| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0006379585326954, | |
| "eval_steps": 294, | |
| "global_step": 1176, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0008506113769271664, | |
| "grad_norm": 0.7357208728790283, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8145, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0017012227538543328, | |
| "grad_norm": 0.771299421787262, | |
| "learning_rate": 4e-05, | |
| "loss": 3.195, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002551834130781499, | |
| "grad_norm": 0.7344720363616943, | |
| "learning_rate": 6e-05, | |
| "loss": 2.8861, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0034024455077086655, | |
| "grad_norm": 0.7500324845314026, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7421, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004253056884635832, | |
| "grad_norm": 0.9078495502471924, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9622, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005103668261562998, | |
| "grad_norm": 1.0794708728790283, | |
| "learning_rate": 0.00012, | |
| "loss": 3.1124, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005954279638490165, | |
| "grad_norm": 1.0218361616134644, | |
| "learning_rate": 0.00014, | |
| "loss": 2.7233, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006804891015417331, | |
| "grad_norm": 1.059141755104065, | |
| "learning_rate": 0.00016, | |
| "loss": 2.8784, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.007655502392344498, | |
| "grad_norm": 0.4901650547981262, | |
| "learning_rate": 0.00018, | |
| "loss": 2.6192, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.008506113769271665, | |
| "grad_norm": 0.8344933390617371, | |
| "learning_rate": 0.0002, | |
| "loss": 2.6448, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00935672514619883, | |
| "grad_norm": 1.5278894901275635, | |
| "learning_rate": 0.00019999963702861705, | |
| "loss": 2.7457, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.010207336523125997, | |
| "grad_norm": 1.2650033235549927, | |
| "learning_rate": 0.00019999854811710317, | |
| "loss": 2.7532, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.011057947900053162, | |
| "grad_norm": 0.740222156047821, | |
| "learning_rate": 0.0001999967332733632, | |
| "loss": 2.6836, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01190855927698033, | |
| "grad_norm": 0.49257639050483704, | |
| "learning_rate": 0.0001999941925105719, | |
| "loss": 2.6658, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.012759170653907496, | |
| "grad_norm": 0.3310573399066925, | |
| "learning_rate": 0.00019999092584717374, | |
| "loss": 2.5043, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013609782030834662, | |
| "grad_norm": 0.33361560106277466, | |
| "learning_rate": 0.00019998693330688282, | |
| "loss": 2.6252, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.014460393407761828, | |
| "grad_norm": 0.4449865221977234, | |
| "learning_rate": 0.00019998221491868273, | |
| "loss": 2.648, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.015311004784688996, | |
| "grad_norm": 0.4820970892906189, | |
| "learning_rate": 0.0001999767707168262, | |
| "loss": 2.7337, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01616161616161616, | |
| "grad_norm": 0.5144203901290894, | |
| "learning_rate": 0.0001999706007408351, | |
| "loss": 2.6967, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01701222753854333, | |
| "grad_norm": 0.501557469367981, | |
| "learning_rate": 0.0001999637050354999, | |
| "loss": 2.7318, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.017862838915470493, | |
| "grad_norm": 0.4480394423007965, | |
| "learning_rate": 0.00019995608365087946, | |
| "loss": 2.4126, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01871345029239766, | |
| "grad_norm": 0.4459284842014313, | |
| "learning_rate": 0.00019994773664230064, | |
| "loss": 2.7072, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01956406166932483, | |
| "grad_norm": 0.39909827709198, | |
| "learning_rate": 0.00019993866407035798, | |
| "loss": 2.6358, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.020414673046251993, | |
| "grad_norm": 0.36802783608436584, | |
| "learning_rate": 0.0001999288660009132, | |
| "loss": 2.6751, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02126528442317916, | |
| "grad_norm": 0.43287962675094604, | |
| "learning_rate": 0.0001999183425050946, | |
| "loss": 2.7518, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.022115895800106325, | |
| "grad_norm": 0.4289425313472748, | |
| "learning_rate": 0.00019990709365929677, | |
| "loss": 2.7535, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.022966507177033493, | |
| "grad_norm": 0.4627043604850769, | |
| "learning_rate": 0.00019989511954517992, | |
| "loss": 2.8111, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02381711855396066, | |
| "grad_norm": 0.4823961853981018, | |
| "learning_rate": 0.00019988242024966923, | |
| "loss": 2.9493, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.024667729930887825, | |
| "grad_norm": 0.4622437059879303, | |
| "learning_rate": 0.00019986899586495432, | |
| "loss": 2.788, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.025518341307814992, | |
| "grad_norm": 0.4963669776916504, | |
| "learning_rate": 0.00019985484648848853, | |
| "loss": 2.8304, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02636895268474216, | |
| "grad_norm": 0.47957557439804077, | |
| "learning_rate": 0.00019983997222298828, | |
| "loss": 2.7323, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.027219564061669324, | |
| "grad_norm": 0.445528507232666, | |
| "learning_rate": 0.00019982437317643217, | |
| "loss": 3.015, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.028070175438596492, | |
| "grad_norm": 0.46085312962532043, | |
| "learning_rate": 0.00019980804946206036, | |
| "loss": 2.8556, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.028920786815523656, | |
| "grad_norm": 0.5078282356262207, | |
| "learning_rate": 0.0001997910011983737, | |
| "loss": 2.8472, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.029771398192450824, | |
| "grad_norm": 0.4612430930137634, | |
| "learning_rate": 0.00019977322850913283, | |
| "loss": 2.6399, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03062200956937799, | |
| "grad_norm": 0.499965101480484, | |
| "learning_rate": 0.00019975473152335726, | |
| "loss": 2.9121, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03147262094630516, | |
| "grad_norm": 0.5101069808006287, | |
| "learning_rate": 0.0001997355103753246, | |
| "loss": 2.8488, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03232323232323232, | |
| "grad_norm": 0.5065872669219971, | |
| "learning_rate": 0.00019971556520456929, | |
| "loss": 2.8311, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03317384370015949, | |
| "grad_norm": 0.5324426889419556, | |
| "learning_rate": 0.00019969489615588189, | |
| "loss": 2.7454, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03402445507708666, | |
| "grad_norm": 0.5128815770149231, | |
| "learning_rate": 0.0001996735033793079, | |
| "loss": 2.8116, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03487506645401382, | |
| "grad_norm": 0.5330538153648376, | |
| "learning_rate": 0.00019965138703014655, | |
| "loss": 2.7584, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03572567783094099, | |
| "grad_norm": 0.556816577911377, | |
| "learning_rate": 0.00019962854726894997, | |
| "loss": 2.8902, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03657628920786816, | |
| "grad_norm": 0.5452866554260254, | |
| "learning_rate": 0.0001996049842615217, | |
| "loss": 2.7984, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.03742690058479532, | |
| "grad_norm": 0.5836021304130554, | |
| "learning_rate": 0.0001995806981789157, | |
| "loss": 2.803, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03827751196172249, | |
| "grad_norm": 0.5968561172485352, | |
| "learning_rate": 0.00019955568919743507, | |
| "loss": 2.8592, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03912812333864966, | |
| "grad_norm": 0.6416970491409302, | |
| "learning_rate": 0.0001995299574986306, | |
| "loss": 2.7488, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03997873471557682, | |
| "grad_norm": 0.704325795173645, | |
| "learning_rate": 0.0001995035032692998, | |
| "loss": 2.6983, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.040829346092503986, | |
| "grad_norm": 0.7766572833061218, | |
| "learning_rate": 0.00019947632670148517, | |
| "loss": 2.9677, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04167995746943115, | |
| "grad_norm": 0.7186003923416138, | |
| "learning_rate": 0.00019944842799247308, | |
| "loss": 3.0728, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04253056884635832, | |
| "grad_norm": 0.7572959065437317, | |
| "learning_rate": 0.00019941980734479214, | |
| "loss": 3.0345, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.043381180223285486, | |
| "grad_norm": 0.48461732268333435, | |
| "learning_rate": 0.00019939046496621194, | |
| "loss": 2.6307, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04423179160021265, | |
| "grad_norm": 0.468675434589386, | |
| "learning_rate": 0.0001993604010697413, | |
| "loss": 2.4616, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.04508240297713982, | |
| "grad_norm": 0.3815957009792328, | |
| "learning_rate": 0.0001993296158736269, | |
| "loss": 2.7479, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.045933014354066985, | |
| "grad_norm": 0.3313361704349518, | |
| "learning_rate": 0.00019929810960135172, | |
| "loss": 2.4983, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04678362573099415, | |
| "grad_norm": 0.32521429657936096, | |
| "learning_rate": 0.00019926588248163316, | |
| "loss": 2.5446, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04763423710792132, | |
| "grad_norm": 0.2972453236579895, | |
| "learning_rate": 0.00019923293474842174, | |
| "loss": 2.5472, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.048484848484848485, | |
| "grad_norm": 0.2972238063812256, | |
| "learning_rate": 0.00019919926664089909, | |
| "loss": 2.5389, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.04933545986177565, | |
| "grad_norm": 0.27498453855514526, | |
| "learning_rate": 0.00019916487840347644, | |
| "loss": 2.571, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.05018607123870282, | |
| "grad_norm": 0.2938655614852905, | |
| "learning_rate": 0.00019912977028579268, | |
| "loss": 2.7134, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.051036682615629984, | |
| "grad_norm": 0.26742392778396606, | |
| "learning_rate": 0.0001990939425427127, | |
| "loss": 2.5632, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05188729399255715, | |
| "grad_norm": 0.28117692470550537, | |
| "learning_rate": 0.00019905739543432536, | |
| "loss": 2.5297, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.05273790536948432, | |
| "grad_norm": 0.28916725516319275, | |
| "learning_rate": 0.00019902012922594177, | |
| "loss": 2.7096, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.053588516746411484, | |
| "grad_norm": 0.32468459010124207, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 2.6192, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.05443912812333865, | |
| "grad_norm": 0.2806537449359894, | |
| "learning_rate": 0.0001989434405965295, | |
| "loss": 2.6747, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05528973950026582, | |
| "grad_norm": 0.2876998782157898, | |
| "learning_rate": 0.0001989040187322164, | |
| "loss": 2.7443, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.056140350877192984, | |
| "grad_norm": 0.27619123458862305, | |
| "learning_rate": 0.00019886387888133413, | |
| "loss": 2.7379, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05699096225412015, | |
| "grad_norm": 0.31479549407958984, | |
| "learning_rate": 0.000198823021335275, | |
| "loss": 2.4039, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05784157363104731, | |
| "grad_norm": 0.300857812166214, | |
| "learning_rate": 0.00019878144639064144, | |
| "loss": 2.5705, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05869218500797448, | |
| "grad_norm": 0.3776433765888214, | |
| "learning_rate": 0.00019873915434924375, | |
| "loss": 2.863, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.05954279638490165, | |
| "grad_norm": 0.30585938692092896, | |
| "learning_rate": 0.00019869614551809795, | |
| "loss": 2.5312, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06039340776182881, | |
| "grad_norm": 0.3163856267929077, | |
| "learning_rate": 0.00019865242020942353, | |
| "loss": 2.8491, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.06124401913875598, | |
| "grad_norm": 0.30077147483825684, | |
| "learning_rate": 0.00019860797874064122, | |
| "loss": 2.7777, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.06209463051568315, | |
| "grad_norm": 0.4153176248073578, | |
| "learning_rate": 0.0001985628214343706, | |
| "loss": 2.7499, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.06294524189261032, | |
| "grad_norm": 0.35611122846603394, | |
| "learning_rate": 0.00019851694861842793, | |
| "loss": 2.7089, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06379585326953748, | |
| "grad_norm": 0.3143812417984009, | |
| "learning_rate": 0.00019847036062582357, | |
| "loss": 2.758, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06464646464646465, | |
| "grad_norm": 0.32024794816970825, | |
| "learning_rate": 0.00019842305779475968, | |
| "loss": 2.4616, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06549707602339182, | |
| "grad_norm": 0.3146126866340637, | |
| "learning_rate": 0.00019837504046862775, | |
| "loss": 2.6104, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06634768740031897, | |
| "grad_norm": 0.32578444480895996, | |
| "learning_rate": 0.00019832630899600608, | |
| "loss": 2.6297, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.06719829877724615, | |
| "grad_norm": 0.36873045563697815, | |
| "learning_rate": 0.00019827686373065728, | |
| "loss": 2.6358, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.06804891015417332, | |
| "grad_norm": 0.3558378517627716, | |
| "learning_rate": 0.00019822670503152567, | |
| "loss": 2.6308, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06889952153110047, | |
| "grad_norm": 0.37967684864997864, | |
| "learning_rate": 0.00019817583326273467, | |
| "loss": 2.7577, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06975013290802765, | |
| "grad_norm": 0.3737669885158539, | |
| "learning_rate": 0.00019812424879358425, | |
| "loss": 2.9207, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.07060074428495482, | |
| "grad_norm": 0.39410829544067383, | |
| "learning_rate": 0.0001980719519985481, | |
| "loss": 2.9544, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.07145135566188197, | |
| "grad_norm": 0.3863750696182251, | |
| "learning_rate": 0.00019801894325727104, | |
| "loss": 2.7794, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.07230196703880915, | |
| "grad_norm": 0.4226458966732025, | |
| "learning_rate": 0.0001979652229545662, | |
| "loss": 2.7491, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07315257841573632, | |
| "grad_norm": 0.42758506536483765, | |
| "learning_rate": 0.0001979107914804122, | |
| "loss": 2.8524, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07400318979266347, | |
| "grad_norm": 0.4379200041294098, | |
| "learning_rate": 0.0001978556492299504, | |
| "loss": 2.6526, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07485380116959064, | |
| "grad_norm": 0.44331902265548706, | |
| "learning_rate": 0.000197799796603482, | |
| "loss": 2.8028, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.07570441254651782, | |
| "grad_norm": 0.4358711540699005, | |
| "learning_rate": 0.0001977432340064651, | |
| "loss": 2.5426, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.07655502392344497, | |
| "grad_norm": 0.45511335134506226, | |
| "learning_rate": 0.00019768596184951173, | |
| "loss": 2.7067, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07740563530037214, | |
| "grad_norm": 0.5394377112388611, | |
| "learning_rate": 0.00019762798054838502, | |
| "loss": 2.8189, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.07825624667729932, | |
| "grad_norm": 0.5124706625938416, | |
| "learning_rate": 0.00019756929052399603, | |
| "loss": 2.7702, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07910685805422647, | |
| "grad_norm": 0.5025349855422974, | |
| "learning_rate": 0.00019750989220240073, | |
| "loss": 2.6872, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07995746943115364, | |
| "grad_norm": 0.5144663453102112, | |
| "learning_rate": 0.00019744978601479694, | |
| "loss": 2.6366, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 0.5908443927764893, | |
| "learning_rate": 0.00019738897239752118, | |
| "loss": 2.7918, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08165869218500797, | |
| "grad_norm": 0.6398508548736572, | |
| "learning_rate": 0.00019732745179204552, | |
| "loss": 2.9972, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.08250930356193514, | |
| "grad_norm": 0.6032273173332214, | |
| "learning_rate": 0.00019726522464497435, | |
| "loss": 2.7638, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.0833599149388623, | |
| "grad_norm": 0.6310097575187683, | |
| "learning_rate": 0.0001972022914080411, | |
| "loss": 2.9328, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.08421052631578947, | |
| "grad_norm": 0.7050711512565613, | |
| "learning_rate": 0.00019713865253810506, | |
| "loss": 2.8143, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.08506113769271664, | |
| "grad_norm": 0.755136251449585, | |
| "learning_rate": 0.00019707430849714807, | |
| "loss": 3.036, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0859117490696438, | |
| "grad_norm": 0.35153907537460327, | |
| "learning_rate": 0.00019700925975227096, | |
| "loss": 2.4444, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.08676236044657097, | |
| "grad_norm": 0.40153488516807556, | |
| "learning_rate": 0.0001969435067756904, | |
| "loss": 2.6068, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.08761297182349814, | |
| "grad_norm": 0.3474213480949402, | |
| "learning_rate": 0.00019687705004473545, | |
| "loss": 2.4261, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.0884635832004253, | |
| "grad_norm": 0.3283519744873047, | |
| "learning_rate": 0.00019680989004184382, | |
| "loss": 2.6736, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.08931419457735247, | |
| "grad_norm": 0.29034170508384705, | |
| "learning_rate": 0.00019674202725455877, | |
| "loss": 2.5551, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.09016480595427964, | |
| "grad_norm": 0.2918970584869385, | |
| "learning_rate": 0.00019667346217552527, | |
| "loss": 2.6039, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0910154173312068, | |
| "grad_norm": 0.2852106988430023, | |
| "learning_rate": 0.00019660419530248655, | |
| "loss": 2.5432, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.09186602870813397, | |
| "grad_norm": 0.30997323989868164, | |
| "learning_rate": 0.0001965342271382805, | |
| "loss": 2.7324, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.09271664008506114, | |
| "grad_norm": 0.34156399965286255, | |
| "learning_rate": 0.00019646355819083589, | |
| "loss": 2.6548, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.0935672514619883, | |
| "grad_norm": 0.2763843238353729, | |
| "learning_rate": 0.00019639218897316883, | |
| "loss": 2.5254, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09441786283891547, | |
| "grad_norm": 0.2835611402988434, | |
| "learning_rate": 0.00019632012000337908, | |
| "loss": 2.5677, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.09526847421584264, | |
| "grad_norm": 0.2940271198749542, | |
| "learning_rate": 0.00019624735180464602, | |
| "loss": 2.5976, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0961190855927698, | |
| "grad_norm": 0.2714485824108124, | |
| "learning_rate": 0.00019617388490522517, | |
| "loss": 2.6087, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.09696969696969697, | |
| "grad_norm": 0.30371204018592834, | |
| "learning_rate": 0.00019609971983844412, | |
| "loss": 2.6129, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.09782030834662414, | |
| "grad_norm": 0.2762625813484192, | |
| "learning_rate": 0.0001960248571426989, | |
| "loss": 2.5759, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0986709197235513, | |
| "grad_norm": 0.2702981233596802, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 2.5443, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.09952153110047847, | |
| "grad_norm": 0.29210978746414185, | |
| "learning_rate": 0.00019587304104321746, | |
| "loss": 2.6425, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.10037214247740564, | |
| "grad_norm": 0.31620749831199646, | |
| "learning_rate": 0.00019579608874157928, | |
| "loss": 2.703, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1012227538543328, | |
| "grad_norm": 0.2803102433681488, | |
| "learning_rate": 0.00019571844101516484, | |
| "loss": 2.6886, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.10207336523125997, | |
| "grad_norm": 0.30169349908828735, | |
| "learning_rate": 0.00019564009842765225, | |
| "loss": 2.8221, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.10292397660818714, | |
| "grad_norm": 0.297553151845932, | |
| "learning_rate": 0.00019556106154776379, | |
| "loss": 2.6897, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1037745879851143, | |
| "grad_norm": 0.30721086263656616, | |
| "learning_rate": 0.000195481330949262, | |
| "loss": 2.6551, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.10462519936204147, | |
| "grad_norm": 0.29124605655670166, | |
| "learning_rate": 0.00019540090721094542, | |
| "loss": 2.6292, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.10547581073896864, | |
| "grad_norm": 0.31037285923957825, | |
| "learning_rate": 0.0001953197909166443, | |
| "loss": 2.5459, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1063264221158958, | |
| "grad_norm": 0.3543750047683716, | |
| "learning_rate": 0.00019523798265521654, | |
| "loss": 2.5622, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.10717703349282297, | |
| "grad_norm": 0.3356544077396393, | |
| "learning_rate": 0.00019515548302054335, | |
| "loss": 2.7272, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.10802764486975014, | |
| "grad_norm": 0.34296396374702454, | |
| "learning_rate": 0.00019507229261152476, | |
| "loss": 2.6629, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1088782562466773, | |
| "grad_norm": 0.34629112482070923, | |
| "learning_rate": 0.0001949884120320756, | |
| "loss": 2.6371, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.10972886762360447, | |
| "grad_norm": 0.34170377254486084, | |
| "learning_rate": 0.00019490384189112082, | |
| "loss": 2.7218, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.11057947900053164, | |
| "grad_norm": 0.38438230752944946, | |
| "learning_rate": 0.0001948185828025913, | |
| "loss": 2.7096, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1114300903774588, | |
| "grad_norm": 0.40347060561180115, | |
| "learning_rate": 0.00019473263538541914, | |
| "loss": 2.8129, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.11228070175438597, | |
| "grad_norm": 0.3742891848087311, | |
| "learning_rate": 0.00019464600026353348, | |
| "loss": 2.7916, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.11313131313131314, | |
| "grad_norm": 0.4015231430530548, | |
| "learning_rate": 0.0001945586780658557, | |
| "loss": 2.6099, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.1139819245082403, | |
| "grad_norm": 0.40618133544921875, | |
| "learning_rate": 0.00019447066942629491, | |
| "loss": 2.6669, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.11483253588516747, | |
| "grad_norm": 0.4171842932701111, | |
| "learning_rate": 0.00019438197498374357, | |
| "loss": 2.6272, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.11568314726209462, | |
| "grad_norm": 0.443013995885849, | |
| "learning_rate": 0.0001942925953820725, | |
| "loss": 2.5722, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1165337586390218, | |
| "grad_norm": 0.4636158347129822, | |
| "learning_rate": 0.00019420253127012645, | |
| "loss": 2.8075, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.11738437001594897, | |
| "grad_norm": 0.4271916151046753, | |
| "learning_rate": 0.00019411178330171937, | |
| "loss": 2.6875, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.11823498139287612, | |
| "grad_norm": 0.47826603055000305, | |
| "learning_rate": 0.00019402035213562954, | |
| "loss": 2.7042, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1190855927698033, | |
| "grad_norm": 0.46729791164398193, | |
| "learning_rate": 0.0001939282384355949, | |
| "loss": 2.6663, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11993620414673047, | |
| "grad_norm": 0.4689824879169464, | |
| "learning_rate": 0.0001938354428703082, | |
| "loss": 2.6138, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.12078681552365762, | |
| "grad_norm": 0.526096522808075, | |
| "learning_rate": 0.0001937419661134121, | |
| "loss": 2.9258, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1216374269005848, | |
| "grad_norm": 0.5075511932373047, | |
| "learning_rate": 0.0001936478088434944, | |
| "loss": 2.8021, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.12248803827751197, | |
| "grad_norm": 0.5048439502716064, | |
| "learning_rate": 0.00019355297174408298, | |
| "loss": 2.6274, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.12333864965443912, | |
| "grad_norm": 0.5787357687950134, | |
| "learning_rate": 0.00019345745550364087, | |
| "loss": 2.851, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1241892610313663, | |
| "grad_norm": 0.5641311407089233, | |
| "learning_rate": 0.00019336126081556134, | |
| "loss": 2.7681, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.12503987240829345, | |
| "grad_norm": 0.5504147410392761, | |
| "learning_rate": 0.00019326438837816276, | |
| "loss": 2.6905, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.12589048378522064, | |
| "grad_norm": 0.6101283431053162, | |
| "learning_rate": 0.00019316683889468358, | |
| "loss": 2.589, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1267410951621478, | |
| "grad_norm": 0.7153661847114563, | |
| "learning_rate": 0.00019306861307327725, | |
| "loss": 2.9563, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.12759170653907495, | |
| "grad_norm": 0.7049738168716431, | |
| "learning_rate": 0.00019296971162700694, | |
| "loss": 2.8023, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12844231791600214, | |
| "grad_norm": 0.3282754421234131, | |
| "learning_rate": 0.00019287013527384062, | |
| "loss": 2.4278, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1292929292929293, | |
| "grad_norm": 0.350577712059021, | |
| "learning_rate": 0.00019276988473664557, | |
| "loss": 2.5845, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.13014354066985645, | |
| "grad_norm": 0.32433176040649414, | |
| "learning_rate": 0.00019266896074318334, | |
| "loss": 2.6126, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.13099415204678364, | |
| "grad_norm": 0.31844663619995117, | |
| "learning_rate": 0.00019256736402610436, | |
| "loss": 2.527, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.1318447634237108, | |
| "grad_norm": 0.2559802830219269, | |
| "learning_rate": 0.00019246509532294266, | |
| "loss": 2.2437, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.13269537480063795, | |
| "grad_norm": 0.28512275218963623, | |
| "learning_rate": 0.00019236215537611046, | |
| "loss": 2.5739, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.13354598617756513, | |
| "grad_norm": 0.26634740829467773, | |
| "learning_rate": 0.00019225854493289286, | |
| "loss": 2.4485, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1343965975544923, | |
| "grad_norm": 0.2785400450229645, | |
| "learning_rate": 0.0001921542647454424, | |
| "loss": 2.7944, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.13524720893141945, | |
| "grad_norm": 0.27485981583595276, | |
| "learning_rate": 0.00019204931557077355, | |
| "loss": 2.6518, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.13609782030834663, | |
| "grad_norm": 0.2687318027019501, | |
| "learning_rate": 0.00019194369817075724, | |
| "loss": 2.6595, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1369484316852738, | |
| "grad_norm": 0.26418977975845337, | |
| "learning_rate": 0.00019183741331211537, | |
| "loss": 2.7045, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.13779904306220095, | |
| "grad_norm": 0.28258347511291504, | |
| "learning_rate": 0.00019173046176641513, | |
| "loss": 2.5896, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.13864965443912813, | |
| "grad_norm": 0.27390146255493164, | |
| "learning_rate": 0.00019162284431006358, | |
| "loss": 2.5566, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1395002658160553, | |
| "grad_norm": 0.2916048765182495, | |
| "learning_rate": 0.00019151456172430183, | |
| "loss": 2.609, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 0.30684247612953186, | |
| "learning_rate": 0.00019140561479519955, | |
| "loss": 2.5222, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.14120148856990963, | |
| "grad_norm": 0.26836761832237244, | |
| "learning_rate": 0.00019129600431364897, | |
| "loss": 2.5891, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1420520999468368, | |
| "grad_norm": 0.2658300995826721, | |
| "learning_rate": 0.00019118573107535953, | |
| "loss": 2.644, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.14290271132376395, | |
| "grad_norm": 0.2789425551891327, | |
| "learning_rate": 0.00019107479588085182, | |
| "loss": 2.5641, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.14375332270069113, | |
| "grad_norm": 0.2909972071647644, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 2.5982, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1446039340776183, | |
| "grad_norm": 0.3741363286972046, | |
| "learning_rate": 0.0001908509428492852, | |
| "loss": 2.6293, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 0.2989426851272583, | |
| "learning_rate": 0.0001907380266372712, | |
| "loss": 2.7364, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.14630515683147263, | |
| "grad_norm": 0.28862622380256653, | |
| "learning_rate": 0.00019062445171911686, | |
| "loss": 2.5656, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.1471557682083998, | |
| "grad_norm": 0.3215920329093933, | |
| "learning_rate": 0.0001905102189193112, | |
| "loss": 2.8443, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.14800637958532695, | |
| "grad_norm": 0.2994636595249176, | |
| "learning_rate": 0.00019039532906711882, | |
| "loss": 2.7014, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.14885699096225413, | |
| "grad_norm": 0.32109183073043823, | |
| "learning_rate": 0.00019027978299657436, | |
| "loss": 2.8364, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1497076023391813, | |
| "grad_norm": 0.30813783407211304, | |
| "learning_rate": 0.00019016358154647618, | |
| "loss": 2.5102, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.15055821371610845, | |
| "grad_norm": 0.32674533128738403, | |
| "learning_rate": 0.00019004672556038028, | |
| "loss": 2.757, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.15140882509303563, | |
| "grad_norm": 0.34680357575416565, | |
| "learning_rate": 0.00018992921588659422, | |
| "loss": 2.5228, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.1522594364699628, | |
| "grad_norm": 0.35170817375183105, | |
| "learning_rate": 0.00018981105337817104, | |
| "loss": 2.6148, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.15311004784688995, | |
| "grad_norm": 0.3741483986377716, | |
| "learning_rate": 0.00018969223889290284, | |
| "loss": 2.8025, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15396065922381713, | |
| "grad_norm": 0.4156269431114197, | |
| "learning_rate": 0.00018957277329331485, | |
| "loss": 2.72, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.1548112706007443, | |
| "grad_norm": 0.3726477324962616, | |
| "learning_rate": 0.00018945265744665886, | |
| "loss": 2.6197, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.15566188197767145, | |
| "grad_norm": 0.4135706424713135, | |
| "learning_rate": 0.00018933189222490726, | |
| "loss": 2.7176, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.15651249335459863, | |
| "grad_norm": 0.38799911737442017, | |
| "learning_rate": 0.00018921047850474642, | |
| "loss": 2.5641, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1573631047315258, | |
| "grad_norm": 0.4622843265533447, | |
| "learning_rate": 0.00018908841716757042, | |
| "loss": 2.7626, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.15821371610845295, | |
| "grad_norm": 0.4251146912574768, | |
| "learning_rate": 0.00018896570909947475, | |
| "loss": 2.6842, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.15906432748538013, | |
| "grad_norm": 0.4628697335720062, | |
| "learning_rate": 0.00018884235519124972, | |
| "loss": 2.9476, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.1599149388623073, | |
| "grad_norm": 0.5052159428596497, | |
| "learning_rate": 0.0001887183563383741, | |
| "loss": 2.769, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.16076555023923444, | |
| "grad_norm": 0.4817435145378113, | |
| "learning_rate": 0.00018859371344100864, | |
| "loss": 2.6266, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 0.4751468598842621, | |
| "learning_rate": 0.0001884684274039894, | |
| "loss": 2.877, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1624667729930888, | |
| "grad_norm": 0.5826165676116943, | |
| "learning_rate": 0.00018834249913682132, | |
| "loss": 2.7308, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.16331738437001594, | |
| "grad_norm": 0.5441760420799255, | |
| "learning_rate": 0.00018821592955367154, | |
| "loss": 2.6764, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1641679957469431, | |
| "grad_norm": 0.5005947947502136, | |
| "learning_rate": 0.00018808871957336275, | |
| "loss": 2.664, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.1650186071238703, | |
| "grad_norm": 0.5205551981925964, | |
| "learning_rate": 0.00018796087011936665, | |
| "loss": 2.6192, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.16586921850079744, | |
| "grad_norm": 0.5489931106567383, | |
| "learning_rate": 0.0001878323821197971, | |
| "loss": 2.5061, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.1667198298777246, | |
| "grad_norm": 0.5525840520858765, | |
| "learning_rate": 0.00018770325650740345, | |
| "loss": 2.7474, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1675704412546518, | |
| "grad_norm": 0.5978725552558899, | |
| "learning_rate": 0.0001875734942195637, | |
| "loss": 2.6055, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.16842105263157894, | |
| "grad_norm": 0.6148700714111328, | |
| "learning_rate": 0.0001874430961982778, | |
| "loss": 2.8352, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1692716640085061, | |
| "grad_norm": 0.5956620573997498, | |
| "learning_rate": 0.0001873120633901608, | |
| "loss": 2.7367, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.17012227538543329, | |
| "grad_norm": 0.7082740664482117, | |
| "learning_rate": 0.0001871803967464358, | |
| "loss": 2.9437, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17097288676236044, | |
| "grad_norm": 0.32244405150413513, | |
| "learning_rate": 0.00018704809722292737, | |
| "loss": 2.3835, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.1718234981392876, | |
| "grad_norm": 0.3367772102355957, | |
| "learning_rate": 0.00018691516578005427, | |
| "loss": 2.601, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.17267410951621479, | |
| "grad_norm": 0.31732872128486633, | |
| "learning_rate": 0.00018678160338282272, | |
| "loss": 2.5894, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.17352472089314194, | |
| "grad_norm": 0.27467650175094604, | |
| "learning_rate": 0.0001866474110008193, | |
| "loss": 2.4369, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1743753322700691, | |
| "grad_norm": 0.29726937413215637, | |
| "learning_rate": 0.00018651258960820385, | |
| "loss": 2.6123, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.17522594364699628, | |
| "grad_norm": 0.27499106526374817, | |
| "learning_rate": 0.00018637714018370253, | |
| "loss": 2.5141, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.17607655502392344, | |
| "grad_norm": 0.27535390853881836, | |
| "learning_rate": 0.00018624106371060067, | |
| "loss": 2.5148, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.1769271664008506, | |
| "grad_norm": 0.2687024176120758, | |
| "learning_rate": 0.00018610436117673555, | |
| "loss": 2.6057, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.31320950388908386, | |
| "learning_rate": 0.00018596703357448934, | |
| "loss": 2.6813, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.17862838915470494, | |
| "grad_norm": 0.25832033157348633, | |
| "learning_rate": 0.00018582908190078185, | |
| "loss": 2.4898, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1794790005316321, | |
| "grad_norm": 0.2806166410446167, | |
| "learning_rate": 0.00018569050715706325, | |
| "loss": 2.5762, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.18032961190855928, | |
| "grad_norm": 0.26099708676338196, | |
| "learning_rate": 0.00018555131034930685, | |
| "loss": 2.5386, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.18118022328548644, | |
| "grad_norm": 0.26140880584716797, | |
| "learning_rate": 0.00018541149248800184, | |
| "loss": 2.7159, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.1820308346624136, | |
| "grad_norm": 0.2698177695274353, | |
| "learning_rate": 0.0001852710545881459, | |
| "loss": 2.5942, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.18288144603934078, | |
| "grad_norm": 0.27240726351737976, | |
| "learning_rate": 0.00018512999766923772, | |
| "loss": 2.5377, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.18373205741626794, | |
| "grad_norm": 0.2780822813510895, | |
| "learning_rate": 0.00018498832275526988, | |
| "loss": 2.6185, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.1845826687931951, | |
| "grad_norm": 0.2713901400566101, | |
| "learning_rate": 0.00018484603087472109, | |
| "loss": 2.5802, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.18543328017012228, | |
| "grad_norm": 0.2843954265117645, | |
| "learning_rate": 0.000184703123060549, | |
| "loss": 2.6404, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.18628389154704944, | |
| "grad_norm": 0.2679051160812378, | |
| "learning_rate": 0.0001845596003501826, | |
| "loss": 2.6688, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.1871345029239766, | |
| "grad_norm": 0.292568176984787, | |
| "learning_rate": 0.00018441546378551458, | |
| "loss": 2.6505, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18798511430090378, | |
| "grad_norm": 0.282326877117157, | |
| "learning_rate": 0.00018427071441289388, | |
| "loss": 2.6299, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.18883572567783094, | |
| "grad_norm": 0.2853985130786896, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 2.8143, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1896863370547581, | |
| "grad_norm": 0.2786814868450165, | |
| "learning_rate": 0.00018397938145142591, | |
| "loss": 2.6007, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.19053694843168528, | |
| "grad_norm": 0.42460358142852783, | |
| "learning_rate": 0.0001838327999774892, | |
| "loss": 2.7891, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.19138755980861244, | |
| "grad_norm": 0.30478838086128235, | |
| "learning_rate": 0.00018368560992540562, | |
| "loss": 2.4551, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1922381711855396, | |
| "grad_norm": 0.3402044177055359, | |
| "learning_rate": 0.00018353781236369064, | |
| "loss": 2.9191, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.19308878256246678, | |
| "grad_norm": 0.33662521839141846, | |
| "learning_rate": 0.00018338940836527004, | |
| "loss": 2.5606, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.19393939393939394, | |
| "grad_norm": 0.34461426734924316, | |
| "learning_rate": 0.0001832403990074719, | |
| "loss": 2.714, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.1947900053163211, | |
| "grad_norm": 0.342184454202652, | |
| "learning_rate": 0.0001830907853720188, | |
| "loss": 2.6936, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.19564061669324828, | |
| "grad_norm": 0.3557281494140625, | |
| "learning_rate": 0.0001829405685450202, | |
| "loss": 2.6663, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19649122807017544, | |
| "grad_norm": 0.38674700260162354, | |
| "learning_rate": 0.0001827897496169642, | |
| "loss": 2.7257, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.1973418394471026, | |
| "grad_norm": 0.3849089741706848, | |
| "learning_rate": 0.00018263832968271, | |
| "loss": 2.7178, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.19819245082402978, | |
| "grad_norm": 0.4508901834487915, | |
| "learning_rate": 0.00018248630984147955, | |
| "loss": 2.7947, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.19904306220095694, | |
| "grad_norm": 0.39502936601638794, | |
| "learning_rate": 0.00018233369119684996, | |
| "loss": 2.5885, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.1998936735778841, | |
| "grad_norm": 0.4287837743759155, | |
| "learning_rate": 0.00018218047485674523, | |
| "loss": 2.6911, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.20074428495481128, | |
| "grad_norm": 0.4257849454879761, | |
| "learning_rate": 0.00018202666193342833, | |
| "loss": 2.8803, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.20159489633173844, | |
| "grad_norm": 0.4459477961063385, | |
| "learning_rate": 0.00018187225354349295, | |
| "loss": 2.8352, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.2024455077086656, | |
| "grad_norm": 0.4430312514305115, | |
| "learning_rate": 0.0001817172508078557, | |
| "loss": 2.7517, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.20329611908559278, | |
| "grad_norm": 0.4465429484844208, | |
| "learning_rate": 0.00018156165485174773, | |
| "loss": 2.7119, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.20414673046251994, | |
| "grad_norm": 0.4532601833343506, | |
| "learning_rate": 0.00018140546680470659, | |
| "loss": 2.7346, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2049973418394471, | |
| "grad_norm": 0.4750036299228668, | |
| "learning_rate": 0.00018124868780056814, | |
| "loss": 2.6113, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.20584795321637428, | |
| "grad_norm": 0.5072234272956848, | |
| "learning_rate": 0.00018109131897745822, | |
| "loss": 2.844, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.20669856459330144, | |
| "grad_norm": 0.5094662308692932, | |
| "learning_rate": 0.00018093336147778438, | |
| "loss": 2.7737, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2075491759702286, | |
| "grad_norm": 0.606842577457428, | |
| "learning_rate": 0.00018077481644822768, | |
| "loss": 2.6153, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.20839978734715578, | |
| "grad_norm": 0.5311163067817688, | |
| "learning_rate": 0.00018061568503973435, | |
| "loss": 2.6038, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.20925039872408294, | |
| "grad_norm": 0.5758761167526245, | |
| "learning_rate": 0.00018045596840750723, | |
| "loss": 2.6446, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2101010101010101, | |
| "grad_norm": 0.598297119140625, | |
| "learning_rate": 0.00018029566771099776, | |
| "loss": 2.7002, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.21095162147793728, | |
| "grad_norm": 0.6635774970054626, | |
| "learning_rate": 0.00018013478411389716, | |
| "loss": 2.8011, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.21180223285486444, | |
| "grad_norm": 0.6850919723510742, | |
| "learning_rate": 0.00017997331878412835, | |
| "loss": 2.8903, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.2126528442317916, | |
| "grad_norm": 0.7298348546028137, | |
| "learning_rate": 0.00017981127289383716, | |
| "loss": 2.9483, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21350345560871878, | |
| "grad_norm": 0.33354559540748596, | |
| "learning_rate": 0.00017964864761938404, | |
| "loss": 2.4727, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.21435406698564594, | |
| "grad_norm": 0.3557465374469757, | |
| "learning_rate": 0.00017948544414133534, | |
| "loss": 2.5058, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2152046783625731, | |
| "grad_norm": 0.3230442702770233, | |
| "learning_rate": 0.00017932166364445498, | |
| "loss": 2.5422, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.21605528973950028, | |
| "grad_norm": 0.28668278455734253, | |
| "learning_rate": 0.0001791573073176956, | |
| "loss": 2.3173, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.21690590111642744, | |
| "grad_norm": 0.30019721388816833, | |
| "learning_rate": 0.00017899237635419002, | |
| "loss": 2.6444, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2177565124933546, | |
| "grad_norm": 0.285314679145813, | |
| "learning_rate": 0.0001788268719512427, | |
| "loss": 2.5319, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.21860712387028178, | |
| "grad_norm": 0.27584996819496155, | |
| "learning_rate": 0.00017866079531032088, | |
| "loss": 2.6496, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.21945773524720893, | |
| "grad_norm": 0.2874069809913635, | |
| "learning_rate": 0.0001784941476370459, | |
| "loss": 2.5156, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2203083466241361, | |
| "grad_norm": 0.26786255836486816, | |
| "learning_rate": 0.00017832693014118448, | |
| "loss": 2.6211, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.22115895800106328, | |
| "grad_norm": 0.2633914351463318, | |
| "learning_rate": 0.0001781591440366399, | |
| "loss": 2.5811, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22200956937799043, | |
| "grad_norm": 0.2724866569042206, | |
| "learning_rate": 0.00017799079054144334, | |
| "loss": 2.5904, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.2228601807549176, | |
| "grad_norm": 0.29333001375198364, | |
| "learning_rate": 0.00017782187087774477, | |
| "loss": 2.7581, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.22371079213184478, | |
| "grad_norm": 0.2735550105571747, | |
| "learning_rate": 0.00017765238627180424, | |
| "loss": 2.7114, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.22456140350877193, | |
| "grad_norm": 0.2721397280693054, | |
| "learning_rate": 0.00017748233795398307, | |
| "loss": 2.5991, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2254120148856991, | |
| "grad_norm": 0.25755858421325684, | |
| "learning_rate": 0.0001773117271587346, | |
| "loss": 2.5786, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.22626262626262628, | |
| "grad_norm": 0.25772804021835327, | |
| "learning_rate": 0.00017714055512459565, | |
| "loss": 2.488, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.22711323763955343, | |
| "grad_norm": 0.2766227424144745, | |
| "learning_rate": 0.0001769688230941772, | |
| "loss": 2.8924, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2279638490164806, | |
| "grad_norm": 0.26846593618392944, | |
| "learning_rate": 0.00017679653231415552, | |
| "loss": 2.5783, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.22881446039340775, | |
| "grad_norm": 0.26374372839927673, | |
| "learning_rate": 0.00017662368403526302, | |
| "loss": 2.4675, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.22966507177033493, | |
| "grad_norm": 0.28237268328666687, | |
| "learning_rate": 0.0001764502795122793, | |
| "loss": 2.5994, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2305156831472621, | |
| "grad_norm": 0.2786102890968323, | |
| "learning_rate": 0.00017627632000402193, | |
| "loss": 2.514, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.23136629452418925, | |
| "grad_norm": 0.27646180987358093, | |
| "learning_rate": 0.00017610180677333739, | |
| "loss": 2.5673, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.23221690590111643, | |
| "grad_norm": 0.3052549660205841, | |
| "learning_rate": 0.00017592674108709186, | |
| "loss": 2.5345, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2330675172780436, | |
| "grad_norm": 0.30554690957069397, | |
| "learning_rate": 0.00017575112421616202, | |
| "loss": 2.709, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.23391812865497075, | |
| "grad_norm": 0.3219161331653595, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 2.6825, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.23476874003189793, | |
| "grad_norm": 0.31834957003593445, | |
| "learning_rate": 0.0001753982420237533, | |
| "loss": 2.7017, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.2356193514088251, | |
| "grad_norm": 0.30264872312545776, | |
| "learning_rate": 0.00017522097926399722, | |
| "loss": 2.3725, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.23646996278575225, | |
| "grad_norm": 0.3283548951148987, | |
| "learning_rate": 0.00017504317044298367, | |
| "loss": 2.6217, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.23732057416267943, | |
| "grad_norm": 0.33564746379852295, | |
| "learning_rate": 0.00017486481685150302, | |
| "loss": 2.5738, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2381711855396066, | |
| "grad_norm": 0.37258434295654297, | |
| "learning_rate": 0.0001746859197843002, | |
| "loss": 2.783, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23902179691653375, | |
| "grad_norm": 0.3897363245487213, | |
| "learning_rate": 0.0001745064805400656, | |
| "loss": 2.7908, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.23987240829346093, | |
| "grad_norm": 0.3756699562072754, | |
| "learning_rate": 0.00017432650042142536, | |
| "loss": 2.5944, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.2407230196703881, | |
| "grad_norm": 0.3787755072116852, | |
| "learning_rate": 0.00017414598073493216, | |
| "loss": 2.7574, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.24157363104731525, | |
| "grad_norm": 0.38891106843948364, | |
| "learning_rate": 0.0001739649227910556, | |
| "loss": 2.8635, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 0.40293633937835693, | |
| "learning_rate": 0.00017378332790417273, | |
| "loss": 2.729, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2432748538011696, | |
| "grad_norm": 0.414109468460083, | |
| "learning_rate": 0.00017360119739255852, | |
| "loss": 2.6077, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.24412546517809675, | |
| "grad_norm": 0.42549028992652893, | |
| "learning_rate": 0.0001734185325783762, | |
| "loss": 2.7812, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.24497607655502393, | |
| "grad_norm": 0.42882055044174194, | |
| "learning_rate": 0.00017323533478766777, | |
| "loss": 2.7653, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.2458266879319511, | |
| "grad_norm": 0.42119139432907104, | |
| "learning_rate": 0.00017305160535034436, | |
| "loss": 2.5355, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.24667729930887825, | |
| "grad_norm": 0.4749990999698639, | |
| "learning_rate": 0.0001728673456001766, | |
| "loss": 2.7885, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24752791068580543, | |
| "grad_norm": 0.4682268500328064, | |
| "learning_rate": 0.00017268255687478469, | |
| "loss": 2.6402, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.2483785220627326, | |
| "grad_norm": 0.4854019284248352, | |
| "learning_rate": 0.00017249724051562906, | |
| "loss": 2.7255, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.24922913343965974, | |
| "grad_norm": 0.5112527012825012, | |
| "learning_rate": 0.00017231139786800042, | |
| "loss": 2.8374, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.2500797448165869, | |
| "grad_norm": 0.5242344737052917, | |
| "learning_rate": 0.0001721250302810101, | |
| "loss": 2.9178, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2500797448165869, | |
| "eval_loss": 2.688343048095703, | |
| "eval_runtime": 80.6326, | |
| "eval_samples_per_second": 12.278, | |
| "eval_steps_per_second": 6.139, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2509303561935141, | |
| "grad_norm": 0.6918848156929016, | |
| "learning_rate": 0.00017193813910758, | |
| "loss": 2.6556, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.25178096757044127, | |
| "grad_norm": 0.5682982802391052, | |
| "learning_rate": 0.00017175072570443312, | |
| "loss": 2.6581, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.25263157894736843, | |
| "grad_norm": 0.6087559461593628, | |
| "learning_rate": 0.00017156279143208352, | |
| "loss": 2.5665, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.2534821903242956, | |
| "grad_norm": 0.6545628309249878, | |
| "learning_rate": 0.00017137433765482642, | |
| "loss": 2.8215, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.25433280170122274, | |
| "grad_norm": 0.6754540801048279, | |
| "learning_rate": 0.00017118536574072842, | |
| "loss": 2.7991, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.2551834130781499, | |
| "grad_norm": 0.6926846504211426, | |
| "learning_rate": 0.0001709958770616174, | |
| "loss": 2.7371, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2560340244550771, | |
| "grad_norm": 0.32997071743011475, | |
| "learning_rate": 0.00017080587299307283, | |
| "loss": 2.7739, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.25688463583200427, | |
| "grad_norm": 0.3260290324687958, | |
| "learning_rate": 0.0001706153549144154, | |
| "loss": 2.5971, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.25773524720893143, | |
| "grad_norm": 0.3378421366214752, | |
| "learning_rate": 0.00017042432420869732, | |
| "loss": 2.7588, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.2585858585858586, | |
| "grad_norm": 0.27844691276550293, | |
| "learning_rate": 0.0001702327822626922, | |
| "loss": 2.6493, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.25943646996278574, | |
| "grad_norm": 0.29100462794303894, | |
| "learning_rate": 0.00017004073046688497, | |
| "loss": 2.6397, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2602870813397129, | |
| "grad_norm": 0.2764577865600586, | |
| "learning_rate": 0.00016984817021546177, | |
| "loss": 2.4199, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.26113769271664006, | |
| "grad_norm": 0.26870644092559814, | |
| "learning_rate": 0.00016965510290629972, | |
| "loss": 2.2552, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.26198830409356727, | |
| "grad_norm": 0.2770349979400635, | |
| "learning_rate": 0.00016946152994095704, | |
| "loss": 2.6118, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.2628389154704944, | |
| "grad_norm": 0.27041003108024597, | |
| "learning_rate": 0.00016926745272466268, | |
| "loss": 2.5329, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.2636895268474216, | |
| "grad_norm": 0.25608015060424805, | |
| "learning_rate": 0.00016907287266630614, | |
| "loss": 2.3411, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26454013822434874, | |
| "grad_norm": 0.2750420570373535, | |
| "learning_rate": 0.00016887779117842725, | |
| "loss": 2.6393, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2653907496012759, | |
| "grad_norm": 0.26276537775993347, | |
| "learning_rate": 0.00016868220967720604, | |
| "loss": 2.3616, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.26624136097820306, | |
| "grad_norm": 0.2735307514667511, | |
| "learning_rate": 0.00016848612958245216, | |
| "loss": 2.5156, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.26709197235513027, | |
| "grad_norm": 0.32951095700263977, | |
| "learning_rate": 0.00016828955231759497, | |
| "loss": 2.5329, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.2679425837320574, | |
| "grad_norm": 0.2762184143066406, | |
| "learning_rate": 0.00016809247930967282, | |
| "loss": 2.6873, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2687931951089846, | |
| "grad_norm": 0.2802570164203644, | |
| "learning_rate": 0.000167894911989323, | |
| "loss": 2.5532, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.26964380648591174, | |
| "grad_norm": 0.26968276500701904, | |
| "learning_rate": 0.0001676968517907712, | |
| "loss": 2.602, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2704944178628389, | |
| "grad_norm": 0.27560874819755554, | |
| "learning_rate": 0.00016749830015182107, | |
| "loss": 2.5003, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.27134502923976606, | |
| "grad_norm": 0.288411021232605, | |
| "learning_rate": 0.00016729925851384386, | |
| "loss": 2.6859, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.27219564061669327, | |
| "grad_norm": 0.2999224364757538, | |
| "learning_rate": 0.00016709972832176797, | |
| "loss": 2.8356, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2730462519936204, | |
| "grad_norm": 0.2956329882144928, | |
| "learning_rate": 0.0001668997110240684, | |
| "loss": 2.6157, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2738968633705476, | |
| "grad_norm": 0.30924192070961, | |
| "learning_rate": 0.00016669920807275623, | |
| "loss": 2.8421, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.27474747474747474, | |
| "grad_norm": 0.31185418367385864, | |
| "learning_rate": 0.00016649822092336812, | |
| "loss": 2.758, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2755980861244019, | |
| "grad_norm": 0.3129124045372009, | |
| "learning_rate": 0.0001662967510349558, | |
| "loss": 2.6861, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.27644869750132905, | |
| "grad_norm": 0.3300238251686096, | |
| "learning_rate": 0.00016609479987007527, | |
| "loss": 2.8284, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.27729930887825627, | |
| "grad_norm": 0.3459511399269104, | |
| "learning_rate": 0.00016589236889477646, | |
| "loss": 2.6454, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.2781499202551834, | |
| "grad_norm": 0.3566714823246002, | |
| "learning_rate": 0.00016568945957859236, | |
| "loss": 2.399, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2790005316321106, | |
| "grad_norm": 0.3650771379470825, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 2.7529, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.27985114300903774, | |
| "grad_norm": 0.35940608382225037, | |
| "learning_rate": 0.00016528221181905217, | |
| "loss": 2.5894, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 0.4018422067165375, | |
| "learning_rate": 0.0001650778763320817, | |
| "loss": 2.7001, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.28155236576289205, | |
| "grad_norm": 0.3867095708847046, | |
| "learning_rate": 0.00016487306841697578, | |
| "loss": 2.9919, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.28240297713981927, | |
| "grad_norm": 0.5714160799980164, | |
| "learning_rate": 0.0001646677895605227, | |
| "loss": 2.9366, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.2832535885167464, | |
| "grad_norm": 0.3697023391723633, | |
| "learning_rate": 0.00016446204125292942, | |
| "loss": 2.653, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.2841041998936736, | |
| "grad_norm": 0.4222877323627472, | |
| "learning_rate": 0.00016425582498781087, | |
| "loss": 2.8557, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.28495481127060074, | |
| "grad_norm": 0.40899163484573364, | |
| "learning_rate": 0.0001640491422621792, | |
| "loss": 2.7875, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2858054226475279, | |
| "grad_norm": 0.44694027304649353, | |
| "learning_rate": 0.00016384199457643262, | |
| "loss": 2.8616, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.28665603402445505, | |
| "grad_norm": 0.45921215415000916, | |
| "learning_rate": 0.00016363438343434483, | |
| "loss": 2.5821, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.28750664540138227, | |
| "grad_norm": 0.4156991243362427, | |
| "learning_rate": 0.00016342631034305384, | |
| "loss": 2.7228, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.2883572567783094, | |
| "grad_norm": 0.43912479281425476, | |
| "learning_rate": 0.00016321777681305125, | |
| "loss": 2.7119, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2892078681552366, | |
| "grad_norm": 0.4638505280017853, | |
| "learning_rate": 0.00016300878435817113, | |
| "loss": 2.6832, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.29005847953216374, | |
| "grad_norm": 0.4912424087524414, | |
| "learning_rate": 0.00016279933449557906, | |
| "loss": 2.6583, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 0.5067815780639648, | |
| "learning_rate": 0.00016258942874576118, | |
| "loss": 2.7723, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.29175970228601805, | |
| "grad_norm": 0.49635544419288635, | |
| "learning_rate": 0.0001623790686325131, | |
| "loss": 2.7351, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.29261031366294527, | |
| "grad_norm": 0.49465370178222656, | |
| "learning_rate": 0.00016216825568292885, | |
| "loss": 2.7251, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.2934609250398724, | |
| "grad_norm": 0.558665931224823, | |
| "learning_rate": 0.00016195699142738975, | |
| "loss": 2.5475, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.2943115364167996, | |
| "grad_norm": 0.6477727890014648, | |
| "learning_rate": 0.00016174527739955342, | |
| "loss": 2.7069, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.29516214779372674, | |
| "grad_norm": 0.6656542420387268, | |
| "learning_rate": 0.00016153311513634257, | |
| "loss": 3.0185, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.2960127591706539, | |
| "grad_norm": 0.7209298610687256, | |
| "learning_rate": 0.0001613205061779337, | |
| "loss": 2.8225, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.29686337054758105, | |
| "grad_norm": 0.6285322308540344, | |
| "learning_rate": 0.0001611074520677462, | |
| "loss": 2.9088, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.29771398192450826, | |
| "grad_norm": 0.7612189054489136, | |
| "learning_rate": 0.00016089395435243105, | |
| "loss": 2.9744, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2985645933014354, | |
| "grad_norm": 0.3265992999076843, | |
| "learning_rate": 0.00016068001458185936, | |
| "loss": 2.3931, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.2994152046783626, | |
| "grad_norm": 0.3179806172847748, | |
| "learning_rate": 0.00016046563430911146, | |
| "loss": 2.6212, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.30026581605528974, | |
| "grad_norm": 0.31296634674072266, | |
| "learning_rate": 0.00016025081509046544, | |
| "loss": 2.5008, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.3011164274322169, | |
| "grad_norm": 0.283408522605896, | |
| "learning_rate": 0.00016003555848538586, | |
| "loss": 2.3946, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.30196703880914405, | |
| "grad_norm": 0.2742927074432373, | |
| "learning_rate": 0.00015981986605651248, | |
| "loss": 2.5154, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.30281765018607126, | |
| "grad_norm": 0.29107552766799927, | |
| "learning_rate": 0.00015960373936964892, | |
| "loss": 2.505, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.3036682615629984, | |
| "grad_norm": 0.3244473338127136, | |
| "learning_rate": 0.0001593871799937512, | |
| "loss": 2.5244, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.3045188729399256, | |
| "grad_norm": 0.27494120597839355, | |
| "learning_rate": 0.0001591701895009164, | |
| "loss": 2.4161, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.30536948431685274, | |
| "grad_norm": 0.28784480690956116, | |
| "learning_rate": 0.00015895276946637136, | |
| "loss": 2.6434, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.3062200956937799, | |
| "grad_norm": 0.28096505999565125, | |
| "learning_rate": 0.00015873492146846108, | |
| "loss": 2.6784, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.30707070707070705, | |
| "grad_norm": 0.2602795362472534, | |
| "learning_rate": 0.00015851664708863735, | |
| "loss": 2.2712, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.30792131844763426, | |
| "grad_norm": 0.28796225786209106, | |
| "learning_rate": 0.0001582979479114472, | |
| "loss": 2.8673, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.3087719298245614, | |
| "grad_norm": 0.273456871509552, | |
| "learning_rate": 0.00015807882552452154, | |
| "loss": 2.6931, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3096225412014886, | |
| "grad_norm": 0.29190319776535034, | |
| "learning_rate": 0.00015785928151856347, | |
| "loss": 2.5932, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.31047315257841573, | |
| "grad_norm": 0.290829598903656, | |
| "learning_rate": 0.0001576393174873368, | |
| "loss": 2.6889, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3113237639553429, | |
| "grad_norm": 0.27631402015686035, | |
| "learning_rate": 0.0001574189350276545, | |
| "loss": 2.6508, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.31217437533227005, | |
| "grad_norm": 0.30464449524879456, | |
| "learning_rate": 0.00015719813573936712, | |
| "loss": 2.4918, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.31302498670919726, | |
| "grad_norm": 0.2622525095939636, | |
| "learning_rate": 0.00015697692122535107, | |
| "loss": 2.4657, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.3138755980861244, | |
| "grad_norm": 0.2790607810020447, | |
| "learning_rate": 0.0001567552930914972, | |
| "loss": 2.78, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.3147262094630516, | |
| "grad_norm": 0.27601122856140137, | |
| "learning_rate": 0.00015653325294669884, | |
| "loss": 2.6908, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.31557682083997873, | |
| "grad_norm": 0.3054703176021576, | |
| "learning_rate": 0.0001563108024028404, | |
| "loss": 2.6602, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3164274322169059, | |
| "grad_norm": 0.2979956567287445, | |
| "learning_rate": 0.00015608794307478546, | |
| "loss": 2.8034, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.31727804359383305, | |
| "grad_norm": 0.3022500276565552, | |
| "learning_rate": 0.00015586467658036524, | |
| "loss": 2.6491, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.31812865497076026, | |
| "grad_norm": 0.3278852701187134, | |
| "learning_rate": 0.0001556410045403667, | |
| "loss": 2.5928, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.3189792663476874, | |
| "grad_norm": 0.3023948073387146, | |
| "learning_rate": 0.0001554169285785208, | |
| "loss": 2.6165, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3198298777246146, | |
| "grad_norm": 0.33025047183036804, | |
| "learning_rate": 0.00015519245032149083, | |
| "loss": 2.7669, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.32068048910154173, | |
| "grad_norm": 0.3081379532814026, | |
| "learning_rate": 0.0001549675713988604, | |
| "loss": 2.5991, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.3215311004784689, | |
| "grad_norm": 0.35036811232566833, | |
| "learning_rate": 0.0001547422934431218, | |
| "loss": 2.6891, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.32238171185539605, | |
| "grad_norm": 0.399915486574173, | |
| "learning_rate": 0.00015451661808966405, | |
| "loss": 2.8271, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 0.3355950713157654, | |
| "learning_rate": 0.00015429054697676107, | |
| "loss": 2.3574, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3240829346092504, | |
| "grad_norm": 0.35686787962913513, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 2.6926, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3249335459861776, | |
| "grad_norm": 0.3730961084365845, | |
| "learning_rate": 0.00015383722404006806, | |
| "loss": 2.7418, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.32578415736310473, | |
| "grad_norm": 0.36210712790489197, | |
| "learning_rate": 0.00015360997550714305, | |
| "loss": 2.7188, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3266347687400319, | |
| "grad_norm": 0.4164154827594757, | |
| "learning_rate": 0.0001533823377964791, | |
| "loss": 2.8182, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.32748538011695905, | |
| "grad_norm": 0.3890111446380615, | |
| "learning_rate": 0.0001531543125605956, | |
| "loss": 2.6873, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3283359914938862, | |
| "grad_norm": 0.3929746747016907, | |
| "learning_rate": 0.0001529259014548253, | |
| "loss": 2.6169, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.3291866028708134, | |
| "grad_norm": 0.4354581832885742, | |
| "learning_rate": 0.0001526971061373021, | |
| "loss": 2.681, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3300372142477406, | |
| "grad_norm": 0.4014440178871155, | |
| "learning_rate": 0.00015246792826894906, | |
| "loss": 2.6601, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.33088782562466773, | |
| "grad_norm": 0.41521206498146057, | |
| "learning_rate": 0.00015223836951346634, | |
| "loss": 2.7763, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3317384370015949, | |
| "grad_norm": 0.43811067938804626, | |
| "learning_rate": 0.00015200843153731906, | |
| "loss": 2.7373, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.33258904837852205, | |
| "grad_norm": 0.4687197208404541, | |
| "learning_rate": 0.0001517781160097254, | |
| "loss": 2.6432, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3334396597554492, | |
| "grad_norm": 0.4998420178890228, | |
| "learning_rate": 0.00015154742460264425, | |
| "loss": 2.6434, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.3342902711323764, | |
| "grad_norm": 0.4983424246311188, | |
| "learning_rate": 0.0001513163589907632, | |
| "loss": 2.7619, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3351408825093036, | |
| "grad_norm": 0.5066902041435242, | |
| "learning_rate": 0.00015108492085148632, | |
| "loss": 2.6515, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.33599149388623073, | |
| "grad_norm": 0.5106616020202637, | |
| "learning_rate": 0.00015085311186492206, | |
| "loss": 2.6288, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3368421052631579, | |
| "grad_norm": 0.6036600470542908, | |
| "learning_rate": 0.00015062093371387097, | |
| "loss": 2.6391, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.33769271664008504, | |
| "grad_norm": 0.6260906457901001, | |
| "learning_rate": 0.00015038838808381354, | |
| "loss": 2.8623, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3385433280170122, | |
| "grad_norm": 0.6070663332939148, | |
| "learning_rate": 0.00015015547666289797, | |
| "loss": 2.615, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3393939393939394, | |
| "grad_norm": 0.66013503074646, | |
| "learning_rate": 0.00014992220114192785, | |
| "loss": 2.6865, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.34024455077086657, | |
| "grad_norm": 0.7371327877044678, | |
| "learning_rate": 0.00014968856321434998, | |
| "loss": 2.6959, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.34109516214779373, | |
| "grad_norm": 0.3081108629703522, | |
| "learning_rate": 0.00014945456457624197, | |
| "loss": 2.321, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.3419457735247209, | |
| "grad_norm": 0.37835538387298584, | |
| "learning_rate": 0.0001492202069263, | |
| "loss": 2.5144, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.34279638490164804, | |
| "grad_norm": 0.3184291422367096, | |
| "learning_rate": 0.00014898549196582645, | |
| "loss": 2.5783, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.3436469962785752, | |
| "grad_norm": 0.3378361165523529, | |
| "learning_rate": 0.00014875042139871766, | |
| "loss": 2.667, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.3444976076555024, | |
| "grad_norm": 0.285756379365921, | |
| "learning_rate": 0.00014851499693145135, | |
| "loss": 2.5316, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.34534821903242957, | |
| "grad_norm": 0.28837427496910095, | |
| "learning_rate": 0.00014827922027307451, | |
| "loss": 2.5012, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.34619883040935673, | |
| "grad_norm": 0.30004727840423584, | |
| "learning_rate": 0.0001480430931351906, | |
| "loss": 2.5237, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.3470494417862839, | |
| "grad_norm": 0.28527382016181946, | |
| "learning_rate": 0.00014780661723194757, | |
| "loss": 2.5067, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.34790005316321104, | |
| "grad_norm": 0.26983842253685, | |
| "learning_rate": 0.00014756979428002514, | |
| "loss": 2.5577, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.3487506645401382, | |
| "grad_norm": 0.2686617076396942, | |
| "learning_rate": 0.00014733262599862234, | |
| "loss": 2.5267, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3496012759170654, | |
| "grad_norm": 0.26147857308387756, | |
| "learning_rate": 0.00014709511410944523, | |
| "loss": 2.4459, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.35045188729399257, | |
| "grad_norm": 0.29142996668815613, | |
| "learning_rate": 0.00014685726033669412, | |
| "loss": 2.7246, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.3513024986709197, | |
| "grad_norm": 0.29214030504226685, | |
| "learning_rate": 0.00014661906640705129, | |
| "loss": 2.6422, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.3521531100478469, | |
| "grad_norm": 0.2727803885936737, | |
| "learning_rate": 0.00014638053404966836, | |
| "loss": 2.6416, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.35300372142477404, | |
| "grad_norm": 0.2637098431587219, | |
| "learning_rate": 0.0001461416649961537, | |
| "loss": 2.5555, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3538543328017012, | |
| "grad_norm": 0.2736997902393341, | |
| "learning_rate": 0.00014590246098055996, | |
| "loss": 2.7423, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.3547049441786284, | |
| "grad_norm": 0.2691054940223694, | |
| "learning_rate": 0.0001456629237393713, | |
| "loss": 2.7818, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.28304579854011536, | |
| "learning_rate": 0.0001454230550114911, | |
| "loss": 2.554, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.3564061669324827, | |
| "grad_norm": 0.2698013484477997, | |
| "learning_rate": 0.00014518285653822898, | |
| "loss": 2.5155, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.3572567783094099, | |
| "grad_norm": 0.25874024629592896, | |
| "learning_rate": 0.00014494233006328837, | |
| "loss": 2.4841, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.35810738968633704, | |
| "grad_norm": 0.2918216586112976, | |
| "learning_rate": 0.00014470147733275387, | |
| "loss": 2.7489, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.3589580010632642, | |
| "grad_norm": 0.28201210498809814, | |
| "learning_rate": 0.0001444603000950784, | |
| "loss": 2.5709, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.3598086124401914, | |
| "grad_norm": 0.3077748119831085, | |
| "learning_rate": 0.0001442188001010707, | |
| "loss": 2.6089, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.36065922381711857, | |
| "grad_norm": 0.2970117926597595, | |
| "learning_rate": 0.00014397697910388248, | |
| "loss": 2.6171, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.3615098351940457, | |
| "grad_norm": 0.302947074174881, | |
| "learning_rate": 0.00014373483885899582, | |
| "loss": 2.4033, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3623604465709729, | |
| "grad_norm": 0.31938815116882324, | |
| "learning_rate": 0.00014349238112421024, | |
| "loss": 2.7042, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.36321105794790004, | |
| "grad_norm": 0.3481922447681427, | |
| "learning_rate": 0.00014324960765963018, | |
| "loss": 2.6479, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.3640616693248272, | |
| "grad_norm": 0.3142179846763611, | |
| "learning_rate": 0.00014300652022765207, | |
| "loss": 2.5285, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.3649122807017544, | |
| "grad_norm": 0.3226439952850342, | |
| "learning_rate": 0.00014276312059295147, | |
| "loss": 2.5389, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.36576289207868157, | |
| "grad_norm": 0.401301771402359, | |
| "learning_rate": 0.00014251941052247045, | |
| "loss": 2.8749, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3666135034556087, | |
| "grad_norm": 0.3587849736213684, | |
| "learning_rate": 0.00014227539178540463, | |
| "loss": 2.7549, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.3674641148325359, | |
| "grad_norm": 0.38098907470703125, | |
| "learning_rate": 0.00014203106615319038, | |
| "loss": 2.6159, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.36831472620946304, | |
| "grad_norm": 0.402971088886261, | |
| "learning_rate": 0.00014178643539949196, | |
| "loss": 2.6626, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3691653375863902, | |
| "grad_norm": 0.3960564136505127, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 2.8099, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.3700159489633174, | |
| "grad_norm": 0.41555696725845337, | |
| "learning_rate": 0.00014129626563336178, | |
| "loss": 2.7282, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.37086656034024457, | |
| "grad_norm": 0.3991285264492035, | |
| "learning_rate": 0.000141050730179282, | |
| "loss": 2.5909, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.3717171717171717, | |
| "grad_norm": 0.442220002412796, | |
| "learning_rate": 0.00014080489672039606, | |
| "loss": 2.8671, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.3725677830940989, | |
| "grad_norm": 0.4263823628425598, | |
| "learning_rate": 0.0001405587670413143, | |
| "loss": 2.5901, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.37341839447102604, | |
| "grad_norm": 0.45683711767196655, | |
| "learning_rate": 0.00014031234292879725, | |
| "loss": 2.7801, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.3742690058479532, | |
| "grad_norm": 0.4672732353210449, | |
| "learning_rate": 0.00014006562617174294, | |
| "loss": 2.696, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3751196172248804, | |
| "grad_norm": 0.45676231384277344, | |
| "learning_rate": 0.0001398186185611738, | |
| "loss": 2.6197, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.37597022860180757, | |
| "grad_norm": 0.4717809855937958, | |
| "learning_rate": 0.00013957132189022374, | |
| "loss": 2.7676, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.3768208399787347, | |
| "grad_norm": 0.4931057393550873, | |
| "learning_rate": 0.00013932373795412503, | |
| "loss": 2.7968, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.3776714513556619, | |
| "grad_norm": 0.5112429857254028, | |
| "learning_rate": 0.0001390758685501954, | |
| "loss": 2.5947, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.37852206273258904, | |
| "grad_norm": 0.5512686371803284, | |
| "learning_rate": 0.00013882771547782475, | |
| "loss": 2.6041, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3793726741095162, | |
| "grad_norm": 0.5506839752197266, | |
| "learning_rate": 0.0001385792805384625, | |
| "loss": 2.7389, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.3802232854864434, | |
| "grad_norm": 0.5831094980239868, | |
| "learning_rate": 0.00013833056553560398, | |
| "loss": 2.8867, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.38107389686337056, | |
| "grad_norm": 0.6127706170082092, | |
| "learning_rate": 0.00013808157227477788, | |
| "loss": 2.689, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.3819245082402977, | |
| "grad_norm": 0.6911614537239075, | |
| "learning_rate": 0.00013783230256353266, | |
| "loss": 2.9826, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.3827751196172249, | |
| "grad_norm": 0.7095205187797546, | |
| "learning_rate": 0.00013758275821142382, | |
| "loss": 2.7725, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.38362573099415204, | |
| "grad_norm": 0.31148967146873474, | |
| "learning_rate": 0.00013733294103000055, | |
| "loss": 2.5203, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.3844763423710792, | |
| "grad_norm": 0.3002786338329315, | |
| "learning_rate": 0.00013708285283279252, | |
| "loss": 2.5911, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.3853269537480064, | |
| "grad_norm": 0.29820379614830017, | |
| "learning_rate": 0.00013683249543529696, | |
| "loss": 2.3441, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.38617756512493356, | |
| "grad_norm": 0.29162371158599854, | |
| "learning_rate": 0.00013658187065496532, | |
| "loss": 2.4791, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.3870281765018607, | |
| "grad_norm": 0.2905353307723999, | |
| "learning_rate": 0.00013633098031119002, | |
| "loss": 2.4208, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3878787878787879, | |
| "grad_norm": 0.28394779562950134, | |
| "learning_rate": 0.00013607982622529133, | |
| "loss": 2.6468, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.38872939925571504, | |
| "grad_norm": 0.2683579921722412, | |
| "learning_rate": 0.00013582841022050424, | |
| "loss": 2.5199, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.3895800106326422, | |
| "grad_norm": 0.2719745934009552, | |
| "learning_rate": 0.00013557673412196503, | |
| "loss": 2.4813, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.39043062200956935, | |
| "grad_norm": 0.26781052350997925, | |
| "learning_rate": 0.00013532479975669808, | |
| "loss": 2.5296, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.39128123338649656, | |
| "grad_norm": 0.26786506175994873, | |
| "learning_rate": 0.00013507260895360274, | |
| "loss": 2.5254, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3921318447634237, | |
| "grad_norm": 0.27148234844207764, | |
| "learning_rate": 0.0001348201635434399, | |
| "loss": 2.6415, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.3929824561403509, | |
| "grad_norm": 0.29129862785339355, | |
| "learning_rate": 0.00013456746535881871, | |
| "loss": 2.6098, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.39383306751727803, | |
| "grad_norm": 0.2636375427246094, | |
| "learning_rate": 0.00013431451623418343, | |
| "loss": 2.4875, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.3946836788942052, | |
| "grad_norm": 0.2701190114021301, | |
| "learning_rate": 0.00013406131800579985, | |
| "loss": 2.7151, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.39553429027113235, | |
| "grad_norm": 0.2739466428756714, | |
| "learning_rate": 0.00013380787251174225, | |
| "loss": 2.6465, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.39638490164805956, | |
| "grad_norm": 0.2672434151172638, | |
| "learning_rate": 0.00013355418159187985, | |
| "loss": 2.6823, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.3972355130249867, | |
| "grad_norm": 0.26863133907318115, | |
| "learning_rate": 0.00013330024708786353, | |
| "loss": 2.5357, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.3980861244019139, | |
| "grad_norm": 0.2704436779022217, | |
| "learning_rate": 0.00013304607084311244, | |
| "loss": 2.6887, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.39893673577884103, | |
| "grad_norm": 0.2772809863090515, | |
| "learning_rate": 0.00013279165470280065, | |
| "loss": 2.6992, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.3997873471557682, | |
| "grad_norm": 0.3024834394454956, | |
| "learning_rate": 0.0001325370005138437, | |
| "loss": 2.753, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.40063795853269535, | |
| "grad_norm": 0.2810865342617035, | |
| "learning_rate": 0.00013228211012488532, | |
| "loss": 2.7921, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.40148856990962256, | |
| "grad_norm": 0.28023043274879456, | |
| "learning_rate": 0.00013202698538628376, | |
| "loss": 2.2935, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4023391812865497, | |
| "grad_norm": 0.30379030108451843, | |
| "learning_rate": 0.0001317716281500987, | |
| "loss": 2.5575, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4031897926634769, | |
| "grad_norm": 0.29648759961128235, | |
| "learning_rate": 0.00013151604027007745, | |
| "loss": 2.4384, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 0.32276931405067444, | |
| "learning_rate": 0.00013126022360164172, | |
| "loss": 2.7529, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4048910154173312, | |
| "grad_norm": 0.3146274983882904, | |
| "learning_rate": 0.00013100418000187419, | |
| "loss": 2.473, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.40574162679425835, | |
| "grad_norm": 0.33162757754325867, | |
| "learning_rate": 0.00013074791132950485, | |
| "loss": 2.5832, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.40659223817118556, | |
| "grad_norm": 0.33521875739097595, | |
| "learning_rate": 0.00013049141944489748, | |
| "loss": 2.4807, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4074428495481127, | |
| "grad_norm": 0.3724415898323059, | |
| "learning_rate": 0.00013023470621003643, | |
| "loss": 2.8245, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4082934609250399, | |
| "grad_norm": 0.3634830117225647, | |
| "learning_rate": 0.00012997777348851288, | |
| "loss": 2.5921, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.40914407230196703, | |
| "grad_norm": 0.38656124472618103, | |
| "learning_rate": 0.0001297206231455113, | |
| "loss": 2.619, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.4099946836788942, | |
| "grad_norm": 0.3939076066017151, | |
| "learning_rate": 0.00012946325704779602, | |
| "loss": 2.7466, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.41084529505582135, | |
| "grad_norm": 0.37405261397361755, | |
| "learning_rate": 0.00012920567706369758, | |
| "loss": 2.764, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.41169590643274856, | |
| "grad_norm": 0.3894766569137573, | |
| "learning_rate": 0.0001289478850630993, | |
| "loss": 2.7373, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.4125465178096757, | |
| "grad_norm": 0.44012099504470825, | |
| "learning_rate": 0.00012868988291742347, | |
| "loss": 2.6475, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4133971291866029, | |
| "grad_norm": 0.41175583004951477, | |
| "learning_rate": 0.0001284316724996181, | |
| "loss": 2.7573, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.41424774056353003, | |
| "grad_norm": 0.4406805634498596, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 2.5429, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.4150983519404572, | |
| "grad_norm": 0.4783489406108856, | |
| "learning_rate": 0.0001279146343469563, | |
| "loss": 2.8275, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.41594896331738435, | |
| "grad_norm": 0.44115763902664185, | |
| "learning_rate": 0.00012765581036550095, | |
| "loss": 2.6858, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.41679957469431156, | |
| "grad_norm": 0.5237467885017395, | |
| "learning_rate": 0.0001273967856186909, | |
| "loss": 2.8638, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4176501860712387, | |
| "grad_norm": 0.5071147680282593, | |
| "learning_rate": 0.00012713756198689757, | |
| "loss": 2.8603, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.4185007974481659, | |
| "grad_norm": 0.5125464797019958, | |
| "learning_rate": 0.00012687814135193612, | |
| "loss": 3.0048, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.41935140882509303, | |
| "grad_norm": 0.5373572707176208, | |
| "learning_rate": 0.0001266185255970519, | |
| "loss": 2.5727, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.4202020202020202, | |
| "grad_norm": 0.4913314878940582, | |
| "learning_rate": 0.00012635871660690676, | |
| "loss": 2.6501, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.5953575968742371, | |
| "learning_rate": 0.00012609871626756522, | |
| "loss": 2.8674, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.42190324295587456, | |
| "grad_norm": 0.5852685570716858, | |
| "learning_rate": 0.00012583852646648095, | |
| "loss": 2.6855, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.4227538543328017, | |
| "grad_norm": 0.6449065804481506, | |
| "learning_rate": 0.00012557814909248296, | |
| "loss": 2.7688, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4236044657097289, | |
| "grad_norm": 0.6615833640098572, | |
| "learning_rate": 0.000125317586035762, | |
| "loss": 2.7446, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.42445507708665603, | |
| "grad_norm": 0.6561222076416016, | |
| "learning_rate": 0.0001250568391878567, | |
| "loss": 2.801, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.4253056884635832, | |
| "grad_norm": 0.8370924592018127, | |
| "learning_rate": 0.00012479591044163997, | |
| "loss": 2.9991, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.42615629984051034, | |
| "grad_norm": 0.3273813724517822, | |
| "learning_rate": 0.0001245348016913051, | |
| "loss": 2.5914, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.42700691121743756, | |
| "grad_norm": 0.3103164732456207, | |
| "learning_rate": 0.00012427351483235223, | |
| "loss": 2.6331, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.4278575225943647, | |
| "grad_norm": 0.3061564862728119, | |
| "learning_rate": 0.00012401205176157447, | |
| "loss": 2.5927, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.42870813397129187, | |
| "grad_norm": 0.2943616509437561, | |
| "learning_rate": 0.00012375041437704393, | |
| "loss": 2.4734, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.42955874534821903, | |
| "grad_norm": 0.29883497953414917, | |
| "learning_rate": 0.00012348860457809838, | |
| "loss": 2.4734, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4304093567251462, | |
| "grad_norm": 0.295578271150589, | |
| "learning_rate": 0.00012322662426532708, | |
| "loss": 2.3561, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.43125996810207334, | |
| "grad_norm": 0.3073442578315735, | |
| "learning_rate": 0.00012296447534055716, | |
| "loss": 2.5489, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.43211057947900056, | |
| "grad_norm": 0.3207886517047882, | |
| "learning_rate": 0.00012270215970683977, | |
| "loss": 2.6984, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.4329611908559277, | |
| "grad_norm": 0.26681843400001526, | |
| "learning_rate": 0.00012243967926843627, | |
| "loss": 2.4313, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.43381180223285487, | |
| "grad_norm": 0.29009976983070374, | |
| "learning_rate": 0.00012217703593080445, | |
| "loss": 2.6128, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.43466241360978203, | |
| "grad_norm": 0.27712225914001465, | |
| "learning_rate": 0.00012191423160058462, | |
| "loss": 2.4976, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.4355130249867092, | |
| "grad_norm": 0.30184635519981384, | |
| "learning_rate": 0.00012165126818558572, | |
| "loss": 2.6912, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 0.28122004866600037, | |
| "learning_rate": 0.00012138814759477176, | |
| "loss": 2.6935, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.43721424774056356, | |
| "grad_norm": 0.2680952250957489, | |
| "learning_rate": 0.00012112487173824753, | |
| "loss": 2.5607, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.4380648591174907, | |
| "grad_norm": 0.26283374428749084, | |
| "learning_rate": 0.00012086144252724513, | |
| "loss": 2.5001, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.43891547049441787, | |
| "grad_norm": 0.26960310339927673, | |
| "learning_rate": 0.00012059786187410984, | |
| "loss": 2.609, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.439766081871345, | |
| "grad_norm": 0.2755014896392822, | |
| "learning_rate": 0.00012033413169228635, | |
| "loss": 2.6356, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.4406166932482722, | |
| "grad_norm": 0.2706344723701477, | |
| "learning_rate": 0.00012007025389630484, | |
| "loss": 2.6909, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.44146730462519934, | |
| "grad_norm": 0.2639751434326172, | |
| "learning_rate": 0.00011980623040176704, | |
| "loss": 2.542, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.44231791600212655, | |
| "grad_norm": 0.2747778594493866, | |
| "learning_rate": 0.00011954206312533245, | |
| "loss": 2.4773, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4431685273790537, | |
| "grad_norm": 0.2990424931049347, | |
| "learning_rate": 0.0001192777539847043, | |
| "loss": 2.6881, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.44401913875598087, | |
| "grad_norm": 0.27671095728874207, | |
| "learning_rate": 0.00011901330489861564, | |
| "loss": 2.5495, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.444869750132908, | |
| "grad_norm": 0.3051941394805908, | |
| "learning_rate": 0.00011874871778681555, | |
| "loss": 2.7591, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.4457203615098352, | |
| "grad_norm": 0.2990604043006897, | |
| "learning_rate": 0.00011848399457005495, | |
| "loss": 2.5765, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.44657097288676234, | |
| "grad_norm": 0.3488616943359375, | |
| "learning_rate": 0.00011821913717007298, | |
| "loss": 2.634, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.44742158426368955, | |
| "grad_norm": 0.3207804262638092, | |
| "learning_rate": 0.00011795414750958265, | |
| "loss": 2.5632, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.4482721956406167, | |
| "grad_norm": 0.3244103491306305, | |
| "learning_rate": 0.0001176890275122573, | |
| "loss": 2.5784, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.44912280701754387, | |
| "grad_norm": 0.33527323603630066, | |
| "learning_rate": 0.00011742377910271639, | |
| "loss": 2.7016, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.449973418394471, | |
| "grad_norm": 0.3452233672142029, | |
| "learning_rate": 0.00011715840420651152, | |
| "loss": 2.5495, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.4508240297713982, | |
| "grad_norm": 0.36391714215278625, | |
| "learning_rate": 0.00011689290475011259, | |
| "loss": 2.662, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.45167464114832534, | |
| "grad_norm": 0.41914063692092896, | |
| "learning_rate": 0.00011662728266089364, | |
| "loss": 2.8876, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.45252525252525255, | |
| "grad_norm": 0.4124680459499359, | |
| "learning_rate": 0.00011636153986711906, | |
| "loss": 2.7271, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.4533758639021797, | |
| "grad_norm": 0.3957962393760681, | |
| "learning_rate": 0.00011609567829792944, | |
| "loss": 2.5991, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.45422647527910687, | |
| "grad_norm": 0.401443213224411, | |
| "learning_rate": 0.00011582969988332757, | |
| "loss": 2.7777, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.455077086656034, | |
| "grad_norm": 0.4349088966846466, | |
| "learning_rate": 0.00011556360655416457, | |
| "loss": 2.524, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4559276980329612, | |
| "grad_norm": 0.44450250267982483, | |
| "learning_rate": 0.00011529740024212565, | |
| "loss": 2.8296, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.45677830940988834, | |
| "grad_norm": 0.44510000944137573, | |
| "learning_rate": 0.00011503108287971626, | |
| "loss": 2.8425, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.4576289207868155, | |
| "grad_norm": 0.4766497313976288, | |
| "learning_rate": 0.00011476465640024814, | |
| "loss": 2.7471, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.4584795321637427, | |
| "grad_norm": 0.458278089761734, | |
| "learning_rate": 0.00011449812273782492, | |
| "loss": 2.7274, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.45933014354066987, | |
| "grad_norm": 0.46523377299308777, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 2.4967, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.460180754917597, | |
| "grad_norm": 0.4741576611995697, | |
| "learning_rate": 0.00011396474160440478, | |
| "loss": 2.6591, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.4610313662945242, | |
| "grad_norm": 0.5007473230361938, | |
| "learning_rate": 0.00011369789800544959, | |
| "loss": 2.8139, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.46188197767145134, | |
| "grad_norm": 0.45440518856048584, | |
| "learning_rate": 0.00011343095496759476, | |
| "loss": 2.6233, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.4627325890483785, | |
| "grad_norm": 0.5588196516036987, | |
| "learning_rate": 0.00011316391442869394, | |
| "loss": 2.6369, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.4635832004253057, | |
| "grad_norm": 0.5833231806755066, | |
| "learning_rate": 0.00011289677832730862, | |
| "loss": 2.8079, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.46443381180223287, | |
| "grad_norm": 0.6029292345046997, | |
| "learning_rate": 0.00011262954860269399, | |
| "loss": 2.8007, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.46528442317916, | |
| "grad_norm": 0.5825842022895813, | |
| "learning_rate": 0.00011236222719478491, | |
| "loss": 2.7767, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.4661350345560872, | |
| "grad_norm": 0.6685728430747986, | |
| "learning_rate": 0.00011209481604418181, | |
| "loss": 2.6621, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.46698564593301434, | |
| "grad_norm": 0.6266542077064514, | |
| "learning_rate": 0.00011182731709213659, | |
| "loss": 2.5234, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.4678362573099415, | |
| "grad_norm": 0.7756956815719604, | |
| "learning_rate": 0.00011155973228053853, | |
| "loss": 2.9975, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4686868686868687, | |
| "grad_norm": 0.27434810996055603, | |
| "learning_rate": 0.00011129206355190025, | |
| "loss": 2.3867, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.46953748006379586, | |
| "grad_norm": 0.28944891691207886, | |
| "learning_rate": 0.00011102431284934345, | |
| "loss": 2.6283, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.470388091440723, | |
| "grad_norm": 0.2992161810398102, | |
| "learning_rate": 0.00011075648211658505, | |
| "loss": 2.5818, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.4712387028176502, | |
| "grad_norm": 0.2830989956855774, | |
| "learning_rate": 0.00011048857329792284, | |
| "loss": 2.5738, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.47208931419457734, | |
| "grad_norm": 0.2766299545764923, | |
| "learning_rate": 0.00011022058833822158, | |
| "loss": 2.5138, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.4729399255715045, | |
| "grad_norm": 0.27325817942619324, | |
| "learning_rate": 0.0001099525291828986, | |
| "loss": 2.5166, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.4737905369484317, | |
| "grad_norm": 0.27071428298950195, | |
| "learning_rate": 0.00010968439777790999, | |
| "loss": 2.4935, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.47464114832535886, | |
| "grad_norm": 0.28781965374946594, | |
| "learning_rate": 0.00010941619606973632, | |
| "loss": 2.5851, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.475491759702286, | |
| "grad_norm": 0.2613832354545593, | |
| "learning_rate": 0.00010914792600536843, | |
| "loss": 2.5289, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.4763423710792132, | |
| "grad_norm": 0.2806464433670044, | |
| "learning_rate": 0.00010887958953229349, | |
| "loss": 2.659, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.47719298245614034, | |
| "grad_norm": 0.2767505347728729, | |
| "learning_rate": 0.00010861118859848067, | |
| "loss": 2.6562, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.4780435938330675, | |
| "grad_norm": 0.26740705966949463, | |
| "learning_rate": 0.0001083427251523672, | |
| "loss": 2.5689, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.4788942052099947, | |
| "grad_norm": 0.2635597884654999, | |
| "learning_rate": 0.000108074201142844, | |
| "loss": 2.3973, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.47974481658692186, | |
| "grad_norm": 0.2559509575366974, | |
| "learning_rate": 0.00010780561851924167, | |
| "loss": 2.4662, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.480595427963849, | |
| "grad_norm": 0.2819572687149048, | |
| "learning_rate": 0.0001075369792313164, | |
| "loss": 2.7678, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.4814460393407762, | |
| "grad_norm": 0.2618950605392456, | |
| "learning_rate": 0.00010726828522923562, | |
| "loss": 2.6463, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.48229665071770333, | |
| "grad_norm": 0.2766074240207672, | |
| "learning_rate": 0.000106999538463564, | |
| "loss": 2.7133, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.4831472620946305, | |
| "grad_norm": 0.280367910861969, | |
| "learning_rate": 0.00010673074088524926, | |
| "loss": 2.61, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.4839978734715577, | |
| "grad_norm": 0.3028632402420044, | |
| "learning_rate": 0.00010646189444560799, | |
| "loss": 2.5465, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 0.2950790822505951, | |
| "learning_rate": 0.00010619300109631145, | |
| "loss": 2.6517, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.485699096225412, | |
| "grad_norm": 0.2803073227405548, | |
| "learning_rate": 0.00010592406278937144, | |
| "loss": 2.6062, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.4865497076023392, | |
| "grad_norm": 0.29916471242904663, | |
| "learning_rate": 0.00010565508147712617, | |
| "loss": 2.5532, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.48740031897926633, | |
| "grad_norm": 0.32185062766075134, | |
| "learning_rate": 0.00010538605911222603, | |
| "loss": 2.722, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.4882509303561935, | |
| "grad_norm": 0.3155268728733063, | |
| "learning_rate": 0.00010511699764761936, | |
| "loss": 2.6655, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.4891015417331207, | |
| "grad_norm": 0.3241024315357208, | |
| "learning_rate": 0.00010484789903653846, | |
| "loss": 2.7093, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.48995215311004786, | |
| "grad_norm": 0.3311263620853424, | |
| "learning_rate": 0.00010457876523248518, | |
| "loss": 2.6485, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.490802764486975, | |
| "grad_norm": 0.34630128741264343, | |
| "learning_rate": 0.00010430959818921694, | |
| "loss": 2.8315, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.4916533758639022, | |
| "grad_norm": 0.34841713309288025, | |
| "learning_rate": 0.00010404039986073244, | |
| "loss": 2.7484, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.49250398724082933, | |
| "grad_norm": 0.35180601477622986, | |
| "learning_rate": 0.00010377117220125741, | |
| "loss": 2.6745, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.4933545986177565, | |
| "grad_norm": 0.38369500637054443, | |
| "learning_rate": 0.00010350191716523059, | |
| "loss": 2.6623, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4942052099946837, | |
| "grad_norm": 0.3735206425189972, | |
| "learning_rate": 0.00010323263670728946, | |
| "loss": 2.5805, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.49505582137161086, | |
| "grad_norm": 0.3994956314563751, | |
| "learning_rate": 0.00010296333278225599, | |
| "loss": 2.5622, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.495906432748538, | |
| "grad_norm": 0.39151209592819214, | |
| "learning_rate": 0.00010269400734512256, | |
| "loss": 2.805, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.4967570441254652, | |
| "grad_norm": 0.40469613671302795, | |
| "learning_rate": 0.0001024246623510377, | |
| "loss": 2.7588, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.49760765550239233, | |
| "grad_norm": 0.4307393431663513, | |
| "learning_rate": 0.0001021552997552919, | |
| "loss": 2.8039, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.4984582668793195, | |
| "grad_norm": 0.41100749373435974, | |
| "learning_rate": 0.00010188592151330343, | |
| "loss": 2.6842, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.4993088782562467, | |
| "grad_norm": 0.45486176013946533, | |
| "learning_rate": 0.00010161652958060417, | |
| "loss": 2.8073, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.5001594896331738, | |
| "grad_norm": 0.43648669123649597, | |
| "learning_rate": 0.00010134712591282538, | |
| "loss": 2.638, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.5001594896331738, | |
| "eval_loss": 2.66949725151062, | |
| "eval_runtime": 80.5199, | |
| "eval_samples_per_second": 12.295, | |
| "eval_steps_per_second": 6.148, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.501010101010101, | |
| "grad_norm": 0.44425809383392334, | |
| "learning_rate": 0.00010107771246568345, | |
| "loss": 2.5832, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.5018607123870282, | |
| "grad_norm": 0.4519883990287781, | |
| "learning_rate": 0.00010080829119496586, | |
| "loss": 2.8142, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5027113237639553, | |
| "grad_norm": 0.4547143280506134, | |
| "learning_rate": 0.0001005388640565168, | |
| "loss": 2.4661, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.5035619351408825, | |
| "grad_norm": 0.5073342323303223, | |
| "learning_rate": 0.00010026943300622313, | |
| "loss": 2.7849, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.5044125465178096, | |
| "grad_norm": 0.5203079581260681, | |
| "learning_rate": 0.0001, | |
| "loss": 2.6077, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.5052631578947369, | |
| "grad_norm": 0.4956808090209961, | |
| "learning_rate": 9.97305669937769e-05, | |
| "loss": 2.7116, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.506113769271664, | |
| "grad_norm": 0.5807839035987854, | |
| "learning_rate": 9.946113594348321e-05, | |
| "loss": 2.8024, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5069643806485912, | |
| "grad_norm": 0.5834396481513977, | |
| "learning_rate": 9.919170880503415e-05, | |
| "loss": 2.6428, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.5078149920255184, | |
| "grad_norm": 0.5668088793754578, | |
| "learning_rate": 9.892228753431657e-05, | |
| "loss": 2.6711, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.5086656034024455, | |
| "grad_norm": 0.6034618616104126, | |
| "learning_rate": 9.865287408717465e-05, | |
| "loss": 2.9129, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.5095162147793727, | |
| "grad_norm": 0.5953810811042786, | |
| "learning_rate": 9.838347041939584e-05, | |
| "loss": 2.7463, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.5103668261562998, | |
| "grad_norm": 0.6790388822555542, | |
| "learning_rate": 9.811407848669657e-05, | |
| "loss": 2.8535, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.511217437533227, | |
| "grad_norm": 0.2997681200504303, | |
| "learning_rate": 9.784470024470812e-05, | |
| "loss": 2.4338, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.5120680489101542, | |
| "grad_norm": 0.28502747416496277, | |
| "learning_rate": 9.757533764896235e-05, | |
| "loss": 2.2975, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.5129186602870813, | |
| "grad_norm": 0.29402679204940796, | |
| "learning_rate": 9.730599265487745e-05, | |
| "loss": 2.6287, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.5137692716640085, | |
| "grad_norm": 0.28246256709098816, | |
| "learning_rate": 9.703666721774402e-05, | |
| "loss": 2.4197, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.5146198830409356, | |
| "grad_norm": 0.29511624574661255, | |
| "learning_rate": 9.676736329271059e-05, | |
| "loss": 2.6028, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.5154704944178629, | |
| "grad_norm": 0.2746260166168213, | |
| "learning_rate": 9.649808283476941e-05, | |
| "loss": 2.5791, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.51632110579479, | |
| "grad_norm": 0.2634756863117218, | |
| "learning_rate": 9.622882779874263e-05, | |
| "loss": 2.4199, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.5171717171717172, | |
| "grad_norm": 0.2797803282737732, | |
| "learning_rate": 9.595960013926761e-05, | |
| "loss": 2.5637, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.5180223285486444, | |
| "grad_norm": 0.2671017646789551, | |
| "learning_rate": 9.569040181078306e-05, | |
| "loss": 2.6811, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.5188729399255715, | |
| "grad_norm": 0.26546764373779297, | |
| "learning_rate": 9.542123476751483e-05, | |
| "loss": 2.5613, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5197235513024987, | |
| "grad_norm": 0.2660638988018036, | |
| "learning_rate": 9.515210096346155e-05, | |
| "loss": 2.4644, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.5205741626794258, | |
| "grad_norm": 0.2697810232639313, | |
| "learning_rate": 9.488300235238067e-05, | |
| "loss": 2.5643, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.521424774056353, | |
| "grad_norm": 0.26712700724601746, | |
| "learning_rate": 9.461394088777402e-05, | |
| "loss": 2.6993, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.5222753854332801, | |
| "grad_norm": 0.2802102565765381, | |
| "learning_rate": 9.434491852287385e-05, | |
| "loss": 2.5723, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.5231259968102073, | |
| "grad_norm": 0.2690255641937256, | |
| "learning_rate": 9.407593721062859e-05, | |
| "loss": 2.5136, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5239766081871345, | |
| "grad_norm": 0.2789754271507263, | |
| "learning_rate": 9.38069989036886e-05, | |
| "loss": 2.5901, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.5248272195640616, | |
| "grad_norm": 0.2947288751602173, | |
| "learning_rate": 9.353810555439203e-05, | |
| "loss": 2.5661, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.5256778309409889, | |
| "grad_norm": 0.268081396818161, | |
| "learning_rate": 9.326925911475075e-05, | |
| "loss": 2.603, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.526528442317916, | |
| "grad_norm": 0.2749037444591522, | |
| "learning_rate": 9.300046153643602e-05, | |
| "loss": 2.7176, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.5273790536948432, | |
| "grad_norm": 0.267333984375, | |
| "learning_rate": 9.27317147707644e-05, | |
| "loss": 2.5775, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5282296650717704, | |
| "grad_norm": 0.2688322961330414, | |
| "learning_rate": 9.246302076868363e-05, | |
| "loss": 2.4796, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.5290802764486975, | |
| "grad_norm": 0.2832041382789612, | |
| "learning_rate": 9.219438148075832e-05, | |
| "loss": 2.5764, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.5299308878256247, | |
| "grad_norm": 0.3032709062099457, | |
| "learning_rate": 9.192579885715602e-05, | |
| "loss": 2.7559, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.5307814992025518, | |
| "grad_norm": 0.3065055012702942, | |
| "learning_rate": 9.165727484763282e-05, | |
| "loss": 2.7058, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.531632110579479, | |
| "grad_norm": 0.314698189496994, | |
| "learning_rate": 9.138881140151931e-05, | |
| "loss": 2.6227, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5324827219564061, | |
| "grad_norm": 0.31485408544540405, | |
| "learning_rate": 9.112041046770653e-05, | |
| "loss": 2.5687, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.3281068801879883, | |
| "learning_rate": 9.085207399463162e-05, | |
| "loss": 2.5957, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.5341839447102605, | |
| "grad_norm": 0.3461722731590271, | |
| "learning_rate": 9.058380393026369e-05, | |
| "loss": 2.8056, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.5350345560871876, | |
| "grad_norm": 0.3617733418941498, | |
| "learning_rate": 9.031560222209002e-05, | |
| "loss": 2.8171, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.5358851674641149, | |
| "grad_norm": 0.34867244958877563, | |
| "learning_rate": 9.00474708171014e-05, | |
| "loss": 2.6481, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.536735778841042, | |
| "grad_norm": 0.3704431653022766, | |
| "learning_rate": 8.977941166177845e-05, | |
| "loss": 2.7771, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.5375863902179692, | |
| "grad_norm": 0.3661860525608063, | |
| "learning_rate": 8.951142670207717e-05, | |
| "loss": 2.5948, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.5384370015948964, | |
| "grad_norm": 0.3825220763683319, | |
| "learning_rate": 8.924351788341496e-05, | |
| "loss": 2.5843, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.5392876129718235, | |
| "grad_norm": 0.39650052785873413, | |
| "learning_rate": 8.897568715065657e-05, | |
| "loss": 2.7541, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.5401382243487507, | |
| "grad_norm": 0.4185117483139038, | |
| "learning_rate": 8.87079364480998e-05, | |
| "loss": 2.791, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5409888357256778, | |
| "grad_norm": 0.41095203161239624, | |
| "learning_rate": 8.844026771946147e-05, | |
| "loss": 2.5707, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.541839447102605, | |
| "grad_norm": 0.4548766016960144, | |
| "learning_rate": 8.817268290786343e-05, | |
| "loss": 2.8622, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.5426900584795321, | |
| "grad_norm": 0.4231926202774048, | |
| "learning_rate": 8.790518395581822e-05, | |
| "loss": 2.5893, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.5435406698564593, | |
| "grad_norm": 0.441785603761673, | |
| "learning_rate": 8.763777280521511e-05, | |
| "loss": 2.7675, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.5443912812333865, | |
| "grad_norm": 0.4583280682563782, | |
| "learning_rate": 8.737045139730605e-05, | |
| "loss": 2.7461, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5452418926103136, | |
| "grad_norm": 0.5172987580299377, | |
| "learning_rate": 8.71032216726914e-05, | |
| "loss": 2.9216, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.5460925039872409, | |
| "grad_norm": 0.49341312050819397, | |
| "learning_rate": 8.683608557130608e-05, | |
| "loss": 2.803, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.546943115364168, | |
| "grad_norm": 0.5035589933395386, | |
| "learning_rate": 8.656904503240527e-05, | |
| "loss": 2.869, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.5477937267410952, | |
| "grad_norm": 0.5143235921859741, | |
| "learning_rate": 8.630210199455041e-05, | |
| "loss": 2.7889, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.5486443381180224, | |
| "grad_norm": 0.5110295414924622, | |
| "learning_rate": 8.603525839559523e-05, | |
| "loss": 2.8229, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5494949494949495, | |
| "grad_norm": 0.555565595626831, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 2.6461, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.5503455608718767, | |
| "grad_norm": 0.56583172082901, | |
| "learning_rate": 8.550187726217507e-05, | |
| "loss": 2.7647, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.5511961722488038, | |
| "grad_norm": 0.6618173718452454, | |
| "learning_rate": 8.523534359975189e-05, | |
| "loss": 2.6992, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.552046783625731, | |
| "grad_norm": 0.6295120120048523, | |
| "learning_rate": 8.496891712028375e-05, | |
| "loss": 2.6303, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.5528973950026581, | |
| "grad_norm": 0.6757770776748657, | |
| "learning_rate": 8.470259975787438e-05, | |
| "loss": 2.8003, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5537480063795853, | |
| "grad_norm": 0.26357659697532654, | |
| "learning_rate": 8.443639344583547e-05, | |
| "loss": 2.4452, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.5545986177565125, | |
| "grad_norm": 0.31750619411468506, | |
| "learning_rate": 8.417030011667241e-05, | |
| "loss": 2.374, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.5554492291334396, | |
| "grad_norm": 0.27070245146751404, | |
| "learning_rate": 8.390432170207057e-05, | |
| "loss": 2.3384, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.5562998405103668, | |
| "grad_norm": 0.28004199266433716, | |
| "learning_rate": 8.363846013288095e-05, | |
| "loss": 2.4357, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.557150451887294, | |
| "grad_norm": 0.323215126991272, | |
| "learning_rate": 8.337271733910637e-05, | |
| "loss": 2.3801, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5580010632642212, | |
| "grad_norm": 0.292837530374527, | |
| "learning_rate": 8.310709524988743e-05, | |
| "loss": 2.3924, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.5588516746411484, | |
| "grad_norm": 0.2810145914554596, | |
| "learning_rate": 8.284159579348851e-05, | |
| "loss": 2.5257, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.5597022860180755, | |
| "grad_norm": 0.2848433554172516, | |
| "learning_rate": 8.257622089728362e-05, | |
| "loss": 2.5743, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.5605528973950027, | |
| "grad_norm": 0.28531527519226074, | |
| "learning_rate": 8.231097248774274e-05, | |
| "loss": 2.5492, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.2882923185825348, | |
| "learning_rate": 8.20458524904174e-05, | |
| "loss": 2.4982, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.562254120148857, | |
| "grad_norm": 0.2807391583919525, | |
| "learning_rate": 8.178086282992705e-05, | |
| "loss": 2.5222, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.5631047315257841, | |
| "grad_norm": 0.29856112599372864, | |
| "learning_rate": 8.151600542994506e-05, | |
| "loss": 2.723, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.5639553429027113, | |
| "grad_norm": 0.2890356481075287, | |
| "learning_rate": 8.125128221318446e-05, | |
| "loss": 2.6931, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.5648059542796385, | |
| "grad_norm": 0.28091031312942505, | |
| "learning_rate": 8.098669510138437e-05, | |
| "loss": 2.6273, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 0.2678775489330292, | |
| "learning_rate": 8.072224601529574e-05, | |
| "loss": 2.5571, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5665071770334928, | |
| "grad_norm": 0.2804068624973297, | |
| "learning_rate": 8.045793687466757e-05, | |
| "loss": 2.4443, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.56735778841042, | |
| "grad_norm": 0.279041051864624, | |
| "learning_rate": 8.0193769598233e-05, | |
| "loss": 2.6947, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.5682083997873472, | |
| "grad_norm": 0.27611520886421204, | |
| "learning_rate": 7.992974610369521e-05, | |
| "loss": 2.782, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.5690590111642744, | |
| "grad_norm": 0.28445136547088623, | |
| "learning_rate": 7.966586830771367e-05, | |
| "loss": 2.5741, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.5699096225412015, | |
| "grad_norm": 0.2607346773147583, | |
| "learning_rate": 7.940213812589018e-05, | |
| "loss": 2.4522, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5707602339181287, | |
| "grad_norm": 0.2881016433238983, | |
| "learning_rate": 7.913855747275489e-05, | |
| "loss": 2.6714, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.5716108452950558, | |
| "grad_norm": 0.3019566237926483, | |
| "learning_rate": 7.887512826175248e-05, | |
| "loss": 2.7117, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.572461456671983, | |
| "grad_norm": 0.3051791489124298, | |
| "learning_rate": 7.861185240522827e-05, | |
| "loss": 2.6867, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.5733120680489101, | |
| "grad_norm": 0.32132992148399353, | |
| "learning_rate": 7.834873181441427e-05, | |
| "loss": 2.5278, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.5741626794258373, | |
| "grad_norm": 0.32348886132240295, | |
| "learning_rate": 7.808576839941542e-05, | |
| "loss": 2.7507, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5750132908027645, | |
| "grad_norm": 0.32297685742378235, | |
| "learning_rate": 7.782296406919557e-05, | |
| "loss": 2.6857, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.5758639021796916, | |
| "grad_norm": 0.34016191959381104, | |
| "learning_rate": 7.756032073156373e-05, | |
| "loss": 2.57, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.5767145135566188, | |
| "grad_norm": 0.349483847618103, | |
| "learning_rate": 7.729784029316025e-05, | |
| "loss": 2.6255, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.577565124933546, | |
| "grad_norm": 0.3523537218570709, | |
| "learning_rate": 7.703552465944287e-05, | |
| "loss": 2.8317, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.5784157363104732, | |
| "grad_norm": 0.38119909167289734, | |
| "learning_rate": 7.677337573467294e-05, | |
| "loss": 2.7303, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5792663476874004, | |
| "grad_norm": 0.4053308367729187, | |
| "learning_rate": 7.651139542190164e-05, | |
| "loss": 2.6776, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.5801169590643275, | |
| "grad_norm": 0.38038370013237, | |
| "learning_rate": 7.624958562295606e-05, | |
| "loss": 2.6411, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.5809675704412547, | |
| "grad_norm": 0.414034903049469, | |
| "learning_rate": 7.598794823842557e-05, | |
| "loss": 2.8368, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 0.41362208127975464, | |
| "learning_rate": 7.572648516764777e-05, | |
| "loss": 2.7608, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.582668793195109, | |
| "grad_norm": 0.41102275252342224, | |
| "learning_rate": 7.54651983086949e-05, | |
| "loss": 2.7317, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.5835194045720361, | |
| "grad_norm": 0.45451274514198303, | |
| "learning_rate": 7.520408955836007e-05, | |
| "loss": 2.7164, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.5843700159489633, | |
| "grad_norm": 0.42562657594680786, | |
| "learning_rate": 7.494316081214334e-05, | |
| "loss": 2.6641, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.5852206273258905, | |
| "grad_norm": 0.46389469504356384, | |
| "learning_rate": 7.468241396423801e-05, | |
| "loss": 2.6883, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.5860712387028176, | |
| "grad_norm": 0.42721521854400635, | |
| "learning_rate": 7.442185090751705e-05, | |
| "loss": 2.5853, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.5869218500797448, | |
| "grad_norm": 0.5020928978919983, | |
| "learning_rate": 7.416147353351909e-05, | |
| "loss": 2.6468, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.587772461456672, | |
| "grad_norm": 0.4806350767612457, | |
| "learning_rate": 7.390128373243479e-05, | |
| "loss": 2.5275, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.5886230728335992, | |
| "grad_norm": 0.5264183282852173, | |
| "learning_rate": 7.364128339309326e-05, | |
| "loss": 2.6082, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.5894736842105263, | |
| "grad_norm": 0.5251814126968384, | |
| "learning_rate": 7.338147440294809e-05, | |
| "loss": 2.6319, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.5903242955874535, | |
| "grad_norm": 0.5503838658332825, | |
| "learning_rate": 7.312185864806391e-05, | |
| "loss": 2.6875, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.5911749069643807, | |
| "grad_norm": 0.5717040300369263, | |
| "learning_rate": 7.286243801310248e-05, | |
| "loss": 2.6859, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.5920255183413078, | |
| "grad_norm": 0.5679376125335693, | |
| "learning_rate": 7.260321438130913e-05, | |
| "loss": 2.6425, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.592876129718235, | |
| "grad_norm": 0.6636802554130554, | |
| "learning_rate": 7.234418963449907e-05, | |
| "loss": 2.7617, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.5937267410951621, | |
| "grad_norm": 0.6101743578910828, | |
| "learning_rate": 7.208536565304373e-05, | |
| "loss": 2.6909, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.5945773524720893, | |
| "grad_norm": 0.6630806922912598, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 2.7461, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.5954279638490165, | |
| "grad_norm": 0.738452136516571, | |
| "learning_rate": 7.156832750038192e-05, | |
| "loss": 2.8512, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5962785752259436, | |
| "grad_norm": 0.2764449715614319, | |
| "learning_rate": 7.131011708257654e-05, | |
| "loss": 2.4965, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.5971291866028708, | |
| "grad_norm": 0.27585679292678833, | |
| "learning_rate": 7.105211493690073e-05, | |
| "loss": 2.5628, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.597979797979798, | |
| "grad_norm": 0.2890099287033081, | |
| "learning_rate": 7.079432293630244e-05, | |
| "loss": 2.6482, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.5988304093567252, | |
| "grad_norm": 0.27924832701683044, | |
| "learning_rate": 7.0536742952204e-05, | |
| "loss": 2.4101, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.5996810207336523, | |
| "grad_norm": 0.30849024653434753, | |
| "learning_rate": 7.02793768544887e-05, | |
| "loss": 2.4998, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6005316321105795, | |
| "grad_norm": 0.286748468875885, | |
| "learning_rate": 7.002222651148714e-05, | |
| "loss": 2.3868, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.6013822434875067, | |
| "grad_norm": 0.2908494472503662, | |
| "learning_rate": 6.976529378996357e-05, | |
| "loss": 2.4456, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.6022328548644338, | |
| "grad_norm": 0.26911187171936035, | |
| "learning_rate": 6.950858055510254e-05, | |
| "loss": 2.633, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.603083466241361, | |
| "grad_norm": 0.33310920000076294, | |
| "learning_rate": 6.925208867049522e-05, | |
| "loss": 2.8533, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.6039340776182881, | |
| "grad_norm": 0.2783631682395935, | |
| "learning_rate": 6.89958199981258e-05, | |
| "loss": 2.4214, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6047846889952153, | |
| "grad_norm": 0.2714787721633911, | |
| "learning_rate": 6.873977639835829e-05, | |
| "loss": 2.6986, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.6056353003721425, | |
| "grad_norm": 0.2905879616737366, | |
| "learning_rate": 6.848395972992261e-05, | |
| "loss": 2.5218, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.6064859117490696, | |
| "grad_norm": 0.288647323846817, | |
| "learning_rate": 6.822837184990132e-05, | |
| "loss": 2.6838, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.6073365231259968, | |
| "grad_norm": 0.2918678820133209, | |
| "learning_rate": 6.797301461371625e-05, | |
| "loss": 2.6513, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.6081871345029239, | |
| "grad_norm": 0.26965370774269104, | |
| "learning_rate": 6.771788987511469e-05, | |
| "loss": 2.5379, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6090377458798512, | |
| "grad_norm": 0.27741289138793945, | |
| "learning_rate": 6.746299948615631e-05, | |
| "loss": 2.6699, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.6098883572567783, | |
| "grad_norm": 0.26440632343292236, | |
| "learning_rate": 6.720834529719939e-05, | |
| "loss": 2.6012, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.6107389686337055, | |
| "grad_norm": 0.27086710929870605, | |
| "learning_rate": 6.695392915688759e-05, | |
| "loss": 2.5854, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.6115895800106327, | |
| "grad_norm": 0.2577430009841919, | |
| "learning_rate": 6.66997529121365e-05, | |
| "loss": 2.4548, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.6124401913875598, | |
| "grad_norm": 0.2519771158695221, | |
| "learning_rate": 6.644581840812018e-05, | |
| "loss": 2.4582, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.613290802764487, | |
| "grad_norm": 0.31657370924949646, | |
| "learning_rate": 6.619212748825776e-05, | |
| "loss": 2.5846, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.6141414141414141, | |
| "grad_norm": 0.28847917914390564, | |
| "learning_rate": 6.593868199420017e-05, | |
| "loss": 2.7865, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.6149920255183413, | |
| "grad_norm": 0.29375341534614563, | |
| "learning_rate": 6.568548376581662e-05, | |
| "loss": 2.3419, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.6158426368952685, | |
| "grad_norm": 0.320289671421051, | |
| "learning_rate": 6.543253464118131e-05, | |
| "loss": 2.7476, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.6166932482721956, | |
| "grad_norm": 0.3045991063117981, | |
| "learning_rate": 6.517983645656014e-05, | |
| "loss": 2.5995, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.6175438596491228, | |
| "grad_norm": 0.3231359124183655, | |
| "learning_rate": 6.492739104639727e-05, | |
| "loss": 2.6869, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.6183944710260499, | |
| "grad_norm": 0.36181551218032837, | |
| "learning_rate": 6.467520024330193e-05, | |
| "loss": 2.8256, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.6192450824029772, | |
| "grad_norm": 0.34118232131004333, | |
| "learning_rate": 6.4423265878035e-05, | |
| "loss": 2.7321, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.6200956937799043, | |
| "grad_norm": 0.3405606746673584, | |
| "learning_rate": 6.417158977949575e-05, | |
| "loss": 2.6993, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.6209463051568315, | |
| "grad_norm": 0.355055570602417, | |
| "learning_rate": 6.392017377470866e-05, | |
| "loss": 2.7063, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6217969165337587, | |
| "grad_norm": 0.368937611579895, | |
| "learning_rate": 6.366901968881002e-05, | |
| "loss": 2.682, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.6226475279106858, | |
| "grad_norm": 0.3755683898925781, | |
| "learning_rate": 6.341812934503469e-05, | |
| "loss": 2.7295, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.623498139287613, | |
| "grad_norm": 0.37153956294059753, | |
| "learning_rate": 6.316750456470303e-05, | |
| "loss": 2.7157, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.6243487506645401, | |
| "grad_norm": 0.4051001965999603, | |
| "learning_rate": 6.291714716720749e-05, | |
| "loss": 2.8429, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.6251993620414673, | |
| "grad_norm": 0.3929100036621094, | |
| "learning_rate": 6.26670589699995e-05, | |
| "loss": 2.7148, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.6260499734183945, | |
| "grad_norm": 0.4600156843662262, | |
| "learning_rate": 6.24172417885762e-05, | |
| "loss": 2.765, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.6269005847953216, | |
| "grad_norm": 0.42697155475616455, | |
| "learning_rate": 6.216769743646733e-05, | |
| "loss": 2.6477, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.6277511961722488, | |
| "grad_norm": 0.4428333640098572, | |
| "learning_rate": 6.191842772522214e-05, | |
| "loss": 2.9439, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.6286018075491759, | |
| "grad_norm": 0.45895466208457947, | |
| "learning_rate": 6.166943446439604e-05, | |
| "loss": 2.6916, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.6294524189261032, | |
| "grad_norm": 0.4783354103565216, | |
| "learning_rate": 6.142071946153751e-05, | |
| "loss": 2.6747, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6303030303030303, | |
| "grad_norm": 0.4584222733974457, | |
| "learning_rate": 6.117228452217525e-05, | |
| "loss": 2.8631, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.6311536416799575, | |
| "grad_norm": 0.5074661374092102, | |
| "learning_rate": 6.092413144980464e-05, | |
| "loss": 2.7687, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.6320042530568847, | |
| "grad_norm": 0.500322699546814, | |
| "learning_rate": 6.0676262045874976e-05, | |
| "loss": 2.6794, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.6328548644338118, | |
| "grad_norm": 0.4991973042488098, | |
| "learning_rate": 6.04286781097763e-05, | |
| "loss": 2.739, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.633705475810739, | |
| "grad_norm": 0.5203631520271301, | |
| "learning_rate": 6.018138143882621e-05, | |
| "loss": 2.5866, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.6345560871876661, | |
| "grad_norm": 0.5206765532493591, | |
| "learning_rate": 5.9934373828257106e-05, | |
| "loss": 2.5869, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.6354066985645933, | |
| "grad_norm": 0.5892865061759949, | |
| "learning_rate": 5.96876570712028e-05, | |
| "loss": 2.6141, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.6362573099415205, | |
| "grad_norm": 0.6528813242912292, | |
| "learning_rate": 5.944123295868573e-05, | |
| "loss": 2.7467, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.6371079213184476, | |
| "grad_norm": 0.6801220178604126, | |
| "learning_rate": 5.9195103279603956e-05, | |
| "loss": 2.8362, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.6379585326953748, | |
| "grad_norm": 0.7539176940917969, | |
| "learning_rate": 5.894926982071804e-05, | |
| "loss": 2.8246, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6388091440723019, | |
| "grad_norm": 0.2966770827770233, | |
| "learning_rate": 5.870373436663823e-05, | |
| "loss": 2.6906, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.6396597554492292, | |
| "grad_norm": 0.27538472414016724, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 2.4334, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.6405103668261563, | |
| "grad_norm": 0.2812007665634155, | |
| "learning_rate": 5.821356460050805e-05, | |
| "loss": 2.4665, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.6413609782030835, | |
| "grad_norm": 0.28760018944740295, | |
| "learning_rate": 5.796893384680964e-05, | |
| "loss": 2.6608, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.6422115895800107, | |
| "grad_norm": 0.2786334455013275, | |
| "learning_rate": 5.772460821459542e-05, | |
| "loss": 2.4717, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.6430622009569378, | |
| "grad_norm": 0.27650853991508484, | |
| "learning_rate": 5.7480589477529545e-05, | |
| "loss": 2.5303, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.643912812333865, | |
| "grad_norm": 0.29573655128479004, | |
| "learning_rate": 5.723687940704856e-05, | |
| "loss": 2.6558, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.6447634237107921, | |
| "grad_norm": 0.2814992070198059, | |
| "learning_rate": 5.699347977234799e-05, | |
| "loss": 2.4268, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.6456140350877193, | |
| "grad_norm": 0.2827305495738983, | |
| "learning_rate": 5.675039234036983e-05, | |
| "loss": 2.5284, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 0.2877061665058136, | |
| "learning_rate": 5.650761887578977e-05, | |
| "loss": 2.5597, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6473152578415736, | |
| "grad_norm": 0.2902846038341522, | |
| "learning_rate": 5.6265161141004244e-05, | |
| "loss": 2.719, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.6481658692185008, | |
| "grad_norm": 0.3014174997806549, | |
| "learning_rate": 5.602302089611755e-05, | |
| "loss": 2.6498, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.6490164805954279, | |
| "grad_norm": 0.2754184603691101, | |
| "learning_rate": 5.578119989892931e-05, | |
| "loss": 2.7288, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.6498670919723551, | |
| "grad_norm": 0.27788618206977844, | |
| "learning_rate": 5.5539699904921635e-05, | |
| "loss": 2.5285, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.6507177033492823, | |
| "grad_norm": 0.27178868651390076, | |
| "learning_rate": 5.529852266724616e-05, | |
| "loss": 2.6693, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6515683147262095, | |
| "grad_norm": 0.2783076763153076, | |
| "learning_rate": 5.505766993671162e-05, | |
| "loss": 2.6003, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.6524189261031367, | |
| "grad_norm": 0.27819153666496277, | |
| "learning_rate": 5.481714346177103e-05, | |
| "loss": 2.5775, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.6532695374800638, | |
| "grad_norm": 0.27801868319511414, | |
| "learning_rate": 5.457694498850891e-05, | |
| "loss": 2.6342, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.654120148856991, | |
| "grad_norm": 0.27529436349868774, | |
| "learning_rate": 5.43370762606287e-05, | |
| "loss": 2.6376, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.6549707602339181, | |
| "grad_norm": 0.2912578582763672, | |
| "learning_rate": 5.409753901944006e-05, | |
| "loss": 2.8169, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6558213716108453, | |
| "grad_norm": 0.28431615233421326, | |
| "learning_rate": 5.385833500384632e-05, | |
| "loss": 2.5608, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.6566719829877724, | |
| "grad_norm": 0.3062645196914673, | |
| "learning_rate": 5.3619465950331646e-05, | |
| "loss": 2.6686, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.6575225943646996, | |
| "grad_norm": 0.29184243083000183, | |
| "learning_rate": 5.3380933592948704e-05, | |
| "loss": 2.4963, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.6583732057416268, | |
| "grad_norm": 0.3256790041923523, | |
| "learning_rate": 5.3142739663305906e-05, | |
| "loss": 2.5206, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.6592238171185539, | |
| "grad_norm": 0.3323804438114166, | |
| "learning_rate": 5.2904885890554836e-05, | |
| "loss": 2.7495, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.6600744284954811, | |
| "grad_norm": 0.3358718454837799, | |
| "learning_rate": 5.266737400137765e-05, | |
| "loss": 2.646, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.6609250398724082, | |
| "grad_norm": 0.34872984886169434, | |
| "learning_rate": 5.24302057199749e-05, | |
| "loss": 2.8041, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.6617756512493355, | |
| "grad_norm": 0.3432627320289612, | |
| "learning_rate": 5.2193382768052436e-05, | |
| "loss": 2.6066, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.6626262626262627, | |
| "grad_norm": 0.3716273009777069, | |
| "learning_rate": 5.19569068648094e-05, | |
| "loss": 2.6634, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.6634768740031898, | |
| "grad_norm": 0.3645707964897156, | |
| "learning_rate": 5.172077972692553e-05, | |
| "loss": 2.5782, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.664327485380117, | |
| "grad_norm": 0.43927833437919617, | |
| "learning_rate": 5.148500306854862e-05, | |
| "loss": 2.6775, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.6651780967570441, | |
| "grad_norm": 0.4325307607650757, | |
| "learning_rate": 5.124957860128237e-05, | |
| "loss": 2.6963, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.6660287081339713, | |
| "grad_norm": 0.39009419083595276, | |
| "learning_rate": 5.101450803417357e-05, | |
| "loss": 2.7308, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.6668793195108984, | |
| "grad_norm": 0.42211011052131653, | |
| "learning_rate": 5.0779793073700044e-05, | |
| "loss": 2.8391, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.6677299308878256, | |
| "grad_norm": 0.4249938130378723, | |
| "learning_rate": 5.054543542375809e-05, | |
| "loss": 2.6562, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.6685805422647528, | |
| "grad_norm": 0.4374600350856781, | |
| "learning_rate": 5.031143678565005e-05, | |
| "loss": 2.6552, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.6694311536416799, | |
| "grad_norm": 0.471137672662735, | |
| "learning_rate": 5.0077798858072156e-05, | |
| "loss": 2.6368, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.6702817650186071, | |
| "grad_norm": 0.4666772186756134, | |
| "learning_rate": 4.984452333710207e-05, | |
| "loss": 2.6859, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.6711323763955342, | |
| "grad_norm": 0.5150439739227295, | |
| "learning_rate": 4.961161191618649e-05, | |
| "loss": 2.9448, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.6719829877724615, | |
| "grad_norm": 0.4853833019733429, | |
| "learning_rate": 4.937906628612905e-05, | |
| "loss": 2.7409, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6728335991493887, | |
| "grad_norm": 0.4760463535785675, | |
| "learning_rate": 4.914688813507797e-05, | |
| "loss": 2.5815, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.6736842105263158, | |
| "grad_norm": 0.5204352140426636, | |
| "learning_rate": 4.89150791485137e-05, | |
| "loss": 2.7374, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.674534821903243, | |
| "grad_norm": 0.5126561522483826, | |
| "learning_rate": 4.86836410092368e-05, | |
| "loss": 2.755, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.6753854332801701, | |
| "grad_norm": 0.55113685131073, | |
| "learning_rate": 4.845257539735577e-05, | |
| "loss": 2.7027, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.6762360446570973, | |
| "grad_norm": 0.5475419759750366, | |
| "learning_rate": 4.822188399027461e-05, | |
| "loss": 2.667, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.6770866560340244, | |
| "grad_norm": 0.6212125420570374, | |
| "learning_rate": 4.799156846268095e-05, | |
| "loss": 2.6865, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.6779372674109516, | |
| "grad_norm": 0.6271610260009766, | |
| "learning_rate": 4.7761630486533694e-05, | |
| "loss": 2.9713, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.6787878787878788, | |
| "grad_norm": 0.6181414723396301, | |
| "learning_rate": 4.7532071731050975e-05, | |
| "loss": 2.8862, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.6796384901648059, | |
| "grad_norm": 0.6996387243270874, | |
| "learning_rate": 4.730289386269792e-05, | |
| "loss": 2.9082, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.6804891015417331, | |
| "grad_norm": 0.7603393793106079, | |
| "learning_rate": 4.70740985451747e-05, | |
| "loss": 3.034, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6813397129186602, | |
| "grad_norm": 0.27659621834754944, | |
| "learning_rate": 4.684568743940444e-05, | |
| "loss": 2.5099, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.6821903242955875, | |
| "grad_norm": 0.29582586884498596, | |
| "learning_rate": 4.661766220352097e-05, | |
| "loss": 2.5086, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.6830409356725147, | |
| "grad_norm": 0.31561413407325745, | |
| "learning_rate": 4.639002449285693e-05, | |
| "loss": 2.5214, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.6838915470494418, | |
| "grad_norm": 0.27388331294059753, | |
| "learning_rate": 4.616277595993196e-05, | |
| "loss": 2.4724, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.684742158426369, | |
| "grad_norm": 0.2823973000049591, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 2.4987, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.6855927698032961, | |
| "grad_norm": 0.3274148106575012, | |
| "learning_rate": 4.57094530232389e-05, | |
| "loss": 2.6108, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.6864433811802233, | |
| "grad_norm": 0.27294856309890747, | |
| "learning_rate": 4.5483381910335955e-05, | |
| "loss": 2.4774, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.6872939925571504, | |
| "grad_norm": 0.2771831750869751, | |
| "learning_rate": 4.525770655687821e-05, | |
| "loss": 2.4953, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.6881446039340776, | |
| "grad_norm": 0.28946414589881897, | |
| "learning_rate": 4.5032428601139644e-05, | |
| "loss": 2.7516, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.6889952153110048, | |
| "grad_norm": 0.2817944884300232, | |
| "learning_rate": 4.48075496785092e-05, | |
| "loss": 2.6658, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6898458266879319, | |
| "grad_norm": 0.2726970613002777, | |
| "learning_rate": 4.4583071421479194e-05, | |
| "loss": 2.6478, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.6906964380648591, | |
| "grad_norm": 0.2892582416534424, | |
| "learning_rate": 4.435899545963332e-05, | |
| "loss": 2.6592, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.6915470494417862, | |
| "grad_norm": 0.2825833559036255, | |
| "learning_rate": 4.4135323419634766e-05, | |
| "loss": 2.4431, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.6923976608187135, | |
| "grad_norm": 0.27093660831451416, | |
| "learning_rate": 4.391205692521453e-05, | |
| "loss": 2.4303, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.6932482721956407, | |
| "grad_norm": 0.2554977834224701, | |
| "learning_rate": 4.368919759715964e-05, | |
| "loss": 2.3259, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.6940988835725678, | |
| "grad_norm": 0.27814123034477234, | |
| "learning_rate": 4.346674705330117e-05, | |
| "loss": 2.5495, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.694949494949495, | |
| "grad_norm": 0.303751140832901, | |
| "learning_rate": 4.32447069085028e-05, | |
| "loss": 2.5834, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.6958001063264221, | |
| "grad_norm": 0.27305588126182556, | |
| "learning_rate": 4.302307877464893e-05, | |
| "loss": 2.479, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.6966507177033493, | |
| "grad_norm": 0.265235036611557, | |
| "learning_rate": 4.280186426063291e-05, | |
| "loss": 2.4847, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.6975013290802764, | |
| "grad_norm": 0.2711983025074005, | |
| "learning_rate": 4.258106497234551e-05, | |
| "loss": 2.3499, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6983519404572036, | |
| "grad_norm": 0.30092155933380127, | |
| "learning_rate": 4.236068251266324e-05, | |
| "loss": 2.7319, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.6992025518341308, | |
| "grad_norm": 0.3007088601589203, | |
| "learning_rate": 4.214071848143655e-05, | |
| "loss": 2.5465, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.7000531632110579, | |
| "grad_norm": 0.3158135712146759, | |
| "learning_rate": 4.192117447547845e-05, | |
| "loss": 2.6333, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.7009037745879851, | |
| "grad_norm": 0.3173391819000244, | |
| "learning_rate": 4.170205208855281e-05, | |
| "loss": 2.8266, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.3236771523952484, | |
| "learning_rate": 4.148335291136267e-05, | |
| "loss": 2.7447, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7026049973418395, | |
| "grad_norm": 0.320290744304657, | |
| "learning_rate": 4.1265078531538916e-05, | |
| "loss": 2.8345, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.7034556087187667, | |
| "grad_norm": 0.33358582854270935, | |
| "learning_rate": 4.104723053362867e-05, | |
| "loss": 2.848, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.7043062200956938, | |
| "grad_norm": 0.35341066122055054, | |
| "learning_rate": 4.082981049908362e-05, | |
| "loss": 2.7216, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.705156831472621, | |
| "grad_norm": 0.35734865069389343, | |
| "learning_rate": 4.061282000624885e-05, | |
| "loss": 2.702, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.7060074428495481, | |
| "grad_norm": 0.3350493013858795, | |
| "learning_rate": 4.0396260630351066e-05, | |
| "loss": 2.4811, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7068580542264753, | |
| "grad_norm": 0.35712873935699463, | |
| "learning_rate": 4.018013394348752e-05, | |
| "loss": 2.7484, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.7077086656034024, | |
| "grad_norm": 0.37364643812179565, | |
| "learning_rate": 3.996444151461417e-05, | |
| "loss": 2.6113, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.7085592769803296, | |
| "grad_norm": 0.41270262002944946, | |
| "learning_rate": 3.9749184909534565e-05, | |
| "loss": 2.6654, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.7094098883572568, | |
| "grad_norm": 0.3864672780036926, | |
| "learning_rate": 3.9534365690888566e-05, | |
| "loss": 2.6718, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.7102604997341839, | |
| "grad_norm": 0.4115494191646576, | |
| "learning_rate": 3.931998541814069e-05, | |
| "loss": 2.6621, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.4606288969516754, | |
| "learning_rate": 3.9106045647569e-05, | |
| "loss": 2.512, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.7119617224880382, | |
| "grad_norm": 0.42326587438583374, | |
| "learning_rate": 3.8892547932253795e-05, | |
| "loss": 2.6212, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.7128123338649655, | |
| "grad_norm": 0.43627670407295227, | |
| "learning_rate": 3.8679493822066314e-05, | |
| "loss": 2.596, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.7136629452418927, | |
| "grad_norm": 0.4475801885128021, | |
| "learning_rate": 3.846688486365748e-05, | |
| "loss": 2.64, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.7145135566188198, | |
| "grad_norm": 0.4370073080062866, | |
| "learning_rate": 3.825472260044658e-05, | |
| "loss": 2.5291, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.715364167995747, | |
| "grad_norm": 0.4832041561603546, | |
| "learning_rate": 3.804300857261025e-05, | |
| "loss": 2.659, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.7162147793726741, | |
| "grad_norm": 0.4905672073364258, | |
| "learning_rate": 3.783174431707119e-05, | |
| "loss": 2.7363, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.7170653907496013, | |
| "grad_norm": 0.5135957598686218, | |
| "learning_rate": 3.762093136748692e-05, | |
| "loss": 2.8243, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.7179160021265284, | |
| "grad_norm": 0.5257413387298584, | |
| "learning_rate": 3.7410571254238834e-05, | |
| "loss": 2.8406, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.7187666135034556, | |
| "grad_norm": 0.5150455236434937, | |
| "learning_rate": 3.7200665504420983e-05, | |
| "loss": 2.6053, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7196172248803828, | |
| "grad_norm": 0.5135449767112732, | |
| "learning_rate": 3.69912156418289e-05, | |
| "loss": 2.4932, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.7204678362573099, | |
| "grad_norm": 0.5775859951972961, | |
| "learning_rate": 3.678222318694875e-05, | |
| "loss": 2.6011, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.7213184476342371, | |
| "grad_norm": 0.6324962973594666, | |
| "learning_rate": 3.657368965694617e-05, | |
| "loss": 2.8145, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.7221690590111642, | |
| "grad_norm": 0.6755786538124084, | |
| "learning_rate": 3.636561656565519e-05, | |
| "loss": 2.7566, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.7230196703880915, | |
| "grad_norm": 0.7153156995773315, | |
| "learning_rate": 3.615800542356738e-05, | |
| "loss": 2.9433, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7238702817650186, | |
| "grad_norm": 0.29322749376296997, | |
| "learning_rate": 3.595085773782083e-05, | |
| "loss": 2.3433, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.7247208931419458, | |
| "grad_norm": 0.28010284900665283, | |
| "learning_rate": 3.574417501218913e-05, | |
| "loss": 2.5661, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.725571504518873, | |
| "grad_norm": 0.27005431056022644, | |
| "learning_rate": 3.55379587470706e-05, | |
| "loss": 2.4129, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.7264221158958001, | |
| "grad_norm": 0.2710054814815521, | |
| "learning_rate": 3.533221043947733e-05, | |
| "loss": 2.4066, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.27026283740997314, | |
| "learning_rate": 3.512693158302421e-05, | |
| "loss": 2.4139, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.7281233386496544, | |
| "grad_norm": 0.2927492558956146, | |
| "learning_rate": 3.492212366791831e-05, | |
| "loss": 2.5517, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.7289739500265816, | |
| "grad_norm": 0.2770216166973114, | |
| "learning_rate": 3.471778818094785e-05, | |
| "loss": 2.6102, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.7298245614035088, | |
| "grad_norm": 0.26914748549461365, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 2.6047, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.7306751727804359, | |
| "grad_norm": 0.27307602763175964, | |
| "learning_rate": 3.4310540421407665e-05, | |
| "loss": 2.5707, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.7315257841573631, | |
| "grad_norm": 0.2949882745742798, | |
| "learning_rate": 3.4107631105223525e-05, | |
| "loss": 2.502, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7323763955342902, | |
| "grad_norm": 0.26683247089385986, | |
| "learning_rate": 3.390520012992474e-05, | |
| "loss": 2.5785, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.7332270069112174, | |
| "grad_norm": 0.2797079384326935, | |
| "learning_rate": 3.370324896504425e-05, | |
| "loss": 2.6119, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.7340776182881446, | |
| "grad_norm": 0.27322161197662354, | |
| "learning_rate": 3.3501779076631864e-05, | |
| "loss": 2.5449, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.7349282296650718, | |
| "grad_norm": 0.2743871212005615, | |
| "learning_rate": 3.330079192724379e-05, | |
| "loss": 2.6722, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.735778841041999, | |
| "grad_norm": 0.27956199645996094, | |
| "learning_rate": 3.3100288975931635e-05, | |
| "loss": 2.5033, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.7366294524189261, | |
| "grad_norm": 0.2711998522281647, | |
| "learning_rate": 3.290027167823204e-05, | |
| "loss": 2.5824, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.7374800637958533, | |
| "grad_norm": 0.277340292930603, | |
| "learning_rate": 3.270074148615615e-05, | |
| "loss": 2.5168, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.7383306751727804, | |
| "grad_norm": 0.26151034235954285, | |
| "learning_rate": 3.250169984817897e-05, | |
| "loss": 2.5749, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.7391812865497076, | |
| "grad_norm": 0.2847073972225189, | |
| "learning_rate": 3.230314820922883e-05, | |
| "loss": 2.4749, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.7400318979266348, | |
| "grad_norm": 0.26492998003959656, | |
| "learning_rate": 3.2105088010677e-05, | |
| "loss": 2.5874, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7408825093035619, | |
| "grad_norm": 0.2851366400718689, | |
| "learning_rate": 3.1907520690327184e-05, | |
| "loss": 2.7108, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.7417331206804891, | |
| "grad_norm": 0.29604560136795044, | |
| "learning_rate": 3.1710447682405076e-05, | |
| "loss": 2.8558, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.7425837320574162, | |
| "grad_norm": 0.2940882444381714, | |
| "learning_rate": 3.151387041754784e-05, | |
| "loss": 2.5961, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.7434343434343434, | |
| "grad_norm": 0.3165973126888275, | |
| "learning_rate": 3.131779032279397e-05, | |
| "loss": 2.5976, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.7442849548112705, | |
| "grad_norm": 0.3157351315021515, | |
| "learning_rate": 3.112220882157275e-05, | |
| "loss": 2.5901, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7451355661881978, | |
| "grad_norm": 0.3334716558456421, | |
| "learning_rate": 3.092712733369387e-05, | |
| "loss": 2.6966, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.745986177565125, | |
| "grad_norm": 0.3281215727329254, | |
| "learning_rate": 3.073254727533732e-05, | |
| "loss": 2.616, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.7468367889420521, | |
| "grad_norm": 0.3478872776031494, | |
| "learning_rate": 3.053847005904298e-05, | |
| "loss": 2.7568, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.7476874003189793, | |
| "grad_norm": 0.35839834809303284, | |
| "learning_rate": 3.034489709370033e-05, | |
| "loss": 2.7552, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.7485380116959064, | |
| "grad_norm": 0.3570314049720764, | |
| "learning_rate": 3.0151829784538254e-05, | |
| "loss": 2.6189, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7493886230728336, | |
| "grad_norm": 0.3771030008792877, | |
| "learning_rate": 2.995926953311504e-05, | |
| "loss": 3.0162, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.7502392344497608, | |
| "grad_norm": 0.38607585430145264, | |
| "learning_rate": 2.9767217737307806e-05, | |
| "loss": 2.8731, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.7502392344497608, | |
| "eval_loss": 2.657837152481079, | |
| "eval_runtime": 80.5036, | |
| "eval_samples_per_second": 12.298, | |
| "eval_steps_per_second": 6.149, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.7510898458266879, | |
| "grad_norm": 0.3817814588546753, | |
| "learning_rate": 2.9575675791302703e-05, | |
| "loss": 2.7678, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.7519404572036151, | |
| "grad_norm": 0.4072461724281311, | |
| "learning_rate": 2.9384645085584663e-05, | |
| "loss": 2.7004, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.7527910685805422, | |
| "grad_norm": 0.3934648036956787, | |
| "learning_rate": 2.9194127006927208e-05, | |
| "loss": 2.4883, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.7536416799574694, | |
| "grad_norm": 0.43659910559654236, | |
| "learning_rate": 2.9004122938382617e-05, | |
| "loss": 2.7447, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.7544922913343965, | |
| "grad_norm": 0.4242823123931885, | |
| "learning_rate": 2.881463425927161e-05, | |
| "loss": 2.7324, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.7553429027113238, | |
| "grad_norm": 0.4546271860599518, | |
| "learning_rate": 2.86256623451736e-05, | |
| "loss": 2.5985, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.756193514088251, | |
| "grad_norm": 0.462896466255188, | |
| "learning_rate": 2.8437208567916517e-05, | |
| "loss": 2.5522, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.7570441254651781, | |
| "grad_norm": 0.47179552912712097, | |
| "learning_rate": 2.8249274295566864e-05, | |
| "loss": 2.7653, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7578947368421053, | |
| "grad_norm": 0.5069417357444763, | |
| "learning_rate": 2.8061860892420012e-05, | |
| "loss": 2.8404, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.7587453482190324, | |
| "grad_norm": 0.5162703990936279, | |
| "learning_rate": 2.787496971898994e-05, | |
| "loss": 2.5679, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.7595959595959596, | |
| "grad_norm": 0.5141138434410095, | |
| "learning_rate": 2.7688602131999565e-05, | |
| "loss": 2.7156, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.7604465709728868, | |
| "grad_norm": 0.5267017483711243, | |
| "learning_rate": 2.7502759484370944e-05, | |
| "loss": 2.8324, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.7612971823498139, | |
| "grad_norm": 0.5354244112968445, | |
| "learning_rate": 2.7317443125215357e-05, | |
| "loss": 2.7099, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.7621477937267411, | |
| "grad_norm": 0.5597330331802368, | |
| "learning_rate": 2.7132654399823444e-05, | |
| "loss": 2.7128, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.7629984051036682, | |
| "grad_norm": 0.6345269680023193, | |
| "learning_rate": 2.6948394649655627e-05, | |
| "loss": 2.8904, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.7638490164805954, | |
| "grad_norm": 0.5837516784667969, | |
| "learning_rate": 2.6764665212332253e-05, | |
| "loss": 2.5579, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.7646996278575225, | |
| "grad_norm": 0.7032825350761414, | |
| "learning_rate": 2.658146742162384e-05, | |
| "loss": 2.7115, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.7655502392344498, | |
| "grad_norm": 0.6780116558074951, | |
| "learning_rate": 2.6398802607441507e-05, | |
| "loss": 2.6573, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.766400850611377, | |
| "grad_norm": 0.2630327641963959, | |
| "learning_rate": 2.6216672095827266e-05, | |
| "loss": 2.5077, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.7672514619883041, | |
| "grad_norm": 0.2665589153766632, | |
| "learning_rate": 2.6035077208944415e-05, | |
| "loss": 2.4456, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.7681020733652313, | |
| "grad_norm": 0.2564515471458435, | |
| "learning_rate": 2.5854019265067853e-05, | |
| "loss": 2.2768, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.7689526847421584, | |
| "grad_norm": 0.290310263633728, | |
| "learning_rate": 2.5673499578574645e-05, | |
| "loss": 2.6339, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.7698032961190856, | |
| "grad_norm": 0.27893102169036865, | |
| "learning_rate": 2.5493519459934423e-05, | |
| "loss": 2.6156, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.7706539074960128, | |
| "grad_norm": 0.27719345688819885, | |
| "learning_rate": 2.531408021569982e-05, | |
| "loss": 2.4212, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.7715045188729399, | |
| "grad_norm": 0.2813560664653778, | |
| "learning_rate": 2.5135183148496978e-05, | |
| "loss": 2.5064, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.7723551302498671, | |
| "grad_norm": 0.28871771693229675, | |
| "learning_rate": 2.4956829557016338e-05, | |
| "loss": 2.505, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.7732057416267942, | |
| "grad_norm": 0.30079928040504456, | |
| "learning_rate": 2.4779020736002834e-05, | |
| "loss": 2.4488, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.7740563530037214, | |
| "grad_norm": 0.27679336071014404, | |
| "learning_rate": 2.4601757976246686e-05, | |
| "loss": 2.6442, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7749069643806485, | |
| "grad_norm": 0.27239468693733215, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 2.4256, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.7757575757575758, | |
| "grad_norm": 0.2782379686832428, | |
| "learning_rate": 2.4248875783837987e-05, | |
| "loss": 2.5383, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.776608187134503, | |
| "grad_norm": 0.28101786971092224, | |
| "learning_rate": 2.407325891290817e-05, | |
| "loss": 2.5509, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.7774587985114301, | |
| "grad_norm": 0.26222068071365356, | |
| "learning_rate": 2.3898193226662634e-05, | |
| "loss": 2.5718, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.7783094098883573, | |
| "grad_norm": 0.28495025634765625, | |
| "learning_rate": 2.3723679995978088e-05, | |
| "loss": 2.7216, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.7791600212652844, | |
| "grad_norm": 0.27799373865127563, | |
| "learning_rate": 2.3549720487720738e-05, | |
| "loss": 2.4825, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.7800106326422116, | |
| "grad_norm": 0.2805662751197815, | |
| "learning_rate": 2.3376315964737004e-05, | |
| "loss": 2.5418, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.7808612440191387, | |
| "grad_norm": 0.2894647717475891, | |
| "learning_rate": 2.3203467685844494e-05, | |
| "loss": 2.4872, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.7817118553960659, | |
| "grad_norm": 0.2725624740123749, | |
| "learning_rate": 2.3031176905822805e-05, | |
| "loss": 2.6821, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.7825624667729931, | |
| "grad_norm": 0.2858252227306366, | |
| "learning_rate": 2.2859444875404347e-05, | |
| "loss": 2.5256, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7834130781499202, | |
| "grad_norm": 0.27502021193504333, | |
| "learning_rate": 2.268827284126539e-05, | |
| "loss": 2.6097, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.7842636895268474, | |
| "grad_norm": 0.30405184626579285, | |
| "learning_rate": 2.2517662046016975e-05, | |
| "loss": 2.5872, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.7851143009037745, | |
| "grad_norm": 0.3032589256763458, | |
| "learning_rate": 2.234761372819577e-05, | |
| "loss": 2.6253, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.7859649122807018, | |
| "grad_norm": 0.3149045705795288, | |
| "learning_rate": 2.2178129122255253e-05, | |
| "loss": 2.5631, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.786815523657629, | |
| "grad_norm": 0.3490796387195587, | |
| "learning_rate": 2.200920945855669e-05, | |
| "loss": 2.7784, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.7876661350345561, | |
| "grad_norm": 0.332727313041687, | |
| "learning_rate": 2.184085596336011e-05, | |
| "loss": 2.7369, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.7885167464114833, | |
| "grad_norm": 0.345241516828537, | |
| "learning_rate": 2.1673069858815554e-05, | |
| "loss": 2.734, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.7893673577884104, | |
| "grad_norm": 0.356146901845932, | |
| "learning_rate": 2.150585236295415e-05, | |
| "loss": 2.5664, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.7902179691653376, | |
| "grad_norm": 0.34294942021369934, | |
| "learning_rate": 2.133920468967915e-05, | |
| "loss": 2.7841, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.7910685805422647, | |
| "grad_norm": 0.3757612407207489, | |
| "learning_rate": 2.1173128048757306e-05, | |
| "loss": 2.8389, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7919191919191919, | |
| "grad_norm": 0.37201401591300964, | |
| "learning_rate": 2.1007623645810003e-05, | |
| "loss": 2.7424, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.7927698032961191, | |
| "grad_norm": 0.37049031257629395, | |
| "learning_rate": 2.0842692682304444e-05, | |
| "loss": 2.7447, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.7936204146730462, | |
| "grad_norm": 0.40525051951408386, | |
| "learning_rate": 2.0678336355545048e-05, | |
| "loss": 2.7354, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.7944710260499734, | |
| "grad_norm": 0.37740442156791687, | |
| "learning_rate": 2.0514555858664663e-05, | |
| "loss": 2.5756, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.7953216374269005, | |
| "grad_norm": 0.399179607629776, | |
| "learning_rate": 2.0351352380616008e-05, | |
| "loss": 2.606, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.7961722488038278, | |
| "grad_norm": 0.4020417630672455, | |
| "learning_rate": 2.0188727106162874e-05, | |
| "loss": 2.5945, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.797022860180755, | |
| "grad_norm": 0.4364849627017975, | |
| "learning_rate": 2.0026681215871656e-05, | |
| "loss": 2.7867, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.7978734715576821, | |
| "grad_norm": 0.4335819482803345, | |
| "learning_rate": 1.986521588610285e-05, | |
| "loss": 2.6452, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.7987240829346093, | |
| "grad_norm": 0.44533243775367737, | |
| "learning_rate": 1.9704332289002293e-05, | |
| "loss": 2.7037, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.7995746943115364, | |
| "grad_norm": 0.46524372696876526, | |
| "learning_rate": 1.9544031592492763e-05, | |
| "loss": 2.7772, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8004253056884636, | |
| "grad_norm": 0.4814035892486572, | |
| "learning_rate": 1.9384314960265692e-05, | |
| "loss": 2.7603, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.8012759170653907, | |
| "grad_norm": 0.5147087574005127, | |
| "learning_rate": 1.922518355177232e-05, | |
| "loss": 2.8118, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.8021265284423179, | |
| "grad_norm": 0.5196573734283447, | |
| "learning_rate": 1.906663852221565e-05, | |
| "loss": 2.7226, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.8029771398192451, | |
| "grad_norm": 0.538750171661377, | |
| "learning_rate": 1.890868102254182e-05, | |
| "loss": 2.8252, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.8038277511961722, | |
| "grad_norm": 0.5554125308990479, | |
| "learning_rate": 1.875131219943187e-05, | |
| "loss": 2.7985, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8046783625730994, | |
| "grad_norm": 0.5610653162002563, | |
| "learning_rate": 1.8594533195293427e-05, | |
| "loss": 2.7006, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.8055289739500265, | |
| "grad_norm": 0.5772664546966553, | |
| "learning_rate": 1.843834514825229e-05, | |
| "loss": 2.7453, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.8063795853269538, | |
| "grad_norm": 0.7284188866615295, | |
| "learning_rate": 1.82827491921443e-05, | |
| "loss": 2.7826, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.807230196703881, | |
| "grad_norm": 0.6359315514564514, | |
| "learning_rate": 1.8127746456507077e-05, | |
| "loss": 2.5094, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 0.7618624567985535, | |
| "learning_rate": 1.797333806657171e-05, | |
| "loss": 2.9762, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8089314194577353, | |
| "grad_norm": 0.2761584520339966, | |
| "learning_rate": 1.7819525143254755e-05, | |
| "loss": 2.3495, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.8097820308346624, | |
| "grad_norm": 0.2787644565105438, | |
| "learning_rate": 1.7666308803150043e-05, | |
| "loss": 2.6593, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.8106326422115896, | |
| "grad_norm": 0.2583043873310089, | |
| "learning_rate": 1.751369015852046e-05, | |
| "loss": 2.3744, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.8114832535885167, | |
| "grad_norm": 0.27526992559432983, | |
| "learning_rate": 1.7361670317290012e-05, | |
| "loss": 2.424, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.8123338649654439, | |
| "grad_norm": 0.2713683843612671, | |
| "learning_rate": 1.7210250383035807e-05, | |
| "loss": 2.5069, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8131844763423711, | |
| "grad_norm": 0.28076109290122986, | |
| "learning_rate": 1.7059431454979824e-05, | |
| "loss": 2.4204, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.8140350877192982, | |
| "grad_norm": 0.2972472012042999, | |
| "learning_rate": 1.6909214627981197e-05, | |
| "loss": 2.6109, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.8148856990962254, | |
| "grad_norm": 0.31532159447669983, | |
| "learning_rate": 1.6759600992528147e-05, | |
| "loss": 2.6416, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.8157363104731525, | |
| "grad_norm": 0.2699948847293854, | |
| "learning_rate": 1.6610591634729965e-05, | |
| "loss": 2.4951, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.8165869218500797, | |
| "grad_norm": 0.27630969882011414, | |
| "learning_rate": 1.6462187636309345e-05, | |
| "loss": 2.6336, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.817437533227007, | |
| "grad_norm": 0.27848508954048157, | |
| "learning_rate": 1.631439007459441e-05, | |
| "loss": 2.6848, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.8182881446039341, | |
| "grad_norm": 0.2589823007583618, | |
| "learning_rate": 1.61672000225108e-05, | |
| "loss": 2.6077, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.8191387559808613, | |
| "grad_norm": 0.27338507771492004, | |
| "learning_rate": 1.6020618548574108e-05, | |
| "loss": 2.4618, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.8199893673577884, | |
| "grad_norm": 0.26967307925224304, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 2.4733, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.8208399787347156, | |
| "grad_norm": 0.27979278564453125, | |
| "learning_rate": 1.5729285587106136e-05, | |
| "loss": 2.6789, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.8216905901116427, | |
| "grad_norm": 0.263261616230011, | |
| "learning_rate": 1.5584536214485457e-05, | |
| "loss": 2.5471, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.8225412014885699, | |
| "grad_norm": 0.27126452326774597, | |
| "learning_rate": 1.5440399649817385e-05, | |
| "loss": 2.6185, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.8233918128654971, | |
| "grad_norm": 0.2681058943271637, | |
| "learning_rate": 1.5296876939450978e-05, | |
| "loss": 2.567, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.8242424242424242, | |
| "grad_norm": 0.25772204995155334, | |
| "learning_rate": 1.5153969125278934e-05, | |
| "loss": 2.4961, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.8250930356193514, | |
| "grad_norm": 0.27491864562034607, | |
| "learning_rate": 1.5011677244730161e-05, | |
| "loss": 2.6572, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8259436469962785, | |
| "grad_norm": 0.3069652318954468, | |
| "learning_rate": 1.4870002330762289e-05, | |
| "loss": 2.6633, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.8267942583732057, | |
| "grad_norm": 0.29441124200820923, | |
| "learning_rate": 1.4728945411854133e-05, | |
| "loss": 2.5631, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.827644869750133, | |
| "grad_norm": 0.29318681359291077, | |
| "learning_rate": 1.4588507511998162e-05, | |
| "loss": 2.7143, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.8284954811270601, | |
| "grad_norm": 0.3085382282733917, | |
| "learning_rate": 1.4448689650693147e-05, | |
| "loss": 2.6399, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.8293460925039873, | |
| "grad_norm": 0.3353934586048126, | |
| "learning_rate": 1.4309492842936768e-05, | |
| "loss": 2.7392, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.8301967038809144, | |
| "grad_norm": 0.35874679684638977, | |
| "learning_rate": 1.4170918099218166e-05, | |
| "loss": 2.6478, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.8310473152578416, | |
| "grad_norm": 0.3231167793273926, | |
| "learning_rate": 1.4032966425510663e-05, | |
| "loss": 2.5808, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.8318979266347687, | |
| "grad_norm": 0.3558177351951599, | |
| "learning_rate": 1.3895638823264446e-05, | |
| "loss": 2.8192, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.8327485380116959, | |
| "grad_norm": 0.3447312116622925, | |
| "learning_rate": 1.3758936289399348e-05, | |
| "loss": 2.439, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.8335991493886231, | |
| "grad_norm": 0.35401877760887146, | |
| "learning_rate": 1.3622859816297473e-05, | |
| "loss": 2.6874, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8344497607655502, | |
| "grad_norm": 0.3697657585144043, | |
| "learning_rate": 1.3487410391796162e-05, | |
| "loss": 2.6217, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.8353003721424774, | |
| "grad_norm": 0.39387422800064087, | |
| "learning_rate": 1.3352588999180726e-05, | |
| "loss": 2.9086, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.8361509835194045, | |
| "grad_norm": 0.4256942570209503, | |
| "learning_rate": 1.3218396617177287e-05, | |
| "loss": 2.6578, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.8370015948963317, | |
| "grad_norm": 0.40965893864631653, | |
| "learning_rate": 1.308483421994573e-05, | |
| "loss": 2.6781, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.837852206273259, | |
| "grad_norm": 0.4041212797164917, | |
| "learning_rate": 1.2951902777072655e-05, | |
| "loss": 2.8038, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.8387028176501861, | |
| "grad_norm": 0.416715532541275, | |
| "learning_rate": 1.2819603253564205e-05, | |
| "loss": 2.5885, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.8395534290271133, | |
| "grad_norm": 0.4417300522327423, | |
| "learning_rate": 1.2687936609839235e-05, | |
| "loss": 2.8096, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.8404040404040404, | |
| "grad_norm": 0.47883424162864685, | |
| "learning_rate": 1.2556903801722219e-05, | |
| "loss": 2.8092, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.8412546517809676, | |
| "grad_norm": 0.4706938862800598, | |
| "learning_rate": 1.2426505780436326e-05, | |
| "loss": 2.6285, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.4550638496875763, | |
| "learning_rate": 1.2296743492596586e-05, | |
| "loss": 2.5213, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8429558745348219, | |
| "grad_norm": 0.48749181628227234, | |
| "learning_rate": 1.2167617880202908e-05, | |
| "loss": 2.791, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.8438064859117491, | |
| "grad_norm": 0.46483585238456726, | |
| "learning_rate": 1.2039129880633349e-05, | |
| "loss": 2.7672, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.8446570972886762, | |
| "grad_norm": 0.510725200176239, | |
| "learning_rate": 1.1911280426637273e-05, | |
| "loss": 2.9072, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.8455077086656034, | |
| "grad_norm": 0.5274655222892761, | |
| "learning_rate": 1.1784070446328476e-05, | |
| "loss": 2.8219, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.8463583200425305, | |
| "grad_norm": 0.5416675806045532, | |
| "learning_rate": 1.1657500863178694e-05, | |
| "loss": 2.7237, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.8472089314194577, | |
| "grad_norm": 0.5472414493560791, | |
| "learning_rate": 1.153157259601062e-05, | |
| "loss": 2.68, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.8480595427963848, | |
| "grad_norm": 0.5868435502052307, | |
| "learning_rate": 1.1406286558991375e-05, | |
| "loss": 2.7562, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.8489101541733121, | |
| "grad_norm": 0.6191973090171814, | |
| "learning_rate": 1.1281643661625895e-05, | |
| "loss": 2.7655, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.8497607655502393, | |
| "grad_norm": 0.6745545268058777, | |
| "learning_rate": 1.1157644808750312e-05, | |
| "loss": 2.8605, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.8506113769271664, | |
| "grad_norm": 0.697556734085083, | |
| "learning_rate": 1.103429090052528e-05, | |
| "loss": 2.6958, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8514619883040936, | |
| "grad_norm": 0.26258981227874756, | |
| "learning_rate": 1.0911582832429589e-05, | |
| "loss": 2.4457, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.8523125996810207, | |
| "grad_norm": 0.26820749044418335, | |
| "learning_rate": 1.0789521495253618e-05, | |
| "loss": 2.6249, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.8531632110579479, | |
| "grad_norm": 0.26697129011154175, | |
| "learning_rate": 1.0668107775092751e-05, | |
| "loss": 2.4464, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.8540138224348751, | |
| "grad_norm": 0.2953515350818634, | |
| "learning_rate": 1.0547342553341144e-05, | |
| "loss": 2.5413, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.8548644338118022, | |
| "grad_norm": 0.264387309551239, | |
| "learning_rate": 1.0427226706685178e-05, | |
| "loss": 2.6889, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.8557150451887294, | |
| "grad_norm": 0.2656628489494324, | |
| "learning_rate": 1.030776110709718e-05, | |
| "loss": 2.5115, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.8565656565656565, | |
| "grad_norm": 0.25700512528419495, | |
| "learning_rate": 1.0188946621828976e-05, | |
| "loss": 2.5474, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.8574162679425837, | |
| "grad_norm": 0.28423216938972473, | |
| "learning_rate": 1.0070784113405763e-05, | |
| "loss": 2.5926, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.8582668793195108, | |
| "grad_norm": 0.2653867304325104, | |
| "learning_rate": 9.953274439619741e-06, | |
| "loss": 2.4022, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.8591174906964381, | |
| "grad_norm": 0.27558040618896484, | |
| "learning_rate": 9.836418453523833e-06, | |
| "loss": 2.6529, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8599681020733653, | |
| "grad_norm": 0.25471609830856323, | |
| "learning_rate": 9.720217003425647e-06, | |
| "loss": 2.5865, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.8608187134502924, | |
| "grad_norm": 0.2609470784664154, | |
| "learning_rate": 9.60467093288121e-06, | |
| "loss": 2.5492, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.8616693248272196, | |
| "grad_norm": 0.29608237743377686, | |
| "learning_rate": 9.489781080688865e-06, | |
| "loss": 2.5104, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.8625199362041467, | |
| "grad_norm": 0.29410475492477417, | |
| "learning_rate": 9.375548280883128e-06, | |
| "loss": 2.6142, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.8633705475810739, | |
| "grad_norm": 0.269397497177124, | |
| "learning_rate": 9.261973362728827e-06, | |
| "loss": 2.5652, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.8642211589580011, | |
| "grad_norm": 0.28159505128860474, | |
| "learning_rate": 9.149057150714801e-06, | |
| "loss": 2.7499, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.8650717703349282, | |
| "grad_norm": 0.27257442474365234, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 2.6305, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.8659223817118554, | |
| "grad_norm": 0.26392441987991333, | |
| "learning_rate": 8.92520411914819e-06, | |
| "loss": 2.4767, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.8667729930887825, | |
| "grad_norm": 0.2779422998428345, | |
| "learning_rate": 8.814268924640468e-06, | |
| "loss": 2.6056, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.8676236044657097, | |
| "grad_norm": 0.27914363145828247, | |
| "learning_rate": 8.70399568635104e-06, | |
| "loss": 2.5612, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8684742158426368, | |
| "grad_norm": 0.28274431824684143, | |
| "learning_rate": 8.594385204800482e-06, | |
| "loss": 2.5673, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.8693248272195641, | |
| "grad_norm": 0.30289795994758606, | |
| "learning_rate": 8.485438275698154e-06, | |
| "loss": 2.6624, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.8701754385964913, | |
| "grad_norm": 0.3084239363670349, | |
| "learning_rate": 8.377155689936434e-06, | |
| "loss": 2.5199, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.8710260499734184, | |
| "grad_norm": 0.32038623094558716, | |
| "learning_rate": 8.269538233584883e-06, | |
| "loss": 2.5174, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.8718766613503456, | |
| "grad_norm": 0.32201388478279114, | |
| "learning_rate": 8.162586687884654e-06, | |
| "loss": 2.5764, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 0.31848224997520447, | |
| "learning_rate": 8.056301829242784e-06, | |
| "loss": 2.6327, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.8735778841041999, | |
| "grad_norm": 0.34866663813591003, | |
| "learning_rate": 7.950684429226463e-06, | |
| "loss": 2.8354, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.8744284954811271, | |
| "grad_norm": 0.3639064431190491, | |
| "learning_rate": 7.845735254557606e-06, | |
| "loss": 2.7546, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.8752791068580542, | |
| "grad_norm": 0.35733523964881897, | |
| "learning_rate": 7.741455067107162e-06, | |
| "loss": 2.69, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.8761297182349814, | |
| "grad_norm": 0.35862669348716736, | |
| "learning_rate": 7.637844623889556e-06, | |
| "loss": 2.5966, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.8769803296119085, | |
| "grad_norm": 0.36205005645751953, | |
| "learning_rate": 7.534904677057353e-06, | |
| "loss": 2.4966, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.8778309409888357, | |
| "grad_norm": 0.3843689560890198, | |
| "learning_rate": 7.4326359738956515e-06, | |
| "loss": 2.7942, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.8786815523657628, | |
| "grad_norm": 0.40694114565849304, | |
| "learning_rate": 7.331039256816663e-06, | |
| "loss": 2.739, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.87953216374269, | |
| "grad_norm": 0.40859633684158325, | |
| "learning_rate": 7.230115263354431e-06, | |
| "loss": 2.6879, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.8803827751196173, | |
| "grad_norm": 0.41907498240470886, | |
| "learning_rate": 7.129864726159408e-06, | |
| "loss": 2.5804, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.8812333864965444, | |
| "grad_norm": 0.4525774121284485, | |
| "learning_rate": 7.030288372993066e-06, | |
| "loss": 2.6814, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.8820839978734716, | |
| "grad_norm": 0.45679178833961487, | |
| "learning_rate": 6.931386926722772e-06, | |
| "loss": 2.7909, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.8829346092503987, | |
| "grad_norm": 0.47028636932373047, | |
| "learning_rate": 6.833161105316421e-06, | |
| "loss": 2.8758, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.8837852206273259, | |
| "grad_norm": 0.4830760359764099, | |
| "learning_rate": 6.7356116218372566e-06, | |
| "loss": 2.6979, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.8846358320042531, | |
| "grad_norm": 0.49996817111968994, | |
| "learning_rate": 6.63873918443868e-06, | |
| "loss": 2.8513, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8854864433811802, | |
| "grad_norm": 0.4843011200428009, | |
| "learning_rate": 6.542544496359138e-06, | |
| "loss": 2.553, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.8863370547581074, | |
| "grad_norm": 0.5989816188812256, | |
| "learning_rate": 6.447028255917054e-06, | |
| "loss": 2.5985, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.8871876661350345, | |
| "grad_norm": 0.5418705940246582, | |
| "learning_rate": 6.352191156505627e-06, | |
| "loss": 2.7277, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.8880382775119617, | |
| "grad_norm": 0.5338729619979858, | |
| "learning_rate": 6.258033886587911e-06, | |
| "loss": 2.9324, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.5353463292121887, | |
| "learning_rate": 6.164557129691828e-06, | |
| "loss": 2.6506, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.889739500265816, | |
| "grad_norm": 0.6170979142189026, | |
| "learning_rate": 6.0717615644051206e-06, | |
| "loss": 2.7699, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.8905901116427433, | |
| "grad_norm": 0.6048348546028137, | |
| "learning_rate": 5.979647864370486e-06, | |
| "loss": 2.7457, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.8914407230196704, | |
| "grad_norm": 0.6213808059692383, | |
| "learning_rate": 5.888216698280647e-06, | |
| "loss": 2.671, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.8922913343965976, | |
| "grad_norm": 0.6980977654457092, | |
| "learning_rate": 5.7974687298735454e-06, | |
| "loss": 2.9322, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.8931419457735247, | |
| "grad_norm": 0.762441873550415, | |
| "learning_rate": 5.7074046179275255e-06, | |
| "loss": 2.9242, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8939925571504519, | |
| "grad_norm": 0.2550997734069824, | |
| "learning_rate": 5.6180250162564455e-06, | |
| "loss": 2.3309, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.8948431685273791, | |
| "grad_norm": 0.2799069285392761, | |
| "learning_rate": 5.5293305737050825e-06, | |
| "loss": 2.6428, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.8956937799043062, | |
| "grad_norm": 0.25860702991485596, | |
| "learning_rate": 5.441321934144339e-06, | |
| "loss": 2.2591, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.8965443912812334, | |
| "grad_norm": 0.262991726398468, | |
| "learning_rate": 5.35399973646653e-06, | |
| "loss": 2.3974, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.8973950026581605, | |
| "grad_norm": 0.25796017050743103, | |
| "learning_rate": 5.267364614580861e-06, | |
| "loss": 2.4487, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.8982456140350877, | |
| "grad_norm": 0.26405441761016846, | |
| "learning_rate": 5.181417197408734e-06, | |
| "loss": 2.3894, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.8990962254120148, | |
| "grad_norm": 0.27741700410842896, | |
| "learning_rate": 5.09615810887919e-06, | |
| "loss": 2.5089, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.899946836788942, | |
| "grad_norm": 0.27698126435279846, | |
| "learning_rate": 5.011587967924414e-06, | |
| "loss": 2.4868, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.9007974481658693, | |
| "grad_norm": 0.2699805796146393, | |
| "learning_rate": 4.927707388475255e-06, | |
| "loss": 2.4822, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.9016480595427964, | |
| "grad_norm": 0.2688996493816376, | |
| "learning_rate": 4.84451697945667e-06, | |
| "loss": 2.574, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9024986709197236, | |
| "grad_norm": 0.2651573419570923, | |
| "learning_rate": 4.7620173447834425e-06, | |
| "loss": 2.5219, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.9033492822966507, | |
| "grad_norm": 0.27622607350349426, | |
| "learning_rate": 4.680209083355713e-06, | |
| "loss": 2.6289, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.9041998936735779, | |
| "grad_norm": 0.27151504158973694, | |
| "learning_rate": 4.5990927890545935e-06, | |
| "loss": 2.625, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.9050505050505051, | |
| "grad_norm": 0.2963936924934387, | |
| "learning_rate": 4.518669050737989e-06, | |
| "loss": 2.4124, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.9059011164274322, | |
| "grad_norm": 0.2703079879283905, | |
| "learning_rate": 4.438938452236219e-06, | |
| "loss": 2.482, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9067517278043594, | |
| "grad_norm": 0.25973039865493774, | |
| "learning_rate": 4.359901572347758e-06, | |
| "loss": 2.6443, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.9076023391812865, | |
| "grad_norm": 0.27102747559547424, | |
| "learning_rate": 4.281558984835143e-06, | |
| "loss": 2.6452, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.9084529505582137, | |
| "grad_norm": 0.2813124656677246, | |
| "learning_rate": 4.203911258420712e-06, | |
| "loss": 2.6012, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.9093035619351408, | |
| "grad_norm": 0.274431437253952, | |
| "learning_rate": 4.126958956782545e-06, | |
| "loss": 2.6887, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.910154173312068, | |
| "grad_norm": 0.2668261229991913, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 2.5839, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9110047846889953, | |
| "grad_norm": 0.2745119631290436, | |
| "learning_rate": 3.975142857301117e-06, | |
| "loss": 2.4371, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.9118553960659224, | |
| "grad_norm": 0.31567510962486267, | |
| "learning_rate": 3.900280161555881e-06, | |
| "loss": 2.4209, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.9127060074428496, | |
| "grad_norm": 0.28338539600372314, | |
| "learning_rate": 3.826115094774863e-06, | |
| "loss": 2.47, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.9135566188197767, | |
| "grad_norm": 0.3005933463573456, | |
| "learning_rate": 3.7526481953539915e-06, | |
| "loss": 2.5939, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.9144072301967039, | |
| "grad_norm": 0.31587204337120056, | |
| "learning_rate": 3.6798799966209497e-06, | |
| "loss": 2.6447, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.915257841573631, | |
| "grad_norm": 0.3279857635498047, | |
| "learning_rate": 3.607811026831176e-06, | |
| "loss": 2.5845, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.9161084529505582, | |
| "grad_norm": 0.3389846980571747, | |
| "learning_rate": 3.5364418091641373e-06, | |
| "loss": 2.7449, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.9169590643274854, | |
| "grad_norm": 0.3381056487560272, | |
| "learning_rate": 3.4657728617195295e-06, | |
| "loss": 2.407, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.9178096757044125, | |
| "grad_norm": 0.3725632429122925, | |
| "learning_rate": 3.3958046975134495e-06, | |
| "loss": 2.8884, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.9186602870813397, | |
| "grad_norm": 0.3629099726676941, | |
| "learning_rate": 3.32653782447474e-06, | |
| "loss": 2.6872, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9195108984582668, | |
| "grad_norm": 0.37425747513771057, | |
| "learning_rate": 3.25797274544124e-06, | |
| "loss": 2.659, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.920361509835194, | |
| "grad_norm": 0.3709467649459839, | |
| "learning_rate": 3.1901099581561845e-06, | |
| "loss": 2.5593, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.9212121212121213, | |
| "grad_norm": 0.37324512004852295, | |
| "learning_rate": 3.122949955264587e-06, | |
| "loss": 2.5461, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.9220627325890484, | |
| "grad_norm": 0.39722326397895813, | |
| "learning_rate": 3.0564932243095866e-06, | |
| "loss": 2.8043, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.9229133439659756, | |
| "grad_norm": 0.40863049030303955, | |
| "learning_rate": 2.9907402477290514e-06, | |
| "loss": 2.5307, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.9237639553429027, | |
| "grad_norm": 0.4645911157131195, | |
| "learning_rate": 2.9256915028519573e-06, | |
| "loss": 2.3838, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.9246145667198299, | |
| "grad_norm": 0.42394137382507324, | |
| "learning_rate": 2.8613474618949366e-06, | |
| "loss": 2.7368, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.925465178096757, | |
| "grad_norm": 0.44541221857070923, | |
| "learning_rate": 2.7977085919589254e-06, | |
| "loss": 2.6973, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.9263157894736842, | |
| "grad_norm": 0.45734384655952454, | |
| "learning_rate": 2.7347753550256872e-06, | |
| "loss": 2.8454, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.9271664008506114, | |
| "grad_norm": 0.4581248164176941, | |
| "learning_rate": 2.672548207954495e-06, | |
| "loss": 2.7356, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9280170122275385, | |
| "grad_norm": 0.45908254384994507, | |
| "learning_rate": 2.6110276024788214e-06, | |
| "loss": 2.5562, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.9288676236044657, | |
| "grad_norm": 0.4975565969944, | |
| "learning_rate": 2.550213985203076e-06, | |
| "loss": 2.8292, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.9297182349813928, | |
| "grad_norm": 0.5445611476898193, | |
| "learning_rate": 2.4901077975992838e-06, | |
| "loss": 2.9136, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.93056884635832, | |
| "grad_norm": 0.5516471862792969, | |
| "learning_rate": 2.4307094760039785e-06, | |
| "loss": 3.0927, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.9314194577352473, | |
| "grad_norm": 0.5213038921356201, | |
| "learning_rate": 2.3720194516149818e-06, | |
| "loss": 2.705, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.9322700691121744, | |
| "grad_norm": 0.5873907208442688, | |
| "learning_rate": 2.3140381504882737e-06, | |
| "loss": 2.8415, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.9331206804891016, | |
| "grad_norm": 0.6056773662567139, | |
| "learning_rate": 2.2567659935349372e-06, | |
| "loss": 2.6933, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.9339712918660287, | |
| "grad_norm": 0.6297914981842041, | |
| "learning_rate": 2.200203396517997e-06, | |
| "loss": 2.8197, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.9348219032429559, | |
| "grad_norm": 0.6759224534034729, | |
| "learning_rate": 2.144350770049597e-06, | |
| "loss": 2.7496, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.935672514619883, | |
| "grad_norm": 0.7379116415977478, | |
| "learning_rate": 2.0892085195878154e-06, | |
| "loss": 2.926, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9365231259968102, | |
| "grad_norm": 0.25760582089424133, | |
| "learning_rate": 2.034777045433811e-06, | |
| "loss": 2.3794, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.9373737373737374, | |
| "grad_norm": 0.2775615155696869, | |
| "learning_rate": 1.9810567427289595e-06, | |
| "loss": 2.5169, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.9382243487506645, | |
| "grad_norm": 0.26485174894332886, | |
| "learning_rate": 1.92804800145191e-06, | |
| "loss": 2.4342, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.9390749601275917, | |
| "grad_norm": 0.24917316436767578, | |
| "learning_rate": 1.8757512064157656e-06, | |
| "loss": 2.4177, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.9399255715045188, | |
| "grad_norm": 0.25125351548194885, | |
| "learning_rate": 1.8241667372653316e-06, | |
| "loss": 2.2352, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.940776182881446, | |
| "grad_norm": 0.2972431182861328, | |
| "learning_rate": 1.7732949684743594e-06, | |
| "loss": 2.4803, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.9416267942583733, | |
| "grad_norm": 0.2540951073169708, | |
| "learning_rate": 1.7231362693427288e-06, | |
| "loss": 2.5117, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.9424774056353004, | |
| "grad_norm": 0.26120516657829285, | |
| "learning_rate": 1.6736910039939157e-06, | |
| "loss": 2.3744, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.9433280170122276, | |
| "grad_norm": 0.2617899477481842, | |
| "learning_rate": 1.62495953137225e-06, | |
| "loss": 2.4657, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.9441786283891547, | |
| "grad_norm": 0.29227423667907715, | |
| "learning_rate": 1.576942205240317e-06, | |
| "loss": 2.4297, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9450292397660819, | |
| "grad_norm": 0.2886480689048767, | |
| "learning_rate": 1.5296393741764391e-06, | |
| "loss": 2.5686, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.945879851143009, | |
| "grad_norm": 0.27778005599975586, | |
| "learning_rate": 1.4830513815720759e-06, | |
| "loss": 2.5925, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.9467304625199362, | |
| "grad_norm": 0.2743171155452728, | |
| "learning_rate": 1.4371785656294046e-06, | |
| "loss": 2.6989, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.9475810738968634, | |
| "grad_norm": 0.27193981409072876, | |
| "learning_rate": 1.392021259358811e-06, | |
| "loss": 2.457, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.9484316852737905, | |
| "grad_norm": 0.2713122069835663, | |
| "learning_rate": 1.3475797905764809e-06, | |
| "loss": 2.6029, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.9492822966507177, | |
| "grad_norm": 0.25593551993370056, | |
| "learning_rate": 1.303854481902067e-06, | |
| "loss": 2.6974, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.9501329080276448, | |
| "grad_norm": 0.26953616738319397, | |
| "learning_rate": 1.2608456507562705e-06, | |
| "loss": 2.6234, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.950983519404572, | |
| "grad_norm": 0.28958868980407715, | |
| "learning_rate": 1.2185536093585747e-06, | |
| "loss": 2.6315, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.9518341307814993, | |
| "grad_norm": 0.263473242521286, | |
| "learning_rate": 1.1769786647250147e-06, | |
| "loss": 2.5595, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.9526847421584264, | |
| "grad_norm": 0.26081758737564087, | |
| "learning_rate": 1.1361211186658894e-06, | |
| "loss": 2.4885, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9535353535353536, | |
| "grad_norm": 0.275305837392807, | |
| "learning_rate": 1.0959812677835968e-06, | |
| "loss": 2.794, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.9543859649122807, | |
| "grad_norm": 0.28377169370651245, | |
| "learning_rate": 1.0565594034704918e-06, | |
| "loss": 2.7399, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.9552365762892079, | |
| "grad_norm": 0.282060831785202, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 2.4815, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.956087187666135, | |
| "grad_norm": 0.2975423038005829, | |
| "learning_rate": 9.798707740582447e-07, | |
| "loss": 2.5839, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "grad_norm": 0.3122364580631256, | |
| "learning_rate": 9.42604565674654e-07, | |
| "loss": 2.6381, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.9577884104199894, | |
| "grad_norm": 0.3378431797027588, | |
| "learning_rate": 9.060574572873237e-07, | |
| "loss": 2.5897, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.9586390217969165, | |
| "grad_norm": 0.33718162775039673, | |
| "learning_rate": 8.702297142073379e-07, | |
| "loss": 2.689, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 0.9594896331738437, | |
| "grad_norm": 0.34144076704978943, | |
| "learning_rate": 8.351215965235915e-07, | |
| "loss": 2.4692, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.9603402445507708, | |
| "grad_norm": 0.3324032425880432, | |
| "learning_rate": 8.007333591009358e-07, | |
| "loss": 2.653, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 0.961190855927698, | |
| "grad_norm": 0.3601853847503662, | |
| "learning_rate": 7.670652515782917e-07, | |
| "loss": 2.5995, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.9620414673046253, | |
| "grad_norm": 0.3701609969139099, | |
| "learning_rate": 7.341175183668503e-07, | |
| "loss": 2.6668, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 0.9628920786815524, | |
| "grad_norm": 0.37852516770362854, | |
| "learning_rate": 7.018903986483083e-07, | |
| "loss": 2.7223, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 0.9637426900584796, | |
| "grad_norm": 0.391939640045166, | |
| "learning_rate": 6.703841263730914e-07, | |
| "loss": 2.7923, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 0.9645933014354067, | |
| "grad_norm": 0.38707369565963745, | |
| "learning_rate": 6.395989302587113e-07, | |
| "loss": 2.6445, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.9654439128123339, | |
| "grad_norm": 0.4049709141254425, | |
| "learning_rate": 6.095350337880667e-07, | |
| "loss": 2.7038, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.966294524189261, | |
| "grad_norm": 0.41350454092025757, | |
| "learning_rate": 5.801926552078563e-07, | |
| "loss": 2.6472, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 0.9671451355661882, | |
| "grad_norm": 0.42945706844329834, | |
| "learning_rate": 5.515720075269348e-07, | |
| "loss": 2.7802, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 0.9679957469431154, | |
| "grad_norm": 0.4451653063297272, | |
| "learning_rate": 5.236732985148374e-07, | |
| "loss": 2.7633, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 0.9688463583200425, | |
| "grad_norm": 0.46683037281036377, | |
| "learning_rate": 4.964967307002244e-07, | |
| "loss": 2.8216, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.4562697112560272, | |
| "learning_rate": 4.7004250136940543e-07, | |
| "loss": 2.7924, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9705475810738968, | |
| "grad_norm": 0.48352983593940735, | |
| "learning_rate": 4.443108025649623e-07, | |
| "loss": 2.5207, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 0.971398192450824, | |
| "grad_norm": 0.4667017459869385, | |
| "learning_rate": 4.193018210843058e-07, | |
| "loss": 2.6552, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 0.9722488038277513, | |
| "grad_norm": 0.49360570311546326, | |
| "learning_rate": 3.950157384783104e-07, | |
| "loss": 2.7523, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 0.9730994152046784, | |
| "grad_norm": 0.526669442653656, | |
| "learning_rate": 3.714527310500371e-07, | |
| "loss": 2.7332, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 0.9739500265816056, | |
| "grad_norm": 0.5592300891876221, | |
| "learning_rate": 3.486129698534457e-07, | |
| "loss": 2.7264, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.9748006379585327, | |
| "grad_norm": 0.5670251250267029, | |
| "learning_rate": 3.264966206921294e-07, | |
| "loss": 2.8538, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.9756512493354599, | |
| "grad_norm": 0.5744861960411072, | |
| "learning_rate": 3.0510384411812644e-07, | |
| "loss": 2.6884, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 0.976501860712387, | |
| "grad_norm": 0.6117346882820129, | |
| "learning_rate": 2.844347954307325e-07, | |
| "loss": 2.8192, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 0.9773524720893142, | |
| "grad_norm": 0.7534440159797668, | |
| "learning_rate": 2.644896246754236e-07, | |
| "loss": 2.7148, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 0.9782030834662414, | |
| "grad_norm": 0.7108234167098999, | |
| "learning_rate": 2.452684766427349e-07, | |
| "loss": 2.8751, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9790536948431685, | |
| "grad_norm": 0.264021635055542, | |
| "learning_rate": 2.2677149086718364e-07, | |
| "loss": 2.6035, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 0.9799043062200957, | |
| "grad_norm": 0.2746794819831848, | |
| "learning_rate": 2.0899880162630336e-07, | |
| "loss": 2.5935, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.9807549175970228, | |
| "grad_norm": 0.2742723226547241, | |
| "learning_rate": 1.9195053793964468e-07, | |
| "loss": 2.4199, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 0.98160552897395, | |
| "grad_norm": 0.2900630235671997, | |
| "learning_rate": 1.7562682356786487e-07, | |
| "loss": 2.5906, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.26273003220558167, | |
| "learning_rate": 1.6002777701175086e-07, | |
| "loss": 2.4287, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.9833067517278044, | |
| "grad_norm": 0.28278976678848267, | |
| "learning_rate": 1.451535115114866e-07, | |
| "loss": 2.6875, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 0.9841573631047316, | |
| "grad_norm": 0.2644931674003601, | |
| "learning_rate": 1.310041350457092e-07, | |
| "loss": 2.5806, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 0.9850079744816587, | |
| "grad_norm": 0.26039785146713257, | |
| "learning_rate": 1.1757975033078739e-07, | |
| "loss": 2.2813, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.9858585858585859, | |
| "grad_norm": 0.27614063024520874, | |
| "learning_rate": 1.0488045482008879e-07, | |
| "loss": 2.6836, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 0.986709197235513, | |
| "grad_norm": 0.27048394083976746, | |
| "learning_rate": 9.29063407032249e-08, | |
| "loss": 2.5862, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9875598086124402, | |
| "grad_norm": 0.28808289766311646, | |
| "learning_rate": 8.16574949054072e-08, | |
| "loss": 2.765, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 0.9884104199893674, | |
| "grad_norm": 0.323574423789978, | |
| "learning_rate": 7.113399908681429e-08, | |
| "loss": 2.6432, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 0.9892610313662945, | |
| "grad_norm": 0.3291761875152588, | |
| "learning_rate": 6.133592964201463e-08, | |
| "loss": 2.6037, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 0.9901116427432217, | |
| "grad_norm": 0.345300555229187, | |
| "learning_rate": 5.226335769936697e-08, | |
| "loss": 2.6076, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.9909622541201488, | |
| "grad_norm": 0.37476396560668945, | |
| "learning_rate": 4.391634912056519e-08, | |
| "loss": 2.8919, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.991812865497076, | |
| "grad_norm": 0.36915716528892517, | |
| "learning_rate": 3.629496450011649e-08, | |
| "loss": 2.5514, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 0.9926634768740031, | |
| "grad_norm": 0.3973197937011719, | |
| "learning_rate": 2.9399259164897274e-08, | |
| "loss": 2.8279, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 0.9935140882509303, | |
| "grad_norm": 0.42460885643959045, | |
| "learning_rate": 2.322928317378681e-08, | |
| "loss": 2.5969, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 0.9943646996278576, | |
| "grad_norm": 0.4290263056755066, | |
| "learning_rate": 1.778508131728973e-08, | |
| "loss": 2.7255, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 0.9952153110047847, | |
| "grad_norm": 0.4642498195171356, | |
| "learning_rate": 1.3066693117191886e-08, | |
| "loss": 2.7869, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9960659223817119, | |
| "grad_norm": 0.5059999823570251, | |
| "learning_rate": 9.074152826271665e-09, | |
| "loss": 2.582, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 0.996916533758639, | |
| "grad_norm": 0.49888238310813904, | |
| "learning_rate": 5.807489428111268e-09, | |
| "loss": 2.8126, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 0.9977671451355662, | |
| "grad_norm": 0.5143745541572571, | |
| "learning_rate": 3.2667266368080484e-09, | |
| "loss": 2.7024, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 0.9986177565124934, | |
| "grad_norm": 0.5905907154083252, | |
| "learning_rate": 1.4518828968523857e-09, | |
| "loss": 2.7198, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 0.9994683678894205, | |
| "grad_norm": 0.6912564635276794, | |
| "learning_rate": 3.629713829500503e-10, | |
| "loss": 2.6901, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.0006379585326954, | |
| "grad_norm": 1.383741021156311, | |
| "learning_rate": 0.0, | |
| "loss": 4.1919, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.0006379585326954, | |
| "eval_loss": 2.6547293663024902, | |
| "eval_runtime": 80.6568, | |
| "eval_samples_per_second": 12.274, | |
| "eval_steps_per_second": 6.137, | |
| "step": 1176 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1176, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 294, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.199442414922629e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |