Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -6947,3 +6947,31 @@
 {"timestamp": 1787553363.683678, "elapsed_hours": 7.427720880773332, "session_id": 2, "event": "train_step", "step": 61500, "epoch": 1, "loss": 1.8962210810184479, "loss_raw": 2.2880043983459473, "perplexity": 6.660676666837458, "muon_lr": 0.006863749742959461, "adamw_lr": 0.00020591249228878378, "tokens_per_sec": 150792.42870690985, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 4030464000, "tokens_processed_billions": 4.030464}
 {"timestamp": 1787553363.6871634, "elapsed_hours": 7.42772184934881, "session_id": 2, "event": "lambda_stats", "step": 61500, "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": 1787553377.3721118, "elapsed_hours": 7.431523223784235, "session_id": 2, "event": "validation", "step": 61500, "epoch": 1, "val_loss": 1.8882437086105346, "val_perplexity": 6.607753343768968, "is_best": true, "best_loss": 1.8882437086105346, "best_perplexity": 6.607753343768968, "train_loss": 2.2880043983459473, "train_loss_avg": 1.8962210810184479}
+{"timestamp": 1787553382.2236836, "elapsed_hours": 7.432870882484648, "session_id": 2, "event": "train_step", "step": 61510, "epoch": 1, "loss": 1.8802126729488373, "loss_raw": 2.253920555114746, "perplexity": 6.554898763612719, "muon_lr": 0.0068630214888615185, "adamw_lr": 0.0002058906446658455, "tokens_per_sec": 150712.40760229967, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4031119360, "tokens_processed_billions": 4.03111936}
+{"timestamp": 1787553386.3887177, "elapsed_hours": 7.43402783619033, "session_id": 2, "event": "train_step", "step": 61520, "epoch": 1, "loss": 1.8837075161933898, "loss_raw": 2.4867055416107178, "perplexity": 6.577847184484758, "muon_lr": 0.006861564862768598, "adamw_lr": 0.00020584694588305787, "tokens_per_sec": 150713.4404225692, "vram_gb": 27.299458026885986, "grad_norm": 0.15625, "tokens_processed": 4031774720, "tokens_processed_billions": 4.03177472}
+{"timestamp": 1787553390.3758008, "elapsed_hours": 7.435135359366734, "session_id": 2, "event": "train_step", "step": 61530, "epoch": 1, "loss": 1.883083040714264, "loss_raw": 2.6988778114318848, "perplexity": 6.573740762525969, "muon_lr": 0.006860836490815624, "adamw_lr": 0.00020582509472446867, "tokens_per_sec": 150715.47561080597, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 4032430080, "tokens_processed_billions": 4.03243008}
+{"timestamp": 1787553394.442213, "elapsed_hours": 7.436264918247859, "session_id": 2, "event": "train_step", "step": 61540, "epoch": 1, "loss": 1.7905553078651428, "loss_raw": 2.0866639614105225, "perplexity": 5.992779380090777, "muon_lr": 0.006860108079619646, "adamw_lr": 0.00020580324238858932, "tokens_per_sec": 150717.06399976538, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4033085440, "tokens_processed_billions": 4.03308544}
+{"timestamp": 1787553398.428715, "elapsed_hours": 7.437372279961904, "session_id": 2, "event": "train_step", "step": 61550, "epoch": 1, "loss": 1.8063808512687682, "loss_raw": 1.3220467567443848, "perplexity": 6.088372783971524, "muon_lr": 0.006859379629201672, "adamw_lr": 0.0002057813888760501, "tokens_per_sec": 150719.10104871084, "vram_gb": 27.710740089416504, "grad_norm": 0.146484375, "tokens_processed": 4033740800, "tokens_processed_billions": 4.0337408}
+{"timestamp": 1787553402.5676181, "elapsed_hours": 7.438521975345082, "session_id": 2, "event": "train_step", "step": 61560, "epoch": 1, "loss": 1.8421988677978516, "loss_raw": 2.8212389945983887, "perplexity": 6.310398749352847, "muon_lr": 0.006857922610783759, "adamw_lr": 0.0002057376783235127, "tokens_per_sec": 150720.27915698598, "vram_gb": 27.299458026885986, "grad_norm": 0.1494140625, "tokens_processed": 4034396160, "tokens_processed_billions": 4.03439616}
+{"timestamp": 1787553406.6653204, "elapsed_hours": 7.439660225841735, "session_id": 2, "event": "train_step", "step": 61570, "epoch": 1, "loss": 1.7714339339733123, "loss_raw": 1.4493687152862549, "perplexity": 5.879277816082215, "muon_lr": 0.006857194042825837, "adamw_lr": 0.00020571582128477506, "tokens_per_sec": 150721.68991876522, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 4035051520, "tokens_processed_billions": 4.03505152}
+{"timestamp": 1787553410.6607301, "elapsed_hours": 7.440770061943266, "session_id": 2, "event": "train_step", "step": 61580, "epoch": 1, "loss": 1.6941853094100952, "loss_raw": 2.64863920211792, "perplexity": 5.44221044103363, "muon_lr": 0.00685646543572995, "adamw_lr": 0.00020569396307189845, "tokens_per_sec": 150723.675476758, "vram_gb": 27.710740089416504, "grad_norm": 0.1357421875, "tokens_processed": 4035706880, "tokens_processed_billions": 4.03570688}
+{"timestamp": 1787553414.7875535, "elapsed_hours": 7.441916401783625, "session_id": 2, "event": "train_step", "step": 61590, "epoch": 1, "loss": 1.8191975450515747, "loss_raw": 1.9030307531356812, "perplexity": 6.166907798132619, "muon_lr": 0.006855736789517111, "adamw_lr": 0.00020567210368551328, "tokens_per_sec": 150724.92043189477, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 4036362240, "tokens_processed_billions": 4.03636224}
+{"timestamp": 1787553419.0342805, "elapsed_hours": 7.443096048169666, "session_id": 2, "event": "train_step", "step": 61600, "epoch": 1, "loss": 1.8627235794067383, "loss_raw": 2.6419172286987305, "perplexity": 6.441256175528475, "muon_lr": 0.006854279379824623, "adamw_lr": 0.00020562838139473864, "tokens_per_sec": 150725.49061358525, "vram_gb": 27.299458026885986, "grad_norm": 0.166015625, "tokens_processed": 4037017600, "tokens_processed_billions": 4.0370176}
+{"timestamp": 1787553419.0393665, "elapsed_hours": 7.4430974611971115, "session_id": 2, "event": "lambda_stats", "step": 61600, "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": 1787553423.0573215, "elapsed_hours": 7.444213559561305, "session_id": 2, "event": "train_step", "step": 61610, "epoch": 1, "loss": 1.867603520154953, "loss_raw": 2.043064832687378, "perplexity": 6.472765944380979, "muon_lr": 0.006853550616387001, "adamw_lr": 0.00020560651849161, "tokens_per_sec": 150727.3197275899, "vram_gb": 27.710740089416504, "grad_norm": 0.13671875, "tokens_processed": 4037672960, "tokens_processed_billions": 4.03767296}
+{"timestamp": 1787553427.0423918, "elapsed_hours": 7.445320523513688, "session_id": 2, "event": "train_step", "step": 61620, "epoch": 1, "loss": 1.836540412902832, "loss_raw": 1.7916620969772339, "perplexity": 6.274792475909926, "muon_lr": 0.006852821813916483, "adamw_lr": 0.00020558465441749445, "tokens_per_sec": 150729.36208586616, "vram_gb": 27.710740089416504, "grad_norm": 0.138671875, "tokens_processed": 4038328320, "tokens_processed_billions": 4.03832832}
+{"timestamp": 1787553431.1437309, "elapsed_hours": 7.446459784309069, "session_id": 2, "event": "train_step", "step": 61630, "epoch": 1, "loss": 1.955612713098526, "loss_raw": 1.805769920349121, "perplexity": 7.06824850564366, "muon_lr": 0.006852092972434084, "adamw_lr": 0.00020556278917302248, "tokens_per_sec": 150730.74844917184, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 4038983680, "tokens_processed_billions": 4.03898368}
+{"timestamp": 1787553435.2656293, "elapsed_hours": 7.447604756090376, "session_id": 2, "event": "train_step", "step": 61640, "epoch": 1, "loss": 1.834051958322525, "loss_raw": 1.6915830373764038, "perplexity": 6.259197351759445, "muon_lr": 0.006850635172517718, "adamw_lr": 0.0002055190551755315, "tokens_per_sec": 150732.01885018862, "vram_gb": 27.299458026885986, "grad_norm": 0.1396484375, "tokens_processed": 4039639040, "tokens_processed_billions": 4.03963904}
+{"timestamp": 1787553439.2280114, "elapsed_hours": 7.44870541771253, "session_id": 2, "event": "train_step", "step": 61650, "epoch": 1, "loss": 1.7669844019412995, "loss_raw": 1.1666673421859741, "perplexity": 5.853175894846308, "muon_lr": 0.00684990621412579, "adamw_lr": 0.00020549718642377365, "tokens_per_sec": 150734.1865895799, "vram_gb": 27.710740089416504, "grad_norm": 0.1357421875, "tokens_processed": 4040294400, "tokens_processed_billions": 4.0402944}
+{"timestamp": 1787553443.1935008, "elapsed_hours": 7.449806942608621, "session_id": 2, "event": "train_step", "step": 61660, "epoch": 1, "loss": 1.8393376350402832, "loss_raw": 2.4332633018493652, "perplexity": 6.292369035641868, "muon_lr": 0.006849177216806059, "adamw_lr": 0.00020547531650418172, "tokens_per_sec": 150736.3365622611, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 4040949760, "tokens_processed_billions": 4.04094976}
+{"timestamp": 1787553447.169875, "elapsed_hours": 7.450911490983433, "session_id": 2, "event": "train_step", "step": 61670, "epoch": 1, "loss": 1.9366798973083497, "loss_raw": 1.7858681678771973, "perplexity": 6.935685513839432, "muon_lr": 0.00684844818057955, "adamw_lr": 0.00020545344541738643, "tokens_per_sec": 150738.42417492642, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 4041605120, "tokens_processed_billions": 4.04160512}
+{"timestamp": 1787553451.3938425, "elapsed_hours": 7.452084815369712, "session_id": 2, "event": "train_step", "step": 61680, "epoch": 1, "loss": 1.9389102792739867, "loss_raw": 2.585500955581665, "perplexity": 6.951172005704985, "muon_lr": 0.006846989991490285, "adamw_lr": 0.0002054096997447085, "tokens_per_sec": 150739.11896988517, "vram_gb": 27.299458026885986, "grad_norm": 0.1533203125, "tokens_processed": 4042260480, "tokens_processed_billions": 4.04226048}
+{"timestamp": 1787553455.361635, "elapsed_hours": 7.453186979889869, "session_id": 2, "event": "train_step", "step": 61690, "epoch": 1, "loss": 2.0050202298164366, "loss_raw": 1.4293407201766968, "perplexity": 7.426244126793755, "muon_lr": 0.0068462608386695815, "adamw_lr": 0.00020538782516008738, "tokens_per_sec": 150741.25440607753, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 4042915840, "tokens_processed_billions": 4.04291584}
+{"timestamp": 1787553459.3382907, "elapsed_hours": 7.454291606479221, "session_id": 2, "event": "train_step", "step": 61700, "epoch": 1, "loss": 2.071452708244324, "loss_raw": 2.074390411376953, "perplexity": 7.936343939898091, "muon_lr": 0.006845531647026198, "adamw_lr": 0.00020536594941078588, "tokens_per_sec": 150743.33858888852, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 4043571200, "tokens_processed_billions": 4.0435712}
+{"timestamp": 1787553459.3418806, "elapsed_hours": 7.4542926037973825, "session_id": 2, "event": "lambda_stats", "step": 61700, "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": 1787553463.3164158, "elapsed_hours": 7.455396641294161, "session_id": 2, "event": "train_step", "step": 61710, "epoch": 1, "loss": 1.9908077609539032, "loss_raw": 2.2961883544921875, "perplexity": 7.321445352093991, "muon_lr": 0.006844802416581165, "adamw_lr": 0.00020534407249743489, "tokens_per_sec": 150745.41433962286, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 4044226560, "tokens_processed_billions": 4.04422656}
+{"timestamp": 1787553467.442592, "elapsed_hours": 7.456542801393403, "session_id": 2, "event": "train_step", "step": 61720, "epoch": 1, "loss": 1.9183655202388763, "loss_raw": 1.3028589487075806, "perplexity": 6.809818856936677, "muon_lr": 0.0068433438393702625, "adamw_lr": 0.00020530031518110783, "tokens_per_sec": 150746.6574769743, "vram_gb": 27.299458026885986, "grad_norm": 0.138671875, "tokens_processed": 4044881920, "tokens_processed_billions": 4.04488192}
+{"timestamp": 1787553471.5143025, "elapsed_hours": 7.4576738320456615, "session_id": 2, "event": "train_step", "step": 61730, "epoch": 1, "loss": 1.887573629617691, "loss_raw": 1.5054792165756226, "perplexity": 6.603327110192102, "muon_lr": 0.006842614492646455, "adamw_lr": 0.0002052784347793936, "tokens_per_sec": 150748.20689885368, "vram_gb": 27.710740089416504, "grad_norm": 0.1611328125, "tokens_processed": 4045537280, "tokens_processed_billions": 4.04553728}
+{"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}