| { | |
| "best_metric": 2.4975922107696533, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.011764705882353, | |
| "eval_steps": 50, | |
| "global_step": 192, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01568627450980392, | |
| "grad_norm": 0.9290716052055359, | |
| "learning_rate": 1e-05, | |
| "loss": 2.8736, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01568627450980392, | |
| "eval_loss": 3.324666738510132, | |
| "eval_runtime": 8.0111, | |
| "eval_samples_per_second": 13.481, | |
| "eval_steps_per_second": 3.37, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03137254901960784, | |
| "grad_norm": 1.0439985990524292, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8739, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.047058823529411764, | |
| "grad_norm": 1.0311352014541626, | |
| "learning_rate": 3e-05, | |
| "loss": 2.9072, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06274509803921569, | |
| "grad_norm": 0.9413182735443115, | |
| "learning_rate": 4e-05, | |
| "loss": 2.9037, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 0.9702916741371155, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8073, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09411764705882353, | |
| "grad_norm": 0.9648849964141846, | |
| "learning_rate": 6e-05, | |
| "loss": 2.9957, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10980392156862745, | |
| "grad_norm": 0.8518514037132263, | |
| "learning_rate": 7e-05, | |
| "loss": 2.9002, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12549019607843137, | |
| "grad_norm": 0.901975691318512, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7139, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1411764705882353, | |
| "grad_norm": 0.8090202808380127, | |
| "learning_rate": 9e-05, | |
| "loss": 2.7318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 0.8015203475952148, | |
| "learning_rate": 0.0001, | |
| "loss": 2.6717, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17254901960784313, | |
| "grad_norm": 0.9336282014846802, | |
| "learning_rate": 9.999255120204248e-05, | |
| "loss": 2.6711, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.18823529411764706, | |
| "grad_norm": 1.0132514238357544, | |
| "learning_rate": 9.997020702755353e-05, | |
| "loss": 2.7361, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20392156862745098, | |
| "grad_norm": 1.0461229085922241, | |
| "learning_rate": 9.99329741340228e-05, | |
| "loss": 2.6794, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2196078431372549, | |
| "grad_norm": 1.241179347038269, | |
| "learning_rate": 9.98808636150624e-05, | |
| "loss": 2.6331, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 1.49308180809021, | |
| "learning_rate": 9.981389099710132e-05, | |
| "loss": 3.0748, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25098039215686274, | |
| "grad_norm": 0.5110021829605103, | |
| "learning_rate": 9.973207623475965e-05, | |
| "loss": 2.409, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.6003811955451965, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 2.5082, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2823529411764706, | |
| "grad_norm": 0.5671645998954773, | |
| "learning_rate": 9.952402219937816e-05, | |
| "loss": 2.5027, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2980392156862745, | |
| "grad_norm": 0.5409471988677979, | |
| "learning_rate": 9.939784491643734e-05, | |
| "loss": 2.46, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 0.5076523423194885, | |
| "learning_rate": 9.92569494508437e-05, | |
| "loss": 2.6268, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.32941176470588235, | |
| "grad_norm": 0.47187337279319763, | |
| "learning_rate": 9.910137778267152e-05, | |
| "loss": 2.5395, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.34509803921568627, | |
| "grad_norm": 0.4689958393573761, | |
| "learning_rate": 9.893117626479777e-05, | |
| "loss": 2.5222, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3607843137254902, | |
| "grad_norm": 0.4607944190502167, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 2.4532, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3764705882352941, | |
| "grad_norm": 0.5923854112625122, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 2.4775, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.39215686274509803, | |
| "grad_norm": 0.6515561938285828, | |
| "learning_rate": 9.833332143466099e-05, | |
| "loss": 2.5327, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.40784313725490196, | |
| "grad_norm": 0.5666857361793518, | |
| "learning_rate": 9.810515099218003e-05, | |
| "loss": 2.715, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4235294117647059, | |
| "grad_norm": 0.6389107704162598, | |
| "learning_rate": 9.78626475276808e-05, | |
| "loss": 2.5737, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4392156862745098, | |
| "grad_norm": 0.7637258172035217, | |
| "learning_rate": 9.760588329553571e-05, | |
| "loss": 2.6324, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4549019607843137, | |
| "grad_norm": 0.8146405816078186, | |
| "learning_rate": 9.73349347991403e-05, | |
| "loss": 2.5954, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 1.183030605316162, | |
| "learning_rate": 9.704988276811883e-05, | |
| "loss": 2.9496, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48627450980392156, | |
| "grad_norm": 0.2972618639469147, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 2.3553, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5019607843137255, | |
| "grad_norm": 0.4887743294239044, | |
| "learning_rate": 9.643781200626511e-05, | |
| "loss": 2.4257, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5176470588235295, | |
| "grad_norm": 0.4153500497341156, | |
| "learning_rate": 9.611097564309053e-05, | |
| "loss": 2.5037, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.41760119795799255, | |
| "learning_rate": 9.577040042626833e-05, | |
| "loss": 2.3801, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5490196078431373, | |
| "grad_norm": 0.4032338857650757, | |
| "learning_rate": 9.54161878308377e-05, | |
| "loss": 2.4549, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5647058823529412, | |
| "grad_norm": 0.4308023452758789, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 2.4362, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5803921568627451, | |
| "grad_norm": 0.44621264934539795, | |
| "learning_rate": 9.466727668927816e-05, | |
| "loss": 2.5759, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.596078431372549, | |
| "grad_norm": 0.4338065981864929, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 2.4042, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.611764705882353, | |
| "grad_norm": 0.4249041974544525, | |
| "learning_rate": 9.38651347099721e-05, | |
| "loss": 2.3852, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 0.46373647451400757, | |
| "learning_rate": 9.344439843625034e-05, | |
| "loss": 2.4953, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6431372549019608, | |
| "grad_norm": 0.4927481412887573, | |
| "learning_rate": 9.301071782067504e-05, | |
| "loss": 2.4796, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6588235294117647, | |
| "grad_norm": 0.6068108081817627, | |
| "learning_rate": 9.256422207921757e-05, | |
| "loss": 2.5241, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6745098039215687, | |
| "grad_norm": 0.5682300925254822, | |
| "learning_rate": 9.210504424614059e-05, | |
| "loss": 2.4522, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6901960784313725, | |
| "grad_norm": 0.6877283453941345, | |
| "learning_rate": 9.163332113436032e-05, | |
| "loss": 2.5321, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 1.1677114963531494, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 2.8537, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7215686274509804, | |
| "grad_norm": 0.38572782278060913, | |
| "learning_rate": 9.065280497392663e-05, | |
| "loss": 2.4599, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7372549019607844, | |
| "grad_norm": 0.3719371557235718, | |
| "learning_rate": 9.014430407194413e-05, | |
| "loss": 2.4086, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7529411764705882, | |
| "grad_norm": 0.3547096252441406, | |
| "learning_rate": 8.962384209755452e-05, | |
| "loss": 2.5402, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7686274509803922, | |
| "grad_norm": 0.385681688785553, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 2.4216, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 0.39732497930526733, | |
| "learning_rate": 8.854765873974898e-05, | |
| "loss": 2.597, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "eval_loss": 2.5103862285614014, | |
| "eval_runtime": 8.1694, | |
| "eval_samples_per_second": 13.22, | |
| "eval_steps_per_second": 3.305, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.382482647895813, | |
| "learning_rate": 8.799225800722895e-05, | |
| "loss": 2.4412, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8156862745098039, | |
| "grad_norm": 0.4492253065109253, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.4853, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8313725490196079, | |
| "grad_norm": 0.40429365634918213, | |
| "learning_rate": 8.684766579921684e-05, | |
| "loss": 2.4598, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8470588235294118, | |
| "grad_norm": 0.4485275447368622, | |
| "learning_rate": 8.625881535716883e-05, | |
| "loss": 2.4755, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8627450980392157, | |
| "grad_norm": 0.44603973627090454, | |
| "learning_rate": 8.565916153152983e-05, | |
| "loss": 2.4749, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8784313725490196, | |
| "grad_norm": 0.4761582612991333, | |
| "learning_rate": 8.504888299030747e-05, | |
| "loss": 2.6098, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8941176470588236, | |
| "grad_norm": 0.510277271270752, | |
| "learning_rate": 8.442816156716385e-05, | |
| "loss": 2.5405, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9098039215686274, | |
| "grad_norm": 0.5629740357398987, | |
| "learning_rate": 8.379718220723773e-05, | |
| "loss": 2.5641, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9254901960784314, | |
| "grad_norm": 0.6572622060775757, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 2.5844, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.9117251634597778, | |
| "learning_rate": 8.250520468343722e-05, | |
| "loss": 2.9142, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9568627450980393, | |
| "grad_norm": 0.30293720960617065, | |
| "learning_rate": 8.184459146674446e-05, | |
| "loss": 2.3538, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9725490196078431, | |
| "grad_norm": 0.37605491280555725, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 2.3623, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9882352941176471, | |
| "grad_norm": 0.5062958002090454, | |
| "learning_rate": 8.049510022000364e-05, | |
| "loss": 2.5813, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.003921568627451, | |
| "grad_norm": 0.8249017596244812, | |
| "learning_rate": 7.980662427346127e-05, | |
| "loss": 3.606, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0196078431372548, | |
| "grad_norm": 0.24899208545684814, | |
| "learning_rate": 7.910926738603854e-05, | |
| "loss": 2.1644, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.035294117647059, | |
| "grad_norm": 0.2910781502723694, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 2.2541, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0509803921568628, | |
| "grad_norm": 0.3324323892593384, | |
| "learning_rate": 7.768874448802665e-05, | |
| "loss": 2.4114, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.3214768171310425, | |
| "learning_rate": 7.696600172495997e-05, | |
| "loss": 2.2752, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0823529411764705, | |
| "grad_norm": 0.34132134914398193, | |
| "learning_rate": 7.62352243899504e-05, | |
| "loss": 2.3824, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0980392156862746, | |
| "grad_norm": 0.3731231689453125, | |
| "learning_rate": 7.54966302195068e-05, | |
| "loss": 2.4044, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1137254901960785, | |
| "grad_norm": 0.3671027421951294, | |
| "learning_rate": 7.475043927917907e-05, | |
| "loss": 2.352, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1294117647058823, | |
| "grad_norm": 0.3711940348148346, | |
| "learning_rate": 7.399687389798933e-05, | |
| "loss": 2.2605, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1450980392156862, | |
| "grad_norm": 0.41749900579452515, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 2.36, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1607843137254903, | |
| "grad_norm": 0.4730680286884308, | |
| "learning_rate": 7.246852004835807e-05, | |
| "loss": 2.3771, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.482975035905838, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 2.3348, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.192156862745098, | |
| "grad_norm": 0.5463060140609741, | |
| "learning_rate": 7.091339003877826e-05, | |
| "loss": 2.318, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2078431372549019, | |
| "grad_norm": 0.6054916381835938, | |
| "learning_rate": 7.012636193699837e-05, | |
| "loss": 2.1102, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.223529411764706, | |
| "grad_norm": 0.719634473323822, | |
| "learning_rate": 6.933333714707094e-05, | |
| "loss": 2.3898, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2392156862745098, | |
| "grad_norm": 0.8080974221229553, | |
| "learning_rate": 6.853455195225338e-05, | |
| "loss": 2.6727, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2549019607843137, | |
| "grad_norm": 0.33123213052749634, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 2.1137, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2705882352941176, | |
| "grad_norm": 0.416637659072876, | |
| "learning_rate": 6.692065399168352e-05, | |
| "loss": 2.3176, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2862745098039214, | |
| "grad_norm": 0.415536493062973, | |
| "learning_rate": 6.610602208992454e-05, | |
| "loss": 2.2767, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3019607843137255, | |
| "grad_norm": 0.42290326952934265, | |
| "learning_rate": 6.528659136798764e-05, | |
| "loss": 2.277, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3176470588235294, | |
| "grad_norm": 0.4520817697048187, | |
| "learning_rate": 6.446260597682839e-05, | |
| "loss": 2.3805, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.46147623658180237, | |
| "learning_rate": 6.363431142447469e-05, | |
| "loss": 2.3688, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3490196078431373, | |
| "grad_norm": 0.48055508732795715, | |
| "learning_rate": 6.280195450287736e-05, | |
| "loss": 2.3444, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3647058823529412, | |
| "grad_norm": 0.477186381816864, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 2.4343, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.380392156862745, | |
| "grad_norm": 0.4650111794471741, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 2.2705, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.396078431372549, | |
| "grad_norm": 0.5212112069129944, | |
| "learning_rate": 6.028299515429683e-05, | |
| "loss": 2.3152, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.5310537815093994, | |
| "learning_rate": 5.943687977264584e-05, | |
| "loss": 2.3124, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.427450980392157, | |
| "grad_norm": 0.60441654920578, | |
| "learning_rate": 5.8587952654563817e-05, | |
| "loss": 2.3926, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4431372549019608, | |
| "grad_norm": 0.6553325057029724, | |
| "learning_rate": 5.773646673951406e-05, | |
| "loss": 2.2566, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4588235294117646, | |
| "grad_norm": 0.8707178235054016, | |
| "learning_rate": 5.688267572935842e-05, | |
| "loss": 2.5228, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.4745098039215687, | |
| "grad_norm": 0.844674825668335, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 2.5466, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4901960784313726, | |
| "grad_norm": 0.3334199786186218, | |
| "learning_rate": 5.5169196589418504e-05, | |
| "loss": 2.0263, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5058823529411764, | |
| "grad_norm": 0.4047907888889313, | |
| "learning_rate": 5.431001899403098e-05, | |
| "loss": 2.3, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5215686274509803, | |
| "grad_norm": 0.42536845803260803, | |
| "learning_rate": 5.344955722021624e-05, | |
| "loss": 2.3235, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5372549019607842, | |
| "grad_norm": 0.42733073234558105, | |
| "learning_rate": 5.258806764421048e-05, | |
| "loss": 2.2457, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5529411764705883, | |
| "grad_norm": 0.47130250930786133, | |
| "learning_rate": 5.172580694848541e-05, | |
| "loss": 2.3569, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "grad_norm": 0.49199551343917847, | |
| "learning_rate": 5.086303204526943e-05, | |
| "loss": 2.3505, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "eval_loss": 2.4975922107696533, | |
| "eval_runtime": 8.1691, | |
| "eval_samples_per_second": 13.221, | |
| "eval_steps_per_second": 3.305, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5843137254901962, | |
| "grad_norm": 0.5012519359588623, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2786, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5001832842826843, | |
| "learning_rate": 4.913696795473058e-05, | |
| "loss": 2.2554, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.615686274509804, | |
| "grad_norm": 0.5830171704292297, | |
| "learning_rate": 4.827419305151461e-05, | |
| "loss": 2.298, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6313725490196078, | |
| "grad_norm": 0.5844489336013794, | |
| "learning_rate": 4.741193235578952e-05, | |
| "loss": 2.4365, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.6321450471878052, | |
| "learning_rate": 4.655044277978375e-05, | |
| "loss": 2.3374, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6627450980392156, | |
| "grad_norm": 0.6509003639221191, | |
| "learning_rate": 4.568998100596903e-05, | |
| "loss": 2.3349, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6784313725490196, | |
| "grad_norm": 0.7752017378807068, | |
| "learning_rate": 4.48308034105815e-05, | |
| "loss": 2.2024, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6941176470588235, | |
| "grad_norm": 0.8786823153495789, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 2.4999, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7098039215686276, | |
| "grad_norm": 0.808777928352356, | |
| "learning_rate": 4.31173242706416e-05, | |
| "loss": 2.6242, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7254901960784315, | |
| "grad_norm": 0.3257434666156769, | |
| "learning_rate": 4.226353326048593e-05, | |
| "loss": 1.9046, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7411764705882353, | |
| "grad_norm": 0.4144049882888794, | |
| "learning_rate": 4.1412047345436195e-05, | |
| "loss": 2.2418, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7568627450980392, | |
| "grad_norm": 0.4396112859249115, | |
| "learning_rate": 4.056312022735417e-05, | |
| "loss": 2.3271, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.772549019607843, | |
| "grad_norm": 0.4944157302379608, | |
| "learning_rate": 3.971700484570318e-05, | |
| "loss": 2.3628, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.788235294117647, | |
| "grad_norm": 0.4805627465248108, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 2.3454, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.803921568627451, | |
| "grad_norm": 0.5068280100822449, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 2.213, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8196078431372549, | |
| "grad_norm": 0.533676028251648, | |
| "learning_rate": 3.719804549712265e-05, | |
| "loss": 2.2378, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.835294117647059, | |
| "grad_norm": 0.5281825065612793, | |
| "learning_rate": 3.6365688575525315e-05, | |
| "loss": 2.2165, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.8509803921568628, | |
| "grad_norm": 0.5699700713157654, | |
| "learning_rate": 3.553739402317162e-05, | |
| "loss": 2.3169, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 0.590506911277771, | |
| "learning_rate": 3.471340863201237e-05, | |
| "loss": 2.2439, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.6972747445106506, | |
| "learning_rate": 3.389397791007548e-05, | |
| "loss": 2.2499, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8980392156862744, | |
| "grad_norm": 0.7098711133003235, | |
| "learning_rate": 3.307934600831648e-05, | |
| "loss": 2.1955, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9137254901960783, | |
| "grad_norm": 0.8078807592391968, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.3609, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9294117647058824, | |
| "grad_norm": 0.9425795674324036, | |
| "learning_rate": 3.146544804774663e-05, | |
| "loss": 2.2705, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9450980392156862, | |
| "grad_norm": 0.9871799349784851, | |
| "learning_rate": 3.066666285292906e-05, | |
| "loss": 2.5466, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9607843137254903, | |
| "grad_norm": 0.38264453411102295, | |
| "learning_rate": 2.9873638063001628e-05, | |
| "loss": 1.9602, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9764705882352942, | |
| "grad_norm": 0.5518760085105896, | |
| "learning_rate": 2.9086609961221755e-05, | |
| "loss": 2.3357, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.992156862745098, | |
| "grad_norm": 0.9333047866821289, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 2.5903, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.007843137254902, | |
| "grad_norm": 1.0014146566390991, | |
| "learning_rate": 2.7531479951641924e-05, | |
| "loss": 2.6325, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.023529411764706, | |
| "grad_norm": 0.38848909735679626, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 2.0197, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0392156862745097, | |
| "grad_norm": 0.4253176748752594, | |
| "learning_rate": 2.6003126102010695e-05, | |
| "loss": 2.1717, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0549019607843135, | |
| "grad_norm": 0.44695788621902466, | |
| "learning_rate": 2.5249560720820932e-05, | |
| "loss": 2.2973, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.070588235294118, | |
| "grad_norm": 0.4659847319126129, | |
| "learning_rate": 2.450336978049322e-05, | |
| "loss": 2.1405, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0862745098039217, | |
| "grad_norm": 0.4852655529975891, | |
| "learning_rate": 2.37647756100496e-05, | |
| "loss": 2.1411, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.1019607843137256, | |
| "grad_norm": 0.5167310833930969, | |
| "learning_rate": 2.3033998275040046e-05, | |
| "loss": 2.2418, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.5107938647270203, | |
| "learning_rate": 2.2311255511973345e-05, | |
| "loss": 2.0234, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.5627231597900391, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 2.2439, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.149019607843137, | |
| "grad_norm": 0.5920283794403076, | |
| "learning_rate": 2.0890732613961478e-05, | |
| "loss": 2.1049, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.164705882352941, | |
| "grad_norm": 0.6250417232513428, | |
| "learning_rate": 2.0193375726538737e-05, | |
| "loss": 2.0543, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.180392156862745, | |
| "grad_norm": 0.7671651840209961, | |
| "learning_rate": 1.9504899779996355e-05, | |
| "loss": 2.013, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.196078431372549, | |
| "grad_norm": 0.7899158000946045, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 1.9573, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.211764705882353, | |
| "grad_norm": 0.935232937335968, | |
| "learning_rate": 1.8155408533255553e-05, | |
| "loss": 2.1216, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.227450980392157, | |
| "grad_norm": 1.1227110624313354, | |
| "learning_rate": 1.749479531656279e-05, | |
| "loss": 2.005, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.243137254901961, | |
| "grad_norm": 0.7702217698097229, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 2.2587, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2588235294117647, | |
| "grad_norm": 0.4433023929595947, | |
| "learning_rate": 1.6202817792762282e-05, | |
| "loss": 2.1471, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.2745098039215685, | |
| "grad_norm": 0.48942095041275024, | |
| "learning_rate": 1.557183843283614e-05, | |
| "loss": 2.2903, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.2901960784313724, | |
| "grad_norm": 0.5090365409851074, | |
| "learning_rate": 1.4951117009692528e-05, | |
| "loss": 2.2312, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.3058823529411763, | |
| "grad_norm": 0.5861947536468506, | |
| "learning_rate": 1.4340838468470197e-05, | |
| "loss": 2.1935, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3215686274509806, | |
| "grad_norm": 0.6123110055923462, | |
| "learning_rate": 1.3741184642831189e-05, | |
| "loss": 2.3119, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.3372549019607844, | |
| "grad_norm": 0.6334862112998962, | |
| "learning_rate": 1.3152334200783167e-05, | |
| "loss": 2.2283, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.6538926362991333, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 2.1705, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "eval_loss": 2.53293514251709, | |
| "eval_runtime": 8.1691, | |
| "eval_samples_per_second": 13.221, | |
| "eval_steps_per_second": 3.305, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.368627450980392, | |
| "grad_norm": 0.6534598469734192, | |
| "learning_rate": 1.2007741992771065e-05, | |
| "loss": 2.0549, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.384313725490196, | |
| "grad_norm": 0.7089661359786987, | |
| "learning_rate": 1.145234126025102e-05, | |
| "loss": 2.11, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.7214992642402649, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 2.0214, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.4156862745098038, | |
| "grad_norm": 0.8083807826042175, | |
| "learning_rate": 1.0376157902445488e-05, | |
| "loss": 1.9978, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.431372549019608, | |
| "grad_norm": 0.8732563853263855, | |
| "learning_rate": 9.85569592805588e-06, | |
| "loss": 2.0557, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.447058823529412, | |
| "grad_norm": 0.9465264678001404, | |
| "learning_rate": 9.347195026073369e-06, | |
| "loss": 1.9518, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.462745098039216, | |
| "grad_norm": 1.2992017269134521, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 2.0409, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.4784313725490197, | |
| "grad_norm": 0.8976895809173584, | |
| "learning_rate": 8.366678865639688e-06, | |
| "loss": 2.1384, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.4941176470588236, | |
| "grad_norm": 0.4451674520969391, | |
| "learning_rate": 7.894955753859413e-06, | |
| "loss": 1.9495, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.5098039215686274, | |
| "grad_norm": 0.5084771513938904, | |
| "learning_rate": 7.435777920782444e-06, | |
| "loss": 2.1625, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5254901960784313, | |
| "grad_norm": 0.5429627895355225, | |
| "learning_rate": 6.989282179324963e-06, | |
| "loss": 2.1547, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.541176470588235, | |
| "grad_norm": 0.5458650588989258, | |
| "learning_rate": 6.555601563749675e-06, | |
| "loss": 2.1326, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.556862745098039, | |
| "grad_norm": 0.6028746366500854, | |
| "learning_rate": 6.1348652900279025e-06, | |
| "loss": 2.2729, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.572549019607843, | |
| "grad_norm": 0.6017650365829468, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 2.1029, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.6778936982154846, | |
| "learning_rate": 5.332723310721854e-06, | |
| "loss": 2.155, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.603921568627451, | |
| "grad_norm": 0.716666042804718, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 2.1131, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.619607843137255, | |
| "grad_norm": 0.7246668338775635, | |
| "learning_rate": 4.5838121691623e-06, | |
| "loss": 2.0944, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.635294117647059, | |
| "grad_norm": 0.8063449263572693, | |
| "learning_rate": 4.229599573731685e-06, | |
| "loss": 2.1857, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.6509803921568627, | |
| "grad_norm": 0.8714430332183838, | |
| "learning_rate": 3.8890243569094874e-06, | |
| "loss": 2.2817, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.9104402661323547, | |
| "learning_rate": 3.5621879937348836e-06, | |
| "loss": 2.0639, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.682352941176471, | |
| "grad_norm": 0.9760646820068359, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 1.984, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.6980392156862747, | |
| "grad_norm": 1.343449354171753, | |
| "learning_rate": 2.950117231881183e-06, | |
| "loss": 2.0159, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.7137254901960786, | |
| "grad_norm": 0.9933635592460632, | |
| "learning_rate": 2.6650652008597068e-06, | |
| "loss": 2.3773, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.7294117647058824, | |
| "grad_norm": 0.46657246351242065, | |
| "learning_rate": 2.3941167044642944e-06, | |
| "loss": 1.8924, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.7450980392156863, | |
| "grad_norm": 0.540969967842102, | |
| "learning_rate": 2.137352472319215e-06, | |
| "loss": 2.2537, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.76078431372549, | |
| "grad_norm": 0.5452009439468384, | |
| "learning_rate": 1.8948490078199764e-06, | |
| "loss": 2.2074, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.776470588235294, | |
| "grad_norm": 0.5477240681648254, | |
| "learning_rate": 1.6666785653390249e-06, | |
| "loss": 2.1253, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.792156862745098, | |
| "grad_norm": 0.6009814739227295, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 2.2472, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.8078431372549018, | |
| "grad_norm": 0.6575978994369507, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 2.2573, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.6617922782897949, | |
| "learning_rate": 1.0688237352022345e-06, | |
| "loss": 2.2366, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.83921568627451, | |
| "grad_norm": 0.6877896785736084, | |
| "learning_rate": 8.986222173284875e-07, | |
| "loss": 2.1161, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.854901960784314, | |
| "grad_norm": 0.7269079089164734, | |
| "learning_rate": 7.4305054915631e-07, | |
| "loss": 2.1263, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.8705882352941177, | |
| "grad_norm": 0.7962198257446289, | |
| "learning_rate": 6.021550835626777e-07, | |
| "loss": 2.1707, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.8862745098039215, | |
| "grad_norm": 0.856226921081543, | |
| "learning_rate": 4.7597780062184073e-07, | |
| "loss": 2.1721, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.9019607843137254, | |
| "grad_norm": 0.9060380458831787, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 1.9191, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.9176470588235293, | |
| "grad_norm": 0.9597678184509277, | |
| "learning_rate": 2.6792376524036877e-07, | |
| "loss": 1.8294, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 1.1927050352096558, | |
| "learning_rate": 1.8610900289867673e-07, | |
| "loss": 2.0034, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.9490196078431374, | |
| "grad_norm": 0.9366448521614075, | |
| "learning_rate": 1.191363849376237e-07, | |
| "loss": 2.2449, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.9647058823529413, | |
| "grad_norm": 0.531377375125885, | |
| "learning_rate": 6.702586597719385e-08, | |
| "loss": 1.838, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.980392156862745, | |
| "grad_norm": 0.7152462005615234, | |
| "learning_rate": 2.9792972446479605e-08, | |
| "loss": 2.2187, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.996078431372549, | |
| "grad_norm": 1.2911417484283447, | |
| "learning_rate": 7.448797957526621e-09, | |
| "loss": 2.4433, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.011764705882353, | |
| "grad_norm": 0.6130143404006958, | |
| "learning_rate": 0.0, | |
| "loss": 2.2335, | |
| "step": 192 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 192, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.62298517110784e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |