Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -6975,3 +6975,31 @@
 {"timestamp": 1787553475.478936, "elapsed_hours": 7.45877511911922, "session_id": 2, "event": "train_step", "step": 61740, "epoch": 1, "loss": 1.8567187225818633, "loss_raw": 2.36076021194458, "perplexity": 6.40269325271328, "muon_lr": 0.006841885107205121, "adamw_lr": 0.0002052565532161536, "tokens_per_sec": 150750.35592325442, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4046192640, "tokens_processed_billions": 4.04619264}
 {"timestamp": 1787553479.4555717, "elapsed_hours": 7.459879740013017, "session_id": 2, "event": "train_step", "step": 61750, "epoch": 1, "loss": 1.7446818947792053, "loss_raw": 1.9710958003997803, "perplexity": 5.724080325942154, "muon_lr": 0.006841155683067292, "adamw_lr": 0.0002052346704920187, "tokens_per_sec": 150752.43749176932, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 4046848000, "tokens_processed_billions": 4.046848}
 {"timestamp": 1787553483.5825102, "elapsed_hours": 7.4610261119074295, "session_id": 2, "event": "train_step", "step": 61760, "epoch": 1, "loss": 1.6756164562702178, "loss_raw": 1.9585497379302979, "perplexity": 5.342087297446728, "muon_lr": 0.006839696718786292, "adamw_lr": 0.00020519090156358868, "tokens_per_sec": 150753.67446103718, "vram_gb": 27.299458026885986, "grad_norm": 0.1318359375, "tokens_processed": 4047503360, "tokens_processed_billions": 4.04750336}
+{"timestamp": 1787553488.6914327, "elapsed_hours": 7.462445257107417, "session_id": 2, "event": "train_step", "step": 61770, "epoch": 1, "loss": 1.6931996142864227, "loss_raw": 0.7947161197662354, "perplexity": 5.436848723683346, "muon_lr": 0.006838967178685192, "adamw_lr": 0.0002051690153605557, "tokens_per_sec": 150749.39906292252, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 4048158720, "tokens_processed_billions": 4.04815872}
+{"timestamp": 1787553492.7745078, "elapsed_hours": 7.463579444554117, "session_id": 2, "event": "train_step", "step": 61780, "epoch": 1, "loss": 1.7593056070804596, "loss_raw": 2.1753721237182617, "perplexity": 5.808402680031311, "muon_lr": 0.006838237599971744, "adamw_lr": 0.00020514712799915224, "tokens_per_sec": 150750.88164991135, "vram_gb": 27.710740089416504, "grad_norm": 0.14453125, "tokens_processed": 4048814080, "tokens_processed_billions": 4.04881408}
+{"timestamp": 1787553496.9029582, "elapsed_hours": 7.464726236330138, "session_id": 2, "event": "train_step", "step": 61790, "epoch": 1, "loss": 1.7735970151424407, "loss_raw": 2.751047134399414, "perplexity": 5.892008935472575, "muon_lr": 0.0068375079826669865, "adamw_lr": 0.00020512523948000954, "tokens_per_sec": 150752.11016045127, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 4049469440, "tokens_processed_billions": 4.04946944}
+{"timestamp": 1787553501.122957, "elapsed_hours": 7.465898458361625, "session_id": 2, "event": "train_step", "step": 61800, "epoch": 1, "loss": 1.863937805891037, "loss_raw": 0.8885688185691833, "perplexity": 6.449082069611378, "muon_lr": 0.006836048632367707, "adamw_lr": 0.00020508145897103115, "tokens_per_sec": 150752.82410945004, "vram_gb": 27.299458026885986, "grad_norm": 0.13671875, "tokens_processed": 4050124800, "tokens_processed_billions": 4.0501248}
+{"timestamp": 1787553501.1311002, "elapsed_hours": 7.46590072049035, "session_id": 2, "event": "lambda_stats", "step": 61800, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787553505.1106524, "elapsed_hours": 7.467006151411269, "session_id": 2, "event": "train_step", "step": 61810, "epoch": 1, "loss": 2.0025165736675263, "loss_raw": 1.3506616353988647, "perplexity": 7.4076746205515684, "muon_lr": 0.006835318899415266, "adamw_lr": 0.00020505956698245792, "tokens_per_sec": 150754.84176719855, "vram_gb": 27.710740089416504, "grad_norm": 0.154296875, "tokens_processed": 4050780160, "tokens_processed_billions": 4.05078016}
+{"timestamp": 1787553509.0823622, "elapsed_hours": 7.468109404179785, "session_id": 2, "event": "train_step", "step": 61820, "epoch": 1, "loss": 2.04408452630043, "loss_raw": 1.7278276681900024, "perplexity": 7.722085934433811, "muon_lr": 0.006834589127955685, "adamw_lr": 0.0002050376738386705, "tokens_per_sec": 150756.9482711366, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 4051435520, "tokens_processed_billions": 4.05143552}
+{"timestamp": 1787553513.1601532, "elapsed_hours": 7.469242123895222, "session_id": 2, "event": "train_step", "step": 61830, "epoch": 1, "loss": 1.9243993580341339, "loss_raw": 1.836055040359497, "perplexity": 6.851032412249282, "muon_lr": 0.0068338593180100075, "adamw_lr": 0.00020501577954030016, "tokens_per_sec": 150758.4591073906, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 4052090880, "tokens_processed_billions": 4.05209088}
+{"timestamp": 1787553517.2936282, "elapsed_hours": 7.470390311413341, "session_id": 2, "event": "train_step", "step": 61840, "epoch": 1, "loss": 1.946780380010605, "loss_raw": 2.3237972259521484, "perplexity": 7.006094268012778, "muon_lr": 0.006832399582744543, "adamw_lr": 0.00020497198748233624, "tokens_per_sec": 150759.65624700306, "vram_gb": 27.299458026885986, "grad_norm": 0.150390625, "tokens_processed": 4052746240, "tokens_processed_billions": 4.05274624}
+{"timestamp": 1787553521.3777118, "elapsed_hours": 7.471524779068099, "session_id": 2, "event": "train_step", "step": 61850, "epoch": 1, "loss": 1.9998998737335205, "loss_raw": 1.814191222190857, "perplexity": 7.388316297368059, "muon_lr": 0.006831669657466854, "adamw_lr": 0.00020495008972400556, "tokens_per_sec": 150761.13124312786, "vram_gb": 27.710740089416504, "grad_norm": 0.1572265625, "tokens_processed": 4053401600, "tokens_processed_billions": 4.0534016}
+{"timestamp": 1787553525.352648, "elapsed_hours": 7.472628927959336, "session_id": 2, "event": "train_step", "step": 61860, "epoch": 1, "loss": 1.8856453335285186, "loss_raw": 3.0661919116973877, "perplexity": 6.590606209123587, "muon_lr": 0.006830939693787258, "adamw_lr": 0.00020492819081361767, "tokens_per_sec": 150763.21742853906, "vram_gb": 27.710740089416504, "grad_norm": 0.134765625, "tokens_processed": 4054056960, "tokens_processed_billions": 4.05405696}
+{"timestamp": 1787553529.4101417, "elapsed_hours": 7.473756009538969, "session_id": 2, "event": "train_step", "step": 61870, "epoch": 1, "loss": 1.8436147367954254, "loss_raw": 1.5863964557647705, "perplexity": 6.319339775471566, "muon_lr": 0.006830209691726805, "adamw_lr": 0.00020490629075180407, "tokens_per_sec": 150764.84014325202, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 4054712320, "tokens_processed_billions": 4.05471232}
+{"timestamp": 1787553533.531663, "elapsed_hours": 7.47490087668101, "session_id": 2, "event": "train_step", "step": 61880, "epoch": 1, "loss": 1.9116361010074616, "loss_raw": 1.5562570095062256, "perplexity": 6.764146577272865, "muon_lr": 0.006828749572547533, "adamw_lr": 0.00020486248717642593, "tokens_per_sec": 150766.10286510832, "vram_gb": 27.299458026885986, "grad_norm": 0.142578125, "tokens_processed": 4055367680, "tokens_processed_billions": 4.05536768}
+{"timestamp": 1787553537.5354762, "elapsed_hours": 7.476013046966659, "session_id": 2, "event": "train_step", "step": 61890, "epoch": 1, "loss": 1.8626719653606414, "loss_raw": 2.015446424484253, "perplexity": 6.4409237248149465, "muon_lr": 0.006828019455470822, "adamw_lr": 0.0002048405836641246, "tokens_per_sec": 150768.02556974924, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 4056023040, "tokens_processed_billions": 4.05602304}
+{"timestamp": 1787553541.5235035, "elapsed_hours": 7.477120832535956, "session_id": 2, "event": "train_step", "step": 61900, "epoch": 1, "loss": 1.8629288375377655, "loss_raw": 1.820663571357727, "perplexity": 6.442578431429777, "muon_lr": 0.006827289300097466, "adamw_lr": 0.00020481867900292393, "tokens_per_sec": 150770.03630958943, "vram_gb": 27.710740089416504, "grad_norm": 0.1474609375, "tokens_processed": 4056678400, "tokens_processed_billions": 4.0566784}
+{"timestamp": 1787553541.5297081, "elapsed_hours": 7.477122555971146, "session_id": 2, "event": "lambda_stats", "step": 61900, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787553545.5906866, "elapsed_hours": 7.478250605331527, "session_id": 2, "event": "train_step", "step": 61910, "epoch": 1, "loss": 1.884461257457733, "loss_raw": 2.07759952545166, "perplexity": 6.582807048330452, "muon_lr": 0.006826559106448521, "adamw_lr": 0.00020479677319345559, "tokens_per_sec": 150771.60239654692, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4057333760, "tokens_processed_billions": 4.05733376}
+{"timestamp": 1787553549.7343237, "elapsed_hours": 7.479401615659396, "session_id": 2, "event": "train_step", "step": 61920, "epoch": 1, "loss": 1.9448426020145417, "loss_raw": 1.7672866582870483, "perplexity": 6.992531158093962, "muon_lr": 0.006825098604408094, "adamw_lr": 0.00020475295813224276, "tokens_per_sec": 150772.73970760478, "vram_gb": 27.299458026885986, "grad_norm": 0.1533203125, "tokens_processed": 4057989120, "tokens_processed_billions": 4.05798912}
+{"timestamp": 1787553553.8325405, "elapsed_hours": 7.480540009273423, "session_id": 2, "event": "train_step", "step": 61930, "epoch": 1, "loss": 1.920338863134384, "loss_raw": 2.3944954872131348, "perplexity": 6.823270232320669, "muon_lr": 0.006824368296058729, "adamw_lr": 0.0002047310488817618, "tokens_per_sec": 150774.1320281005, "vram_gb": 27.710740089416504, "grad_norm": 0.1416015625, "tokens_processed": 4058644480, "tokens_processed_billions": 4.05864448}
+{"timestamp": 1787553557.9156048, "elapsed_hours": 7.481674193806119, "session_id": 2, "event": "train_step", "step": 61940, "epoch": 1, "loss": 1.9080090487003327, "loss_raw": 1.6022257804870605, "perplexity": 6.739657102973756, "muon_lr": 0.006823637949518008, "adamw_lr": 0.00020470913848554017, "tokens_per_sec": 150775.60768137113, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 4059299840, "tokens_processed_billions": 4.05929984}
+{"timestamp": 1787553561.966096, "elapsed_hours": 7.482799330155054, "session_id": 2, "event": "train_step", "step": 61950, "epoch": 1, "loss": 1.8574263906478883, "loss_raw": 0.9616332650184631, "perplexity": 6.407225837858503, "muon_lr": 0.006822907564806994, "adamw_lr": 0.00020468722694420976, "tokens_per_sec": 150777.26574442693, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4059955200, "tokens_processed_billions": 4.0599552}
+{"timestamp": 1787553566.143642, "elapsed_hours": 7.483959759540028, "session_id": 2, "event": "train_step", "step": 61960, "epoch": 1, "loss": 1.8051075959205627, "loss_raw": 1.7494616508483887, "perplexity": 6.080625663840156, "muon_lr": 0.006821446680958338, "adamw_lr": 0.00020464340042875007, "tokens_per_sec": 150778.2114859591, "vram_gb": 27.299458026885986, "grad_norm": 0.134765625, "tokens_processed": 4060610560, "tokens_processed_billions": 4.06061056}
+{"timestamp": 1787553570.1323674, "elapsed_hours": 7.48506773882442, "session_id": 2, "event": "train_step", "step": 61970, "epoch": 1, "loss": 1.806875511407852, "loss_raw": 2.2316887378692627, "perplexity": 6.091385204300342, "muon_lr": 0.006820716181862823, "adamw_lr": 0.00020462148545588463, "tokens_per_sec": 150780.21491842173, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 4061265920, "tokens_processed_billions": 4.06126592}
+{"timestamp": 1787553574.1221783, "elapsed_hours": 7.486176019708315, "session_id": 2, "event": "train_step", "step": 61980, "epoch": 1, "loss": 1.7975877475738526, "loss_raw": 1.8524448871612549, "perplexity": 6.0350717748288085, "muon_lr": 0.006819985644681272, "adamw_lr": 0.0002045995693404381, "tokens_per_sec": 150782.21151096828, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 4061921280, "tokens_processed_billions": 4.06192128}
+{"timestamp": 1787553578.1688304, "elapsed_hours": 7.487300089796384, "session_id": 2, "event": "train_step", "step": 61990, "epoch": 1, "loss": 1.8612461161613465, "loss_raw": 2.8734285831451416, "perplexity": 6.431746483115377, "muon_lr": 0.0068192550694347485, "adamw_lr": 0.00020457765208304238, "tokens_per_sec": 150783.88831867598, "vram_gb": 27.710740089416504, "grad_norm": 0.12158203125, "tokens_processed": 4062576640, "tokens_processed_billions": 4.06257664}
+{"timestamp": 1787553582.3549106, "elapsed_hours": 7.488462889790535, "session_id": 2, "event": "train_step", "step": 62000, "epoch": 1, "loss": 1.8511174249649047, "loss_raw": 1.6314724683761597, "perplexity": 6.366930115745943, "muon_lr": 0.006817793804831066, "adamw_lr": 0.00020453381414493193, "tokens_per_sec": 150784.7851341746, "vram_gb": 27.299458026885986, "grad_norm": 0.14453125, "tokens_processed": 4063232000, "tokens_processed_billions": 4.063232}
+{"timestamp": 1787553582.358314, "elapsed_hours": 7.488463835318883, "session_id": 2, "event": "lambda_stats", "step": 62000, "epoch": 1, "lambda_min": 1.078125, "lambda_max": 2.015625, "lambda_mean": 1.1892361111111112, "lambda_std": 0.20446532211957086, "lambda_all": [2.015625, 1.25, 1.1875, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.1171875, 1.140625, 1.109375, 1.109375, 1.078125, 1.1015625, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787553595.8974645, "elapsed_hours": 7.492224710716141, "session_id": 2, "event": "validation", "step": 62000, "epoch": 1, "val_loss": 1.8865914559364318, "val_perplexity": 6.596844680052977, "is_best": true, "best_loss": 1.8865914559364318, "best_perplexity": 6.596844680052977, "train_loss": 1.6314724683761597, "train_loss_avg": 1.8511174249649047}