| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.557374764857521, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0027868738242876052, | |
| "grad_norm": 4.701911449432373, | |
| "learning_rate": 2.2222222222222225e-06, | |
| "loss": 3.5703, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0055737476485752105, | |
| "grad_norm": 4.517332077026367, | |
| "learning_rate": 5e-06, | |
| "loss": 3.5228, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008360621472862817, | |
| "grad_norm": 4.366923809051514, | |
| "learning_rate": 7.777777777777777e-06, | |
| "loss": 3.4713, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.011147495297150421, | |
| "grad_norm": 4.379341125488281, | |
| "learning_rate": 1.0555555555555555e-05, | |
| "loss": 3.5622, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013934369121438027, | |
| "grad_norm": 5.151899814605713, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 3.4422, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.016721242945725634, | |
| "grad_norm": 4.80206298828125, | |
| "learning_rate": 1.6111111111111115e-05, | |
| "loss": 3.3216, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.019508116770013236, | |
| "grad_norm": 3.687617778778076, | |
| "learning_rate": 1.888888888888889e-05, | |
| "loss": 2.942, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.022294990594300842, | |
| "grad_norm": 3.1759581565856934, | |
| "learning_rate": 2.1666666666666667e-05, | |
| "loss": 2.6761, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.025081864418588447, | |
| "grad_norm": 2.588658332824707, | |
| "learning_rate": 2.4444444444444445e-05, | |
| "loss": 2.6441, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.027868738242876053, | |
| "grad_norm": 2.1114368438720703, | |
| "learning_rate": 2.7222222222222223e-05, | |
| "loss": 2.4923, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03065561206716366, | |
| "grad_norm": 2.0633840560913086, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2875, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03344248589145127, | |
| "grad_norm": 1.8759537935256958, | |
| "learning_rate": 3.277777777777778e-05, | |
| "loss": 2.408, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03622935971573887, | |
| "grad_norm": 1.9867842197418213, | |
| "learning_rate": 3.555555555555556e-05, | |
| "loss": 2.3362, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03901623354002647, | |
| "grad_norm": 1.781015157699585, | |
| "learning_rate": 3.8333333333333334e-05, | |
| "loss": 2.1793, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04180310736431408, | |
| "grad_norm": 2.1073060035705566, | |
| "learning_rate": 4.111111111111111e-05, | |
| "loss": 2.2342, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.044589981188601684, | |
| "grad_norm": 1.8073269128799438, | |
| "learning_rate": 4.388888888888889e-05, | |
| "loss": 2.1447, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04737685501288929, | |
| "grad_norm": 1.8754196166992188, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 2.3169, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.050163728837176895, | |
| "grad_norm": 1.9674766063690186, | |
| "learning_rate": 4.9444444444444446e-05, | |
| "loss": 2.1966, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0529506026614645, | |
| "grad_norm": 2.0507538318634033, | |
| "learning_rate": 5.222222222222223e-05, | |
| "loss": 2.1436, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.055737476485752106, | |
| "grad_norm": 2.1478066444396973, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.0947, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05852435031003971, | |
| "grad_norm": 1.8122990131378174, | |
| "learning_rate": 5.7777777777777776e-05, | |
| "loss": 2.1266, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06131122413432732, | |
| "grad_norm": 2.30039119720459, | |
| "learning_rate": 6.055555555555555e-05, | |
| "loss": 2.175, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06409809795861493, | |
| "grad_norm": 2.1259849071502686, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 2.0622, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.06688497178290254, | |
| "grad_norm": 2.159100294113159, | |
| "learning_rate": 6.611111111111111e-05, | |
| "loss": 2.078, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06967184560719013, | |
| "grad_norm": 2.2886006832122803, | |
| "learning_rate": 6.88888888888889e-05, | |
| "loss": 2.1515, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.07245871943147773, | |
| "grad_norm": 2.1963438987731934, | |
| "learning_rate": 7.166666666666667e-05, | |
| "loss": 2.0886, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07524559325576534, | |
| "grad_norm": 1.9774861335754395, | |
| "learning_rate": 7.444444444444444e-05, | |
| "loss": 2.0426, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07803246708005294, | |
| "grad_norm": 2.2124135494232178, | |
| "learning_rate": 7.722222222222223e-05, | |
| "loss": 2.0686, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08081934090434055, | |
| "grad_norm": 2.0250704288482666, | |
| "learning_rate": 8e-05, | |
| "loss": 2.1701, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08360621472862816, | |
| "grad_norm": 2.2717087268829346, | |
| "learning_rate": 8.277777777777778e-05, | |
| "loss": 2.1313, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08639308855291576, | |
| "grad_norm": 2.090191602706909, | |
| "learning_rate": 8.555555555555556e-05, | |
| "loss": 2.0572, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08917996237720337, | |
| "grad_norm": 2.09201979637146, | |
| "learning_rate": 8.833333333333333e-05, | |
| "loss": 1.982, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09196683620149097, | |
| "grad_norm": 2.206078290939331, | |
| "learning_rate": 9.111111111111112e-05, | |
| "loss": 2.0063, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.09475371002577858, | |
| "grad_norm": 2.1212692260742188, | |
| "learning_rate": 9.388888888888889e-05, | |
| "loss": 2.0597, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.09754058385006618, | |
| "grad_norm": 2.215811252593994, | |
| "learning_rate": 9.666666666666667e-05, | |
| "loss": 2.0566, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.10032745767435379, | |
| "grad_norm": 2.1642961502075195, | |
| "learning_rate": 9.944444444444446e-05, | |
| "loss": 2.0843, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1031143314986414, | |
| "grad_norm": 2.315004348754883, | |
| "learning_rate": 9.999848639521432e-05, | |
| "loss": 2.0011, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.105901205322929, | |
| "grad_norm": 2.2057459354400635, | |
| "learning_rate": 9.999233753283091e-05, | |
| "loss": 1.9906, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1086880791472166, | |
| "grad_norm": 2.516157865524292, | |
| "learning_rate": 9.998145939378577e-05, | |
| "loss": 1.9941, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.11147495297150421, | |
| "grad_norm": 2.2229058742523193, | |
| "learning_rate": 9.996585300715116e-05, | |
| "loss": 1.9935, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11426182679579182, | |
| "grad_norm": 2.1316332817077637, | |
| "learning_rate": 9.994551984929175e-05, | |
| "loss": 1.9692, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11704870062007942, | |
| "grad_norm": 1.8723924160003662, | |
| "learning_rate": 9.992046184372492e-05, | |
| "loss": 2.0208, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11983557444436703, | |
| "grad_norm": 2.1844356060028076, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 1.9584, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.12262244826865464, | |
| "grad_norm": 2.2411093711853027, | |
| "learning_rate": 9.985618121816779e-05, | |
| "loss": 2.0796, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12540932209294225, | |
| "grad_norm": 2.0734896659851074, | |
| "learning_rate": 9.981696467912664e-05, | |
| "loss": 1.9712, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12819619591722986, | |
| "grad_norm": 2.1242079734802246, | |
| "learning_rate": 9.97730354537011e-05, | |
| "loss": 1.9637, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13098306974151747, | |
| "grad_norm": 2.026093006134033, | |
| "learning_rate": 9.972439769759722e-05, | |
| "loss": 1.8615, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.13376994356580507, | |
| "grad_norm": 2.029449701309204, | |
| "learning_rate": 9.967105601194823e-05, | |
| "loss": 2.11, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13655681739009268, | |
| "grad_norm": 2.0343263149261475, | |
| "learning_rate": 9.961301544287922e-05, | |
| "loss": 1.9325, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13934369121438026, | |
| "grad_norm": 2.2704412937164307, | |
| "learning_rate": 9.955028148102979e-05, | |
| "loss": 1.8849, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.14213056503866786, | |
| "grad_norm": 2.563328742980957, | |
| "learning_rate": 9.948286006103466e-05, | |
| "loss": 2.1103, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.14491743886295547, | |
| "grad_norm": 2.4252612590789795, | |
| "learning_rate": 9.941075756096226e-05, | |
| "loss": 1.9867, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14770431268724307, | |
| "grad_norm": 2.1968955993652344, | |
| "learning_rate": 9.933398080171123e-05, | |
| "loss": 1.998, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.15049118651153068, | |
| "grad_norm": 2.1625869274139404, | |
| "learning_rate": 9.925253704636543e-05, | |
| "loss": 1.9043, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15327806033581828, | |
| "grad_norm": 2.349252700805664, | |
| "learning_rate": 9.916643399950656e-05, | |
| "loss": 1.9854, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.1560649341601059, | |
| "grad_norm": 2.2877767086029053, | |
| "learning_rate": 9.907567980648549e-05, | |
| "loss": 1.9119, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1588518079843935, | |
| "grad_norm": 2.0312581062316895, | |
| "learning_rate": 9.898028305265169e-05, | |
| "loss": 2.0732, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1616386818086811, | |
| "grad_norm": 2.036317825317383, | |
| "learning_rate": 9.888025276254096e-05, | |
| "loss": 2.0147, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1644255556329687, | |
| "grad_norm": 2.345585823059082, | |
| "learning_rate": 9.877559839902184e-05, | |
| "loss": 1.9635, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1672124294572563, | |
| "grad_norm": 2.371412992477417, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 1.9034, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16999930328154392, | |
| "grad_norm": 2.3761579990386963, | |
| "learning_rate": 9.855245748948326e-05, | |
| "loss": 1.9687, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.17278617710583152, | |
| "grad_norm": 2.1475725173950195, | |
| "learning_rate": 9.843399205260068e-05, | |
| "loss": 1.9903, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.17557305093011913, | |
| "grad_norm": 2.40266752243042, | |
| "learning_rate": 9.831094475858652e-05, | |
| "loss": 1.9729, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.17835992475440673, | |
| "grad_norm": 2.187901258468628, | |
| "learning_rate": 9.818332724771857e-05, | |
| "loss": 2.0002, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.18114679857869434, | |
| "grad_norm": 2.374246597290039, | |
| "learning_rate": 9.805115159261726e-05, | |
| "loss": 1.8938, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.18393367240298195, | |
| "grad_norm": 2.2676525115966797, | |
| "learning_rate": 9.791443029710361e-05, | |
| "loss": 1.8728, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18672054622726955, | |
| "grad_norm": 2.175401449203491, | |
| "learning_rate": 9.777317629501636e-05, | |
| "loss": 1.9195, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.18950742005155716, | |
| "grad_norm": 2.104386329650879, | |
| "learning_rate": 9.762740294898846e-05, | |
| "loss": 1.8618, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.19229429387584476, | |
| "grad_norm": 2.2644734382629395, | |
| "learning_rate": 9.747712404918286e-05, | |
| "loss": 1.8384, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.19508116770013237, | |
| "grad_norm": 2.6918885707855225, | |
| "learning_rate": 9.732235381198813e-05, | |
| "loss": 1.9305, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19786804152441997, | |
| "grad_norm": 2.374101161956787, | |
| "learning_rate": 9.716310687867342e-05, | |
| "loss": 1.7964, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.20065491534870758, | |
| "grad_norm": 2.401925802230835, | |
| "learning_rate": 9.699939831400351e-05, | |
| "loss": 2.0372, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.20344178917299519, | |
| "grad_norm": 2.3377015590667725, | |
| "learning_rate": 9.683124360481364e-05, | |
| "loss": 1.9354, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2062286629972828, | |
| "grad_norm": 2.669875144958496, | |
| "learning_rate": 9.665865865854445e-05, | |
| "loss": 1.9243, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2090155368215704, | |
| "grad_norm": 2.251633882522583, | |
| "learning_rate": 9.648165980173712e-05, | |
| "loss": 1.9379, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.211802410645858, | |
| "grad_norm": 2.413219928741455, | |
| "learning_rate": 9.630026377848892e-05, | |
| "loss": 1.9232, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.2145892844701456, | |
| "grad_norm": 2.4033939838409424, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 1.9206, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.2173761582944332, | |
| "grad_norm": 2.7085165977478027, | |
| "learning_rate": 9.592434928729616e-05, | |
| "loss": 1.965, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.22016303211872082, | |
| "grad_norm": 2.440049409866333, | |
| "learning_rate": 9.572986638087396e-05, | |
| "loss": 1.938, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.22294990594300843, | |
| "grad_norm": 2.2908005714416504, | |
| "learning_rate": 9.553105742769154e-05, | |
| "loss": 1.8841, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.22573677976729603, | |
| "grad_norm": 2.0851612091064453, | |
| "learning_rate": 9.532794123508197e-05, | |
| "loss": 1.7699, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.22852365359158364, | |
| "grad_norm": 2.466432809829712, | |
| "learning_rate": 9.512053701784329e-05, | |
| "loss": 1.9162, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.23131052741587124, | |
| "grad_norm": 2.130993127822876, | |
| "learning_rate": 9.490886439642081e-05, | |
| "loss": 1.9023, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.23409740124015885, | |
| "grad_norm": 2.220752716064453, | |
| "learning_rate": 9.469294339505098e-05, | |
| "loss": 1.975, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.23688427506444645, | |
| "grad_norm": 2.1913466453552246, | |
| "learning_rate": 9.447279443986716e-05, | |
| "loss": 1.8632, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.23967114888873406, | |
| "grad_norm": 2.321277379989624, | |
| "learning_rate": 9.424843835696724e-05, | |
| "loss": 1.9294, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.24245802271302166, | |
| "grad_norm": 2.286395788192749, | |
| "learning_rate": 9.401989637044355e-05, | |
| "loss": 1.8428, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.24524489653730927, | |
| "grad_norm": 2.3919999599456787, | |
| "learning_rate": 9.3787190100375e-05, | |
| "loss": 1.8813, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.24803177036159688, | |
| "grad_norm": 2.1720130443573, | |
| "learning_rate": 9.355034156078188e-05, | |
| "loss": 1.8569, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.2508186441858845, | |
| "grad_norm": 2.086071014404297, | |
| "learning_rate": 9.330937315754329e-05, | |
| "loss": 1.8796, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2536055180101721, | |
| "grad_norm": 2.447908639907837, | |
| "learning_rate": 9.306430768627753e-05, | |
| "loss": 1.851, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.2563923918344597, | |
| "grad_norm": 2.631868362426758, | |
| "learning_rate": 9.281516833018571e-05, | |
| "loss": 1.8706, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2591792656587473, | |
| "grad_norm": 2.198009729385376, | |
| "learning_rate": 9.256197865785854e-05, | |
| "loss": 1.9986, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.26196613948303493, | |
| "grad_norm": 2.151010513305664, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 1.8895, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.26475301330732254, | |
| "grad_norm": 2.335463047027588, | |
| "learning_rate": 9.204354455239539e-05, | |
| "loss": 1.9415, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.26753988713161014, | |
| "grad_norm": 2.4173576831817627, | |
| "learning_rate": 9.177834916314165e-05, | |
| "loss": 1.8676, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.27032676095589775, | |
| "grad_norm": 2.409341812133789, | |
| "learning_rate": 9.150920154077754e-05, | |
| "loss": 1.879, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.27311363478018535, | |
| "grad_norm": 2.496391534805298, | |
| "learning_rate": 9.123612714667634e-05, | |
| "loss": 1.9162, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.2759005086044729, | |
| "grad_norm": 2.2236454486846924, | |
| "learning_rate": 9.095915181368412e-05, | |
| "loss": 1.9825, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2786873824287605, | |
| "grad_norm": 2.4536592960357666, | |
| "learning_rate": 9.067830174367586e-05, | |
| "loss": 1.933, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2814742562530481, | |
| "grad_norm": 2.130772829055786, | |
| "learning_rate": 9.039360350507679e-05, | |
| "loss": 1.901, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.2842611300773357, | |
| "grad_norm": 2.2794065475463867, | |
| "learning_rate": 9.010508403034898e-05, | |
| "loss": 1.8599, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2870480039016233, | |
| "grad_norm": 2.342435598373413, | |
| "learning_rate": 8.98127706134436e-05, | |
| "loss": 1.8981, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.28983487772591093, | |
| "grad_norm": 2.3479907512664795, | |
| "learning_rate": 8.951669090721881e-05, | |
| "loss": 1.9312, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.29262175155019854, | |
| "grad_norm": 2.5833990573883057, | |
| "learning_rate": 8.921687292082393e-05, | |
| "loss": 1.8288, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.29540862537448614, | |
| "grad_norm": 2.2673983573913574, | |
| "learning_rate": 8.891334501704962e-05, | |
| "loss": 1.9026, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.29819549919877375, | |
| "grad_norm": 2.6991782188415527, | |
| "learning_rate": 8.86061359096449e-05, | |
| "loss": 1.8369, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.30098237302306136, | |
| "grad_norm": 2.9269566535949707, | |
| "learning_rate": 8.829527466060072e-05, | |
| "loss": 1.7069, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.30376924684734896, | |
| "grad_norm": 2.2337591648101807, | |
| "learning_rate": 8.798079067740077e-05, | |
| "loss": 1.994, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.30655612067163657, | |
| "grad_norm": 2.3603134155273438, | |
| "learning_rate": 8.766271371023948e-05, | |
| "loss": 1.8576, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3093429944959242, | |
| "grad_norm": 2.4910824298858643, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 1.8134, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.3121298683202118, | |
| "grad_norm": 2.3701236248016357, | |
| "learning_rate": 8.701590152144612e-05, | |
| "loss": 1.8629, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3149167421444994, | |
| "grad_norm": 2.594458818435669, | |
| "learning_rate": 8.668722748826693e-05, | |
| "loss": 1.8344, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.317703615968787, | |
| "grad_norm": 2.4443206787109375, | |
| "learning_rate": 8.635508284224371e-05, | |
| "loss": 2.0025, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3204904897930746, | |
| "grad_norm": 2.4144582748413086, | |
| "learning_rate": 8.601949900427016e-05, | |
| "loss": 1.8186, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3232773636173622, | |
| "grad_norm": 2.4722774028778076, | |
| "learning_rate": 8.568050772058762e-05, | |
| "loss": 1.8776, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3260642374416498, | |
| "grad_norm": 2.769221544265747, | |
| "learning_rate": 8.533814105978191e-05, | |
| "loss": 1.86, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3288511112659374, | |
| "grad_norm": 2.3548312187194824, | |
| "learning_rate": 8.499243140974966e-05, | |
| "loss": 1.8534, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.331637985090225, | |
| "grad_norm": 2.4437191486358643, | |
| "learning_rate": 8.464341147463431e-05, | |
| "loss": 1.865, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.3344248589145126, | |
| "grad_norm": 2.4781646728515625, | |
| "learning_rate": 8.429111427173241e-05, | |
| "loss": 1.7592, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.33721173273880023, | |
| "grad_norm": 2.605130910873413, | |
| "learning_rate": 8.393557312837018e-05, | |
| "loss": 1.7903, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.33999860656308784, | |
| "grad_norm": 2.6218454837799072, | |
| "learning_rate": 8.357682167875062e-05, | |
| "loss": 1.8664, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.34278548038737544, | |
| "grad_norm": 2.7091429233551025, | |
| "learning_rate": 8.321489386077192e-05, | |
| "loss": 1.87, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.34557235421166305, | |
| "grad_norm": 2.571959972381592, | |
| "learning_rate": 8.28498239128167e-05, | |
| "loss": 1.852, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.34835922803595065, | |
| "grad_norm": 2.72074031829834, | |
| "learning_rate": 8.248164637051321e-05, | |
| "loss": 1.8048, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.35114610186023826, | |
| "grad_norm": 2.7408368587493896, | |
| "learning_rate": 8.211039606346826e-05, | |
| "loss": 1.9272, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.35393297568452586, | |
| "grad_norm": 2.6123206615448, | |
| "learning_rate": 8.173610811197226e-05, | |
| "loss": 1.8601, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.35671984950881347, | |
| "grad_norm": 2.283332347869873, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 1.8755, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3595067233331011, | |
| "grad_norm": 2.5847232341766357, | |
| "learning_rate": 8.097856119024545e-05, | |
| "loss": 1.9465, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3622935971573887, | |
| "grad_norm": 2.7280638217926025, | |
| "learning_rate": 8.059537388397665e-05, | |
| "loss": 1.8577, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3650804709816763, | |
| "grad_norm": 2.4492626190185547, | |
| "learning_rate": 8.020929225440137e-05, | |
| "loss": 1.8689, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.3678673448059639, | |
| "grad_norm": 2.558868169784546, | |
| "learning_rate": 7.98203528248536e-05, | |
| "loss": 1.8553, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3706542186302515, | |
| "grad_norm": 2.6071197986602783, | |
| "learning_rate": 7.942859238901528e-05, | |
| "loss": 1.9035, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3734410924545391, | |
| "grad_norm": 2.6189825534820557, | |
| "learning_rate": 7.903404800743564e-05, | |
| "loss": 1.9362, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3762279662788267, | |
| "grad_norm": 2.8328700065612793, | |
| "learning_rate": 7.863675700402526e-05, | |
| "loss": 1.8079, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3790148401031143, | |
| "grad_norm": 2.458878755569458, | |
| "learning_rate": 7.823675696252524e-05, | |
| "loss": 1.9774, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3818017139274019, | |
| "grad_norm": 2.524718761444092, | |
| "learning_rate": 7.783408572295174e-05, | |
| "loss": 1.9575, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.3845885877516895, | |
| "grad_norm": 2.4117393493652344, | |
| "learning_rate": 7.742878137801639e-05, | |
| "loss": 1.8446, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.38737546157597713, | |
| "grad_norm": 2.6879658699035645, | |
| "learning_rate": 7.702088226952258e-05, | |
| "loss": 1.8512, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.39016233540026474, | |
| "grad_norm": 2.5234720706939697, | |
| "learning_rate": 7.661042698473853e-05, | |
| "loss": 1.778, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.39294920922455234, | |
| "grad_norm": 2.4943764209747314, | |
| "learning_rate": 7.619745435274667e-05, | |
| "loss": 1.8087, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.39573608304883995, | |
| "grad_norm": 2.897895574569702, | |
| "learning_rate": 7.578200344077073e-05, | |
| "loss": 1.8764, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.39852295687312755, | |
| "grad_norm": 2.5875868797302246, | |
| "learning_rate": 7.536411355047964e-05, | |
| "loss": 1.8782, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.40130983069741516, | |
| "grad_norm": 2.7180755138397217, | |
| "learning_rate": 7.494382421426984e-05, | |
| "loss": 1.8987, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.40409670452170277, | |
| "grad_norm": 2.798145294189453, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 1.8426, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.40688357834599037, | |
| "grad_norm": 3.262134552001953, | |
| "learning_rate": 7.409620646485685e-05, | |
| "loss": 1.7929, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.409670452170278, | |
| "grad_norm": 2.6406302452087402, | |
| "learning_rate": 7.36689582363187e-05, | |
| "loss": 1.876, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.4124573259945656, | |
| "grad_norm": 2.56366229057312, | |
| "learning_rate": 7.323947092360649e-05, | |
| "loss": 1.7674, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.4152441998188532, | |
| "grad_norm": 2.5193862915039062, | |
| "learning_rate": 7.280778515623314e-05, | |
| "loss": 1.9506, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.4180310736431408, | |
| "grad_norm": 2.9591546058654785, | |
| "learning_rate": 7.237394177168548e-05, | |
| "loss": 1.8831, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4208179474674284, | |
| "grad_norm": 2.7385146617889404, | |
| "learning_rate": 7.193798181156095e-05, | |
| "loss": 1.9302, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.423604821291716, | |
| "grad_norm": 2.6676039695739746, | |
| "learning_rate": 7.149994651768514e-05, | |
| "loss": 1.7349, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4263916951160036, | |
| "grad_norm": 3.0209970474243164, | |
| "learning_rate": 7.10598773282103e-05, | |
| "loss": 1.8751, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4291785689402912, | |
| "grad_norm": 2.7455992698669434, | |
| "learning_rate": 7.061781587369519e-05, | |
| "loss": 1.9342, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.4319654427645788, | |
| "grad_norm": 2.842349052429199, | |
| "learning_rate": 7.017380397316695e-05, | |
| "loss": 1.8249, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.4347523165888664, | |
| "grad_norm": 2.6208105087280273, | |
| "learning_rate": 6.972788363016497e-05, | |
| "loss": 1.7323, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.43753919041315403, | |
| "grad_norm": 2.757404088973999, | |
| "learning_rate": 6.92800970287674e-05, | |
| "loss": 1.8447, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.44032606423744164, | |
| "grad_norm": 2.525618553161621, | |
| "learning_rate": 6.883048652960038e-05, | |
| "loss": 1.8332, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.44311293806172924, | |
| "grad_norm": 2.5981028079986572, | |
| "learning_rate": 6.837909466583095e-05, | |
| "loss": 1.8783, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.44589981188601685, | |
| "grad_norm": 2.546555280685425, | |
| "learning_rate": 6.792596413914324e-05, | |
| "loss": 1.7139, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.44868668571030446, | |
| "grad_norm": 2.530876398086548, | |
| "learning_rate": 6.747113781569892e-05, | |
| "loss": 1.9029, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.45147355953459206, | |
| "grad_norm": 2.799030303955078, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 1.9114, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.45426043335887967, | |
| "grad_norm": 2.816466808319092, | |
| "learning_rate": 6.655657004122916e-05, | |
| "loss": 1.9019, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.4570473071831673, | |
| "grad_norm": 3.0766985416412354, | |
| "learning_rate": 6.60969151083432e-05, | |
| "loss": 1.809, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4598341810074549, | |
| "grad_norm": 2.733771800994873, | |
| "learning_rate": 6.563573740679496e-05, | |
| "loss": 1.8111, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.4626210548317425, | |
| "grad_norm": 2.653470516204834, | |
| "learning_rate": 6.517308056400917e-05, | |
| "loss": 1.8637, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4654079286560301, | |
| "grad_norm": 2.7409327030181885, | |
| "learning_rate": 6.470898834733731e-05, | |
| "loss": 1.8205, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.4681948024803177, | |
| "grad_norm": 2.6908485889434814, | |
| "learning_rate": 6.42435046599173e-05, | |
| "loss": 1.8235, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4709816763046053, | |
| "grad_norm": 2.7610902786254883, | |
| "learning_rate": 6.377667353652022e-05, | |
| "loss": 1.8789, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.4737685501288929, | |
| "grad_norm": 2.7466371059417725, | |
| "learning_rate": 6.330853913938466e-05, | |
| "loss": 1.8, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4765554239531805, | |
| "grad_norm": 2.6152100563049316, | |
| "learning_rate": 6.283914575403888e-05, | |
| "loss": 1.9228, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.4793422977774681, | |
| "grad_norm": 2.903139114379883, | |
| "learning_rate": 6.236853778511156e-05, | |
| "loss": 1.8079, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4821291716017557, | |
| "grad_norm": 2.941880941390991, | |
| "learning_rate": 6.189675975213094e-05, | |
| "loss": 1.8818, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.48491604542604333, | |
| "grad_norm": 2.928757429122925, | |
| "learning_rate": 6.142385628531342e-05, | |
| "loss": 1.883, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.48770291925033094, | |
| "grad_norm": 3.0800089836120605, | |
| "learning_rate": 6.09498721213414e-05, | |
| "loss": 1.7964, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.49048979307461854, | |
| "grad_norm": 2.787975549697876, | |
| "learning_rate": 6.047485209913137e-05, | |
| "loss": 1.9406, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.49327666689890615, | |
| "grad_norm": 2.5201809406280518, | |
| "learning_rate": 5.999884115559192e-05, | |
| "loss": 1.8103, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.49606354072319375, | |
| "grad_norm": 2.659198760986328, | |
| "learning_rate": 5.952188432137293e-05, | |
| "loss": 1.879, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.49885041454748136, | |
| "grad_norm": 2.3848142623901367, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 1.7912, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.501637288371769, | |
| "grad_norm": 3.1244029998779297, | |
| "learning_rate": 5.8565313546633684e-05, | |
| "loss": 1.8467, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5044241621960566, | |
| "grad_norm": 2.36189866065979, | |
| "learning_rate": 5.8085790097738025e-05, | |
| "loss": 1.8748, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.5072110360203442, | |
| "grad_norm": 2.6590969562530518, | |
| "learning_rate": 5.7605501732851475e-05, | |
| "loss": 1.8892, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5099979098446318, | |
| "grad_norm": 2.493211030960083, | |
| "learning_rate": 5.712449388726807e-05, | |
| "loss": 1.9295, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.5127847836689194, | |
| "grad_norm": 2.8434841632843018, | |
| "learning_rate": 5.664281206434472e-05, | |
| "loss": 1.7492, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.515571657493207, | |
| "grad_norm": 3.066373348236084, | |
| "learning_rate": 5.616050183119663e-05, | |
| "loss": 1.9097, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5183585313174947, | |
| "grad_norm": 2.5913686752319336, | |
| "learning_rate": 5.5677608814386616e-05, | |
| "loss": 1.7054, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5211454051417822, | |
| "grad_norm": 2.7032017707824707, | |
| "learning_rate": 5.519417869560889e-05, | |
| "loss": 1.816, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5239322789660699, | |
| "grad_norm": 3.0254099369049072, | |
| "learning_rate": 5.471025720736747e-05, | |
| "loss": 1.8939, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5267191527903574, | |
| "grad_norm": 3.013810873031616, | |
| "learning_rate": 5.422589012864996e-05, | |
| "loss": 1.8953, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5295060266146451, | |
| "grad_norm": 2.9430110454559326, | |
| "learning_rate": 5.3741123280596864e-05, | |
| "loss": 1.8232, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5322929004389326, | |
| "grad_norm": 2.7250967025756836, | |
| "learning_rate": 5.325600252216685e-05, | |
| "loss": 1.8462, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5350797742632203, | |
| "grad_norm": 2.7945637702941895, | |
| "learning_rate": 5.27705737457985e-05, | |
| "loss": 1.7108, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5378666480875078, | |
| "grad_norm": 2.6481263637542725, | |
| "learning_rate": 5.228488287306896e-05, | |
| "loss": 1.8026, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.5406535219117955, | |
| "grad_norm": 2.819915771484375, | |
| "learning_rate": 5.179897585034963e-05, | |
| "loss": 1.8175, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.543440395736083, | |
| "grad_norm": 3.0015969276428223, | |
| "learning_rate": 5.1312898644459776e-05, | |
| "loss": 1.7393, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5462272695603707, | |
| "grad_norm": 3.3672678470611572, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 1.8442, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5490141433846583, | |
| "grad_norm": 2.963168144226074, | |
| "learning_rate": 5.0340417626592016e-05, | |
| "loss": 1.8713, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5518010172089458, | |
| "grad_norm": 3.0407772064208984, | |
| "learning_rate": 4.9854105811348216e-05, | |
| "loss": 1.8686, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5545878910332335, | |
| "grad_norm": 3.218909978866577, | |
| "learning_rate": 4.936780779769913e-05, | |
| "loss": 1.8486, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.557374764857521, | |
| "grad_norm": 2.7832529544830322, | |
| "learning_rate": 4.888156958945174e-05, | |
| "loss": 1.8146, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1795, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5976966563082240.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |