Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -3783,3 +3783,43 @@
 {"timestamp": 1787541244.0985897, "elapsed_hours": 4.061169467303488, "session_id": 2, "event": "train_step", "step": 33500, "epoch": 1, "loss": 2.0122306108474732, "loss_raw": 2.278702974319458, "perplexity": 7.479983685098726, "muon_lr": 0.00903126342474549, "adamw_lr": 0.0002709379027423646, "tokens_per_sec": 150281.3498870874, "vram_gb": 27.710740089416504, "grad_norm": 0.181640625, "tokens_processed": 2195456000, "tokens_processed_billions": 2.195456}
 {"timestamp": 1787541244.103416, "elapsed_hours": 4.061170808209313, "session_id": 2, "event": "lambda_stats", "step": 33500, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
 {"timestamp": 1787541257.6578157, "elapsed_hours": 4.064935919377539, "session_id": 2, "event": "validation", "step": 33500, "epoch": 1, "val_loss": 2.0166217213869095, "val_perplexity": 7.512901339930947, "is_best": true, "best_loss": 2.0166217213869095, "best_perplexity": 7.512901339930947, "train_loss": 2.278702974319458, "train_loss_avg": 2.0122306108474732}
+{"timestamp": 1787541262.6737814, "elapsed_hours": 4.0663292429182265, "session_id": 2, "event": "train_step", "step": 33510, "epoch": 1, "loss": 1.9290950691699982, "loss_raw": 3.393746852874756, "perplexity": 6.88327853141866, "muon_lr": 0.0090307897070502, "adamw_lr": 0.000270923691211506, "tokens_per_sec": 150135.31409630427, "vram_gb": 27.710740089416504, "grad_norm": 0.1328125, "tokens_processed": 2196111360, "tokens_processed_billions": 2.19611136}
+{"timestamp": 1787541266.8765643, "elapsed_hours": 4.067496682471699, "session_id": 2, "event": "train_step", "step": 33520, "epoch": 1, "loss": 1.9239944458007812, "loss_raw": 1.5421807765960693, "perplexity": 6.848258906965469, "muon_lr": 0.009029841966216777, "adamw_lr": 0.00027089525898650325, "tokens_per_sec": 150136.97951104876, "vram_gb": 27.299458026885986, "grad_norm": 0.154296875, "tokens_processed": 2196766720, "tokens_processed_billions": 2.19676672}
+{"timestamp": 1787541270.8639379, "elapsed_hours": 4.068604286313057, "session_id": 2, "event": "train_step", "step": 33530, "epoch": 1, "loss": 1.9285002660751343, "loss_raw": 1.3963932991027832, "perplexity": 6.879185553424023, "muon_lr": 0.009029367943105968, "adamw_lr": 0.00027088103829317895, "tokens_per_sec": 150140.8547691453, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 2197422080, "tokens_processed_billions": 2.19742208}
+{"timestamp": 1787541274.8991883, "elapsed_hours": 4.0697251893414395, "session_id": 2, "event": "train_step", "step": 33540, "epoch": 1, "loss": 2.0282968151569367, "loss_raw": 2.032637357711792, "perplexity": 7.601129199650663, "muon_lr": 0.00902889381821732, "adamw_lr": 0.00027086681454651957, "tokens_per_sec": 150144.23652234414, "vram_gb": 27.710740089416504, "grad_norm": 0.1455078125, "tokens_processed": 2198077440, "tokens_processed_billions": 2.19807744}
+{"timestamp": 1787541278.889346, "elapsed_hours": 4.070833566188813, "session_id": 2, "event": "train_step", "step": 33550, "epoch": 1, "loss": 2.024489359855652, "loss_raw": 2.9699089527130127, "perplexity": 7.572243265830553, "muon_lr": 0.009028419591564509, "adamw_lr": 0.00027085258774693513, "tokens_per_sec": 150148.07823378546, "vram_gb": 27.710740089416504, "grad_norm": 0.1484375, "tokens_processed": 2198732800, "tokens_processed_billions": 2.1987328}
+{"timestamp": 1787541283.0512757, "elapsed_hours": 4.071989658143785, "session_id": 2, "event": "train_step", "step": 33560, "epoch": 1, "loss": 2.0425743567943573, "loss_raw": 2.118920087814331, "perplexity": 7.710433076842089, "muon_lr": 0.009027470833021085, "adamw_lr": 0.00027082412499063256, "tokens_per_sec": 150150.1570906603, "vram_gb": 27.299458026885986, "grad_norm": 0.1455078125, "tokens_processed": 2199388160, "tokens_processed_billions": 2.19938816}
+{"timestamp": 1787541287.0325954, "elapsed_hours": 4.073095579942067, "session_id": 2, "event": "train_step", "step": 33570, "epoch": 1, "loss": 2.023438050746918, "loss_raw": 0.9491477012634277, "perplexity": 7.564286680659703, "muon_lr": 0.00902699630115784, "adamw_lr": 0.0002708098890347351, "tokens_per_sec": 150154.08587343906, "vram_gb": 27.710740089416504, "grad_norm": 0.150390625, "tokens_processed": 2200043520, "tokens_processed_billions": 2.20004352}
+{"timestamp": 1787541291.0497646, "elapsed_hours": 4.074211460616853, "session_id": 2, "event": "train_step", "step": 33580, "epoch": 1, "loss": 2.0186446785926817, "loss_raw": 0.9651072025299072, "perplexity": 7.528115000937086, "muon_lr": 0.009026521667585146, "adamw_lr": 0.00027079565002755424, "tokens_per_sec": 150157.645613894, "vram_gb": 27.710740089416504, "grad_norm": 0.1376953125, "tokens_processed": 2200698880, "tokens_processed_billions": 2.20069888}
+{"timestamp": 1787541295.1383932, "elapsed_hours": 4.075347190697988, "session_id": 2, "event": "train_step", "step": 33590, "epoch": 1, "loss": 2.0046549367904665, "loss_raw": 2.711989402770996, "perplexity": 7.423531867020036, "muon_lr": 0.009026046932316692, "adamw_lr": 0.0002707814079695007, "tokens_per_sec": 150160.47049004675, "vram_gb": 27.710740089416504, "grad_norm": 0.1396484375, "tokens_processed": 2201354240, "tokens_processed_billions": 2.20135424}
+{"timestamp": 1787541299.3462965, "elapsed_hours": 4.0765160524845125, "session_id": 2, "event": "train_step", "step": 33600, "epoch": 1, "loss": 2.054771933555603, "loss_raw": 3.506903648376465, "perplexity": 7.805057597898486, "muon_lr": 0.009025097156747267, "adamw_lr": 0.000270752914702418, "tokens_per_sec": 150162.07225133164, "vram_gb": 27.299458026885986, "grad_norm": 0.1552734375, "tokens_processed": 2202009600, "tokens_processed_billions": 2.2020096}
+{"timestamp": 1787541299.350681, "elapsed_hours": 4.0765172709359065, "session_id": 2, "event": "lambda_stats", "step": 33600, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541303.410799, "elapsed_hours": 4.077645081149207, "session_id": 2, "event": "train_step", "step": 33610, "epoch": 1, "loss": 2.1214167785644533, "loss_raw": 2.002535820007324, "perplexity": 8.342949230033673, "muon_lr": 0.009024622116473688, "adamw_lr": 0.0002707386634942106, "tokens_per_sec": 150165.14299695133, "vram_gb": 27.710740089416504, "grad_norm": 0.173828125, "tokens_processed": 2202664960, "tokens_processed_billions": 2.20266496}
+{"timestamp": 1787541307.443845, "elapsed_hours": 4.078765371640523, "session_id": 2, "event": "train_step", "step": 33620, "epoch": 1, "loss": 2.0991676235198975, "loss_raw": 2.6710047721862793, "perplexity": 8.15937541298303, "muon_lr": 0.009024146974559128, "adamw_lr": 0.00027072440923677383, "tokens_per_sec": 150168.53248881683, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 2203320320, "tokens_processed_billions": 2.20332032}
+{"timestamp": 1787541311.477518, "elapsed_hours": 4.079885836376084, "session_id": 2, "event": "train_step", "step": 33630, "epoch": 1, "loss": 2.098010505437851, "loss_raw": 0.9868882894515991, "perplexity": 8.149939512432864, "muon_lr": 0.009023671731017289, "adamw_lr": 0.00027071015193051853, "tokens_per_sec": 150171.9149630267, "vram_gb": 27.710740089416504, "grad_norm": 0.1669921875, "tokens_processed": 2203975680, "tokens_processed_billions": 2.20397568}
+{"timestamp": 1787541315.7087357, "elapsed_hours": 4.081061174604628, "session_id": 2, "event": "train_step", "step": 33640, "epoch": 1, "loss": 2.063553205728531, "loss_raw": 1.9788380861282349, "perplexity": 7.8738977426557994, "muon_lr": 0.009022720939106589, "adamw_lr": 0.00027068162817319764, "tokens_per_sec": 150173.27339496877, "vram_gb": 27.299458026885986, "grad_norm": 0.146484375, "tokens_processed": 2204631040, "tokens_processed_billions": 2.20463104}
+{"timestamp": 1787541319.6912339, "elapsed_hours": 4.082167424096001, "session_id": 2, "event": "train_step", "step": 33650, "epoch": 1, "loss": 1.9969485557079316, "loss_raw": 2.064739227294922, "perplexity": 7.3665431718140075, "muon_lr": 0.009022245390765151, "adamw_lr": 0.00027066736172295453, "tokens_per_sec": 150177.17474271855, "vram_gb": 27.710740089416504, "grad_norm": 0.142578125, "tokens_processed": 2205286400, "tokens_processed_billions": 2.2052864}
+{"timestamp": 1787541323.738527, "elapsed_hours": 4.083291672269503, "session_id": 2, "event": "train_step", "step": 33660, "epoch": 1, "loss": 1.939905388355255, "loss_raw": 3.2201201915740967, "perplexity": 6.958092622906742, "muon_lr": 0.009021769740851272, "adamw_lr": 0.00027065309222553806, "tokens_per_sec": 150180.41185065993, "vram_gb": 27.710740089416504, "grad_norm": 0.1318359375, "tokens_processed": 2205941760, "tokens_processed_billions": 2.20594176}
+{"timestamp": 1787541327.822408, "elapsed_hours": 4.084426083962122, "session_id": 2, "event": "train_step", "step": 33670, "epoch": 1, "loss": 1.9090789794921874, "loss_raw": 2.140543222427368, "perplexity": 6.746871928628567, "muon_lr": 0.009021293989378662, "adamw_lr": 0.00027063881968135986, "tokens_per_sec": 150183.274046067, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2206597120, "tokens_processed_billions": 2.20659712}
+{"timestamp": 1787541332.0367985, "elapsed_hours": 4.085596747597059, "session_id": 2, "event": "train_step", "step": 33680, "epoch": 1, "loss": 2.042131587266922, "loss_raw": 3.8626785278320312, "perplexity": 7.707019887716164, "muon_lr": 0.00902034218181215, "adamw_lr": 0.00027061026545436445, "tokens_per_sec": 150184.79974707036, "vram_gb": 27.299458026885986, "grad_norm": 0.1708984375, "tokens_processed": 2207252480, "tokens_processed_billions": 2.20725248}
+{"timestamp": 1787541336.062533, "elapsed_hours": 4.0867150070932174, "session_id": 2, "event": "train_step", "step": 33690, "epoch": 1, "loss": 2.0670627748966215, "loss_raw": 1.798220157623291, "perplexity": 7.901580279881707, "muon_lr": 0.009019866125745693, "adamw_lr": 0.00027059598377237074, "tokens_per_sec": 150188.2525665547, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 2207907840, "tokens_processed_billions": 2.20790784}
+{"timestamp": 1787541340.0750213, "elapsed_hours": 4.087829587393337, "session_id": 2, "event": "train_step", "step": 33700, "epoch": 1, "loss": 2.056183284521103, "loss_raw": 1.7181427478790283, "perplexity": 7.8160810506255745, "muon_lr": 0.009019389968175404, "adamw_lr": 0.0002705816990452621, "tokens_per_sec": 150191.83807552911, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2208563200, "tokens_processed_billions": 2.2085632}
+{"timestamp": 1787541340.078666, "elapsed_hours": 4.087830599745114, "session_id": 2, "event": "lambda_stats", "step": 33700, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541344.049813, "elapsed_hours": 4.088933695952098, "session_id": 2, "event": "train_step", "step": 33710, "epoch": 1, "loss": 2.0371977400779726, "loss_raw": 2.180931568145752, "perplexity": 7.669088280631387, "muon_lr": 0.009018913709115017, "adamw_lr": 0.0002705674112734504, "tokens_per_sec": 150195.80682031697, "vram_gb": 27.710740089416504, "grad_norm": 0.1435546875, "tokens_processed": 2209218560, "tokens_processed_billions": 2.20921856}
+{"timestamp": 1787541348.1979737, "elapsed_hours": 4.090085962944561, "session_id": 2, "event": "train_step", "step": 33720, "epoch": 1, "loss": 2.0440957701206206, "loss_raw": 1.5495548248291016, "perplexity": 7.722172760667681, "muon_lr": 0.009017960886578878, "adamw_lr": 0.0002705388265973663, "tokens_per_sec": 150198.00347760966, "vram_gb": 27.299458026885986, "grad_norm": 0.150390625, "tokens_processed": 2209873920, "tokens_processed_billions": 2.20987392}
+{"timestamp": 1787541352.2100255, "elapsed_hours": 4.091200421717431, "session_id": 2, "event": "train_step", "step": 33730, "epoch": 1, "loss": 1.9735776126384734, "loss_raw": 1.1172428131103516, "perplexity": 7.19637633581559, "muon_lr": 0.009017484323130608, "adamw_lr": 0.0002705245296939182, "tokens_per_sec": 150201.5889996703, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 2210529280, "tokens_processed_billions": 2.21052928}
+{"timestamp": 1787541356.1887841, "elapsed_hours": 4.092305632564757, "session_id": 2, "event": "train_step", "step": 33740, "epoch": 1, "loss": 1.9423173546791077, "loss_raw": 1.7936009168624878, "perplexity": 6.974895563909321, "muon_lr": 0.00901700765824719, "adamw_lr": 0.00027051022974741566, "tokens_per_sec": 150205.5108516693, "vram_gb": 27.710740089416504, "grad_norm": 0.1513671875, "tokens_processed": 2211184640, "tokens_processed_billions": 2.21118464}
+{"timestamp": 1787541360.1603072, "elapsed_hours": 4.09340883327855, "session_id": 2, "event": "train_step", "step": 33750, "epoch": 1, "loss": 1.9665838360786438, "loss_raw": 2.999263286590576, "perplexity": 7.146222075958336, "muon_lr": 0.009016530891942372, "adamw_lr": 0.00027049592675827114, "tokens_per_sec": 150209.50541258254, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 2211840000, "tokens_processed_billions": 2.21184}
+{"timestamp": 1787541364.35232, "elapsed_hours": 4.094573281274902, "session_id": 2, "event": "train_step", "step": 33760, "epoch": 1, "loss": 2.073190145492554, "loss_raw": 3.441413402557373, "perplexity": 7.950144825088314, "muon_lr": 0.009015577055123535, "adamw_lr": 0.000270467311653706, "tokens_per_sec": 150211.24872542464, "vram_gb": 27.299458026885986, "grad_norm": 0.1953125, "tokens_processed": 2212495360, "tokens_processed_billions": 2.21249536}
+{"timestamp": 1787541368.3223999, "elapsed_hours": 4.0956760812468, "session_id": 2, "event": "train_step", "step": 33770, "epoch": 1, "loss": 2.0252203023433686, "loss_raw": 1.8726139068603516, "perplexity": 7.577780163491231, "muon_lr": 0.009015099984637023, "adamw_lr": 0.0002704529995391106, "tokens_per_sec": 150215.25481872677, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2213150720, "tokens_processed_billions": 2.21315072}
+{"timestamp": 1787541372.2891285, "elapsed_hours": 4.096777950459057, "session_id": 2, "event": "train_step", "step": 33780, "epoch": 1, "loss": 1.9257426619529725, "loss_raw": 1.5247001647949219, "perplexity": 6.860241614930965, "muon_lr": 0.009014622812784121, "adamw_lr": 0.00027043868438352355, "tokens_per_sec": 150219.2913117809, "vram_gb": 27.710740089416504, "grad_norm": 0.1806640625, "tokens_processed": 2213806080, "tokens_processed_billions": 2.21380608}
+{"timestamp": 1787541376.2678914, "elapsed_hours": 4.097883162233565, "session_id": 2, "event": "train_step", "step": 33790, "epoch": 1, "loss": 2.0017713367938996, "loss_raw": 1.5714807510375977, "perplexity": 7.402156204795357, "muon_lr": 0.009014145539578591, "adamw_lr": 0.0002704243661873577, "tokens_per_sec": 150223.20396452083, "vram_gb": 27.710740089416504, "grad_norm": 0.1494140625, "tokens_processed": 2214461440, "tokens_processed_billions": 2.21446144}
+{"timestamp": 1787541380.3856604, "elapsed_hours": 4.099026987089051, "session_id": 2, "event": "train_step", "step": 33800, "epoch": 1, "loss": 2.0626386392116545, "loss_raw": 1.8803197145462036, "perplexity": 7.86669983140908, "muon_lr": 0.009013190689164709, "adamw_lr": 0.0002703957206749412, "tokens_per_sec": 150225.6972342037, "vram_gb": 27.299458026885986, "grad_norm": 0.1484375, "tokens_processed": 2215116800, "tokens_processed_billions": 2.2151168}
+{"timestamp": 1787541380.3903396, "elapsed_hours": 4.0990282868014445, "session_id": 2, "event": "lambda_stats", "step": 33800, "epoch": 1, "lambda_min": 1.109375, "lambda_max": 2.015625, "lambda_mean": 1.1966145833333333, "lambda_std": 0.20269468977080954, "lambda_all": [2.015625, 1.265625, 1.2109375, 1.140625, 1.140625, 1.140625, 1.140625, 1.21875, 1.140625, 1.109375, 1.140625, 1.125, 1.140625, 1.109375, 1.140625, 1.109375, 1.125, 1.125], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787541384.4285269, "elapsed_hours": 4.1001500054862765, "session_id": 2, "event": "train_step", "step": 33810, "epoch": 1, "loss": 2.004932893514633, "loss_raw": 2.5764269828796387, "perplexity": 7.425595574416926, "muon_lr": 0.00901271311198389, "adamw_lr": 0.00027038139335951666, "tokens_per_sec": 150228.95369317167, "vram_gb": 27.710740089416504, "grad_norm": 0.1533203125, "tokens_processed": 2215772160, "tokens_processed_billions": 2.21577216}
+{"timestamp": 1787541388.3974178, "elapsed_hours": 4.101252475115988, "session_id": 2, "event": "train_step", "step": 33820, "epoch": 1, "loss": 2.0543827402591703, "loss_raw": 1.9688236713409424, "perplexity": 7.802020512848019, "muon_lr": 0.009012235433505513, "adamw_lr": 0.00027036706300516536, "tokens_per_sec": 150232.96060033704, "vram_gb": 27.710740089416504, "grad_norm": 0.16015625, "tokens_processed": 2216427520, "tokens_processed_billions": 2.21642752}
+{"timestamp": 1787541392.37667, "elapsed_hours": 4.102357822987768, "session_id": 2, "event": "train_step", "step": 33830, "epoch": 1, "loss": 2.136542522907257, "loss_raw": 0.9016726016998291, "perplexity": 8.47010176125199, "muon_lr": 0.009011757653743357, "adamw_lr": 0.00027035272961230066, "tokens_per_sec": 150236.8596826263, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2217082880, "tokens_processed_billions": 2.21708288}
+{"timestamp": 1787541396.5410988, "elapsed_hours": 4.103514608807034, "session_id": 2, "event": "train_step", "step": 33840, "epoch": 1, "loss": 2.0386482775211334, "loss_raw": 0.9342986345291138, "perplexity": 7.68022065234642, "muon_lr": 0.009010801790422813, "adamw_lr": 0.0002703240537126843, "tokens_per_sec": 150238.87331770358, "vram_gb": 27.299458026885986, "grad_norm": 0.13671875, "tokens_processed": 2217738240, "tokens_processed_billions": 2.21773824}
+{"timestamp": 1787541400.6087234, "elapsed_hours": 4.104644504586855, "session_id": 2, "event": "train_step", "step": 33850, "epoch": 1, "loss": 2.051640108823776, "loss_raw": 2.5566468238830566, "perplexity": 7.780651762827354, "muon_lr": 0.00901032370689199, "adamw_lr": 0.0002703097112067597, "tokens_per_sec": 150241.87017006474, "vram_gb": 27.710740089416504, "grad_norm": 0.1552734375, "tokens_processed": 2218393600, "tokens_processed_billions": 2.2183936}
+{"timestamp": 1787541404.6381867, "elapsed_hours": 4.105763800011741, "session_id": 2, "event": "train_step", "step": 33860, "epoch": 1, "loss": 2.045316025018692, "loss_raw": 2.2412502765655518, "perplexity": 7.731601531384495, "muon_lr": 0.009009845522132517, "adamw_lr": 0.00027029536566397543, "tokens_per_sec": 150245.25240108997, "vram_gb": 27.710740089416504, "grad_norm": 0.15234375, "tokens_processed": 2219048960, "tokens_processed_billions": 2.21904896}
+{"timestamp": 1787541408.6118424, "elapsed_hours": 4.106867593063249, "session_id": 2, "event": "train_step", "step": 33870, "epoch": 1, "loss": 2.118738993406296, "loss_raw": 3.0878119468688965, "perplexity": 8.320638489426917, "muon_lr": 0.00900936723615818, "adamw_lr": 0.00027028101708474533, "tokens_per_sec": 150249.20182105686, "vram_gb": 27.710740089416504, "grad_norm": 0.15625, "tokens_processed": 2219704320, "tokens_processed_billions": 2.21970432}