| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.000839118163094, | |
| "eval_steps": 500, | |
| "global_step": 410, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024410710199099855, | |
| "grad_norm": 8.533896446228027, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1655, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004882142039819971, | |
| "grad_norm": 8.988560676574707, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.2516, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0073232130597299565, | |
| "grad_norm": 7.550627708435059, | |
| "learning_rate": 6e-06, | |
| "loss": 1.1895, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009764284079639942, | |
| "grad_norm": 3.6377415657043457, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.0982, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012205355099549928, | |
| "grad_norm": 3.964740753173828, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0622, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014646426119459913, | |
| "grad_norm": 4.8016157150268555, | |
| "learning_rate": 9.999962669988608e-06, | |
| "loss": 1.0653, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.017087497139369898, | |
| "grad_norm": 2.9538488388061523, | |
| "learning_rate": 9.999850680511844e-06, | |
| "loss": 1.026, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.019528568159279884, | |
| "grad_norm": 2.869965076446533, | |
| "learning_rate": 9.999664033241933e-06, | |
| "loss": 1.0349, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02196963917918987, | |
| "grad_norm": 1.8026058673858643, | |
| "learning_rate": 9.999402730965894e-06, | |
| "loss": 1.0421, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.024410710199099857, | |
| "grad_norm": 1.075210452079773, | |
| "learning_rate": 9.999066777585496e-06, | |
| "loss": 1.0008, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02685178121900984, | |
| "grad_norm": 1.4493818283081055, | |
| "learning_rate": 9.998656178117193e-06, | |
| "loss": 0.9347, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.029292852238919826, | |
| "grad_norm": 1.2218502759933472, | |
| "learning_rate": 9.99817093869206e-06, | |
| "loss": 0.9537, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03173392325882981, | |
| "grad_norm": 1.0389800071716309, | |
| "learning_rate": 9.997611066555694e-06, | |
| "loss": 0.9458, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.034174994278739795, | |
| "grad_norm": 0.959168016910553, | |
| "learning_rate": 9.99697657006811e-06, | |
| "loss": 0.9622, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03661606529864978, | |
| "grad_norm": 1.0173426866531372, | |
| "learning_rate": 9.99626745870361e-06, | |
| "loss": 0.9594, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03905713631855977, | |
| "grad_norm": 0.9893942475318909, | |
| "learning_rate": 9.995483743050649e-06, | |
| "loss": 0.9128, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.041498207338469754, | |
| "grad_norm": 0.9174278974533081, | |
| "learning_rate": 9.99462543481167e-06, | |
| "loss": 0.9108, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04393927835837974, | |
| "grad_norm": 0.8355342745780945, | |
| "learning_rate": 9.993692546802943e-06, | |
| "loss": 0.9341, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04638034937828973, | |
| "grad_norm": 0.9482454061508179, | |
| "learning_rate": 9.992685092954347e-06, | |
| "loss": 0.8488, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.048821420398199714, | |
| "grad_norm": 0.8152992129325867, | |
| "learning_rate": 9.991603088309195e-06, | |
| "loss": 0.9388, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05126249141810969, | |
| "grad_norm": 0.7824520468711853, | |
| "learning_rate": 9.990446549023977e-06, | |
| "loss": 0.917, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05370356243801968, | |
| "grad_norm": 0.8396065831184387, | |
| "learning_rate": 9.989215492368152e-06, | |
| "loss": 0.9043, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.056144633457929666, | |
| "grad_norm": 0.7503563761711121, | |
| "learning_rate": 9.98790993672386e-06, | |
| "loss": 0.9368, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05858570447783965, | |
| "grad_norm": 0.846466600894928, | |
| "learning_rate": 9.98652990158566e-06, | |
| "loss": 0.8641, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06102677549774964, | |
| "grad_norm": 0.8216990232467651, | |
| "learning_rate": 9.985075407560247e-06, | |
| "loss": 0.8744, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06346784651765962, | |
| "grad_norm": 0.7758781313896179, | |
| "learning_rate": 9.983546476366133e-06, | |
| "loss": 0.8722, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06590891753756961, | |
| "grad_norm": 0.8065202236175537, | |
| "learning_rate": 9.981943130833323e-06, | |
| "loss": 0.8582, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06834998855747959, | |
| "grad_norm": 0.79361891746521, | |
| "learning_rate": 9.980265394902982e-06, | |
| "loss": 0.8549, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07079105957738958, | |
| "grad_norm": 0.7769683003425598, | |
| "learning_rate": 9.978513293627068e-06, | |
| "loss": 0.8801, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07323213059729956, | |
| "grad_norm": 0.7662413120269775, | |
| "learning_rate": 9.976686853167967e-06, | |
| "loss": 0.849, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07567320161720956, | |
| "grad_norm": 0.7053027153015137, | |
| "learning_rate": 9.974786100798098e-06, | |
| "loss": 0.8925, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07811427263711954, | |
| "grad_norm": 0.7407605051994324, | |
| "learning_rate": 9.9728110648995e-06, | |
| "loss": 0.8623, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08055534365702952, | |
| "grad_norm": 0.7798149585723877, | |
| "learning_rate": 9.970761774963421e-06, | |
| "loss": 0.8711, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08299641467693951, | |
| "grad_norm": 0.7310554385185242, | |
| "learning_rate": 9.968638261589866e-06, | |
| "loss": 0.9071, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08543748569684949, | |
| "grad_norm": 0.8006892204284668, | |
| "learning_rate": 9.966440556487149e-06, | |
| "loss": 0.9026, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08787855671675948, | |
| "grad_norm": 0.7774298787117004, | |
| "learning_rate": 9.96416869247141e-06, | |
| "loss": 0.8512, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09031962773666946, | |
| "grad_norm": 0.7737051844596863, | |
| "learning_rate": 9.961822703466131e-06, | |
| "loss": 0.8629, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09276069875657945, | |
| "grad_norm": 0.8388147950172424, | |
| "learning_rate": 9.959402624501636e-06, | |
| "loss": 0.803, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09520176977648943, | |
| "grad_norm": 0.7394818067550659, | |
| "learning_rate": 9.956908491714552e-06, | |
| "loss": 0.8768, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09764284079639943, | |
| "grad_norm": 0.8373251557350159, | |
| "learning_rate": 9.95434034234728e-06, | |
| "loss": 0.8604, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1000839118163094, | |
| "grad_norm": 0.7941448092460632, | |
| "learning_rate": 9.951698214747441e-06, | |
| "loss": 0.8397, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10252498283621939, | |
| "grad_norm": 0.7676767706871033, | |
| "learning_rate": 9.948982148367294e-06, | |
| "loss": 0.8434, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10496605385612938, | |
| "grad_norm": 0.7958892583847046, | |
| "learning_rate": 9.946192183763155e-06, | |
| "loss": 0.8503, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10740712487603936, | |
| "grad_norm": 0.793487012386322, | |
| "learning_rate": 9.943328362594788e-06, | |
| "loss": 0.8566, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10984819589594935, | |
| "grad_norm": 0.716295599937439, | |
| "learning_rate": 9.940390727624785e-06, | |
| "loss": 0.8128, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11228926691585933, | |
| "grad_norm": 0.7760279178619385, | |
| "learning_rate": 9.937379322717923e-06, | |
| "loss": 0.8409, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11473033793576932, | |
| "grad_norm": 0.8229836821556091, | |
| "learning_rate": 9.934294192840518e-06, | |
| "loss": 0.8429, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1171714089556793, | |
| "grad_norm": 0.6973395347595215, | |
| "learning_rate": 9.931135384059737e-06, | |
| "loss": 0.8542, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11961247997558928, | |
| "grad_norm": 0.7911590933799744, | |
| "learning_rate": 9.927902943542932e-06, | |
| "loss": 0.8554, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12205355099549928, | |
| "grad_norm": 0.6992570757865906, | |
| "learning_rate": 9.924596919556917e-06, | |
| "loss": 0.8706, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12449462201540926, | |
| "grad_norm": 0.7577567100524902, | |
| "learning_rate": 9.921217361467259e-06, | |
| "loss": 0.856, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12693569303531924, | |
| "grad_norm": 0.8022581934928894, | |
| "learning_rate": 9.917764319737533e-06, | |
| "loss": 0.8276, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12937676405522924, | |
| "grad_norm": 0.720230758190155, | |
| "learning_rate": 9.914237845928574e-06, | |
| "loss": 0.8613, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.13181783507513922, | |
| "grad_norm": 0.7254828214645386, | |
| "learning_rate": 9.910637992697707e-06, | |
| "loss": 0.8617, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1342589060950492, | |
| "grad_norm": 0.7254623174667358, | |
| "learning_rate": 9.906964813797955e-06, | |
| "loss": 0.8543, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13669997711495918, | |
| "grad_norm": 0.7306321859359741, | |
| "learning_rate": 9.903218364077242e-06, | |
| "loss": 0.8332, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13914104813486916, | |
| "grad_norm": 0.7202122211456299, | |
| "learning_rate": 9.899398699477573e-06, | |
| "loss": 0.8663, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14158211915477917, | |
| "grad_norm": 0.7067145109176636, | |
| "learning_rate": 9.895505877034198e-06, | |
| "loss": 0.8165, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.14402319017468915, | |
| "grad_norm": 0.7376930713653564, | |
| "learning_rate": 9.891539954874758e-06, | |
| "loss": 0.8267, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.14646426119459913, | |
| "grad_norm": 0.7250686883926392, | |
| "learning_rate": 9.887500992218421e-06, | |
| "loss": 0.8239, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1489053322145091, | |
| "grad_norm": 0.7254573106765747, | |
| "learning_rate": 9.883389049374998e-06, | |
| "loss": 0.8452, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1513464032344191, | |
| "grad_norm": 0.7461521029472351, | |
| "learning_rate": 9.879204187744036e-06, | |
| "loss": 0.803, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1537874742543291, | |
| "grad_norm": 0.7778986096382141, | |
| "learning_rate": 9.874946469813907e-06, | |
| "loss": 0.8287, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15622854527423907, | |
| "grad_norm": 0.7395936846733093, | |
| "learning_rate": 9.870615959160876e-06, | |
| "loss": 0.8781, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.15866961629414905, | |
| "grad_norm": 0.7308329343795776, | |
| "learning_rate": 9.866212720448149e-06, | |
| "loss": 0.807, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16111068731405903, | |
| "grad_norm": 0.7851212620735168, | |
| "learning_rate": 9.861736819424904e-06, | |
| "loss": 0.821, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.16355175833396904, | |
| "grad_norm": 0.7638505697250366, | |
| "learning_rate": 9.857188322925317e-06, | |
| "loss": 0.8273, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.16599282935387902, | |
| "grad_norm": 0.7750548720359802, | |
| "learning_rate": 9.852567298867557e-06, | |
| "loss": 0.8523, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.168433900373789, | |
| "grad_norm": 0.7466771602630615, | |
| "learning_rate": 9.84787381625278e-06, | |
| "loss": 0.8415, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.17087497139369898, | |
| "grad_norm": 0.6956301331520081, | |
| "learning_rate": 9.843107945164086e-06, | |
| "loss": 0.8206, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17331604241360898, | |
| "grad_norm": 0.7392652630805969, | |
| "learning_rate": 9.838269756765483e-06, | |
| "loss": 0.8098, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17575711343351896, | |
| "grad_norm": 0.7311574220657349, | |
| "learning_rate": 9.833359323300827e-06, | |
| "loss": 0.8116, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17819818445342894, | |
| "grad_norm": 0.6983757615089417, | |
| "learning_rate": 9.82837671809273e-06, | |
| "loss": 0.8436, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.18063925547333892, | |
| "grad_norm": 0.7569893598556519, | |
| "learning_rate": 9.823322015541474e-06, | |
| "loss": 0.8058, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1830803264932489, | |
| "grad_norm": 0.7439902424812317, | |
| "learning_rate": 9.818195291123903e-06, | |
| "loss": 0.8424, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1855213975131589, | |
| "grad_norm": 0.7790477275848389, | |
| "learning_rate": 9.81299662139229e-06, | |
| "loss": 0.8483, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1879624685330689, | |
| "grad_norm": 0.7717331051826477, | |
| "learning_rate": 9.807726083973192e-06, | |
| "loss": 0.8214, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.19040353955297887, | |
| "grad_norm": 0.7872374057769775, | |
| "learning_rate": 9.8023837575663e-06, | |
| "loss": 0.7938, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.19284461057288885, | |
| "grad_norm": 0.8018149137496948, | |
| "learning_rate": 9.796969721943257e-06, | |
| "loss": 0.802, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.19528568159279885, | |
| "grad_norm": 0.709600031375885, | |
| "learning_rate": 9.791484057946465e-06, | |
| "loss": 0.7944, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19772675261270883, | |
| "grad_norm": 0.8216169476509094, | |
| "learning_rate": 9.785926847487885e-06, | |
| "loss": 0.8181, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2001678236326188, | |
| "grad_norm": 0.7138919830322266, | |
| "learning_rate": 9.780298173547811e-06, | |
| "loss": 0.8043, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2026088946525288, | |
| "grad_norm": 0.7637642621994019, | |
| "learning_rate": 9.774598120173625e-06, | |
| "loss": 0.8034, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.20504996567243877, | |
| "grad_norm": 0.7272418141365051, | |
| "learning_rate": 9.76882677247855e-06, | |
| "loss": 0.8271, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.20749103669234878, | |
| "grad_norm": 0.7340764999389648, | |
| "learning_rate": 9.762984216640378e-06, | |
| "loss": 0.8508, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20993210771225876, | |
| "grad_norm": 0.7231638431549072, | |
| "learning_rate": 9.75707053990018e-06, | |
| "loss": 0.823, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.21237317873216874, | |
| "grad_norm": 0.7670260071754456, | |
| "learning_rate": 9.751085830561e-06, | |
| "loss": 0.8595, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.21481424975207872, | |
| "grad_norm": 0.7142215371131897, | |
| "learning_rate": 9.74503017798655e-06, | |
| "loss": 0.8325, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2172553207719887, | |
| "grad_norm": 0.7884289026260376, | |
| "learning_rate": 9.738903672599858e-06, | |
| "loss": 0.7751, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2196963917918987, | |
| "grad_norm": 0.7771654725074768, | |
| "learning_rate": 9.732706405881931e-06, | |
| "loss": 0.7827, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22213746281180868, | |
| "grad_norm": 0.7293388247489929, | |
| "learning_rate": 9.726438470370385e-06, | |
| "loss": 0.8724, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.22457853383171866, | |
| "grad_norm": 0.7578020095825195, | |
| "learning_rate": 9.720099959658062e-06, | |
| "loss": 0.8277, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.22701960485162864, | |
| "grad_norm": 0.7896732091903687, | |
| "learning_rate": 9.713690968391634e-06, | |
| "loss": 0.7769, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.22946067587153865, | |
| "grad_norm": 0.6877868175506592, | |
| "learning_rate": 9.707211592270183e-06, | |
| "loss": 0.7938, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.23190174689144863, | |
| "grad_norm": 0.8047687411308289, | |
| "learning_rate": 9.700661928043787e-06, | |
| "loss": 0.7735, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2343428179113586, | |
| "grad_norm": 0.7561459541320801, | |
| "learning_rate": 9.69404207351206e-06, | |
| "loss": 0.8079, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2367838889312686, | |
| "grad_norm": 0.7163955569267273, | |
| "learning_rate": 9.687352127522703e-06, | |
| "loss": 0.8042, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.23922495995117857, | |
| "grad_norm": 0.7289466857910156, | |
| "learning_rate": 9.680592189970015e-06, | |
| "loss": 0.8449, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.24166603097108857, | |
| "grad_norm": 0.6951574087142944, | |
| "learning_rate": 9.673762361793418e-06, | |
| "loss": 0.7988, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.24410710199099855, | |
| "grad_norm": 0.7552266716957092, | |
| "learning_rate": 9.666862744975938e-06, | |
| "loss": 0.8323, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24654817301090853, | |
| "grad_norm": 0.7086972594261169, | |
| "learning_rate": 9.659893442542683e-06, | |
| "loss": 0.8567, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2489892440308185, | |
| "grad_norm": 0.7231544852256775, | |
| "learning_rate": 9.652854558559309e-06, | |
| "loss": 0.8265, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2514303150507285, | |
| "grad_norm": 0.7094722986221313, | |
| "learning_rate": 9.645746198130462e-06, | |
| "loss": 0.7803, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.25387138607063847, | |
| "grad_norm": 0.6969436407089233, | |
| "learning_rate": 9.638568467398215e-06, | |
| "loss": 0.804, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.25631245709054845, | |
| "grad_norm": 0.7204388380050659, | |
| "learning_rate": 9.631321473540476e-06, | |
| "loss": 0.787, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2587535281104585, | |
| "grad_norm": 0.6980841159820557, | |
| "learning_rate": 9.62400532476939e-06, | |
| "loss": 0.8294, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.26119459913036847, | |
| "grad_norm": 0.6793758273124695, | |
| "learning_rate": 9.61662013032972e-06, | |
| "loss": 0.7739, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.26363567015027844, | |
| "grad_norm": 0.7096854448318481, | |
| "learning_rate": 9.60916600049723e-06, | |
| "loss": 0.8035, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2660767411701884, | |
| "grad_norm": 0.6875160932540894, | |
| "learning_rate": 9.601643046577014e-06, | |
| "loss": 0.8567, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2685178121900984, | |
| "grad_norm": 0.7122709155082703, | |
| "learning_rate": 9.59405138090186e-06, | |
| "loss": 0.8153, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2709588832100084, | |
| "grad_norm": 0.695655882358551, | |
| "learning_rate": 9.586391116830549e-06, | |
| "loss": 0.7813, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.27339995422991836, | |
| "grad_norm": 0.674659788608551, | |
| "learning_rate": 9.578662368746183e-06, | |
| "loss": 0.8802, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.27584102524982834, | |
| "grad_norm": 0.7121911644935608, | |
| "learning_rate": 9.570865252054462e-06, | |
| "loss": 0.8017, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2782820962697383, | |
| "grad_norm": 0.7068195939064026, | |
| "learning_rate": 9.562999883181968e-06, | |
| "loss": 0.7817, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.28072316728964836, | |
| "grad_norm": 0.6847429275512695, | |
| "learning_rate": 9.555066379574423e-06, | |
| "loss": 0.801, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.28316423830955834, | |
| "grad_norm": 0.743248462677002, | |
| "learning_rate": 9.547064859694943e-06, | |
| "loss": 0.7978, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2856053093294683, | |
| "grad_norm": 0.7640885710716248, | |
| "learning_rate": 9.538995443022256e-06, | |
| "loss": 0.7913, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2880463803493783, | |
| "grad_norm": 0.7139798402786255, | |
| "learning_rate": 9.530858250048933e-06, | |
| "loss": 0.7994, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2904874513692883, | |
| "grad_norm": 0.7640753388404846, | |
| "learning_rate": 9.52265340227957e-06, | |
| "loss": 0.7946, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.29292852238919825, | |
| "grad_norm": 0.7454321980476379, | |
| "learning_rate": 9.514381022228997e-06, | |
| "loss": 0.809, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.29536959340910823, | |
| "grad_norm": 0.6853974461555481, | |
| "learning_rate": 9.506041233420427e-06, | |
| "loss": 0.8013, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2978106644290182, | |
| "grad_norm": 0.723430335521698, | |
| "learning_rate": 9.497634160383627e-06, | |
| "loss": 0.7923, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3002517354489282, | |
| "grad_norm": 0.7062557935714722, | |
| "learning_rate": 9.489159928653047e-06, | |
| "loss": 0.7702, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3026928064688382, | |
| "grad_norm": 0.6789696216583252, | |
| "learning_rate": 9.480618664765956e-06, | |
| "loss": 0.7748, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3051338774887482, | |
| "grad_norm": 0.7581243515014648, | |
| "learning_rate": 9.472010496260545e-06, | |
| "loss": 0.771, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3075749485086582, | |
| "grad_norm": 0.7822269201278687, | |
| "learning_rate": 9.463335551674024e-06, | |
| "loss": 0.8, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.31001601952856817, | |
| "grad_norm": 0.7157217264175415, | |
| "learning_rate": 9.454593960540709e-06, | |
| "loss": 0.7883, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.31245709054847814, | |
| "grad_norm": 0.7614567875862122, | |
| "learning_rate": 9.445785853390074e-06, | |
| "loss": 0.7929, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3148981615683881, | |
| "grad_norm": 0.7470414042472839, | |
| "learning_rate": 9.436911361744817e-06, | |
| "loss": 0.7826, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3173392325882981, | |
| "grad_norm": 0.7033482193946838, | |
| "learning_rate": 9.427970618118888e-06, | |
| "loss": 0.8359, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3197803036082081, | |
| "grad_norm": 0.7030816674232483, | |
| "learning_rate": 9.418963756015511e-06, | |
| "loss": 0.7966, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.32222137462811806, | |
| "grad_norm": 0.7050835490226746, | |
| "learning_rate": 9.409890909925191e-06, | |
| "loss": 0.7852, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3246624456480281, | |
| "grad_norm": 0.7047673463821411, | |
| "learning_rate": 9.400752215323712e-06, | |
| "loss": 0.8134, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3271035166679381, | |
| "grad_norm": 0.6739450693130493, | |
| "learning_rate": 9.391547808670097e-06, | |
| "loss": 0.8186, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.32954458768784806, | |
| "grad_norm": 0.7166461944580078, | |
| "learning_rate": 9.38227782740459e-06, | |
| "loss": 0.8118, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.33198565870775804, | |
| "grad_norm": 0.6905531287193298, | |
| "learning_rate": 9.372942409946597e-06, | |
| "loss": 0.8092, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.334426729727668, | |
| "grad_norm": 0.7552813291549683, | |
| "learning_rate": 9.36354169569261e-06, | |
| "loss": 0.7405, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.336867800747578, | |
| "grad_norm": 0.6745990514755249, | |
| "learning_rate": 9.35407582501414e-06, | |
| "loss": 0.8397, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.339308871767488, | |
| "grad_norm": 0.7749987840652466, | |
| "learning_rate": 9.344544939255608e-06, | |
| "loss": 0.7979, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.34174994278739795, | |
| "grad_norm": 0.7859154939651489, | |
| "learning_rate": 9.334949180732245e-06, | |
| "loss": 0.8217, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.34419101380730793, | |
| "grad_norm": 0.7111227512359619, | |
| "learning_rate": 9.325288692727963e-06, | |
| "loss": 0.7692, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.34663208482721797, | |
| "grad_norm": 0.824995219707489, | |
| "learning_rate": 9.315563619493209e-06, | |
| "loss": 0.7989, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.34907315584712795, | |
| "grad_norm": 0.7707095742225647, | |
| "learning_rate": 9.305774106242825e-06, | |
| "loss": 0.8115, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3515142268670379, | |
| "grad_norm": 0.7036089301109314, | |
| "learning_rate": 9.295920299153863e-06, | |
| "loss": 0.8119, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3539552978869479, | |
| "grad_norm": 0.7585278153419495, | |
| "learning_rate": 9.286002345363418e-06, | |
| "loss": 0.7853, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3563963689068579, | |
| "grad_norm": 0.7351112961769104, | |
| "learning_rate": 9.276020392966423e-06, | |
| "loss": 0.7974, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.35883743992676786, | |
| "grad_norm": 0.7286148071289062, | |
| "learning_rate": 9.265974591013434e-06, | |
| "loss": 0.8044, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.36127851094667784, | |
| "grad_norm": 0.6930050253868103, | |
| "learning_rate": 9.25586508950841e-06, | |
| "loss": 0.8117, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3637195819665878, | |
| "grad_norm": 0.8765610456466675, | |
| "learning_rate": 9.24569203940648e-06, | |
| "loss": 0.7551, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3661606529864978, | |
| "grad_norm": 0.7214458584785461, | |
| "learning_rate": 9.235455592611667e-06, | |
| "loss": 0.7984, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.36860172400640784, | |
| "grad_norm": 0.7065439820289612, | |
| "learning_rate": 9.225155901974645e-06, | |
| "loss": 0.8106, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3710427950263178, | |
| "grad_norm": 0.7775700092315674, | |
| "learning_rate": 9.214793121290442e-06, | |
| "loss": 0.8211, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3734838660462278, | |
| "grad_norm": 0.7118616700172424, | |
| "learning_rate": 9.204367405296144e-06, | |
| "loss": 0.82, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.3759249370661378, | |
| "grad_norm": 0.7476733326911926, | |
| "learning_rate": 9.193878909668591e-06, | |
| "loss": 0.7584, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.37836600808604776, | |
| "grad_norm": 0.7488994002342224, | |
| "learning_rate": 9.183327791022048e-06, | |
| "loss": 0.7552, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.38080707910595774, | |
| "grad_norm": 0.7086935043334961, | |
| "learning_rate": 9.172714206905866e-06, | |
| "loss": 0.7993, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3832481501258677, | |
| "grad_norm": 0.7513390183448792, | |
| "learning_rate": 9.162038315802132e-06, | |
| "loss": 0.7684, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3856892211457777, | |
| "grad_norm": 0.6983102560043335, | |
| "learning_rate": 9.1513002771233e-06, | |
| "loss": 0.7904, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3881302921656877, | |
| "grad_norm": 0.6591006517410278, | |
| "learning_rate": 9.140500251209813e-06, | |
| "loss": 0.7357, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3905713631855977, | |
| "grad_norm": 0.7491998672485352, | |
| "learning_rate": 9.129638399327707e-06, | |
| "loss": 0.7964, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3930124342055077, | |
| "grad_norm": 0.7312127947807312, | |
| "learning_rate": 9.118714883666204e-06, | |
| "loss": 0.7706, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.39545350522541767, | |
| "grad_norm": 0.7120770215988159, | |
| "learning_rate": 9.107729867335287e-06, | |
| "loss": 0.8367, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.39789457624532765, | |
| "grad_norm": 0.735023021697998, | |
| "learning_rate": 9.096683514363275e-06, | |
| "loss": 0.7832, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4003356472652376, | |
| "grad_norm": 0.7334295511245728, | |
| "learning_rate": 9.085575989694358e-06, | |
| "loss": 0.7977, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4027767182851476, | |
| "grad_norm": 0.7482827305793762, | |
| "learning_rate": 9.074407459186144e-06, | |
| "loss": 0.868, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4052177893050576, | |
| "grad_norm": 0.7395485043525696, | |
| "learning_rate": 9.063178089607183e-06, | |
| "loss": 0.7676, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.40765886032496756, | |
| "grad_norm": 0.6970906257629395, | |
| "learning_rate": 9.051888048634471e-06, | |
| "loss": 0.762, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.41009993134487754, | |
| "grad_norm": 0.7200821042060852, | |
| "learning_rate": 9.040537504850954e-06, | |
| "loss": 0.8067, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4125410023647875, | |
| "grad_norm": 0.7742771506309509, | |
| "learning_rate": 9.029126627743003e-06, | |
| "loss": 0.7767, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.41498207338469756, | |
| "grad_norm": 0.7340243458747864, | |
| "learning_rate": 9.017655587697885e-06, | |
| "loss": 0.7816, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.41742314440460754, | |
| "grad_norm": 0.7570080161094666, | |
| "learning_rate": 9.006124556001223e-06, | |
| "loss": 0.8374, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4198642154245175, | |
| "grad_norm": 0.7807502150535583, | |
| "learning_rate": 8.994533704834435e-06, | |
| "loss": 0.7749, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4223052864444275, | |
| "grad_norm": 0.7137355208396912, | |
| "learning_rate": 8.982883207272164e-06, | |
| "loss": 0.7397, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.4247463574643375, | |
| "grad_norm": 0.7511448860168457, | |
| "learning_rate": 8.971173237279693e-06, | |
| "loss": 0.8006, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.42718742848424746, | |
| "grad_norm": 0.7791663408279419, | |
| "learning_rate": 8.959403969710346e-06, | |
| "loss": 0.7684, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.42962849950415744, | |
| "grad_norm": 0.7711341381072998, | |
| "learning_rate": 8.947575580302879e-06, | |
| "loss": 0.7905, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4320695705240674, | |
| "grad_norm": 0.7793801426887512, | |
| "learning_rate": 8.935688245678859e-06, | |
| "loss": 0.8121, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4345106415439774, | |
| "grad_norm": 0.7082055807113647, | |
| "learning_rate": 8.92374214334002e-06, | |
| "loss": 0.7657, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.43695171256388743, | |
| "grad_norm": 0.735462486743927, | |
| "learning_rate": 8.911737451665616e-06, | |
| "loss": 0.7833, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4393927835837974, | |
| "grad_norm": 0.7432037591934204, | |
| "learning_rate": 8.899674349909759e-06, | |
| "loss": 0.7645, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4418338546037074, | |
| "grad_norm": 0.7552315592765808, | |
| "learning_rate": 8.887553018198738e-06, | |
| "loss": 0.8018, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.44427492562361737, | |
| "grad_norm": 0.677143931388855, | |
| "learning_rate": 8.875373637528336e-06, | |
| "loss": 0.8029, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.44671599664352735, | |
| "grad_norm": 0.7790682911872864, | |
| "learning_rate": 8.863136389761115e-06, | |
| "loss": 0.792, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4491570676634373, | |
| "grad_norm": 0.735373854637146, | |
| "learning_rate": 8.85084145762372e-06, | |
| "loss": 0.78, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4515981386833473, | |
| "grad_norm": 0.7221420407295227, | |
| "learning_rate": 8.838489024704131e-06, | |
| "loss": 0.807, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4540392097032573, | |
| "grad_norm": 0.7021591067314148, | |
| "learning_rate": 8.826079275448934e-06, | |
| "loss": 0.7828, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.45648028072316726, | |
| "grad_norm": 0.7104141712188721, | |
| "learning_rate": 8.81361239516056e-06, | |
| "loss": 0.8051, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4589213517430773, | |
| "grad_norm": 0.749536395072937, | |
| "learning_rate": 8.801088569994523e-06, | |
| "loss": 0.7811, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4613624227629873, | |
| "grad_norm": 0.7570759654045105, | |
| "learning_rate": 8.788507986956639e-06, | |
| "loss": 0.8015, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.46380349378289726, | |
| "grad_norm": 0.6997769474983215, | |
| "learning_rate": 8.775870833900226e-06, | |
| "loss": 0.7816, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.46624456480280724, | |
| "grad_norm": 0.6764109134674072, | |
| "learning_rate": 8.763177299523318e-06, | |
| "loss": 0.7577, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4686856358227172, | |
| "grad_norm": 0.7811216115951538, | |
| "learning_rate": 8.750427573365825e-06, | |
| "loss": 0.7324, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4711267068426272, | |
| "grad_norm": 0.7098534107208252, | |
| "learning_rate": 8.737621845806715e-06, | |
| "loss": 0.7321, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4735677778625372, | |
| "grad_norm": 0.7705920934677124, | |
| "learning_rate": 8.724760308061172e-06, | |
| "loss": 0.7501, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.47600884888244716, | |
| "grad_norm": 0.7170778512954712, | |
| "learning_rate": 8.711843152177735e-06, | |
| "loss": 0.767, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.47844991990235713, | |
| "grad_norm": 0.7175964713096619, | |
| "learning_rate": 8.698870571035436e-06, | |
| "loss": 0.7592, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.48089099092226717, | |
| "grad_norm": 0.7901434898376465, | |
| "learning_rate": 8.685842758340912e-06, | |
| "loss": 0.7921, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.48333206194217715, | |
| "grad_norm": 0.7608402371406555, | |
| "learning_rate": 8.672759908625528e-06, | |
| "loss": 0.8617, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.48577313296208713, | |
| "grad_norm": 0.7593024373054504, | |
| "learning_rate": 8.65962221724245e-06, | |
| "loss": 0.7674, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.4882142039819971, | |
| "grad_norm": 0.7110275626182556, | |
| "learning_rate": 8.646429880363746e-06, | |
| "loss": 0.7521, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4906552750019071, | |
| "grad_norm": 0.7535459399223328, | |
| "learning_rate": 8.633183094977453e-06, | |
| "loss": 0.7296, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.49309634602181707, | |
| "grad_norm": 0.7531000971794128, | |
| "learning_rate": 8.61988205888463e-06, | |
| "loss": 0.7863, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.49553741704172705, | |
| "grad_norm": 0.7889319658279419, | |
| "learning_rate": 8.60652697069641e-06, | |
| "loss": 0.7784, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.497978488061637, | |
| "grad_norm": 0.6903645396232605, | |
| "learning_rate": 8.593118029831025e-06, | |
| "loss": 0.7954, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.500419559081547, | |
| "grad_norm": 0.7375295758247375, | |
| "learning_rate": 8.579655436510847e-06, | |
| "loss": 0.7764, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.502860630101457, | |
| "grad_norm": 0.7218457460403442, | |
| "learning_rate": 8.566139391759378e-06, | |
| "loss": 0.7852, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.505301701121367, | |
| "grad_norm": 0.7074956297874451, | |
| "learning_rate": 8.552570097398262e-06, | |
| "loss": 0.7824, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5077427721412769, | |
| "grad_norm": 0.6844367384910583, | |
| "learning_rate": 8.53894775604426e-06, | |
| "loss": 0.8005, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5101838431611869, | |
| "grad_norm": 0.7443989515304565, | |
| "learning_rate": 8.525272571106242e-06, | |
| "loss": 0.7761, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5126249141810969, | |
| "grad_norm": 0.7639645338058472, | |
| "learning_rate": 8.511544746782124e-06, | |
| "loss": 0.8032, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.515065985201007, | |
| "grad_norm": 0.699748158454895, | |
| "learning_rate": 8.497764488055848e-06, | |
| "loss": 0.7801, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.517507056220917, | |
| "grad_norm": 0.7058794498443604, | |
| "learning_rate": 8.483932000694295e-06, | |
| "loss": 0.7693, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.519948127240827, | |
| "grad_norm": 0.7830145359039307, | |
| "learning_rate": 8.470047491244232e-06, | |
| "loss": 0.7684, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5223891982607369, | |
| "grad_norm": 0.6766949892044067, | |
| "learning_rate": 8.456111167029219e-06, | |
| "loss": 0.8214, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5248302692806469, | |
| "grad_norm": 0.7066507339477539, | |
| "learning_rate": 8.442123236146509e-06, | |
| "loss": 0.7639, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5272713403005569, | |
| "grad_norm": 0.7286085486412048, | |
| "learning_rate": 8.42808390746395e-06, | |
| "loss": 0.7723, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5297124113204669, | |
| "grad_norm": 0.7587203979492188, | |
| "learning_rate": 8.413993390616865e-06, | |
| "loss": 0.8034, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5321534823403768, | |
| "grad_norm": 0.6527595520019531, | |
| "learning_rate": 8.399851896004914e-06, | |
| "loss": 0.7587, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5345945533602868, | |
| "grad_norm": 0.8271955251693726, | |
| "learning_rate": 8.385659634788959e-06, | |
| "loss": 0.7846, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5370356243801968, | |
| "grad_norm": 0.7351842522621155, | |
| "learning_rate": 8.371416818887907e-06, | |
| "loss": 0.8002, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5394766954001068, | |
| "grad_norm": 0.7915340065956116, | |
| "learning_rate": 8.357123660975553e-06, | |
| "loss": 0.7511, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5419177664200168, | |
| "grad_norm": 0.6955085396766663, | |
| "learning_rate": 8.342780374477396e-06, | |
| "loss": 0.7754, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5443588374399267, | |
| "grad_norm": 0.7368732690811157, | |
| "learning_rate": 8.328387173567453e-06, | |
| "loss": 0.7775, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5467999084598367, | |
| "grad_norm": 0.6908881068229675, | |
| "learning_rate": 8.313944273165068e-06, | |
| "loss": 0.7571, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5492409794797467, | |
| "grad_norm": 0.700554370880127, | |
| "learning_rate": 8.299451888931696e-06, | |
| "loss": 0.7714, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5516820504996567, | |
| "grad_norm": 0.7163404822349548, | |
| "learning_rate": 8.284910237267681e-06, | |
| "loss": 0.7767, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5541231215195667, | |
| "grad_norm": 0.7443628311157227, | |
| "learning_rate": 8.270319535309035e-06, | |
| "loss": 0.7709, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5565641925394766, | |
| "grad_norm": 0.691213071346283, | |
| "learning_rate": 8.255680000924184e-06, | |
| "loss": 0.7997, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5590052635593867, | |
| "grad_norm": 0.7387362718582153, | |
| "learning_rate": 8.240991852710724e-06, | |
| "loss": 0.7502, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5614463345792967, | |
| "grad_norm": 0.7051777243614197, | |
| "learning_rate": 8.22625530999215e-06, | |
| "loss": 0.7811, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5638874055992067, | |
| "grad_norm": 0.6610181331634521, | |
| "learning_rate": 8.211470592814586e-06, | |
| "loss": 0.7884, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5663284766191167, | |
| "grad_norm": 0.6575087904930115, | |
| "learning_rate": 8.196637921943496e-06, | |
| "loss": 0.7797, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5687695476390267, | |
| "grad_norm": 0.7363637685775757, | |
| "learning_rate": 8.181757518860387e-06, | |
| "loss": 0.7369, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5712106186589366, | |
| "grad_norm": 0.6955734491348267, | |
| "learning_rate": 8.166829605759507e-06, | |
| "loss": 0.7841, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5736516896788466, | |
| "grad_norm": 0.7019768357276917, | |
| "learning_rate": 8.151854405544526e-06, | |
| "loss": 0.7702, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5760927606987566, | |
| "grad_norm": 0.7041372656822205, | |
| "learning_rate": 8.136832141825197e-06, | |
| "loss": 0.7755, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5785338317186666, | |
| "grad_norm": 0.7138186693191528, | |
| "learning_rate": 8.12176303891403e-06, | |
| "loss": 0.7815, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5809749027385765, | |
| "grad_norm": 0.6682398319244385, | |
| "learning_rate": 8.106647321822943e-06, | |
| "loss": 0.7573, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5834159737584865, | |
| "grad_norm": 0.6600127816200256, | |
| "learning_rate": 8.091485216259886e-06, | |
| "loss": 0.7644, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5858570447783965, | |
| "grad_norm": 0.7157433032989502, | |
| "learning_rate": 8.076276948625495e-06, | |
| "loss": 0.7661, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5882981157983065, | |
| "grad_norm": 0.7180731892585754, | |
| "learning_rate": 8.061022746009687e-06, | |
| "loss": 0.756, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5907391868182165, | |
| "grad_norm": 0.6941264271736145, | |
| "learning_rate": 8.04572283618829e-06, | |
| "loss": 0.7501, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5931802578381264, | |
| "grad_norm": 0.7047881484031677, | |
| "learning_rate": 8.030377447619622e-06, | |
| "loss": 0.7564, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5956213288580364, | |
| "grad_norm": 0.6860742568969727, | |
| "learning_rate": 8.014986809441093e-06, | |
| "loss": 0.8048, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5980623998779464, | |
| "grad_norm": 0.6961259245872498, | |
| "learning_rate": 7.999551151465793e-06, | |
| "loss": 0.8085, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6005034708978564, | |
| "grad_norm": 0.6859815716743469, | |
| "learning_rate": 7.984070704179026e-06, | |
| "loss": 0.7911, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6029445419177665, | |
| "grad_norm": 0.739782452583313, | |
| "learning_rate": 7.968545698734908e-06, | |
| "loss": 0.7981, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6053856129376765, | |
| "grad_norm": 0.7173625230789185, | |
| "learning_rate": 7.952976366952888e-06, | |
| "loss": 0.7738, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6078266839575864, | |
| "grad_norm": 0.7094762921333313, | |
| "learning_rate": 7.9373629413143e-06, | |
| "loss": 0.7802, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6102677549774964, | |
| "grad_norm": 0.6974766254425049, | |
| "learning_rate": 7.921705654958886e-06, | |
| "loss": 0.7956, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6127088259974064, | |
| "grad_norm": 0.7235715389251709, | |
| "learning_rate": 7.906004741681321e-06, | |
| "loss": 0.7581, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6151498970173164, | |
| "grad_norm": 0.68167644739151, | |
| "learning_rate": 7.890260435927709e-06, | |
| "loss": 0.7746, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6175909680372264, | |
| "grad_norm": 0.6965702176094055, | |
| "learning_rate": 7.874472972792097e-06, | |
| "loss": 0.7638, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6200320390571363, | |
| "grad_norm": 0.7000617384910583, | |
| "learning_rate": 7.858642588012957e-06, | |
| "loss": 0.7252, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6224731100770463, | |
| "grad_norm": 0.6918544173240662, | |
| "learning_rate": 7.842769517969665e-06, | |
| "loss": 0.7867, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6249141810969563, | |
| "grad_norm": 0.7168439626693726, | |
| "learning_rate": 7.826853999678978e-06, | |
| "loss": 0.7349, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6273552521168663, | |
| "grad_norm": 0.6823673844337463, | |
| "learning_rate": 7.810896270791484e-06, | |
| "loss": 0.7749, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6297963231367762, | |
| "grad_norm": 0.7399064898490906, | |
| "learning_rate": 7.794896569588066e-06, | |
| "loss": 0.7886, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6322373941566862, | |
| "grad_norm": 0.6988884806632996, | |
| "learning_rate": 7.778855134976334e-06, | |
| "loss": 0.7329, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6346784651765962, | |
| "grad_norm": 0.7028211951255798, | |
| "learning_rate": 7.762772206487066e-06, | |
| "loss": 0.8214, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6371195361965062, | |
| "grad_norm": 0.7346593737602234, | |
| "learning_rate": 7.74664802427062e-06, | |
| "loss": 0.7626, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6395606072164162, | |
| "grad_norm": 0.7249420881271362, | |
| "learning_rate": 7.73048282909336e-06, | |
| "loss": 0.7617, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6420016782363261, | |
| "grad_norm": 0.7126630544662476, | |
| "learning_rate": 7.714276862334051e-06, | |
| "loss": 0.7599, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6444427492562361, | |
| "grad_norm": 0.7718750238418579, | |
| "learning_rate": 7.698030365980265e-06, | |
| "loss": 0.8228, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6468838202761461, | |
| "grad_norm": 0.7375472187995911, | |
| "learning_rate": 7.681743582624761e-06, | |
| "loss": 0.7458, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6493248912960562, | |
| "grad_norm": 0.7044516205787659, | |
| "learning_rate": 7.66541675546186e-06, | |
| "loss": 0.779, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6517659623159662, | |
| "grad_norm": 0.7249746322631836, | |
| "learning_rate": 7.64905012828382e-06, | |
| "loss": 0.7983, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6542070333358762, | |
| "grad_norm": 0.7117093801498413, | |
| "learning_rate": 7.632643945477195e-06, | |
| "loss": 0.7436, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6566481043557861, | |
| "grad_norm": 0.7090557217597961, | |
| "learning_rate": 7.616198452019176e-06, | |
| "loss": 0.7563, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6590891753756961, | |
| "grad_norm": 0.720168948173523, | |
| "learning_rate": 7.59971389347395e-06, | |
| "loss": 0.7487, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6615302463956061, | |
| "grad_norm": 0.6775338053703308, | |
| "learning_rate": 7.583190515989022e-06, | |
| "loss": 0.7708, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6639713174155161, | |
| "grad_norm": 0.711544394493103, | |
| "learning_rate": 7.566628566291537e-06, | |
| "loss": 0.7732, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.666412388435426, | |
| "grad_norm": 0.7197690606117249, | |
| "learning_rate": 7.550028291684603e-06, | |
| "loss": 0.7681, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.668853459455336, | |
| "grad_norm": 0.7002537250518799, | |
| "learning_rate": 7.5333899400435986e-06, | |
| "loss": 0.7414, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.671294530475246, | |
| "grad_norm": 0.7534189820289612, | |
| "learning_rate": 7.516713759812465e-06, | |
| "loss": 0.7756, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.673735601495156, | |
| "grad_norm": 0.6993451714515686, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7494, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.676176672515066, | |
| "grad_norm": 0.7160854339599609, | |
| "learning_rate": 7.483248910176144e-06, | |
| "loss": 0.7727, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.678617743534976, | |
| "grad_norm": 0.7340339422225952, | |
| "learning_rate": 7.466460740468246e-06, | |
| "loss": 0.7641, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6810588145548859, | |
| "grad_norm": 0.6893506050109863, | |
| "learning_rate": 7.44963574155733e-06, | |
| "loss": 0.7859, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6834998855747959, | |
| "grad_norm": 0.7197214365005493, | |
| "learning_rate": 7.432774164674359e-06, | |
| "loss": 0.7645, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6859409565947059, | |
| "grad_norm": 0.722647488117218, | |
| "learning_rate": 7.4158762615964744e-06, | |
| "loss": 0.7614, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6883820276146159, | |
| "grad_norm": 0.7112955451011658, | |
| "learning_rate": 7.398942284643242e-06, | |
| "loss": 0.7565, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6908230986345258, | |
| "grad_norm": 0.7392273545265198, | |
| "learning_rate": 7.381972486672886e-06, | |
| "loss": 0.7474, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6932641696544359, | |
| "grad_norm": 0.691473126411438, | |
| "learning_rate": 7.3649671210785024e-06, | |
| "loss": 0.7392, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6957052406743459, | |
| "grad_norm": 0.764784038066864, | |
| "learning_rate": 7.34792644178429e-06, | |
| "loss": 0.7341, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6981463116942559, | |
| "grad_norm": 0.7281628251075745, | |
| "learning_rate": 7.330850703241751e-06, | |
| "loss": 0.7804, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7005873827141659, | |
| "grad_norm": 0.843350887298584, | |
| "learning_rate": 7.313740160425887e-06, | |
| "loss": 0.7085, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7030284537340759, | |
| "grad_norm": 0.7312772870063782, | |
| "learning_rate": 7.296595068831406e-06, | |
| "loss": 0.7638, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7054695247539858, | |
| "grad_norm": 0.7479636073112488, | |
| "learning_rate": 7.279415684468893e-06, | |
| "loss": 0.7208, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7079105957738958, | |
| "grad_norm": 0.6945940256118774, | |
| "learning_rate": 7.262202263860989e-06, | |
| "loss": 0.7133, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7103516667938058, | |
| "grad_norm": 0.697964608669281, | |
| "learning_rate": 7.244955064038574e-06, | |
| "loss": 0.7478, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7127927378137158, | |
| "grad_norm": 0.7676611542701721, | |
| "learning_rate": 7.227674342536914e-06, | |
| "loss": 0.7778, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7152338088336258, | |
| "grad_norm": 0.6927245259284973, | |
| "learning_rate": 7.210360357391818e-06, | |
| "loss": 0.7041, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7176748798535357, | |
| "grad_norm": 0.7169522643089294, | |
| "learning_rate": 7.1930133671357915e-06, | |
| "loss": 0.7493, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7201159508734457, | |
| "grad_norm": 0.6862355470657349, | |
| "learning_rate": 7.175633630794176e-06, | |
| "loss": 0.7547, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7225570218933557, | |
| "grad_norm": 0.7003543376922607, | |
| "learning_rate": 7.1582214078812715e-06, | |
| "loss": 0.7677, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7249980929132657, | |
| "grad_norm": 0.6878992915153503, | |
| "learning_rate": 7.140776958396468e-06, | |
| "loss": 0.7663, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7274391639331756, | |
| "grad_norm": 0.6947048306465149, | |
| "learning_rate": 7.123300542820367e-06, | |
| "loss": 0.7514, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7298802349530856, | |
| "grad_norm": 0.6599522829055786, | |
| "learning_rate": 7.1057924221108856e-06, | |
| "loss": 0.7363, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7323213059729956, | |
| "grad_norm": 0.6951528787612915, | |
| "learning_rate": 7.08825285769936e-06, | |
| "loss": 0.7247, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7347623769929056, | |
| "grad_norm": 0.7057417035102844, | |
| "learning_rate": 7.0706821114866475e-06, | |
| "loss": 0.7829, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7372034480128157, | |
| "grad_norm": 0.7661730647087097, | |
| "learning_rate": 7.053080445839211e-06, | |
| "loss": 0.7233, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7396445190327257, | |
| "grad_norm": 0.6780954003334045, | |
| "learning_rate": 7.035448123585201e-06, | |
| "loss": 0.7549, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7420855900526356, | |
| "grad_norm": 0.7154073715209961, | |
| "learning_rate": 7.017785408010533e-06, | |
| "loss": 0.7593, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.7445266610725456, | |
| "grad_norm": 0.72113436460495, | |
| "learning_rate": 7.0000925628549595e-06, | |
| "loss": 0.8079, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7469677320924556, | |
| "grad_norm": 0.6903125643730164, | |
| "learning_rate": 6.982369852308124e-06, | |
| "loss": 0.7777, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7494088031123656, | |
| "grad_norm": 0.7365685701370239, | |
| "learning_rate": 6.964617541005617e-06, | |
| "loss": 0.7827, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7518498741322756, | |
| "grad_norm": 0.7428478002548218, | |
| "learning_rate": 6.946835894025037e-06, | |
| "loss": 0.7319, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7542909451521855, | |
| "grad_norm": 0.7224217653274536, | |
| "learning_rate": 6.929025176882016e-06, | |
| "loss": 0.7758, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7567320161720955, | |
| "grad_norm": 0.7415998578071594, | |
| "learning_rate": 6.911185655526263e-06, | |
| "loss": 0.7832, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7591730871920055, | |
| "grad_norm": 0.7322662472724915, | |
| "learning_rate": 6.893317596337592e-06, | |
| "loss": 0.7323, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7616141582119155, | |
| "grad_norm": 0.6983169913291931, | |
| "learning_rate": 6.875421266121946e-06, | |
| "loss": 0.7576, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7640552292318255, | |
| "grad_norm": 0.7458372116088867, | |
| "learning_rate": 6.857496932107407e-06, | |
| "loss": 0.7549, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7664963002517354, | |
| "grad_norm": 0.7724815011024475, | |
| "learning_rate": 6.839544861940214e-06, | |
| "loss": 0.7625, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7689373712716454, | |
| "grad_norm": 0.7590445280075073, | |
| "learning_rate": 6.821565323680759e-06, | |
| "loss": 0.7422, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7713784422915554, | |
| "grad_norm": 0.7100796699523926, | |
| "learning_rate": 6.80355858579959e-06, | |
| "loss": 0.748, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7738195133114654, | |
| "grad_norm": 0.6766054034233093, | |
| "learning_rate": 6.7855249171734e-06, | |
| "loss": 0.7487, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7762605843313753, | |
| "grad_norm": 0.7497526407241821, | |
| "learning_rate": 6.76746458708101e-06, | |
| "loss": 0.7584, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7787016553512853, | |
| "grad_norm": 0.6761816740036011, | |
| "learning_rate": 6.74937786519935e-06, | |
| "loss": 0.7332, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7811427263711954, | |
| "grad_norm": 0.6793827414512634, | |
| "learning_rate": 6.731265021599437e-06, | |
| "loss": 0.7387, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7835837973911054, | |
| "grad_norm": 0.7181971669197083, | |
| "learning_rate": 6.7131263267423305e-06, | |
| "loss": 0.7588, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7860248684110154, | |
| "grad_norm": 0.6722172498703003, | |
| "learning_rate": 6.6949620514751075e-06, | |
| "loss": 0.7264, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7884659394309254, | |
| "grad_norm": 0.6741105914115906, | |
| "learning_rate": 6.676772467026809e-06, | |
| "loss": 0.7806, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7909070104508353, | |
| "grad_norm": 0.6798011660575867, | |
| "learning_rate": 6.65855784500439e-06, | |
| "loss": 0.7281, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7933480814707453, | |
| "grad_norm": 0.6723977327346802, | |
| "learning_rate": 6.640318457388672e-06, | |
| "loss": 0.7358, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7957891524906553, | |
| "grad_norm": 0.6920611262321472, | |
| "learning_rate": 6.622054576530275e-06, | |
| "loss": 0.7754, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7982302235105653, | |
| "grad_norm": 0.7018612623214722, | |
| "learning_rate": 6.603766475145546e-06, | |
| "loss": 0.7714, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8006712945304753, | |
| "grad_norm": 0.7847645282745361, | |
| "learning_rate": 6.585454426312506e-06, | |
| "loss": 0.804, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8031123655503852, | |
| "grad_norm": 0.6560506820678711, | |
| "learning_rate": 6.5671187034667465e-06, | |
| "loss": 0.7768, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8055534365702952, | |
| "grad_norm": 0.720274031162262, | |
| "learning_rate": 6.548759580397363e-06, | |
| "loss": 0.7619, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8079945075902052, | |
| "grad_norm": 0.6853426694869995, | |
| "learning_rate": 6.53037733124287e-06, | |
| "loss": 0.7147, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8104355786101152, | |
| "grad_norm": 0.6448130011558533, | |
| "learning_rate": 6.511972230487091e-06, | |
| "loss": 0.7958, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8128766496300251, | |
| "grad_norm": 0.6405046582221985, | |
| "learning_rate": 6.4935445529550775e-06, | |
| "loss": 0.7659, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8153177206499351, | |
| "grad_norm": 0.7028204798698425, | |
| "learning_rate": 6.475094573808994e-06, | |
| "loss": 0.7609, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8177587916698451, | |
| "grad_norm": 0.7086704969406128, | |
| "learning_rate": 6.456622568544012e-06, | |
| "loss": 0.7719, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8201998626897551, | |
| "grad_norm": 0.6891460418701172, | |
| "learning_rate": 6.438128812984199e-06, | |
| "loss": 0.7667, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8226409337096651, | |
| "grad_norm": 0.7402656674385071, | |
| "learning_rate": 6.419613583278395e-06, | |
| "loss": 0.7833, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.825082004729575, | |
| "grad_norm": 0.7249888777732849, | |
| "learning_rate": 6.401077155896098e-06, | |
| "loss": 0.7031, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8275230757494851, | |
| "grad_norm": 0.6789460182189941, | |
| "learning_rate": 6.3825198076233255e-06, | |
| "loss": 0.7739, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8299641467693951, | |
| "grad_norm": 0.6832931041717529, | |
| "learning_rate": 6.363941815558484e-06, | |
| "loss": 0.7242, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8324052177893051, | |
| "grad_norm": 0.64850252866745, | |
| "learning_rate": 6.345343457108238e-06, | |
| "loss": 0.7378, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8348462888092151, | |
| "grad_norm": 0.7206950783729553, | |
| "learning_rate": 6.32672500998336e-06, | |
| "loss": 0.7718, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.837287359829125, | |
| "grad_norm": 0.690377950668335, | |
| "learning_rate": 6.308086752194586e-06, | |
| "loss": 0.7784, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.839728430849035, | |
| "grad_norm": 0.6793109774589539, | |
| "learning_rate": 6.289428962048467e-06, | |
| "loss": 0.7936, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.842169501868945, | |
| "grad_norm": 0.6914140582084656, | |
| "learning_rate": 6.270751918143213e-06, | |
| "loss": 0.7652, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.844610572888855, | |
| "grad_norm": 0.6733123064041138, | |
| "learning_rate": 6.252055899364525e-06, | |
| "loss": 0.8477, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.847051643908765, | |
| "grad_norm": 0.6884806156158447, | |
| "learning_rate": 6.2333411848814415e-06, | |
| "loss": 0.7544, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.849492714928675, | |
| "grad_norm": 0.6831750273704529, | |
| "learning_rate": 6.214608054142167e-06, | |
| "loss": 0.7333, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8519337859485849, | |
| "grad_norm": 0.6560673713684082, | |
| "learning_rate": 6.195856786869893e-06, | |
| "loss": 0.7252, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.8543748569684949, | |
| "grad_norm": 0.6822741627693176, | |
| "learning_rate": 6.177087663058626e-06, | |
| "loss": 0.7181, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8568159279884049, | |
| "grad_norm": 0.8162235021591187, | |
| "learning_rate": 6.158300962969012e-06, | |
| "loss": 0.7359, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8592569990083149, | |
| "grad_norm": 0.7049196362495422, | |
| "learning_rate": 6.13949696712414e-06, | |
| "loss": 0.7592, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8616980700282248, | |
| "grad_norm": 0.6802664399147034, | |
| "learning_rate": 6.120675956305363e-06, | |
| "loss": 0.7476, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8641391410481348, | |
| "grad_norm": 0.7405847311019897, | |
| "learning_rate": 6.101838211548099e-06, | |
| "loss": 0.7368, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8665802120680448, | |
| "grad_norm": 0.6568160057067871, | |
| "learning_rate": 6.0829840141376385e-06, | |
| "loss": 0.7519, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8690212830879548, | |
| "grad_norm": 0.6947789192199707, | |
| "learning_rate": 6.064113645604945e-06, | |
| "loss": 0.7217, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8714623541078649, | |
| "grad_norm": 0.6590073704719543, | |
| "learning_rate": 6.045227387722445e-06, | |
| "loss": 0.7516, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8739034251277749, | |
| "grad_norm": 0.7120059132575989, | |
| "learning_rate": 6.026325522499829e-06, | |
| "loss": 0.7481, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8763444961476848, | |
| "grad_norm": 0.6988116502761841, | |
| "learning_rate": 6.007408332179836e-06, | |
| "loss": 0.7995, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8787855671675948, | |
| "grad_norm": 0.7005130648612976, | |
| "learning_rate": 5.988476099234033e-06, | |
| "loss": 0.7427, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8812266381875048, | |
| "grad_norm": 0.675496518611908, | |
| "learning_rate": 5.969529106358612e-06, | |
| "loss": 0.7603, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8836677092074148, | |
| "grad_norm": 0.7157605290412903, | |
| "learning_rate": 5.95056763647016e-06, | |
| "loss": 0.7788, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8861087802273248, | |
| "grad_norm": 0.6584900617599487, | |
| "learning_rate": 5.931591972701427e-06, | |
| "loss": 0.7415, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8885498512472347, | |
| "grad_norm": 0.6833528876304626, | |
| "learning_rate": 5.9126023983971114e-06, | |
| "loss": 0.7339, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8909909222671447, | |
| "grad_norm": 0.6946988701820374, | |
| "learning_rate": 5.893599197109625e-06, | |
| "loss": 0.8115, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8934319932870547, | |
| "grad_norm": 0.7004518508911133, | |
| "learning_rate": 5.874582652594855e-06, | |
| "loss": 0.75, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8958730643069647, | |
| "grad_norm": 0.6791942119598389, | |
| "learning_rate": 5.855553048807932e-06, | |
| "loss": 0.7288, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8983141353268747, | |
| "grad_norm": 0.6619113683700562, | |
| "learning_rate": 5.836510669898984e-06, | |
| "loss": 0.7408, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9007552063467846, | |
| "grad_norm": 0.6960625052452087, | |
| "learning_rate": 5.817455800208901e-06, | |
| "loss": 0.7937, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9031962773666946, | |
| "grad_norm": 0.6567366123199463, | |
| "learning_rate": 5.798388724265085e-06, | |
| "loss": 0.7555, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9056373483866046, | |
| "grad_norm": 0.6739965677261353, | |
| "learning_rate": 5.7793097267772e-06, | |
| "loss": 0.7193, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9080784194065146, | |
| "grad_norm": 0.6742777228355408, | |
| "learning_rate": 5.760219092632924e-06, | |
| "loss": 0.7308, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9105194904264245, | |
| "grad_norm": 0.6787696480751038, | |
| "learning_rate": 5.741117106893693e-06, | |
| "loss": 0.7387, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9129605614463345, | |
| "grad_norm": 0.6927146911621094, | |
| "learning_rate": 5.722004054790442e-06, | |
| "loss": 0.7435, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9154016324662446, | |
| "grad_norm": 0.6972612142562866, | |
| "learning_rate": 5.7028802217193565e-06, | |
| "loss": 0.7605, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9178427034861546, | |
| "grad_norm": 0.7299436926841736, | |
| "learning_rate": 5.683745893237598e-06, | |
| "loss": 0.7745, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9202837745060646, | |
| "grad_norm": 0.6617533564567566, | |
| "learning_rate": 5.664601355059044e-06, | |
| "loss": 0.7718, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9227248455259746, | |
| "grad_norm": 0.722768247127533, | |
| "learning_rate": 5.645446893050029e-06, | |
| "loss": 0.783, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9251659165458845, | |
| "grad_norm": 0.686353862285614, | |
| "learning_rate": 5.626282793225066e-06, | |
| "loss": 0.7411, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9276069875657945, | |
| "grad_norm": 0.7211350798606873, | |
| "learning_rate": 5.607109341742579e-06, | |
| "loss": 0.7729, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9300480585857045, | |
| "grad_norm": 0.6945457458496094, | |
| "learning_rate": 5.587926824900637e-06, | |
| "loss": 0.73, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.9324891296056145, | |
| "grad_norm": 0.641502320766449, | |
| "learning_rate": 5.568735529132665e-06, | |
| "loss": 0.7537, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9349302006255245, | |
| "grad_norm": 0.7009978294372559, | |
| "learning_rate": 5.5495357410031805e-06, | |
| "loss": 0.7407, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9373712716454344, | |
| "grad_norm": 0.6966471672058105, | |
| "learning_rate": 5.530327747203507e-06, | |
| "loss": 0.7287, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.9398123426653444, | |
| "grad_norm": 0.7361583709716797, | |
| "learning_rate": 5.511111834547496e-06, | |
| "loss": 0.7132, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9422534136852544, | |
| "grad_norm": 0.8365876078605652, | |
| "learning_rate": 5.491888289967241e-06, | |
| "loss": 0.7517, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9446944847051644, | |
| "grad_norm": 0.7376920580863953, | |
| "learning_rate": 5.472657400508801e-06, | |
| "loss": 0.7354, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.9471355557250744, | |
| "grad_norm": 0.6934507489204407, | |
| "learning_rate": 5.4534194533279e-06, | |
| "loss": 0.7418, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.9495766267449843, | |
| "grad_norm": 0.6778532862663269, | |
| "learning_rate": 5.434174735685658e-06, | |
| "loss": 0.7768, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9520176977648943, | |
| "grad_norm": 0.7122485041618347, | |
| "learning_rate": 5.414923534944283e-06, | |
| "loss": 0.7674, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9544587687848043, | |
| "grad_norm": 0.6961596608161926, | |
| "learning_rate": 5.395666138562794e-06, | |
| "loss": 0.7709, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.9568998398047143, | |
| "grad_norm": 0.7608603239059448, | |
| "learning_rate": 5.376402834092721e-06, | |
| "loss": 0.7787, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.9593409108246242, | |
| "grad_norm": 0.6927947402000427, | |
| "learning_rate": 5.357133909173815e-06, | |
| "loss": 0.7363, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.9617819818445343, | |
| "grad_norm": 0.7136049270629883, | |
| "learning_rate": 5.337859651529747e-06, | |
| "loss": 0.742, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.9642230528644443, | |
| "grad_norm": 0.7149622440338135, | |
| "learning_rate": 5.318580348963826e-06, | |
| "loss": 0.7501, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9666641238843543, | |
| "grad_norm": 0.6613947749137878, | |
| "learning_rate": 5.2992962893546804e-06, | |
| "loss": 0.7045, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9691051949042643, | |
| "grad_norm": 0.7548585534095764, | |
| "learning_rate": 5.280007760651977e-06, | |
| "loss": 0.7447, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9715462659241743, | |
| "grad_norm": 0.6713358759880066, | |
| "learning_rate": 5.260715050872119e-06, | |
| "loss": 0.7356, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9739873369440842, | |
| "grad_norm": 0.7130277156829834, | |
| "learning_rate": 5.241418448093931e-06, | |
| "loss": 0.7523, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9764284079639942, | |
| "grad_norm": 0.7643216252326965, | |
| "learning_rate": 5.222118240454376e-06, | |
| "loss": 0.7581, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9788694789839042, | |
| "grad_norm": 0.6513252258300781, | |
| "learning_rate": 5.202814716144245e-06, | |
| "loss": 0.7635, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.9813105500038142, | |
| "grad_norm": 0.7522091269493103, | |
| "learning_rate": 5.1835081634038455e-06, | |
| "loss": 0.7765, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9837516210237242, | |
| "grad_norm": 0.7359428405761719, | |
| "learning_rate": 5.164198870518714e-06, | |
| "loss": 0.7626, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9861926920436341, | |
| "grad_norm": 0.7049588561058044, | |
| "learning_rate": 5.144887125815301e-06, | |
| "loss": 0.6856, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9886337630635441, | |
| "grad_norm": 0.70652836561203, | |
| "learning_rate": 5.125573217656664e-06, | |
| "loss": 0.7479, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9910748340834541, | |
| "grad_norm": 0.6710291504859924, | |
| "learning_rate": 5.1062574344381686e-06, | |
| "loss": 0.7419, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9935159051033641, | |
| "grad_norm": 0.6925918459892273, | |
| "learning_rate": 5.086940064583179e-06, | |
| "loss": 0.7222, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.995956976123274, | |
| "grad_norm": 0.738444447517395, | |
| "learning_rate": 5.067621396538747e-06, | |
| "loss": 0.738, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.998398047143184, | |
| "grad_norm": 0.6995366811752319, | |
| "learning_rate": 5.048301718771317e-06, | |
| "loss": 0.7986, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.000839118163094, | |
| "grad_norm": 0.7283437848091125, | |
| "learning_rate": 5.028981319762399e-06, | |
| "loss": 0.7439, | |
| "step": 410 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 818, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 205, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.11279028268381e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |