Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -862,3 +862,42 @@
 {"timestamp": 1787529941.055306, "elapsed_hours": 0.9214352218972313, "session_id": 2, "event": "train_step", "step": 7640, "epoch": 1, "loss": 2.4412227296829223, "loss_raw": 1.4502389430999756, "perplexity": 11.487077750248405, "muon_lr": 0.009967800849173248, "adamw_lr": 0.00029903402547519736, "tokens_per_sec": 151453.67341820695, "vram_gb": 27.299458026885986, "grad_norm": 0.1787109375, "tokens_processed": 500695040, "tokens_processed_billions": 0.50069504}
 {"timestamp": 1787529945.0011942, "elapsed_hours": 0.9225313018427955, "session_id": 2, "event": "train_step", "step": 7650, "epoch": 1, "loss": 2.447776403427124, "loss_raw": 3.1908957958221436, "perplexity": 11.562607538555545, "muon_lr": 0.009967709532556424, "adamw_lr": 0.00029903128597669263, "tokens_per_sec": 151471.12471362564, "vram_gb": 27.710740089416504, "grad_norm": 0.203125, "tokens_processed": 501350400, "tokens_processed_billions": 0.5013504}
 {"timestamp": 1787529948.9468982, "elapsed_hours": 0.9236273307932747, "session_id": 2, "event": "train_step", "step": 7660, "epoch": 1, "loss": 2.457958233356476, "loss_raw": 1.7061982154846191, "perplexity": 11.680937427187247, "muon_lr": 0.009967618087102417, "adamw_lr": 0.00029902854261307243, "tokens_per_sec": 151488.5354316091, "vram_gb": 27.710740089416504, "grad_norm": 0.208984375, "tokens_processed": 502005760, "tokens_processed_billions": 0.50200576}
+{"timestamp": 1787529953.4237478, "elapsed_hours": 0.9248709001143773, "session_id": 2, "event": "train_step", "step": 7670, "epoch": 1, "loss": 2.5873436665534975, "loss_raw": 2.9604456424713135, "perplexity": 13.294410271588673, "muon_lr": 0.009967526512813865, "adamw_lr": 0.00029902579538441585, "tokens_per_sec": 151481.65429410094, "vram_gb": 27.710740089416504, "grad_norm": 0.2060546875, "tokens_processed": 502661120, "tokens_processed_billions": 0.50266112}
+{"timestamp": 1787529957.5154293, "elapsed_hours": 0.9260074783696068, "session_id": 2, "event": "train_step", "step": 7680, "epoch": 1, "loss": 2.498497679233551, "loss_raw": 1.3201874494552612, "perplexity": 12.164205687904845, "muon_lr": 0.009967342977743692, "adamw_lr": 0.00029902028933231065, "tokens_per_sec": 151492.35131158476, "vram_gb": 27.299458026885986, "grad_norm": 0.1767578125, "tokens_processed": 503316480, "tokens_processed_billions": 0.50331648}
+{"timestamp": 1787529961.4678745, "elapsed_hours": 0.9271053797668881, "session_id": 2, "event": "train_step", "step": 7690, "epoch": 1, "loss": 2.5831422281265257, "loss_raw": 2.2606849670410156, "perplexity": 13.23867179828316, "muon_lr": 0.009967251016967365, "adamw_lr": 0.00029901753050902083, "tokens_per_sec": 151509.36528104907, "vram_gb": 27.710740089416504, "grad_norm": 0.2236328125, "tokens_processed": 503971840, "tokens_processed_billions": 0.50397184}
+{"timestamp": 1787529965.4252443, "elapsed_hours": 0.9282046492232217, "session_id": 2, "event": "train_step", "step": 7700, "epoch": 1, "loss": 2.434892120361328, "loss_raw": 1.2603949308395386, "perplexity": 11.414587245395353, "muon_lr": 0.009967158927367076, "adamw_lr": 0.00029901476782101224, "tokens_per_sec": 151526.1189030194, "vram_gb": 27.710740089416504, "grad_norm": 0.35546875, "tokens_processed": 504627200, "tokens_processed_billions": 0.5046272}
+{"timestamp": 1787529965.4318888, "elapsed_hours": 0.928206495112843, "session_id": 2, "event": "lambda_stats", "step": 7700, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.796875, "lambda_mean": 1.2183159722222223, "lambda_std": 0.15294845757217546, "lambda_all": [1.796875, 1.375, 1.265625, 1.25, 1.1640625, 1.140625, 1.140625, 1.2421875, 1.171875, 1.140625, 1.140625, 1.140625, 1.140625, 1.1875, 1.140625, 1.1796875, 1.140625, 1.171875], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787529969.3854027, "elapsed_hours": 0.929304693142573, "session_id": 2, "event": "train_step", "step": 7710, "epoch": 1, "loss": 2.4309444975852967, "loss_raw": 3.002472162246704, "perplexity": 11.369615584784437, "muon_lr": 0.009967066708945486, "adamw_lr": 0.00029901200126836446, "tokens_per_sec": 151542.70155223203, "vram_gb": 27.710740089416504, "grad_norm": 0.1748046875, "tokens_processed": 505282560, "tokens_processed_billions": 0.50528256}
+{"timestamp": 1787529973.492313, "elapsed_hours": 0.930445501671897, "session_id": 2, "event": "train_step", "step": 7720, "epoch": 1, "loss": 2.3117968392372132, "loss_raw": 1.9153529405593872, "perplexity": 10.092543049572392, "muon_lr": 0.00996688188564903, "adamw_lr": 0.00029900645656947086, "tokens_per_sec": 151552.58009592045, "vram_gb": 27.299458026885986, "grad_norm": 0.177734375, "tokens_processed": 505937920, "tokens_processed_billions": 0.50593792}
+{"timestamp": 1787529977.4724922, "elapsed_hours": 0.9315511068370608, "session_id": 2, "event": "train_step", "step": 7730, "epoch": 1, "loss": 2.453320188522339, "loss_raw": 2.7090237140655518, "perplexity": 11.626886158691955, "muon_lr": 0.009966789280779497, "adamw_lr": 0.0002990036784233849, "tokens_per_sec": 151568.18493633927, "vram_gb": 27.710740089416504, "grad_norm": 0.1982421875, "tokens_processed": 506593280, "tokens_processed_billions": 0.50659328}
+{"timestamp": 1787529981.425966, "elapsed_hours": 0.9326492942041821, "session_id": 2, "event": "train_step", "step": 7740, "epoch": 1, "loss": 2.4819643187522886, "loss_raw": 1.8125343322753906, "perplexity": 11.9647439205351, "muon_lr": 0.009966696547099321, "adamw_lr": 0.0002990008964129796, "tokens_per_sec": 151584.96167336026, "vram_gb": 27.710740089416504, "grad_norm": 0.2138671875, "tokens_processed": 507248640, "tokens_processed_billions": 0.50724864}
+{"timestamp": 1787529985.3711987, "elapsed_hours": 0.9337451919582155, "session_id": 2, "event": "train_step", "step": 7750, "epoch": 1, "loss": 2.7007857608795165, "loss_raw": 2.4836156368255615, "perplexity": 14.891428230686483, "muon_lr": 0.009966603684611174, "adamw_lr": 0.0002989981105383352, "tokens_per_sec": 151602.07395900684, "vram_gb": 27.710740089416504, "grad_norm": 0.236328125, "tokens_processed": 507904000, "tokens_processed_billions": 0.507904}
+{"timestamp": 1787529989.4836152, "elapsed_hours": 0.9348875300089519, "session_id": 2, "event": "train_step", "step": 7760, "epoch": 1, "loss": 2.6053005933761595, "loss_raw": 2.458993911743164, "perplexity": 13.535293310822485, "muon_lr": 0.009966417573221684, "adamw_lr": 0.00029899252719665047, "tokens_per_sec": 151611.58473906986, "vram_gb": 27.299458026885986, "grad_norm": 0.205078125, "tokens_processed": 508559360, "tokens_processed_billions": 0.50855936}
+{"timestamp": 1787529993.4493282, "elapsed_hours": 0.9359891169601017, "session_id": 2, "event": "train_step", "step": 7770, "epoch": 1, "loss": 2.6401496267318727, "loss_raw": 2.646515369415283, "perplexity": 14.015300514465244, "muon_lr": 0.00996632432432571, "adamw_lr": 0.00029898972972977123, "tokens_per_sec": 151627.70191541064, "vram_gb": 27.710740089416504, "grad_norm": 0.19921875, "tokens_processed": 509214720, "tokens_processed_billions": 0.50921472}
+{"timestamp": 1787529997.3950078, "elapsed_hours": 0.9370851389567058, "session_id": 2, "event": "train_step", "step": 7780, "epoch": 1, "loss": 2.519984004497528, "loss_raw": 2.284712314605713, "perplexity": 12.428397863518846, "muon_lr": 0.009966230946632498, "adamw_lr": 0.0002989869283989749, "tokens_per_sec": 151644.67827691697, "vram_gb": 27.710740089416504, "grad_norm": 0.1806640625, "tokens_processed": 509870080, "tokens_processed_billions": 0.50987008}
+{"timestamp": 1787530001.3367984, "elapsed_hours": 0.9381800808509191, "session_id": 2, "event": "train_step", "step": 7790, "epoch": 1, "loss": 2.430433597564697, "loss_raw": 2.254183530807495, "perplexity": 11.36380833153718, "muon_lr": 0.009966137440144743, "adamw_lr": 0.00029898412320434226, "tokens_per_sec": 151661.78984498105, "vram_gb": 27.710740089416504, "grad_norm": 0.1826171875, "tokens_processed": 510525440, "tokens_processed_billions": 0.51052544}
+{"timestamp": 1787530005.4315834, "elapsed_hours": 0.9393175210555395, "session_id": 2, "event": "train_step", "step": 7800, "epoch": 1, "loss": 2.339638352394104, "loss_raw": 2.5488791465759277, "perplexity": 10.37748289217659, "muon_lr": 0.009965950040796392, "adamw_lr": 0.0002989785012238917, "tokens_per_sec": 151671.97624986924, "vram_gb": 27.299458026885986, "grad_norm": 0.1923828125, "tokens_processed": 511180800, "tokens_processed_billions": 0.5111808}
+{"timestamp": 1787530005.4357393, "elapsed_hours": 0.9393186757299635, "session_id": 2, "event": "lambda_stats", "step": 7800, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.796875, "lambda_mean": 1.2183159722222223, "lambda_std": 0.15294845757217546, "lambda_all": [1.796875, 1.375, 1.265625, 1.25, 1.1640625, 1.140625, 1.140625, 1.2421875, 1.171875, 1.140625, 1.140625, 1.140625, 1.140625, 1.1875, 1.140625, 1.1796875, 1.140625, 1.171875], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787530009.3774076, "elapsed_hours": 0.9404135833846198, "session_id": 2, "event": "train_step", "step": 7810, "epoch": 1, "loss": 2.415656249523163, "loss_raw": 2.9687392711639404, "perplexity": 11.197116049053797, "muon_lr": 0.0099658561479412, "adamw_lr": 0.000298975684438236, "tokens_per_sec": 151688.8389671481, "vram_gb": 27.710740089416504, "grad_norm": 0.173828125, "tokens_processed": 511836160, "tokens_processed_billions": 0.51183616}
+{"timestamp": 1787530013.3162227, "elapsed_hours": 0.941507698694865, "session_id": 2, "event": "train_step", "step": 7820, "epoch": 1, "loss": 2.377842605113983, "loss_raw": 2.4228549003601074, "perplexity": 10.781617548590328, "muon_lr": 0.009965762126302274, "adamw_lr": 0.00029897286378906816, "tokens_per_sec": 151705.97490871942, "vram_gb": 27.710740089416504, "grad_norm": 0.21875, "tokens_processed": 512491520, "tokens_processed_billions": 0.51249152}
+{"timestamp": 1787530017.2782924, "elapsed_hours": 0.9426082736253738, "session_id": 2, "event": "train_step", "step": 7830, "epoch": 1, "loss": 2.498892047405243, "loss_raw": 2.480417490005493, "perplexity": 12.169003809513104, "muon_lr": 0.009965667975882324, "adamw_lr": 0.00029897003927646963, "tokens_per_sec": 151722.0237149424, "vram_gb": 27.710740089416504, "grad_norm": 0.2314453125, "tokens_processed": 513146880, "tokens_processed_billions": 0.51314688}
+{"timestamp": 1787530021.379101, "elapsed_hours": 0.9437473871972826, "session_id": 2, "event": "train_step", "step": 7840, "epoch": 1, "loss": 2.532286581993103, "loss_raw": 2.560755729675293, "perplexity": 12.582243600513275, "muon_lr": 0.009965479288710215, "adamw_lr": 0.00029896437866130634, "tokens_per_sec": 151731.8210097632, "vram_gb": 27.299458026885986, "grad_norm": 0.1875, "tokens_processed": 513802240, "tokens_processed_billions": 0.51380224}
+{"timestamp": 1787530025.3617337, "elapsed_hours": 0.9448536739746729, "session_id": 2, "event": "train_step", "step": 7850, "epoch": 1, "loss": 2.530612499713898, "loss_raw": 1.8942800760269165, "perplexity": 12.561197510827137, "muon_lr": 0.009965384751963496, "adamw_lr": 0.00029896154255890485, "tokens_per_sec": 151746.88612027484, "vram_gb": 27.710740089416504, "grad_norm": 0.203125, "tokens_processed": 514457600, "tokens_processed_billions": 0.5144576}
+{"timestamp": 1787530029.3175416, "elapsed_hours": 0.9459525095091925, "session_id": 2, "event": "train_step", "step": 7860, "epoch": 1, "loss": 2.436538803577423, "loss_raw": 1.6537986993789673, "perplexity": 11.433398938830157, "muon_lr": 0.009965290086446636, "adamw_lr": 0.0002989587025933991, "tokens_per_sec": 151763.11566540273, "vram_gb": 27.710740089416504, "grad_norm": 0.1826171875, "tokens_processed": 515112960, "tokens_processed_billions": 0.51511296}
+{"timestamp": 1787530033.2714975, "elapsed_hours": 0.9470508305894004, "session_id": 2, "event": "train_step", "step": 7870, "epoch": 1, "loss": 2.51214560508728, "loss_raw": 2.576941967010498, "perplexity": 12.331359924933228, "muon_lr": 0.009965195292162367, "adamw_lr": 0.00029895585876487095, "tokens_per_sec": 151779.38456684825, "vram_gb": 27.710740089416504, "grad_norm": 0.193359375, "tokens_processed": 515768320, "tokens_processed_billions": 0.51576832}
+{"timestamp": 1787530037.3639007, "elapsed_hours": 0.9481876093811459, "session_id": 2, "event": "train_step", "step": 7880, "epoch": 1, "loss": 2.4012256598472597, "loss_raw": 1.4361895322799683, "perplexity": 11.03669532844397, "muon_lr": 0.00996500531730253, "adamw_lr": 0.0002989501595190759, "tokens_per_sec": 151789.43916265614, "vram_gb": 27.299458026885986, "grad_norm": 0.181640625, "tokens_processed": 516423680, "tokens_processed_billions": 0.51642368}
+{"timestamp": 1787530041.3505383, "elapsed_hours": 0.9492950086461173, "session_id": 2, "event": "train_step", "step": 7890, "epoch": 1, "loss": 2.4342684507369996, "loss_raw": 2.8144631385803223, "perplexity": 11.40747053352534, "muon_lr": 0.009964910136732442, "adamw_lr": 0.0002989473041019732, "tokens_per_sec": 151804.18893064672, "vram_gb": 27.710740089416504, "grad_norm": 0.2099609375, "tokens_processed": 517079040, "tokens_processed_billions": 0.51707904}
+{"timestamp": 1787530045.319555, "elapsed_hours": 0.9503975131776597, "session_id": 2, "event": "train_step", "step": 7900, "epoch": 1, "loss": 2.488287923336029, "loss_raw": 4.000158786773682, "perplexity": 12.04064395802286, "muon_lr": 0.0099648148274059, "adamw_lr": 0.00029894444482217697, "tokens_per_sec": 151819.68144833352, "vram_gb": 27.710740089416504, "grad_norm": 0.498046875, "tokens_processed": 517734400, "tokens_processed_billions": 0.5177344}
+{"timestamp": 1787530045.3240309, "elapsed_hours": 0.9503987566630046, "session_id": 2, "event": "lambda_stats", "step": 7900, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.796875, "lambda_mean": 1.2183159722222223, "lambda_std": 0.15294845757217546, "lambda_all": [1.796875, 1.375, 1.265625, 1.25, 1.1640625, 1.140625, 1.140625, 1.2421875, 1.171875, 1.140625, 1.140625, 1.140625, 1.140625, 1.1875, 1.140625, 1.1796875, 1.140625, 1.171875], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787530049.2753673, "elapsed_hours": 0.9514963499704997, "session_id": 2, "event": "train_step", "step": 7910, "epoch": 1, "loss": 2.547568426132202, "loss_raw": 1.7611818313598633, "perplexity": 12.776000194444215, "muon_lr": 0.009964719389325652, "adamw_lr": 0.00029894158167976945, "tokens_per_sec": 151835.73546770637, "vram_gb": 27.710740089416504, "grad_norm": 0.2119140625, "tokens_processed": 518389760, "tokens_processed_billions": 0.51838976}
+{"timestamp": 1787530053.3784368, "elapsed_hours": 0.9526360915104548, "session_id": 2, "event": "train_step", "step": 7920, "epoch": 1, "loss": 2.4239144551754, "loss_raw": 1.9082473516464233, "perplexity": 11.289966999534363, "muon_lr": 0.009964528126915048, "adamw_lr": 0.0002989358438074514, "tokens_per_sec": 151845.20012343657, "vram_gb": 27.299458026885986, "grad_norm": 0.185546875, "tokens_processed": 519045120, "tokens_processed_billions": 0.51904512}
+{"timestamp": 1787530057.3110282, "elapsed_hours": 0.9537284780873193, "session_id": 2, "event": "train_step", "step": 7930, "epoch": 1, "loss": 2.4928400647640228, "loss_raw": 2.8454349040985107, "perplexity": 12.095579614778742, "muon_lr": 0.009964432302590205, "adamw_lr": 0.0002989329690777061, "tokens_per_sec": 151862.2172221795, "vram_gb": 27.710740089416504, "grad_norm": 0.23046875, "tokens_processed": 519700480, "tokens_processed_billions": 0.51970048}
+{"timestamp": 1787530061.26002, "elapsed_hours": 0.9548254203134113, "session_id": 2, "event": "train_step", "step": 7940, "epoch": 1, "loss": 2.481046506166458, "loss_raw": 0.7890430688858032, "perplexity": 11.953767565867345, "muon_lr": 0.009964336349522693, "adamw_lr": 0.0002989300904856807, "tokens_per_sec": 151878.462137794, "vram_gb": 27.710740089416504, "grad_norm": 0.1923828125, "tokens_processed": 520355840, "tokens_processed_billions": 0.52035584}
+{"timestamp": 1787530065.214606, "elapsed_hours": 0.9559239163001378, "session_id": 2, "event": "train_step", "step": 7950, "epoch": 1, "loss": 2.381674727201462, "loss_raw": 2.69468092918396, "perplexity": 10.823013289444882, "muon_lr": 0.00996424026771527, "adamw_lr": 0.000298927208031458, "tokens_per_sec": 151894.41702966634, "vram_gb": 27.710740089416504, "grad_norm": 0.205078125, "tokens_processed": 521011200, "tokens_processed_billions": 0.5210112}
+{"timestamp": 1787530069.3240092, "elapsed_hours": 0.957065417236752, "session_id": 2, "event": "train_step", "step": 7960, "epoch": 1, "loss": 2.3996546256542204, "loss_raw": 3.4933552742004395, "perplexity": 11.019369915676235, "muon_lr": 0.009964047717891784, "adamw_lr": 0.0002989214315367535, "tokens_per_sec": 151903.49059270046, "vram_gb": 27.299458026885986, "grad_norm": 0.1953125, "tokens_processed": 521666560, "tokens_processed_billions": 0.52166656}
+{"timestamp": 1787530073.2895784, "elapsed_hours": 0.9581669641865624, "session_id": 2, "event": "train_step", "step": 7970, "epoch": 1, "loss": 2.454969954490662, "loss_raw": 2.706150531768799, "perplexity": 11.646083631108274, "muon_lr": 0.009963951249881274, "adamw_lr": 0.0002989185374964382, "tokens_per_sec": 151918.90317412175, "vram_gb": 27.710740089416504, "grad_norm": 0.212890625, "tokens_processed": 522321920, "tokens_processed_billions": 0.52232192}
+{"timestamp": 1787530077.2661808, "elapsed_hours": 0.9592715760734346, "session_id": 2, "event": "train_step", "step": 7980, "epoch": 1, "loss": 2.4631360006332397, "loss_raw": 1.9081408977508545, "perplexity": 11.741575452078441, "muon_lr": 0.009963854653141965, "adamw_lr": 0.00029891563959425887, "tokens_per_sec": 151933.78734635332, "vram_gb": 27.710740089416504, "grad_norm": 0.2294921875, "tokens_processed": 522977280, "tokens_processed_billions": 0.52297728}
+{"timestamp": 1787530081.2230446, "elapsed_hours": 0.9603707047303518, "session_id": 2, "event": "train_step", "step": 7990, "epoch": 1, "loss": 2.3818616390228273, "loss_raw": 2.428179979324341, "perplexity": 10.825036427639805, "muon_lr": 0.009963757927676637, "adamw_lr": 0.000298912737830299, "tokens_per_sec": 151949.50936505536, "vram_gb": 27.710740089416504, "grad_norm": 0.1748046875, "tokens_processed": 523632640, "tokens_processed_billions": 0.52363264}
+{"timestamp": 1787530085.279525, "elapsed_hours": 0.961497505042288, "session_id": 2, "event": "train_step", "step": 8000, "epoch": 1, "loss": 2.4212454533576966, "loss_raw": 2.854323387145996, "perplexity": 11.259874233778039, "muon_lr": 0.009963564090579087, "adamw_lr": 0.00029890692271737255, "tokens_per_sec": 151960.80343840056, "vram_gb": 27.299458026885986, "grad_norm": 0.197265625, "tokens_processed": 524288000, "tokens_processed_billions": 0.524288}
+{"timestamp": 1787530085.2834525, "elapsed_hours": 0.9614985959397422, "session_id": 2, "event": "lambda_stats", "step": 8000, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.8046875, "lambda_mean": 1.2183159722222223, "lambda_std": 0.15471187865167232, "lambda_all": [1.8046875, 1.375, 1.265625, 1.25, 1.1640625, 1.140625, 1.140625, 1.2421875, 1.171875, 1.140625, 1.140625, 1.140625, 1.140625, 1.1875, 1.140625, 1.171875, 1.140625, 1.171875], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787530098.679286, "elapsed_hours": 0.9652196607324812, "session_id": 2, "event": "validation", "step": 8000, "epoch": 1, "val_loss": 2.5237556743621825, "val_perplexity": 12.475362188497021, "is_best": true, "best_loss": 2.5237556743621825, "best_perplexity": 12.475362188497021, "train_loss": 2.854323387145996, "train_loss_avg": 2.4212454533576966}