Upload metrics_session_002.jsonl with huggingface_hub

metrics_session_002.jsonl

@@ -470,3 +470,40 @@
 {"timestamp": 1787528439.1294062, "elapsed_hours": 0.504233583079444, "session_id": 2, "event": "train_step", "step": 4160, "epoch": 1, "loss": 2.9922337245941164, "loss_raw": 3.2873690128326416, "perplexity": 19.930151276636526, "muon_lr": 0.009995275149150221, "adamw_lr": 0.0002998582544745066, "tokens_per_sec": 151124.91968945725, "vram_gb": 27.299458026885986, "grad_norm": 0.26953125, "tokens_processed": 272629760, "tokens_processed_billions": 0.27262976}
 {"timestamp": 1787528443.1199484, "elapsed_hours": 0.5053420670164956, "session_id": 2, "event": "train_step", "step": 4170, "epoch": 1, "loss": 2.8553618240356444, "loss_raw": 1.2272441387176514, "perplexity": 17.380724834082372, "muon_lr": 0.009995240075357195, "adamw_lr": 0.00029985720226071577, "tokens_per_sec": 151153.84775325947, "vram_gb": 27.710740089416504, "grad_norm": 0.265625, "tokens_processed": 273285120, "tokens_processed_billions": 0.27328512}
 {"timestamp": 1787528447.0973125, "elapsed_hours": 0.5064468903011746, "session_id": 2, "event": "train_step", "step": 4180, "epoch": 1, "loss": 2.888508038520813, "loss_raw": 1.8815553188323975, "perplexity": 17.966484292881297, "muon_lr": 0.009995204871933075, "adamw_lr": 0.0002998561461579922, "tokens_per_sec": 151183.7385355013, "vram_gb": 27.710740089416504, "grad_norm": 0.263671875, "tokens_processed": 273940480, "tokens_processed_billions": 0.27394048}
+{"timestamp": 1787528451.520157, "elapsed_hours": 0.507675458126598, "session_id": 2, "event": "train_step", "step": 4190, "epoch": 1, "loss": 3.018764178752899, "loss_raw": 3.6494743824005127, "perplexity": 20.465983751810324, "muon_lr": 0.00999516953887888, "adamw_lr": 0.0002998550861663664, "tokens_per_sec": 151176.40570243713, "vram_gb": 27.710740089416504, "grad_norm": 0.2470703125, "tokens_processed": 274595840, "tokens_processed_billions": 0.27459584}
+{"timestamp": 1787528455.6279101, "elapsed_hours": 0.5088165007697212, "session_id": 2, "event": "train_step", "step": 4200, "epoch": 1, "loss": 2.8638300704956055, "loss_raw": 2.9332430362701416, "perplexity": 17.52853405478817, "muon_lr": 0.00999509848388434, "adamw_lr": 0.0002998529545165302, "tokens_per_sec": 151195.2833730967, "vram_gb": 27.299458026885986, "grad_norm": 0.2421875, "tokens_processed": 275251200, "tokens_processed_billions": 0.2752512}
+{"timestamp": 1787528455.6334343, "elapsed_hours": 0.5088180357217789, "session_id": 2, "event": "lambda_stats", "step": 4200, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2404513888888888, "lambda_std": 0.12160525803925261, "lambda_all": [1.6484375, 1.390625, 1.328125, 1.328125, 1.203125, 1.2109375, 1.140625, 1.25, 1.203125, 1.15625, 1.171875, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.1796875, 1.265625], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787528459.5861456, "elapsed_hours": 0.5099160106314553, "session_id": 2, "event": "train_step", "step": 4210, "epoch": 1, "loss": 2.8296377801895143, "loss_raw": 1.6732091903686523, "perplexity": 16.939323954447616, "muon_lr": 0.009995062761946043, "adamw_lr": 0.0002998518828583813, "tokens_per_sec": 151226.47283963798, "vram_gb": 27.710740089416504, "grad_norm": 0.265625, "tokens_processed": 275906560, "tokens_processed_billions": 0.27590656}
+{"timestamp": 1787528463.538058, "elapsed_hours": 0.5110137640105353, "session_id": 2, "event": "train_step", "step": 4220, "epoch": 1, "loss": 2.9450096678733826, "loss_raw": 4.90872859954834, "perplexity": 19.01084618003376, "muon_lr": 0.00999502691038177, "adamw_lr": 0.00029985080731145304, "tokens_per_sec": 151258.040768364, "vram_gb": 27.710740089416504, "grad_norm": 0.255859375, "tokens_processed": 276561920, "tokens_processed_billions": 0.27656192}
+{"timestamp": 1787528467.470764, "elapsed_hours": 0.5121061823765437, "session_id": 2, "event": "train_step", "step": 4230, "epoch": 1, "loss": 2.851612913608551, "loss_raw": 2.19277024269104, "perplexity": 17.315688038253146, "muon_lr": 0.009994990929192551, "adamw_lr": 0.00029984972787577645, "tokens_per_sec": 151291.06367067865, "vram_gb": 27.710740089416504, "grad_norm": 0.265625, "tokens_processed": 277217280, "tokens_processed_billions": 0.27721728}
+{"timestamp": 1787528471.5726953, "elapsed_hours": 0.5132456076807446, "session_id": 2, "event": "train_step", "step": 4240, "epoch": 1, "loss": 2.872235312461853, "loss_raw": 3.9665236473083496, "perplexity": 17.676486542013876, "muon_lr": 0.009994918577943435, "adamw_lr": 0.00029984755733830295, "tokens_per_sec": 151309.99619883142, "vram_gb": 27.299458026885986, "grad_norm": 0.267578125, "tokens_processed": 277872640, "tokens_processed_billions": 0.27787264}
+{"timestamp": 1787528475.519067, "elapsed_hours": 0.5143418221341239, "session_id": 2, "event": "train_step", "step": 4250, "epoch": 1, "loss": 2.9803160881996154, "loss_raw": 3.7573137283325195, "perplexity": 19.694040714905743, "muon_lr": 0.009994882207885622, "adamw_lr": 0.0002998464662365686, "tokens_per_sec": 151341.63861548665, "vram_gb": 27.710740089416504, "grad_norm": 0.29296875, "tokens_processed": 278528000, "tokens_processed_billions": 0.278528}
+{"timestamp": 1787528479.478271, "elapsed_hours": 0.5154416010114882, "session_id": 2, "event": "train_step", "step": 4260, "epoch": 1, "loss": 3.0049172949790957, "loss_raw": 2.550346851348877, "perplexity": 20.18454666359939, "muon_lr": 0.009994845708207037, "adamw_lr": 0.0002998453712462111, "tokens_per_sec": 151372.09754146848, "vram_gb": 27.710740089416504, "grad_norm": 0.36328125, "tokens_processed": 279183360, "tokens_processed_billions": 0.27918336}
+{"timestamp": 1787528483.4487064, "elapsed_hours": 0.5165444997284148, "session_id": 2, "event": "train_step", "step": 4270, "epoch": 1, "loss": 3.0365124917030335, "loss_raw": 2.482450485229492, "perplexity": 20.832463010445995, "muon_lr": 0.009994809078908738, "adamw_lr": 0.000299844272367262, "tokens_per_sec": 151401.50101433528, "vram_gb": 27.710740089416504, "grad_norm": 0.294921875, "tokens_processed": 279838720, "tokens_processed_billions": 0.27983872}
+{"timestamp": 1787528487.5359452, "elapsed_hours": 0.5176798439025879, "session_id": 2, "event": "train_step", "step": 4280, "epoch": 1, "loss": 3.110873475074768, "loss_raw": 2.714022636413574, "perplexity": 22.44063717983705, "muon_lr": 0.009994735431457201, "adamw_lr": 0.000299842062943716, "tokens_per_sec": 151421.22538675388, "vram_gb": 27.299458026885986, "grad_norm": 0.287109375, "tokens_processed": 280494080, "tokens_processed_billions": 0.28049408}
+{"timestamp": 1787528491.5061493, "elapsed_hours": 0.5187826783789529, "session_id": 2, "event": "train_step", "step": 4290, "epoch": 1, "loss": 3.016455132961273, "loss_raw": 2.0418758392333984, "perplexity": 20.418781375353454, "muon_lr": 0.009994698413306094, "adamw_lr": 0.00029984095239918274, "tokens_per_sec": 151450.42050996004, "vram_gb": 27.710740089416504, "grad_norm": 0.25, "tokens_processed": 281149440, "tokens_processed_billions": 0.28114944}
+{"timestamp": 1787528495.4822974, "elapsed_hours": 0.5198871638377508, "session_id": 2, "event": "train_step", "step": 4300, "epoch": 1, "loss": 2.9312531566619873, "loss_raw": 3.225945234298706, "perplexity": 18.751113862688147, "muon_lr": 0.009994661265539516, "adamw_lr": 0.00029983983796618536, "tokens_per_sec": 151478.99887869696, "vram_gb": 27.710740089416504, "grad_norm": 0.2314453125, "tokens_processed": 281804800, "tokens_processed_billions": 0.2818048}
+{"timestamp": 1787528495.4858637, "elapsed_hours": 0.519888154996766, "session_id": 2, "event": "lambda_stats", "step": 4300, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2391493055555556, "lambda_std": 0.1220945674205653, "lambda_all": [1.6484375, 1.390625, 1.328125, 1.328125, 1.1953125, 1.203125, 1.140625, 1.25, 1.203125, 1.15625, 1.171875, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.171875, 1.265625], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787528499.4583764, "elapsed_hours": 0.5209916302230623, "session_id": 2, "event": "train_step", "step": 4310, "epoch": 1, "loss": 2.904399826526642, "loss_raw": 1.8812185525894165, "perplexity": 18.254284626471936, "muon_lr": 0.009994623988158535, "adamw_lr": 0.000299838719644756, "tokens_per_sec": 151507.4662345846, "vram_gb": 27.710740089416504, "grad_norm": 0.240234375, "tokens_processed": 282460160, "tokens_processed_billions": 0.28246016}
+{"timestamp": 1787528503.5650697, "elapsed_hours": 0.5221323784854677, "session_id": 2, "event": "train_step", "step": 4320, "epoch": 1, "loss": 2.7131960916519167, "loss_raw": 2.6190876960754395, "perplexity": 15.077387300766315, "muon_lr": 0.009994549044557679, "adamw_lr": 0.00029983647133673027, "tokens_per_sec": 151525.21141288348, "vram_gb": 27.299458026885986, "grad_norm": 0.29296875, "tokens_processed": 283115520, "tokens_processed_billions": 0.28311552}
+{"timestamp": 1787528507.5188496, "elapsed_hours": 0.523230650557412, "session_id": 2, "event": "train_step", "step": 4330, "epoch": 1, "loss": 2.776574318408966, "loss_raw": 2.571624279022217, "perplexity": 16.06389678751951, "muon_lr": 0.009994511378339965, "adamw_lr": 0.00029983534135019884, "tokens_per_sec": 151555.2653224181, "vram_gb": 27.710740089416504, "grad_norm": 0.2734375, "tokens_processed": 283770880, "tokens_processed_billions": 0.28377088}
+{"timestamp": 1787528511.5053215, "elapsed_hours": 0.524338003926807, "session_id": 2, "event": "train_step", "step": 4340, "epoch": 1, "loss": 2.8316357254981996, "loss_raw": 2.5403566360473633, "perplexity": 16.973201628875685, "muon_lr": 0.00999447358251217, "adamw_lr": 0.00029983420747536503, "tokens_per_sec": 151582.54770937513, "vram_gb": 27.710740089416504, "grad_norm": 0.2734375, "tokens_processed": 284426240, "tokens_processed_billions": 0.28442624}
+{"timestamp": 1787528515.4770863, "elapsed_hours": 0.5254412721925311, "session_id": 2, "event": "train_step", "step": 4350, "epoch": 1, "loss": 2.923423686027527, "loss_raw": 2.24885892868042, "perplexity": 18.60487579769402, "muon_lr": 0.00999443565707539, "adamw_lr": 0.0002998330697122616, "tokens_per_sec": 151611.06068481502, "vram_gb": 27.710740089416504, "grad_norm": 0.244140625, "tokens_processed": 285081600, "tokens_processed_billions": 0.2850816}
+{"timestamp": 1787528519.5865521, "elapsed_hours": 0.5265827903482649, "session_id": 2, "event": "train_step", "step": 4360, "epoch": 1, "loss": 2.9587886452674867, "loss_raw": 4.7885260581970215, "perplexity": 19.274609219105148, "muon_lr": 0.00999435941737924, "adamw_lr": 0.00029983078252137713, "tokens_per_sec": 151628.04216974432, "vram_gb": 27.299458026885986, "grad_norm": 0.251953125, "tokens_processed": 285736960, "tokens_processed_billions": 0.28573696}
+{"timestamp": 1787528523.5611525, "elapsed_hours": 0.5276868459913465, "session_id": 2, "event": "train_step", "step": 4370, "epoch": 1, "loss": 3.0189529824256898, "loss_raw": 2.316458225250244, "perplexity": 20.469848169506587, "muon_lr": 0.00999432110312207, "adamw_lr": 0.00029982963309366207, "tokens_per_sec": 151655.95857366925, "vram_gb": 27.710740089416504, "grad_norm": 0.2470703125, "tokens_processed": 286392320, "tokens_processed_billions": 0.28639232}
+{"timestamp": 1787528527.538978, "elapsed_hours": 0.5287917976909213, "session_id": 2, "event": "train_step", "step": 4380, "epoch": 1, "loss": 3.049484574794769, "loss_raw": 4.229621410369873, "perplexity": 21.104463846110654, "muon_lr": 0.009994282659260313, "adamw_lr": 0.0002998284797778093, "tokens_per_sec": 151683.49457006107, "vram_gb": 27.710740089416504, "grad_norm": 0.265625, "tokens_processed": 287047680, "tokens_processed_billions": 0.28704768}
+{"timestamp": 1787528531.5280752, "elapsed_hours": 0.5298998799588945, "session_id": 2, "event": "train_step", "step": 4390, "epoch": 1, "loss": 3.095824863910675, "loss_raw": 3.3022375106811523, "perplexity": 22.10546501965192, "muon_lr": 0.009994244085795072, "adamw_lr": 0.0002998273225738521, "tokens_per_sec": 151709.995963784, "vram_gb": 27.710740089416504, "grad_norm": 0.283203125, "tokens_processed": 287703040, "tokens_processed_billions": 0.28770304}
+{"timestamp": 1787528535.6581626, "elapsed_hours": 0.5310471265845829, "session_id": 2, "event": "train_step", "step": 4400, "epoch": 1, "loss": 2.9704726839065554, "loss_raw": 1.8945057392120361, "perplexity": 19.501135290615704, "muon_lr": 0.009994166550058595, "adamw_lr": 0.00029982499650175776, "tokens_per_sec": 151725.13786553394, "vram_gb": 27.299458026885986, "grad_norm": 0.228515625, "tokens_processed": 288358400, "tokens_processed_billions": 0.2883584}
+{"timestamp": 1787528535.6633065, "elapsed_hours": 0.5310485556390551, "session_id": 2, "event": "lambda_stats", "step": 4400, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2369791666666667, "lambda_std": 0.12217400110145729, "lambda_all": [1.6484375, 1.390625, 1.328125, 1.3203125, 1.1953125, 1.203125, 1.140625, 1.25, 1.203125, 1.15625, 1.171875, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.15625, 1.25], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787528539.6446033, "elapsed_hours": 0.5321544711457359, "session_id": 2, "event": "train_step", "step": 4410, "epoch": 1, "loss": 2.8939284777641294, "loss_raw": 4.169811248779297, "perplexity": 18.064134945029696, "muon_lr": 0.009994127587789593, "adamw_lr": 0.00029982382763368777, "tokens_per_sec": 151751.66792867446, "vram_gb": 27.710740089416504, "grad_norm": 0.3125, "tokens_processed": 289013760, "tokens_processed_billions": 0.28901376}
+{"timestamp": 1787528543.5854175, "elapsed_hours": 0.5332491417063607, "session_id": 2, "event": "train_step", "step": 4420, "epoch": 1, "loss": 2.9860692763328553, "loss_raw": 3.4612958431243896, "perplexity": 19.807670790421675, "muon_lr": 0.009994088495921583, "adamw_lr": 0.0002998226548776475, "tokens_per_sec": 151781.70835881212, "vram_gb": 27.710740089416504, "grad_norm": 0.2431640625, "tokens_processed": 289669120, "tokens_processed_billions": 0.28966912}
+{"timestamp": 1787528547.5424664, "elapsed_hours": 0.5343483218881819, "session_id": 2, "event": "train_step", "step": 4430, "epoch": 1, "loss": 2.931142337322235, "loss_raw": 2.843568801879883, "perplexity": 18.749035991766533, "muon_lr": 0.009994049274455692, "adamw_lr": 0.0002998214782336707, "tokens_per_sec": 151810.33156698837, "vram_gb": 27.710740089416504, "grad_norm": 0.291015625, "tokens_processed": 290324480, "tokens_processed_billions": 0.29032448}
+{"timestamp": 1787528551.643716, "elapsed_hours": 0.5354875579807493, "session_id": 2, "event": "train_step", "step": 4440, "epoch": 1, "loss": 2.8297900652885435, "loss_raw": 3.2307889461517334, "perplexity": 16.941903757501304, "muon_lr": 0.009993970442734784, "adamw_lr": 0.00029981911328204346, "tokens_per_sec": 151827.42801007975, "vram_gb": 27.299458026885986, "grad_norm": 0.2412109375, "tokens_processed": 290979840, "tokens_processed_billions": 0.29097984}
+{"timestamp": 1787528555.5819519, "elapsed_hours": 0.5365815124246809, "session_id": 2, "event": "train_step", "step": 4450, "epoch": 1, "loss": 2.9485987186431886, "loss_raw": 2.8248517513275146, "perplexity": 19.07919966083394, "muon_lr": 0.009993930832482043, "adamw_lr": 0.00029981792497446125, "tokens_per_sec": 151857.33002332246, "vram_gb": 27.710740089416504, "grad_norm": 0.267578125, "tokens_processed": 291635200, "tokens_processed_billions": 0.2916352}
+{"timestamp": 1787528559.5439615, "elapsed_hours": 0.5376820706658894, "session_id": 2, "event": "train_step", "step": 4460, "epoch": 1, "loss": 3.0003969788551332, "loss_raw": 3.725541114807129, "perplexity": 20.09351203951171, "muon_lr": 0.009993891092635966, "adamw_lr": 0.00029981673277907896, "tokens_per_sec": 151885.23275381516, "vram_gb": 27.710740089416504, "grad_norm": 0.25, "tokens_processed": 292290560, "tokens_processed_billions": 0.29229056}
+{"timestamp": 1787528563.5179403, "elapsed_hours": 0.5387859535879559, "session_id": 2, "event": "train_step", "step": 4470, "epoch": 1, "loss": 2.9947130346298216, "loss_raw": 3.110470771789551, "perplexity": 19.97962560646791, "muon_lr": 0.009993851223197698, "adamw_lr": 0.0002998155366959309, "tokens_per_sec": 151912.07876342992, "vram_gb": 27.710740089416504, "grad_norm": 0.2314453125, "tokens_processed": 292945920, "tokens_processed_billions": 0.29294592}
+{"timestamp": 1787528567.6160173, "elapsed_hours": 0.5399243083265093, "session_id": 2, "event": "train_step", "step": 4480, "epoch": 1, "loss": 2.9258391189575197, "loss_raw": 3.078981876373291, "perplexity": 18.649868944442986, "muon_lr": 0.009993771095549192, "adamw_lr": 0.00029981313286647575, "tokens_per_sec": 151929.05584626898, "vram_gb": 27.299458026885986, "grad_norm": 0.2412109375, "tokens_processed": 293601280, "tokens_processed_billions": 0.29360128}
+{"timestamp": 1787528571.5878413, "elapsed_hours": 0.541027592751715, "session_id": 2, "event": "train_step", "step": 4490, "epoch": 1, "loss": 2.9276273441314697, "loss_raw": 2.4901182651519775, "perplexity": 18.68324894616493, "muon_lr": 0.009993730837341269, "adamw_lr": 0.00029981192512023794, "tokens_per_sec": 151955.87413944985, "vram_gb": 27.710740089416504, "grad_norm": 0.2890625, "tokens_processed": 294256640, "tokens_processed_billions": 0.29425664}
+{"timestamp": 1787528575.5635412, "elapsed_hours": 0.5421319540341695, "session_id": 2, "event": "train_step", "step": 4500, "epoch": 1, "loss": 2.8711294078826906, "loss_raw": 2.2729012966156006, "perplexity": 17.656948840011868, "muon_lr": 0.009993690449545772, "adamw_lr": 0.00029981071348637314, "tokens_per_sec": 151982.28489176312, "vram_gb": 27.710740089416504, "grad_norm": 0.251953125, "tokens_processed": 294912000, "tokens_processed_billions": 0.294912}
+{"timestamp": 1787528575.5670547, "elapsed_hours": 0.5421329299608867, "session_id": 2, "event": "lambda_stats", "step": 4500, "epoch": 1, "lambda_min": 1.140625, "lambda_max": 1.6484375, "lambda_mean": 1.2361111111111112, "lambda_std": 0.12279841691156958, "lambda_all": [1.6484375, 1.390625, 1.328125, 1.3203125, 1.1953125, 1.203125, 1.140625, 1.25, 1.203125, 1.140625, 1.171875, 1.140625, 1.140625, 1.1953125, 1.140625, 1.234375, 1.15625, 1.25], "local_layers": 9, "global_layers": 9}
+{"timestamp": 1787528588.9910204, "elapsed_hours": 0.5458618095848295, "session_id": 2, "event": "validation", "step": 4500, "epoch": 1, "val_loss": 2.9183415269851682, "val_perplexity": 18.510562719953008, "is_best": true, "best_loss": 2.9183415269851682, "best_perplexity": 18.510562719953008, "train_loss": 2.2729012966156006, "train_loss_avg": 2.8711294078826906}