sec": 151484.04019180994, "vram_gb": 27.710740089416504, "grad_norm": 0.32421875, "tokens_processed": 337510400, "tokens_processed_billions": 0.3375104}
|
| 582 |
+
{"timestamp": 1787528867.2501369, "elapsed_hours": 0.6231560082568063, "session_id": 2, "event": "train_step", "step": 5160, "epoch": 1, "loss": 2.6997007060050966, "loss_raw": 2.6323206424713135, "perplexity": 14.87527897689541, "muon_lr": 0.009989886642854332, "adamw_lr": 0.0002996965992856299, "tokens_per_sec": 151499.57854898804, "vram_gb": 27.299458026885986, "grad_norm": 0.228515625, "tokens_processed": 338165760, "tokens_processed_billions": 0.33816576}
|
| 583 |
+
{"timestamp": 1787528871.1888456, "elapsed_hours": 0.6242500938971838, "session_id": 2, "event": "train_step", "step": 5170, "epoch": 1, "loss": 2.74244656085968, "loss_raw": 2.1216063499450684, "perplexity": 15.524921335647432, "muon_lr": 0.009989835374229959, "adamw_lr": 0.0002996950612268987, "tokens_per_sec": 151525.81235901316, "vram_gb": 27.710740089416504, "grad_norm": 0.2578125, "tokens_processed": 338821120, "tokens_processed_billions": 0.33882112}
|
| 584 |
+
{"timestamp": 1787528875.1126726, "elapsed_hours": 0.6253400458229913, "session_id": 2, "event": "train_step", "step": 5180, "epoch": 1, "loss": 2.673216452598572, "loss_raw": 2.6850380897521973, "perplexity": 14.486489444221627, "muon_lr": 0.009989783976130352, "adamw_lr": 0.00029969351928391054, "tokens_per_sec": 151552.95020214198, "vram_gb": 27.710740089416504, "grad_norm": 0.2236328125, "tokens_processed": 339476480, "tokens_processed_billions": 0.33947648}
|
| 585 |
+
{"timestamp": 1787528879.054301, "elapsed_hours": 0.6264349426825842, "session_id": 2, "event": "train_step", "step": 5190, "epoch": 1, "loss": 2.7744814419746398, "loss_raw": 2.807326078414917, "perplexity": 16.03031219295147, "muon_lr": 0.009989732448556995, "adamw_lr": 0.00029969197345670976, "tokens_per_sec": 151578.80031532602, "vram_gb": 27.710740089416504, "grad_norm": 0.236328125, "tokens_processed": 340131840, "tokens_processed_billions": 0.34013184}
|
| 586 |
+
{"timestamp": 1787528883.1463723, "elapsed_hours": 0.6275716293520398, "session_id": 2, "event": "train_step", "step": 5200, "epoch": 1, "loss": 2.8235015654563904, "loss_raw": 2.1558587551116943, "perplexity": 16.835698883224055, "muon_lr": 0.00998962900499497, "adamw_lr": 0.000299688870149849, "tokens_per_sec": 151594.40303745828, "vram_gb": 27.299458026885986, "grad_norm": 0.248046875, "tokens_processed": 340787200, "tokens_processed_billions": 0.3407872}
|
| 587 |
+
{"timestamp": 1787528883.1497371, "elapsed_hours": 0.6275725638866425, "session_id": 2, "event": "lambda_stats", "step": 5200, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.671875, "lambda_mean": 1.2300347222222223, "lambda_std": 0.1277959396060139, "lambda_all": [1.671875, 1.390625, 1.3125, 1.3046875, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1953125, 1.140625, 1.21875, 1.140625, 1.234375], "local_layers": 9, "global_layers": 9}
|
| 588 |
+
{"timestamp": 1787528887.1027079, "elapsed_hours": 0.6286706113815308, "session_id": 2, "event": "train_step", "step": 5210, "epoch": 1, "loss": 2.8356874418258666, "loss_raw": 2.1367340087890625, "perplexity": 17.042111734877675, "muon_lr": 0.009989577089009286, "adamw_lr": 0.0002996873126702785, "tokens_per_sec": 151619.09605065515, "vram_gb": 27.710740089416504, "grad_norm": 0.2431640625, "tokens_processed": 341442560, "tokens_processed_billions": 0.34144256}
|
| 589 |
+
{"timestamp": 1787528891.0613654, "elapsed_hours": 0.6297702383332783, "session_id": 2, "event": "train_step", "step": 5220, "epoch": 1, "loss": 2.885050985813141, "loss_raw": 2.501446485519409, "perplexity": 17.90448044674309, "muon_lr": 0.009989525043555814, "adamw_lr": 0.00029968575130667437, "tokens_per_sec": 151643.5460838528, "vram_gb": 27.710740089416504, "grad_norm": 0.255859375, "tokens_processed": 342097920, "tokens_processed_billions": 0.34209792}
|
| 590 |
+
{"timestamp": 1787528895.036709, "elapsed_hours": 0.6308745004733404, "session_id": 2, "event": "train_step", "step": 5230, "epoch": 1, "loss": 2.7751115012168883, "loss_raw": 2.1683425903320312, "perplexity": 16.040415421786893, "muon_lr": 0.009989472868636058, "adamw_lr": 0.0002996841860590817, "tokens_per_sec": 151666.7896650112, "vram_gb": 27.710740089416504, "grad_norm": 0.2314453125, "tokens_processed": 342753280, "tokens_processed_billions": 0.34275328}
|
| 591 |
+
{"timestamp": 1787528899.139629, "elapsed_hours": 0.6320142003562715, "session_id": 2, "event": "train_step", "step": 5240, "epoch": 1, "loss": 2.8116997599601747, "loss_raw": 3.6971373558044434, "perplexity": 16.638175100871877, "muon_lr": 0.009989368130403718, "adamw_lr": 0.00029968104391211146, "tokens_per_sec": 151681.39679877445, "vram_gb": 27.299458026885986, "grad_norm": 0.298828125, "tokens_processed": 343408640, "tokens_processed_billions": 0.34340864}
|
| 592 |
+
{"timestamp": 1787528903.099482, "elapsed_hours": 0.6331141596370273, "session_id": 2, "event": "train_step", "step": 5250, "epoch": 1, "loss": 2.6810176730155946, "loss_raw": 1.913793683052063, "perplexity": 14.59994370690756, "muon_lr": 0.009989315567094151, "adamw_lr": 0.0002996794670128245, "tokens_per_sec": 151705.52958944842, "vram_gb": 27.710740089416504, "grad_norm": 0.2275390625, "tokens_processed": 344064000, "tokens_processed_billions": 0.344064}
|
| 593 |
+
{"timestamp": 1787528907.052977, "elapsed_hours": 0.6342123526334763, "session_id": 2, "event": "train_step", "step": 5260, "epoch": 1, "loss": 2.7800517463684082, "loss_raw": 4.027855396270752, "perplexity": 16.11985507040512, "muon_lr": 0.00998926287432434, "adamw_lr": 0.0002996778862297301, "tokens_per_sec": 151729.99417334565, "vram_gb": 27.710740089416504, "grad_norm": 0.265625, "tokens_processed": 344719360, "tokens_processed_billions": 0.34471936}
|
| 594 |
+
{"timestamp": 1787528910.9918752, "elapsed_hours": 0.635306491057078, "session_id": 2, "event": "train_step", "step": 5270, "epoch": 1, "loss": 2.6843652057647707, "loss_raw": 1.8055270910263062, "perplexity": 14.64889939126454, "muon_lr": 0.009989210052095802, "adamw_lr": 0.000299676301562874, "tokens_per_sec": 151755.3555706946, "vram_gb": 27.710740089416504, "grad_norm": 0.23828125, "tokens_processed": 345374720, "tokens_processed_billions": 0.34537472}
|
| 595 |
+
{"timestamp": 1787528915.1123981, "elapsed_hours": 0.6364510807063845, "session_id": 2, "event": "train_step", "step": 5280, "epoch": 1, "loss": 2.746584997177124, "loss_raw": 4.249378681182861, "perplexity": 15.58930336250002, "muon_lr": 0.009989104019268655, "adamw_lr": 0.00029967312057805956, "tokens_per_sec": 151768.5315979119, "vram_gb": 27.299458026885986, "grad_norm": 0.2216796875, "tokens_processed": 346030080, "tokens_processed_billions": 0.34603008}
|
| 596 |
+
{"timestamp": 1787528919.0306594, "elapsed_hours": 0.6375394866863886, "session_id": 2, "event": "train_step", "step": 5290, "epoch": 1, "loss": 2.779190149307251, "loss_raw": 4.287704944610596, "perplexity": 16.105972232215294, "muon_lr": 0.009989050808673097, "adamw_lr": 0.00029967152426019287, "tokens_per_sec": 151795.108626509, "vram_gb": 27.710740089416504, "grad_norm": 0.2177734375, "tokens_processed": 346685440, "tokens_processed_billions": 0.34668544}
|
| 597 |
+
{"timestamp": 1787528923.0059452, "elapsed_hours": 0.6386437326669693, "session_id": 2, "event": "train_step", "step": 5300, "epoch": 1, "loss": 2.8262028980255125, "loss_raw": 2.4319584369659424, "perplexity": 16.881239187001096, "muon_lr": 0.009988997468624933, "adamw_lr": 0.00029966992405874794, "tokens_per_sec": 151817.80443753337, "vram_gb": 27.710740089416504, "grad_norm": 0.2255859375, "tokens_processed": 347340800, "tokens_processed_billions": 0.3473408}
|
| 598 |
+
{"timestamp": 1787528923.010747, "elapsed_hours": 0.6386450666189194, "session_id": 2, "event": "lambda_stats", "step": 5300, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.671875, "lambda_mean": 1.2291666666666667, "lambda_std": 0.12768386116803399, "lambda_all": [1.671875, 1.390625, 1.3125, 1.296875, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1875, 1.140625, 1.21875, 1.140625, 1.234375], "local_layers": 9, "global_layers": 9}
|
| 599 |
+
{"timestamp": 1787528926.9633498, "elapsed_hours": 0.6397430118587282, "session_id": 2, "event": "train_step", "step": 5310, "epoch": 1, "loss": 2.873494052886963, "loss_raw": 3.8444886207580566, "perplexity": 17.698750659630996, "muon_lr": 0.009988943999125695, "adamw_lr": 0.00029966831997377085, "tokens_per_sec": 151841.61054968726, "vram_gb": 27.710740089416504, "grad_norm": 0.216796875, "tokens_processed": 347996160, "tokens_processed_billions": 0.34799616}
|
| 600 |
+
{"timestamp": 1787528931.0556421, "elapsed_hours": 0.6408797596560584, "session_id": 2, "event": "train_step", "step": 5320, "epoch": 1, "loss": 2.8711449813842775, "loss_raw": 2.8478708267211914, "perplexity": 17.657223822673867, "muon_lr": 0.009988836671780185, "adamw_lr": 0.0002996651001534055, "tokens_per_sec": 151856.40966548194, "vram_gb": 27.299458026885986, "grad_norm": 0.296875, "tokens_processed": 348651520, "tokens_processed_billions": 0.34865152}
|
| 601 |
+
{"timestamp": 1787528935.0033264, "elapsed_hours": 0.6419763386249542, "session_id": 2, "event": "train_step", "step": 5330, "epoch": 1, "loss": 2.89708842754364, "loss_raw": 2.999046802520752, "perplexity": 18.121306987083898, "muon_lr": 0.009988782813937003, "adamw_lr": 0.00029966348441811, "tokens_per_sec": 151880.70909170303, "vram_gb": 27.710740089416504, "grad_norm": 0.224609375, "tokens_processed": 349306880, "tokens_processed_billions": 0.34930688}
|
| 602 |
+
{"timestamp": 1787528938.9390717, "elapsed_hours": 0.6430696011914148, "session_id": 2, "event": "train_step", "step": 5340, "epoch": 1, "loss": 2.703244264125824, "loss_raw": 1.7863348722457886, "perplexity": 14.92808389590679, "muon_lr": 0.009988728826648939, "adamw_lr": 0.0002996618647994681, "tokens_per_sec": 151905.70505571697, "vram_gb": 27.710740089416504, "grad_norm": 0.283203125, "tokens_processed": 349962240, "tokens_processed_billions": 0.34996224}
|
| 603 |
+
{"timestamp": 1787528942.8801684, "elapsed_hours": 0.6441643502977159, "session_id": 2, "event": "train_step", "step": 5350, "epoch": 1, "loss": 2.7047951316833494, "loss_raw": 2.9664103984832764, "perplexity": 14.951253338641589, "muon_lr": 0.009988674709917551, "adamw_lr": 0.0002996602412975265, "tokens_per_sec": 151930.2718861096, "vram_gb": 27.710740089416504, "grad_norm": 0.2109375, "tokens_processed": 350617600, "tokens_processed_billions": 0.3506176}
|
| 604 |
+
{"timestamp": 1787528946.9374135, "elapsed_hours": 0.6452913628021876, "session_id": 2, "event": "train_step", "step": 5360, "epoch": 1, "loss": 2.5537634968757628, "loss_raw": 1.6656100749969482, "perplexity": 12.85539409095056, "muon_lr": 0.009988566088131048, "adamw_lr": 0.0002996569826439314, "tokens_per_sec": 151947.11232967267, "vram_gb": 27.299458026885986, "grad_norm": 0.2265625, "tokens_processed": 351272960, "tokens_processed_billions": 0.35127296}
|
| 605 |
+
{"timestamp": 1787528950.8878245, "elapsed_hours": 0.6463886992798911, "session_id": 2, "event": "train_step", "step": 5370, "epoch": 1, "loss": 2.6436325407028196, "loss_raw": 2.3982813358306885, "perplexity": 14.064199706842508, "muon_lr": 0.009988511583079066, "adamw_lr": 0.00029965534749237197, "tokens_per_sec": 151970.9129820023, "vram_gb": 27.710740089416504, "grad_norm": 0.2138671875, "tokens_processed": 351928320, "tokens_processed_billions": 0.35192832}
|
| 606 |
+
{"timestamp": 1787528954.8443253, "elapsed_hours": 0.647487727271186, "session_id": 2, "event": "train_step", "step": 5380, "epoch": 1, "loss": 2.671121642589569, "loss_raw": 3.0508737564086914, "perplexity": 14.456174763972005, "muon_lr": 0.009988456948590026, "adamw_lr": 0.00029965370845770075, "tokens_per_sec": 151994.2299335789, "vram_gb": 27.710740089416504, "grad_norm": 0.224609375, "tokens_processed": 352583680, "tokens_processed_billions": 0.35258368}
|
| 607 |
+
{"timestamp": 1787528958.8188472, "elapsed_hours": 0.6485917610592312, "session_id": 2, "event": "train_step", "step": 5390, "epoch": 1, "loss": 2.8090044546127317, "loss_raw": 2.5917983055114746, "perplexity": 16.59339051972343, "muon_lr": 0.009988402184665502, "adamw_lr": 0.000299652065539965, "tokens_per_sec": 152016.28714135854, "vram_gb": 27.710740089416504, "grad_norm": 0.2294921875, "tokens_processed": 353239040, "tokens_processed_billions": 0.35323904}
|
| 608 |
+
{"timestamp": 1787528962.93365, "elapsed_hours": 0.6497347618473901, "session_id": 2, "event": "train_step", "step": 5400, "epoch": 1, "loss": 2.8310869240760805, "loss_raw": 2.6949806213378906, "perplexity": 16.96388926723623, "muon_lr": 0.009988292268516323, "adamw_lr": 0.0002996487680554896, "tokens_per_sec": 152029.10059938632, "vram_gb": 27.299458026885986, "grad_norm": 0.220703125, "tokens_processed": 353894400, "tokens_processed_billions": 0.3538944}
|
| 609 |
+
{"timestamp": 1787528962.9397132, "elapsed_hours": 0.6497364464071061, "session_id": 2, "event": "lambda_stats", "step": 5400, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6875, "lambda_mean": 1.2291666666666667, "lambda_std": 0.13072812914594928, "lambda_all": [1.6875, 1.390625, 1.3125, 1.296875, 1.1875, 1.1875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1875, 1.140625, 1.21875, 1.140625, 1.21875], "local_layers": 9, "global_layers": 9}
|
| 610 |
+
{"timestamp": 1787528966.9050379, "elapsed_hours": 0.6508379251427121, "session_id": 2, "event": "train_step", "step": 5410, "epoch": 1, "loss": 2.882933223247528, "loss_raw": 2.4243099689483643, "perplexity": 17.86660313003432, "muon_lr": 0.009988237116294839, "adamw_lr": 0.00029964711348884506, "tokens_per_sec": 152051.2364419711, "vram_gb": 27.710740089416504, "grad_norm": 0.2392578125, "tokens_processed": 354549760, "tokens_processed_billions": 0.35454976}
|
| 611 |
+
{"timestamp": 1787528970.8522484, "elapsed_hours": 0.6519343724515703, "session_id": 2, "event": "train_step", "step": 5420, "epoch": 1, "loss": 2.788612003326416, "loss_raw": 2.8179421424865723, "perplexity": 16.258437476182873, "muon_lr": 0.009988181834644207, "adamw_lr": 0.0002996454550393261, "tokens_per_sec": 152074.8564075762, "vram_gb": 27.710740089416504, "grad_norm": 0.2177734375, "tokens_processed": 355205120, "tokens_processed_billions": 0.35520512}
|
| 612 |
+
{"timestamp": 1787528974.7822182, "elapsed_hours": 0.6530260307259029, "session_id": 2, "event": "train_step", "step": 5430, "epoch": 1, "loss": 2.8787228512763976, "loss_raw": 4.497424602508545, "perplexity": 17.791536225685334, "muon_lr": 0.009988126423566026, "adamw_lr": 0.0002996437927069807, "tokens_per_sec": 152099.5241881428, "vram_gb": 27.710740089416504, "grad_norm": 0.234375, "tokens_processed": 355860480, "tokens_processed_billions": 0.35586048}
|
| 613 |
+
{"timestamp": 1787528978.8807242, "elapsed_hours": 0.6541645046737459, "session_id": 2, "event": "train_step", "step": 5440, "epoch": 1, "loss": 2.7661106133461, "loss_raw": 1.4701323509216309, "perplexity": 15.896685261073348, "muon_lr": 0.00998801521313341, "adamw_lr": 0.00029964045639400223, "tokens_per_sec": 152113.16428753684, "vram_gb": 27.299458026885986, "grad_norm": 0.2353515625, "tokens_processed": 356515840, "tokens_processed_billions": 0.35651584}
|
| 614 |
+
{"timestamp": 1787528982.8260055, "elapsed_hours": 0.6552604161368476, "session_id": 2, "event": "train_step", "step": 5450, "epoch": 1, "loss": 2.7497043132781984, "loss_raw": 2.8641462326049805, "perplexity": 15.638007249401232, "muon_lr": 0.00998795941378218, "adamw_lr": 0.00029963878241346536, "tokens_per_sec": 152136.690614423, "vram_gb": 27.710740089416504, "grad_norm": 0.220703125, "tokens_processed": 357171200, "tokens_processed_billions": 0.3571712}
|
| 615 |
+
{"timestamp": 1787528986.7645538, "elapsed_hours": 0.6563544574711058, "session_id": 2, "event": "train_step", "step": 5460, "epoch": 1, "loss": 2.7103208780288695, "loss_raw": 2.77577543258667, "perplexity": 15.03409885298818, "muon_lr": 0.009987903485009816, "adamw_lr": 0.0002996371045502944, "tokens_per_sec": 152160.57536266375, "vram_gb": 27.710740089416504, "grad_norm": 0.2421875, "tokens_processed": 357826560, "tokens_processed_billions": 0.35782656}
|
| 616 |
+
{"timestamp": 1787528990.718815, "elapsed_hours": 0.6574528632561366, "session_id": 2, "event": "train_step", "step": 5470, "epoch": 1, "loss": 2.8462850069999694, "loss_raw": 4.6618547439575195, "perplexity": 17.223677000287147, "muon_lr": 0.00998784742681793, "adamw_lr": 0.0002996354228045378, "tokens_per_sec": 152183.35888726445, "vram_gb": 27.710740089416504, "grad_norm": 0.2333984375, "tokens_processed": 358481920, "tokens_processed_billions": 0.35848192}
|
| 617 |
+
{"timestamp": 1787528994.7924721, "elapsed_hours": 0.658584434721205, "session_id": 2, "event": "train_step", "step": 5480, "epoch": 1, "loss": 2.669209656715393, "loss_raw": 1.8402153253555298, "perplexity": 14.428561168842812, "muon_lr": 0.009987734922182051, "adamw_lr": 0.00029963204766546146, "tokens_per_sec": 152198.36691253335, "vram_gb": 27.299458026885986, "grad_norm": 0.23828125, "tokens_processed": 359137280, "tokens_processed_billions": 0.35913728}
|
| 618 |
+
{"timestamp": 1787528998.7414734, "elapsed_hours": 0.6596813794639376, "session_id": 2, "event": "train_step", "step": 5490, "epoch": 1, "loss": 2.662991147041321, "loss_raw": 3.270798683166504, "perplexity": 14.339115419497938, "muon_lr": 0.009987678475741307, "adamw_lr": 0.0002996303542722391, "tokens_per_sec": 152221.35323352742, "vram_gb": 27.710740089416504, "grad_norm": 0.21875, "tokens_processed": 359792640, "tokens_processed_billions": 0.35979264}
|
| 619 |
+
{"timestamp": 1787529002.6720486, "elapsed_hours": 0.6607732059558232, "session_id": 2, "event": "train_step", "step": 5500, "epoch": 1, "loss": 2.7483660340309144, "loss_raw": 2.843714714050293, "perplexity": 15.617093226353395, "muon_lr": 0.009987621899887528, "adamw_lr": 0.00029962865699662575, "tokens_per_sec": 152245.44949650436, "vram_gb": 27.710740089416504, "grad_norm": 0.2294921875, "tokens_processed": 360448000, "tokens_processed_billions": 0.360448}
|
| 620 |
+
{"timestamp": 1787529002.6819, "elapsed_hours": 0.6607759425375197, "session_id": 2, "event": "lambda_stats", "step": 5500, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6953125, "lambda_mean": 1.2287326388888888, "lambda_std": 0.13214203612304679, "lambda_all": [1.6953125, 1.390625, 1.3046875, 1.296875, 1.1875, 1.1796875, 1.140625, 1.25, 1.1875, 1.140625, 1.15625, 1.140625, 1.140625, 1.1875, 1.140625, 1.21875, 1.140625, 1.21875], "local_layers": 9, "global_layers": 9}
|
| 621 |
+
{"timestamp": 1787529016.0739782, "elapsed_hours": 0.6644959642489752, "session_id": 2, "event": "validation", "step": 5500, "epoch": 1, "val_loss": 2.756121916770935, "val_perplexity": 15.738688499965537, "is_best": true, "best_loss": 2.756121916770935, "best_perplexity": 15.738688499965537, "train_loss": 2.843714714050293, "train_loss_avg": 2.7483660340309144}
|