| { | |
| "best_metric": 2.497835874557495, | |
| "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": 0.9195303320884705, | |
| "learning_rate": 1e-05, | |
| "loss": 2.8736, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01568627450980392, | |
| "eval_loss": 3.324666738510132, | |
| "eval_runtime": 7.9726, | |
| "eval_samples_per_second": 13.546, | |
| "eval_steps_per_second": 3.387, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03137254901960784, | |
| "grad_norm": 1.0334582328796387, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8739, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.047058823529411764, | |
| "grad_norm": 1.0187386274337769, | |
| "learning_rate": 3e-05, | |
| "loss": 2.9073, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06274509803921569, | |
| "grad_norm": 0.9314672946929932, | |
| "learning_rate": 4e-05, | |
| "loss": 2.9041, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 0.9586840867996216, | |
| "learning_rate": 5e-05, | |
| "loss": 2.807, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09411764705882353, | |
| "grad_norm": 0.9514013528823853, | |
| "learning_rate": 6e-05, | |
| "loss": 2.996, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10980392156862745, | |
| "grad_norm": 0.8477784395217896, | |
| "learning_rate": 7e-05, | |
| "loss": 2.9011, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12549019607843137, | |
| "grad_norm": 0.898993968963623, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7146, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1411764705882353, | |
| "grad_norm": 0.8038886785507202, | |
| "learning_rate": 9e-05, | |
| "loss": 2.7318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 0.8036936521530151, | |
| "learning_rate": 0.0001, | |
| "loss": 2.6728, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17254901960784313, | |
| "grad_norm": 0.9414847493171692, | |
| "learning_rate": 9.999255120204248e-05, | |
| "loss": 2.6703, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.18823529411764706, | |
| "grad_norm": 1.016950249671936, | |
| "learning_rate": 9.997020702755353e-05, | |
| "loss": 2.7351, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20392156862745098, | |
| "grad_norm": 1.0459362268447876, | |
| "learning_rate": 9.99329741340228e-05, | |
| "loss": 2.6783, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2196078431372549, | |
| "grad_norm": 1.2477803230285645, | |
| "learning_rate": 9.98808636150624e-05, | |
| "loss": 2.6327, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 1.4739620685577393, | |
| "learning_rate": 9.981389099710132e-05, | |
| "loss": 3.0748, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25098039215686274, | |
| "grad_norm": 0.5202563405036926, | |
| "learning_rate": 9.973207623475965e-05, | |
| "loss": 2.41, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.6118998527526855, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 2.5087, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2823529411764706, | |
| "grad_norm": 0.5753269791603088, | |
| "learning_rate": 9.952402219937816e-05, | |
| "loss": 2.5023, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2980392156862745, | |
| "grad_norm": 0.548793613910675, | |
| "learning_rate": 9.939784491643734e-05, | |
| "loss": 2.4598, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 0.5158197283744812, | |
| "learning_rate": 9.92569494508437e-05, | |
| "loss": 2.626, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.32941176470588235, | |
| "grad_norm": 0.4742230474948883, | |
| "learning_rate": 9.910137778267152e-05, | |
| "loss": 2.539, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.34509803921568627, | |
| "grad_norm": 0.4577326774597168, | |
| "learning_rate": 9.893117626479777e-05, | |
| "loss": 2.5212, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3607843137254902, | |
| "grad_norm": 0.469974547624588, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 2.4532, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3764705882352941, | |
| "grad_norm": 0.6523585915565491, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 2.4785, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.39215686274509803, | |
| "grad_norm": 0.6335188746452332, | |
| "learning_rate": 9.833332143466099e-05, | |
| "loss": 2.5327, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.40784313725490196, | |
| "grad_norm": 0.5561938881874084, | |
| "learning_rate": 9.810515099218003e-05, | |
| "loss": 2.7133, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4235294117647059, | |
| "grad_norm": 0.6505088210105896, | |
| "learning_rate": 9.78626475276808e-05, | |
| "loss": 2.5733, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4392156862745098, | |
| "grad_norm": 0.7462759613990784, | |
| "learning_rate": 9.760588329553571e-05, | |
| "loss": 2.6319, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4549019607843137, | |
| "grad_norm": 0.8369238376617432, | |
| "learning_rate": 9.73349347991403e-05, | |
| "loss": 2.5938, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 1.1839768886566162, | |
| "learning_rate": 9.704988276811883e-05, | |
| "loss": 2.9516, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48627450980392156, | |
| "grad_norm": 0.29538607597351074, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 2.3553, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5019607843137255, | |
| "grad_norm": 0.40422534942626953, | |
| "learning_rate": 9.643781200626511e-05, | |
| "loss": 2.426, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5176470588235295, | |
| "grad_norm": 0.41036108136177063, | |
| "learning_rate": 9.611097564309053e-05, | |
| "loss": 2.503, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.4156786799430847, | |
| "learning_rate": 9.577040042626833e-05, | |
| "loss": 2.3799, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5490196078431373, | |
| "grad_norm": 0.39962074160575867, | |
| "learning_rate": 9.54161878308377e-05, | |
| "loss": 2.4543, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5647058823529412, | |
| "grad_norm": 0.4289894998073578, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 2.4369, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5803921568627451, | |
| "grad_norm": 0.4545486271381378, | |
| "learning_rate": 9.466727668927816e-05, | |
| "loss": 2.5761, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.596078431372549, | |
| "grad_norm": 0.431194931268692, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 2.4045, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.611764705882353, | |
| "grad_norm": 0.43016883730888367, | |
| "learning_rate": 9.38651347099721e-05, | |
| "loss": 2.3858, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 0.45133844017982483, | |
| "learning_rate": 9.344439843625034e-05, | |
| "loss": 2.4977, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6431372549019608, | |
| "grad_norm": 0.4912121295928955, | |
| "learning_rate": 9.301071782067504e-05, | |
| "loss": 2.4795, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6588235294117647, | |
| "grad_norm": 0.6027780175209045, | |
| "learning_rate": 9.256422207921757e-05, | |
| "loss": 2.5236, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6745098039215687, | |
| "grad_norm": 0.5655380487442017, | |
| "learning_rate": 9.210504424614059e-05, | |
| "loss": 2.4498, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6901960784313725, | |
| "grad_norm": 0.6778488755226135, | |
| "learning_rate": 9.163332113436032e-05, | |
| "loss": 2.5315, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 1.1361249685287476, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 2.8519, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7215686274509804, | |
| "grad_norm": 0.3818633258342743, | |
| "learning_rate": 9.065280497392663e-05, | |
| "loss": 2.4607, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7372549019607844, | |
| "grad_norm": 0.3662630319595337, | |
| "learning_rate": 9.014430407194413e-05, | |
| "loss": 2.4082, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7529411764705882, | |
| "grad_norm": 0.35173290967941284, | |
| "learning_rate": 8.962384209755452e-05, | |
| "loss": 2.5405, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7686274509803922, | |
| "grad_norm": 0.38660287857055664, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 2.4206, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 0.40245190262794495, | |
| "learning_rate": 8.854765873974898e-05, | |
| "loss": 2.5983, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "eval_loss": 2.5114123821258545, | |
| "eval_runtime": 8.157, | |
| "eval_samples_per_second": 13.24, | |
| "eval_steps_per_second": 3.31, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.3828006386756897, | |
| "learning_rate": 8.799225800722895e-05, | |
| "loss": 2.4409, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8156862745098039, | |
| "grad_norm": 0.4464765787124634, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.4858, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8313725490196079, | |
| "grad_norm": 0.4025624692440033, | |
| "learning_rate": 8.684766579921684e-05, | |
| "loss": 2.4596, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8470588235294118, | |
| "grad_norm": 0.4462139904499054, | |
| "learning_rate": 8.625881535716883e-05, | |
| "loss": 2.4744, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8627450980392157, | |
| "grad_norm": 0.4364989995956421, | |
| "learning_rate": 8.565916153152983e-05, | |
| "loss": 2.4725, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8784313725490196, | |
| "grad_norm": 0.4733513593673706, | |
| "learning_rate": 8.504888299030747e-05, | |
| "loss": 2.609, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8941176470588236, | |
| "grad_norm": 0.5065345764160156, | |
| "learning_rate": 8.442816156716385e-05, | |
| "loss": 2.5392, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9098039215686274, | |
| "grad_norm": 0.5636035799980164, | |
| "learning_rate": 8.379718220723773e-05, | |
| "loss": 2.5665, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9254901960784314, | |
| "grad_norm": 0.6569843292236328, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 2.5817, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.907096266746521, | |
| "learning_rate": 8.250520468343722e-05, | |
| "loss": 2.911, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9568627450980393, | |
| "grad_norm": 0.29703453183174133, | |
| "learning_rate": 8.184459146674446e-05, | |
| "loss": 2.3542, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9725490196078431, | |
| "grad_norm": 0.3698834478855133, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 2.3618, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9882352941176471, | |
| "grad_norm": 0.504271388053894, | |
| "learning_rate": 8.049510022000364e-05, | |
| "loss": 2.5812, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.003921568627451, | |
| "grad_norm": 0.8231114745140076, | |
| "learning_rate": 7.980662427346127e-05, | |
| "loss": 3.6064, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0196078431372548, | |
| "grad_norm": 0.24639078974723816, | |
| "learning_rate": 7.910926738603854e-05, | |
| "loss": 2.1655, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.035294117647059, | |
| "grad_norm": 0.29097798466682434, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 2.2555, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0509803921568628, | |
| "grad_norm": 0.3319820165634155, | |
| "learning_rate": 7.768874448802665e-05, | |
| "loss": 2.4114, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.32010725140571594, | |
| "learning_rate": 7.696600172495997e-05, | |
| "loss": 2.2756, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0823529411764705, | |
| "grad_norm": 0.3406822383403778, | |
| "learning_rate": 7.62352243899504e-05, | |
| "loss": 2.3815, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0980392156862746, | |
| "grad_norm": 0.37193605303764343, | |
| "learning_rate": 7.54966302195068e-05, | |
| "loss": 2.406, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1137254901960785, | |
| "grad_norm": 0.36552101373672485, | |
| "learning_rate": 7.475043927917907e-05, | |
| "loss": 2.3517, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1294117647058823, | |
| "grad_norm": 0.3685903251171112, | |
| "learning_rate": 7.399687389798933e-05, | |
| "loss": 2.262, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1450980392156862, | |
| "grad_norm": 0.41463446617126465, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 2.3599, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1607843137254903, | |
| "grad_norm": 0.47038400173187256, | |
| "learning_rate": 7.246852004835807e-05, | |
| "loss": 2.3755, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.4810601770877838, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 2.3339, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.192156862745098, | |
| "grad_norm": 0.5399831533432007, | |
| "learning_rate": 7.091339003877826e-05, | |
| "loss": 2.3184, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2078431372549019, | |
| "grad_norm": 0.6069369316101074, | |
| "learning_rate": 7.012636193699837e-05, | |
| "loss": 2.1125, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.223529411764706, | |
| "grad_norm": 0.7255291938781738, | |
| "learning_rate": 6.933333714707094e-05, | |
| "loss": 2.3926, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2392156862745098, | |
| "grad_norm": 0.819170355796814, | |
| "learning_rate": 6.853455195225338e-05, | |
| "loss": 2.6737, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2549019607843137, | |
| "grad_norm": 0.3163304626941681, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 2.1127, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2705882352941176, | |
| "grad_norm": 0.406746506690979, | |
| "learning_rate": 6.692065399168352e-05, | |
| "loss": 2.3158, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2862745098039214, | |
| "grad_norm": 0.4056803584098816, | |
| "learning_rate": 6.610602208992454e-05, | |
| "loss": 2.2758, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3019607843137255, | |
| "grad_norm": 0.4088509976863861, | |
| "learning_rate": 6.528659136798764e-05, | |
| "loss": 2.2759, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3176470588235294, | |
| "grad_norm": 0.4500698447227478, | |
| "learning_rate": 6.446260597682839e-05, | |
| "loss": 2.3797, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.45922282338142395, | |
| "learning_rate": 6.363431142447469e-05, | |
| "loss": 2.3687, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3490196078431373, | |
| "grad_norm": 0.48098620772361755, | |
| "learning_rate": 6.280195450287736e-05, | |
| "loss": 2.3443, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3647058823529412, | |
| "grad_norm": 0.478667289018631, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 2.4358, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.380392156862745, | |
| "grad_norm": 0.4647475481033325, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 2.2697, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.396078431372549, | |
| "grad_norm": 0.5217877626419067, | |
| "learning_rate": 6.028299515429683e-05, | |
| "loss": 2.3142, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.527504026889801, | |
| "learning_rate": 5.943687977264584e-05, | |
| "loss": 2.3132, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.427450980392157, | |
| "grad_norm": 0.607871413230896, | |
| "learning_rate": 5.8587952654563817e-05, | |
| "loss": 2.3922, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4431372549019608, | |
| "grad_norm": 0.6600663661956787, | |
| "learning_rate": 5.773646673951406e-05, | |
| "loss": 2.2529, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4588235294117646, | |
| "grad_norm": 0.8639673590660095, | |
| "learning_rate": 5.688267572935842e-05, | |
| "loss": 2.5246, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.4745098039215687, | |
| "grad_norm": 0.8424491286277771, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 2.5439, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4901960784313726, | |
| "grad_norm": 0.33690783381462097, | |
| "learning_rate": 5.5169196589418504e-05, | |
| "loss": 2.0268, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5058823529411764, | |
| "grad_norm": 0.40822649002075195, | |
| "learning_rate": 5.431001899403098e-05, | |
| "loss": 2.3005, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5215686274509803, | |
| "grad_norm": 0.42489108443260193, | |
| "learning_rate": 5.344955722021624e-05, | |
| "loss": 2.3231, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5372549019607842, | |
| "grad_norm": 0.43306756019592285, | |
| "learning_rate": 5.258806764421048e-05, | |
| "loss": 2.2458, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5529411764705883, | |
| "grad_norm": 0.4661814868450165, | |
| "learning_rate": 5.172580694848541e-05, | |
| "loss": 2.3549, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "grad_norm": 0.4894493520259857, | |
| "learning_rate": 5.086303204526943e-05, | |
| "loss": 2.3509, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "eval_loss": 2.497835874557495, | |
| "eval_runtime": 8.1646, | |
| "eval_samples_per_second": 13.228, | |
| "eval_steps_per_second": 3.307, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5843137254901962, | |
| "grad_norm": 0.5029212832450867, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2764, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.49811941385269165, | |
| "learning_rate": 4.913696795473058e-05, | |
| "loss": 2.2556, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.615686274509804, | |
| "grad_norm": 0.5826261043548584, | |
| "learning_rate": 4.827419305151461e-05, | |
| "loss": 2.2982, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6313725490196078, | |
| "grad_norm": 0.5830192565917969, | |
| "learning_rate": 4.741193235578952e-05, | |
| "loss": 2.4361, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.6307199597358704, | |
| "learning_rate": 4.655044277978375e-05, | |
| "loss": 2.3384, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6627450980392156, | |
| "grad_norm": 0.652266263961792, | |
| "learning_rate": 4.568998100596903e-05, | |
| "loss": 2.3353, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6784313725490196, | |
| "grad_norm": 0.7673619389533997, | |
| "learning_rate": 4.48308034105815e-05, | |
| "loss": 2.2034, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6941176470588235, | |
| "grad_norm": 0.8802884221076965, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 2.4979, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7098039215686276, | |
| "grad_norm": 0.8109568357467651, | |
| "learning_rate": 4.31173242706416e-05, | |
| "loss": 2.626, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7254901960784315, | |
| "grad_norm": 0.32692885398864746, | |
| "learning_rate": 4.226353326048593e-05, | |
| "loss": 1.9041, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7411764705882353, | |
| "grad_norm": 0.4149419069290161, | |
| "learning_rate": 4.1412047345436195e-05, | |
| "loss": 2.2416, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7568627450980392, | |
| "grad_norm": 0.4411557614803314, | |
| "learning_rate": 4.056312022735417e-05, | |
| "loss": 2.3267, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.772549019607843, | |
| "grad_norm": 0.49479031562805176, | |
| "learning_rate": 3.971700484570318e-05, | |
| "loss": 2.3624, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.788235294117647, | |
| "grad_norm": 0.4794876277446747, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 2.3468, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.803921568627451, | |
| "grad_norm": 0.5065494775772095, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 2.2112, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8196078431372549, | |
| "grad_norm": 0.5268806219100952, | |
| "learning_rate": 3.719804549712265e-05, | |
| "loss": 2.238, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.835294117647059, | |
| "grad_norm": 0.5357680320739746, | |
| "learning_rate": 3.6365688575525315e-05, | |
| "loss": 2.2156, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.8509803921568628, | |
| "grad_norm": 0.5712677240371704, | |
| "learning_rate": 3.553739402317162e-05, | |
| "loss": 2.3148, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 0.5927073955535889, | |
| "learning_rate": 3.471340863201237e-05, | |
| "loss": 2.2461, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.7000199556350708, | |
| "learning_rate": 3.389397791007548e-05, | |
| "loss": 2.2504, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8980392156862744, | |
| "grad_norm": 0.7094666361808777, | |
| "learning_rate": 3.307934600831648e-05, | |
| "loss": 2.198, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9137254901960783, | |
| "grad_norm": 0.8061943650245667, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.3629, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9294117647058824, | |
| "grad_norm": 0.9445766806602478, | |
| "learning_rate": 3.146544804774663e-05, | |
| "loss": 2.2729, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9450980392156862, | |
| "grad_norm": 0.9857886433601379, | |
| "learning_rate": 3.066666285292906e-05, | |
| "loss": 2.5454, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9607843137254903, | |
| "grad_norm": 0.38298577070236206, | |
| "learning_rate": 2.9873638063001628e-05, | |
| "loss": 1.9603, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9764705882352942, | |
| "grad_norm": 0.5521575808525085, | |
| "learning_rate": 2.9086609961221755e-05, | |
| "loss": 2.3363, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.992156862745098, | |
| "grad_norm": 0.9297695755958557, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 2.5893, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.007843137254902, | |
| "grad_norm": 0.9833527207374573, | |
| "learning_rate": 2.7531479951641924e-05, | |
| "loss": 2.6296, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.023529411764706, | |
| "grad_norm": 0.3908270597457886, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 2.0195, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0392156862745097, | |
| "grad_norm": 0.42692381143569946, | |
| "learning_rate": 2.6003126102010695e-05, | |
| "loss": 2.1701, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0549019607843135, | |
| "grad_norm": 0.44662317633628845, | |
| "learning_rate": 2.5249560720820932e-05, | |
| "loss": 2.2962, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.070588235294118, | |
| "grad_norm": 0.4669097363948822, | |
| "learning_rate": 2.450336978049322e-05, | |
| "loss": 2.1407, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0862745098039217, | |
| "grad_norm": 0.48035910725593567, | |
| "learning_rate": 2.37647756100496e-05, | |
| "loss": 2.1402, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.1019607843137256, | |
| "grad_norm": 0.5153262615203857, | |
| "learning_rate": 2.3033998275040046e-05, | |
| "loss": 2.2411, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.511375367641449, | |
| "learning_rate": 2.2311255511973345e-05, | |
| "loss": 2.0221, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.5629892349243164, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 2.2439, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.149019607843137, | |
| "grad_norm": 0.5920019745826721, | |
| "learning_rate": 2.0890732613961478e-05, | |
| "loss": 2.1041, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.164705882352941, | |
| "grad_norm": 0.6213743090629578, | |
| "learning_rate": 2.0193375726538737e-05, | |
| "loss": 2.0556, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.180392156862745, | |
| "grad_norm": 0.7653319239616394, | |
| "learning_rate": 1.9504899779996355e-05, | |
| "loss": 2.0099, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.196078431372549, | |
| "grad_norm": 0.7903580665588379, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 1.9591, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.211764705882353, | |
| "grad_norm": 0.9329378604888916, | |
| "learning_rate": 1.8155408533255553e-05, | |
| "loss": 2.1217, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.227450980392157, | |
| "grad_norm": 1.1085801124572754, | |
| "learning_rate": 1.749479531656279e-05, | |
| "loss": 2.0027, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.243137254901961, | |
| "grad_norm": 0.7666586637496948, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 2.2613, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2588235294117647, | |
| "grad_norm": 0.4413524270057678, | |
| "learning_rate": 1.6202817792762282e-05, | |
| "loss": 2.1461, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.2745098039215685, | |
| "grad_norm": 0.48825109004974365, | |
| "learning_rate": 1.557183843283614e-05, | |
| "loss": 2.2913, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.2901960784313724, | |
| "grad_norm": 0.5957844853401184, | |
| "learning_rate": 1.4951117009692528e-05, | |
| "loss": 2.2318, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.3058823529411763, | |
| "grad_norm": 0.5846164226531982, | |
| "learning_rate": 1.4340838468470197e-05, | |
| "loss": 2.193, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3215686274509806, | |
| "grad_norm": 0.6084363460540771, | |
| "learning_rate": 1.3741184642831189e-05, | |
| "loss": 2.3104, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.3372549019607844, | |
| "grad_norm": 0.630045473575592, | |
| "learning_rate": 1.3152334200783167e-05, | |
| "loss": 2.2282, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.6543457508087158, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 2.1706, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "eval_loss": 2.531811237335205, | |
| "eval_runtime": 8.1637, | |
| "eval_samples_per_second": 13.229, | |
| "eval_steps_per_second": 3.307, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.368627450980392, | |
| "grad_norm": 0.6354141235351562, | |
| "learning_rate": 1.2007741992771065e-05, | |
| "loss": 2.0553, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.384313725490196, | |
| "grad_norm": 0.7069724798202515, | |
| "learning_rate": 1.145234126025102e-05, | |
| "loss": 2.1107, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.7220402359962463, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 2.0186, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.4156862745098038, | |
| "grad_norm": 0.8067901134490967, | |
| "learning_rate": 1.0376157902445488e-05, | |
| "loss": 1.997, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.431372549019608, | |
| "grad_norm": 0.8723448514938354, | |
| "learning_rate": 9.85569592805588e-06, | |
| "loss": 2.0568, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.447058823529412, | |
| "grad_norm": 0.9434618949890137, | |
| "learning_rate": 9.347195026073369e-06, | |
| "loss": 1.9484, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.462745098039216, | |
| "grad_norm": 1.2919942140579224, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 2.0413, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.4784313725490197, | |
| "grad_norm": 0.8914510607719421, | |
| "learning_rate": 8.366678865639688e-06, | |
| "loss": 2.136, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.4941176470588236, | |
| "grad_norm": 0.45241811871528625, | |
| "learning_rate": 7.894955753859413e-06, | |
| "loss": 1.95, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.5098039215686274, | |
| "grad_norm": 0.5093364715576172, | |
| "learning_rate": 7.435777920782444e-06, | |
| "loss": 2.1615, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5254901960784313, | |
| "grad_norm": 0.5468050837516785, | |
| "learning_rate": 6.989282179324963e-06, | |
| "loss": 2.1543, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.541176470588235, | |
| "grad_norm": 0.5462315082550049, | |
| "learning_rate": 6.555601563749675e-06, | |
| "loss": 2.1315, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.556862745098039, | |
| "grad_norm": 0.6020755767822266, | |
| "learning_rate": 6.1348652900279025e-06, | |
| "loss": 2.2727, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.572549019607843, | |
| "grad_norm": 0.6015859246253967, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 2.1035, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.6677731871604919, | |
| "learning_rate": 5.332723310721854e-06, | |
| "loss": 2.1539, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.603921568627451, | |
| "grad_norm": 0.7184847593307495, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 2.1104, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.619607843137255, | |
| "grad_norm": 0.7267415523529053, | |
| "learning_rate": 4.5838121691623e-06, | |
| "loss": 2.0935, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.635294117647059, | |
| "grad_norm": 0.8061967492103577, | |
| "learning_rate": 4.229599573731685e-06, | |
| "loss": 2.186, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.6509803921568627, | |
| "grad_norm": 0.872096836566925, | |
| "learning_rate": 3.8890243569094874e-06, | |
| "loss": 2.2811, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.9103512167930603, | |
| "learning_rate": 3.5621879937348836e-06, | |
| "loss": 2.0644, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.682352941176471, | |
| "grad_norm": 0.9733991622924805, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 1.9821, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.6980392156862747, | |
| "grad_norm": 1.33379065990448, | |
| "learning_rate": 2.950117231881183e-06, | |
| "loss": 2.0174, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.7137254901960786, | |
| "grad_norm": 0.9886175394058228, | |
| "learning_rate": 2.6650652008597068e-06, | |
| "loss": 2.3795, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.7294117647058824, | |
| "grad_norm": 0.4592796266078949, | |
| "learning_rate": 2.3941167044642944e-06, | |
| "loss": 1.892, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.7450980392156863, | |
| "grad_norm": 0.5413078665733337, | |
| "learning_rate": 2.137352472319215e-06, | |
| "loss": 2.2523, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.76078431372549, | |
| "grad_norm": 0.5450495481491089, | |
| "learning_rate": 1.8948490078199764e-06, | |
| "loss": 2.207, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.776470588235294, | |
| "grad_norm": 0.5484597086906433, | |
| "learning_rate": 1.6666785653390249e-06, | |
| "loss": 2.1261, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.792156862745098, | |
| "grad_norm": 0.6012923121452332, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 2.2482, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.8078431372549018, | |
| "grad_norm": 0.6597626805305481, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 2.2572, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.6608418822288513, | |
| "learning_rate": 1.0688237352022345e-06, | |
| "loss": 2.2363, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.83921568627451, | |
| "grad_norm": 0.6887637376785278, | |
| "learning_rate": 8.986222173284875e-07, | |
| "loss": 2.1163, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.854901960784314, | |
| "grad_norm": 0.728211522102356, | |
| "learning_rate": 7.4305054915631e-07, | |
| "loss": 2.1268, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.8705882352941177, | |
| "grad_norm": 0.796875536441803, | |
| "learning_rate": 6.021550835626777e-07, | |
| "loss": 2.1706, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.8862745098039215, | |
| "grad_norm": 0.855240523815155, | |
| "learning_rate": 4.7597780062184073e-07, | |
| "loss": 2.1735, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.9019607843137254, | |
| "grad_norm": 0.899287760257721, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 1.9215, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.9176470588235293, | |
| "grad_norm": 0.957107663154602, | |
| "learning_rate": 2.6792376524036877e-07, | |
| "loss": 1.8281, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 1.1885648965835571, | |
| "learning_rate": 1.8610900289867673e-07, | |
| "loss": 2.0065, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.9490196078431374, | |
| "grad_norm": 0.9298829436302185, | |
| "learning_rate": 1.191363849376237e-07, | |
| "loss": 2.2486, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.9647058823529413, | |
| "grad_norm": 0.5273903608322144, | |
| "learning_rate": 6.702586597719385e-08, | |
| "loss": 1.838, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.980392156862745, | |
| "grad_norm": 0.7140518426895142, | |
| "learning_rate": 2.9792972446479605e-08, | |
| "loss": 2.2171, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.996078431372549, | |
| "grad_norm": 1.283398151397705, | |
| "learning_rate": 7.448797957526621e-09, | |
| "loss": 2.4425, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.011764705882353, | |
| "grad_norm": 0.6145690083503723, | |
| "learning_rate": 0.0, | |
| "loss": 2.2341, | |
| "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 | |
| } | |