| { | |
| "best_global_step": 25000, | |
| "best_metric": 2.073448419570923, | |
| "best_model_checkpoint": "./qwen3-150m-multilingual-checkpoints\\checkpoint-25000", | |
| "epoch": 0.2157269242302324, | |
| "eval_steps": 1000, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0004314538484604648, | |
| "grad_norm": 6.4089813232421875, | |
| "learning_rate": 9.8e-07, | |
| "loss": 10.257, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0008629076969209296, | |
| "grad_norm": 3.5916900634765625, | |
| "learning_rate": 1.98e-06, | |
| "loss": 9.1849, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0012943615453813945, | |
| "grad_norm": 3.186964988708496, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 8.3015, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0017258153938418592, | |
| "grad_norm": 2.7763795852661133, | |
| "learning_rate": 3.98e-06, | |
| "loss": 7.9443, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.002157269242302324, | |
| "grad_norm": 3.0714266300201416, | |
| "learning_rate": 4.98e-06, | |
| "loss": 7.3653, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.002588723090762789, | |
| "grad_norm": 2.2576541900634766, | |
| "learning_rate": 5.98e-06, | |
| "loss": 6.6621, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0030201769392232537, | |
| "grad_norm": 2.4825663566589355, | |
| "learning_rate": 6.98e-06, | |
| "loss": 6.2397, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0034516307876837184, | |
| "grad_norm": 1.9788450002670288, | |
| "learning_rate": 7.98e-06, | |
| "loss": 5.8553, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.003883084636144183, | |
| "grad_norm": 2.507742166519165, | |
| "learning_rate": 8.98e-06, | |
| "loss": 5.5195, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.004314538484604648, | |
| "grad_norm": 3.946438789367676, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 5.2475, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.004745992333065113, | |
| "grad_norm": 2.7235612869262695, | |
| "learning_rate": 1.098e-05, | |
| "loss": 4.9971, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.005177446181525578, | |
| "grad_norm": 2.88761830329895, | |
| "learning_rate": 1.198e-05, | |
| "loss": 4.7636, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.005608900029986042, | |
| "grad_norm": 2.373903512954712, | |
| "learning_rate": 1.2980000000000001e-05, | |
| "loss": 4.5525, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.006040353878446507, | |
| "grad_norm": 2.945600748062134, | |
| "learning_rate": 1.3980000000000002e-05, | |
| "loss": 4.447, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.006471807726906972, | |
| "grad_norm": 3.239373207092285, | |
| "learning_rate": 1.4979999999999999e-05, | |
| "loss": 4.3253, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.006903261575367437, | |
| "grad_norm": 3.2961957454681396, | |
| "learning_rate": 1.598e-05, | |
| "loss": 4.1988, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.007334715423827902, | |
| "grad_norm": 3.4848122596740723, | |
| "learning_rate": 1.698e-05, | |
| "loss": 4.1261, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.007766169272288366, | |
| "grad_norm": 5.849564075469971, | |
| "learning_rate": 1.798e-05, | |
| "loss": 4.063, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.00819762312074883, | |
| "grad_norm": 3.391148567199707, | |
| "learning_rate": 1.898e-05, | |
| "loss": 3.973, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.008629076969209296, | |
| "grad_norm": 2.368360757827759, | |
| "learning_rate": 1.9980000000000002e-05, | |
| "loss": 3.8773, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.008629076969209296, | |
| "eval_loss": 3.885983943939209, | |
| "eval_runtime": 2846.8484, | |
| "eval_samples_per_second": 26.566, | |
| "eval_steps_per_second": 13.283, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.00906053081766976, | |
| "grad_norm": 3.211322069168091, | |
| "learning_rate": 2.098e-05, | |
| "loss": 3.8554, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.009491984666130226, | |
| "grad_norm": 3.2583465576171875, | |
| "learning_rate": 2.198e-05, | |
| "loss": 3.771, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.009923438514590691, | |
| "grad_norm": 3.918938636779785, | |
| "learning_rate": 2.298e-05, | |
| "loss": 3.7751, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.010354892363051156, | |
| "grad_norm": 3.7250285148620605, | |
| "learning_rate": 2.398e-05, | |
| "loss": 3.6291, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.01078634621151162, | |
| "grad_norm": 3.9162580966949463, | |
| "learning_rate": 2.498e-05, | |
| "loss": 3.6013, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.011217800059972085, | |
| "grad_norm": 3.402956008911133, | |
| "learning_rate": 2.598e-05, | |
| "loss": 3.5911, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.01164925390843255, | |
| "grad_norm": 3.117097854614258, | |
| "learning_rate": 2.698e-05, | |
| "loss": 3.556, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.012080707756893015, | |
| "grad_norm": 2.198148012161255, | |
| "learning_rate": 2.798e-05, | |
| "loss": 3.4642, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.01251216160535348, | |
| "grad_norm": 2.386687994003296, | |
| "learning_rate": 2.898e-05, | |
| "loss": 3.4827, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.012943615453813943, | |
| "grad_norm": 2.13498854637146, | |
| "learning_rate": 2.998e-05, | |
| "loss": 3.4887, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.013375069302274408, | |
| "grad_norm": 2.7259521484375, | |
| "learning_rate": 3.0980000000000005e-05, | |
| "loss": 3.4152, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.013806523150734874, | |
| "grad_norm": 2.1376330852508545, | |
| "learning_rate": 3.198e-05, | |
| "loss": 3.3847, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.014237976999195339, | |
| "grad_norm": 2.23085355758667, | |
| "learning_rate": 3.298e-05, | |
| "loss": 3.3698, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.014669430847655804, | |
| "grad_norm": 2.2056055068969727, | |
| "learning_rate": 3.398e-05, | |
| "loss": 3.3082, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.015100884696116269, | |
| "grad_norm": 2.404055595397949, | |
| "learning_rate": 3.498e-05, | |
| "loss": 3.218, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.015532338544576732, | |
| "grad_norm": 2.5828022956848145, | |
| "learning_rate": 3.5980000000000004e-05, | |
| "loss": 3.2422, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.015963792393037197, | |
| "grad_norm": 1.9683724641799927, | |
| "learning_rate": 3.698e-05, | |
| "loss": 3.1683, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.01639524624149766, | |
| "grad_norm": 1.9096944332122803, | |
| "learning_rate": 3.7980000000000006e-05, | |
| "loss": 3.2061, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.016826700089958128, | |
| "grad_norm": 1.851697325706482, | |
| "learning_rate": 3.898e-05, | |
| "loss": 3.1872, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.01725815393841859, | |
| "grad_norm": 1.6863641738891602, | |
| "learning_rate": 3.998e-05, | |
| "loss": 3.1296, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.01725815393841859, | |
| "eval_loss": 3.1464343070983887, | |
| "eval_runtime": 2765.1571, | |
| "eval_samples_per_second": 27.35, | |
| "eval_steps_per_second": 13.675, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.017689607786879058, | |
| "grad_norm": 1.947570562362671, | |
| "learning_rate": 4.0980000000000004e-05, | |
| "loss": 3.1589, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.01812106163533952, | |
| "grad_norm": 1.683140516281128, | |
| "learning_rate": 4.198e-05, | |
| "loss": 3.1179, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.018552515483799988, | |
| "grad_norm": 1.7471286058425903, | |
| "learning_rate": 4.2980000000000005e-05, | |
| "loss": 3.1115, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.01898396933226045, | |
| "grad_norm": 1.4706114530563354, | |
| "learning_rate": 4.398e-05, | |
| "loss": 3.061, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.019415423180720915, | |
| "grad_norm": 1.6069599390029907, | |
| "learning_rate": 4.498e-05, | |
| "loss": 3.0233, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.019846877029181382, | |
| "grad_norm": 1.2190749645233154, | |
| "learning_rate": 4.5980000000000004e-05, | |
| "loss": 3.0544, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.020278330877641845, | |
| "grad_norm": 1.2313108444213867, | |
| "learning_rate": 4.698e-05, | |
| "loss": 3.0468, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.020709784726102312, | |
| "grad_norm": 1.1253961324691772, | |
| "learning_rate": 4.7980000000000005e-05, | |
| "loss": 2.9753, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.021141238574562776, | |
| "grad_norm": 1.1460264921188354, | |
| "learning_rate": 4.898e-05, | |
| "loss": 2.955, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.02157269242302324, | |
| "grad_norm": 1.2542543411254883, | |
| "learning_rate": 4.9980000000000006e-05, | |
| "loss": 2.9418, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.022004146271483706, | |
| "grad_norm": 1.0879753828048706, | |
| "learning_rate": 5.098e-05, | |
| "loss": 2.9779, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.02243560011994417, | |
| "grad_norm": 1.0592762231826782, | |
| "learning_rate": 5.198000000000001e-05, | |
| "loss": 2.9218, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.022867053968404636, | |
| "grad_norm": 1.1562455892562866, | |
| "learning_rate": 5.2980000000000004e-05, | |
| "loss": 2.9204, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.0232985078168651, | |
| "grad_norm": 1.013615369796753, | |
| "learning_rate": 5.3979999999999995e-05, | |
| "loss": 2.9179, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.023729961665325563, | |
| "grad_norm": 1.0790969133377075, | |
| "learning_rate": 5.498e-05, | |
| "loss": 2.9417, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.02416141551378603, | |
| "grad_norm": 0.9028209447860718, | |
| "learning_rate": 5.5979999999999996e-05, | |
| "loss": 2.8743, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.024592869362246493, | |
| "grad_norm": 0.8891520500183105, | |
| "learning_rate": 5.698e-05, | |
| "loss": 2.8934, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.02502432321070696, | |
| "grad_norm": 1.0406967401504517, | |
| "learning_rate": 5.7980000000000004e-05, | |
| "loss": 2.8679, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.025455777059167423, | |
| "grad_norm": 0.938779354095459, | |
| "learning_rate": 5.898e-05, | |
| "loss": 2.8523, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.025887230907627887, | |
| "grad_norm": 0.8158487677574158, | |
| "learning_rate": 5.9980000000000005e-05, | |
| "loss": 2.827, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.025887230907627887, | |
| "eval_loss": 2.8239262104034424, | |
| "eval_runtime": 2782.2115, | |
| "eval_samples_per_second": 27.183, | |
| "eval_steps_per_second": 13.591, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.026318684756088354, | |
| "grad_norm": 0.8945357203483582, | |
| "learning_rate": 6.098e-05, | |
| "loss": 2.8226, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.026750138604548817, | |
| "grad_norm": 0.8776882886886597, | |
| "learning_rate": 6.198e-05, | |
| "loss": 2.7725, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.027181592453009284, | |
| "grad_norm": 0.8972533941268921, | |
| "learning_rate": 6.298000000000001e-05, | |
| "loss": 2.795, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.027613046301469747, | |
| "grad_norm": 0.7877374887466431, | |
| "learning_rate": 6.398000000000001e-05, | |
| "loss": 2.8148, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.028044500149930214, | |
| "grad_norm": 0.7548953890800476, | |
| "learning_rate": 6.498e-05, | |
| "loss": 2.7789, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.028475953998390677, | |
| "grad_norm": 0.7239817380905151, | |
| "learning_rate": 6.598e-05, | |
| "loss": 2.7506, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.02890740784685114, | |
| "grad_norm": 0.8170995116233826, | |
| "learning_rate": 6.698e-05, | |
| "loss": 2.7415, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.029338861695311608, | |
| "grad_norm": 0.7359878420829773, | |
| "learning_rate": 6.798e-05, | |
| "loss": 2.7643, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.02977031554377207, | |
| "grad_norm": 0.7392879128456116, | |
| "learning_rate": 6.898e-05, | |
| "loss": 2.7541, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.030201769392232538, | |
| "grad_norm": 0.7366235852241516, | |
| "learning_rate": 6.998e-05, | |
| "loss": 2.7391, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.030633223240693, | |
| "grad_norm": 0.6895750164985657, | |
| "learning_rate": 7.098e-05, | |
| "loss": 2.7625, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.031064677089153465, | |
| "grad_norm": 0.6986274719238281, | |
| "learning_rate": 7.198e-05, | |
| "loss": 2.714, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.03149613093761393, | |
| "grad_norm": 0.7208825945854187, | |
| "learning_rate": 7.298000000000001e-05, | |
| "loss": 2.6966, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.031927584786074395, | |
| "grad_norm": 0.7059122323989868, | |
| "learning_rate": 7.398e-05, | |
| "loss": 2.7436, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.03235903863453486, | |
| "grad_norm": 0.6853424310684204, | |
| "learning_rate": 7.498e-05, | |
| "loss": 2.696, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.03279049248299532, | |
| "grad_norm": 0.6768801212310791, | |
| "learning_rate": 7.598e-05, | |
| "loss": 2.7349, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.03322194633145579, | |
| "grad_norm": 0.7065775990486145, | |
| "learning_rate": 7.698000000000001e-05, | |
| "loss": 2.6729, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.033653400179916255, | |
| "grad_norm": 0.6956036686897278, | |
| "learning_rate": 7.798000000000001e-05, | |
| "loss": 2.6356, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.03408485402837672, | |
| "grad_norm": 0.6357026696205139, | |
| "learning_rate": 7.897999999999999e-05, | |
| "loss": 2.5888, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.03451630787683718, | |
| "grad_norm": 0.6445960998535156, | |
| "learning_rate": 7.998e-05, | |
| "loss": 2.6392, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.03451630787683718, | |
| "eval_loss": 2.6430516242980957, | |
| "eval_runtime": 2751.3719, | |
| "eval_samples_per_second": 27.487, | |
| "eval_steps_per_second": 13.744, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.03494776172529765, | |
| "grad_norm": 0.6698920130729675, | |
| "learning_rate": 8.098e-05, | |
| "loss": 2.6621, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.035379215573758116, | |
| "grad_norm": 0.6329808235168457, | |
| "learning_rate": 8.198e-05, | |
| "loss": 2.6276, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.035810669422218576, | |
| "grad_norm": 0.6281711459159851, | |
| "learning_rate": 8.298000000000001e-05, | |
| "loss": 2.6093, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.03624212327067904, | |
| "grad_norm": 0.6180399060249329, | |
| "learning_rate": 8.398e-05, | |
| "loss": 2.645, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.03667357711913951, | |
| "grad_norm": 0.620257556438446, | |
| "learning_rate": 8.498e-05, | |
| "loss": 2.6221, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.037105030967599976, | |
| "grad_norm": 0.6216643452644348, | |
| "learning_rate": 8.598e-05, | |
| "loss": 2.5944, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.037536484816060436, | |
| "grad_norm": 0.6922804117202759, | |
| "learning_rate": 8.698000000000001e-05, | |
| "loss": 2.5989, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.0379679386645209, | |
| "grad_norm": 0.5947348475456238, | |
| "learning_rate": 8.798000000000001e-05, | |
| "loss": 2.5872, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.03839939251298137, | |
| "grad_norm": 0.6008750796318054, | |
| "learning_rate": 8.898e-05, | |
| "loss": 2.5942, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.03883084636144183, | |
| "grad_norm": 0.604884922504425, | |
| "learning_rate": 8.998e-05, | |
| "loss": 2.5764, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.0392623002099023, | |
| "grad_norm": 0.6107255816459656, | |
| "learning_rate": 9.098000000000001e-05, | |
| "loss": 2.5419, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.039693754058362764, | |
| "grad_norm": 0.5564243793487549, | |
| "learning_rate": 9.198e-05, | |
| "loss": 2.565, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.040125207906823224, | |
| "grad_norm": 0.5812260508537292, | |
| "learning_rate": 9.298e-05, | |
| "loss": 2.5734, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.04055666175528369, | |
| "grad_norm": 0.5526607632637024, | |
| "learning_rate": 9.398e-05, | |
| "loss": 2.586, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.04098811560374416, | |
| "grad_norm": 0.6254459023475647, | |
| "learning_rate": 9.498e-05, | |
| "loss": 2.4797, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.041419569452204624, | |
| "grad_norm": 0.5473371148109436, | |
| "learning_rate": 9.598e-05, | |
| "loss": 2.5454, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.041851023300665084, | |
| "grad_norm": 0.5498418807983398, | |
| "learning_rate": 9.698000000000001e-05, | |
| "loss": 2.5581, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.04228247714912555, | |
| "grad_norm": 0.565069317817688, | |
| "learning_rate": 9.798000000000001e-05, | |
| "loss": 2.5014, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.04271393099758602, | |
| "grad_norm": 0.5747342705726624, | |
| "learning_rate": 9.898e-05, | |
| "loss": 2.5312, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.04314538484604648, | |
| "grad_norm": 0.5452004075050354, | |
| "learning_rate": 9.998e-05, | |
| "loss": 2.5331, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.04314538484604648, | |
| "eval_loss": 2.526031494140625, | |
| "eval_runtime": 3108.7919, | |
| "eval_samples_per_second": 24.327, | |
| "eval_steps_per_second": 12.164, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.043576838694506945, | |
| "grad_norm": 0.5333559513092041, | |
| "learning_rate": 9.999995182044957e-05, | |
| "loss": 2.5314, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.04400829254296741, | |
| "grad_norm": 0.5360190272331238, | |
| "learning_rate": 9.99998033288046e-05, | |
| "loss": 2.5069, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.04443974639142787, | |
| "grad_norm": 0.5400934219360352, | |
| "learning_rate": 9.999955450529664e-05, | |
| "loss": 2.5327, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.04487120023988834, | |
| "grad_norm": 0.5095759630203247, | |
| "learning_rate": 9.999920535042497e-05, | |
| "loss": 2.496, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.045302654088348805, | |
| "grad_norm": 0.5121382474899292, | |
| "learning_rate": 9.999875586489024e-05, | |
| "loss": 2.4837, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.04573410793680927, | |
| "grad_norm": 0.5363543033599854, | |
| "learning_rate": 9.999820604959439e-05, | |
| "loss": 2.513, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.04616556178526973, | |
| "grad_norm": 0.5149393081665039, | |
| "learning_rate": 9.99975559056407e-05, | |
| "loss": 2.5031, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.0465970156337302, | |
| "grad_norm": 0.5039998888969421, | |
| "learning_rate": 9.999680543433381e-05, | |
| "loss": 2.488, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.047028469482190666, | |
| "grad_norm": 0.5018443465232849, | |
| "learning_rate": 9.999595463717962e-05, | |
| "loss": 2.4613, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.047459923330651126, | |
| "grad_norm": 0.49239611625671387, | |
| "learning_rate": 9.999500351588537e-05, | |
| "loss": 2.4413, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.04789137717911159, | |
| "grad_norm": 0.4872280955314636, | |
| "learning_rate": 9.999395207235966e-05, | |
| "loss": 2.454, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.04832283102757206, | |
| "grad_norm": 0.47833895683288574, | |
| "learning_rate": 9.999280030871234e-05, | |
| "loss": 2.4903, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.048754284876032526, | |
| "grad_norm": 0.472711980342865, | |
| "learning_rate": 9.99915482272546e-05, | |
| "loss": 2.4238, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.049185738724492986, | |
| "grad_norm": 0.47533369064331055, | |
| "learning_rate": 9.99901958304989e-05, | |
| "loss": 2.4662, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.04961719257295345, | |
| "grad_norm": 0.47553250193595886, | |
| "learning_rate": 9.998874312115904e-05, | |
| "loss": 2.4653, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.05004864642141392, | |
| "grad_norm": 0.5022785663604736, | |
| "learning_rate": 9.99871901021501e-05, | |
| "loss": 2.4652, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.05048010026987438, | |
| "grad_norm": 0.4645737409591675, | |
| "learning_rate": 9.998553677658843e-05, | |
| "loss": 2.4522, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.05091155411833485, | |
| "grad_norm": 0.47363319993019104, | |
| "learning_rate": 9.998378314779166e-05, | |
| "loss": 2.4609, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.051343007966795313, | |
| "grad_norm": 0.4770340323448181, | |
| "learning_rate": 9.998192921927872e-05, | |
| "loss": 2.3911, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.05177446181525577, | |
| "grad_norm": 0.48888033628463745, | |
| "learning_rate": 9.997997499476975e-05, | |
| "loss": 2.4673, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.05177446181525577, | |
| "eval_loss": 2.4369618892669678, | |
| "eval_runtime": 2769.3132, | |
| "eval_samples_per_second": 27.309, | |
| "eval_steps_per_second": 13.655, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.05220591566371624, | |
| "grad_norm": 0.4705820083618164, | |
| "learning_rate": 9.997792047818622e-05, | |
| "loss": 2.465, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.05263736951217671, | |
| "grad_norm": 0.47672560811042786, | |
| "learning_rate": 9.99757656736508e-05, | |
| "loss": 2.4048, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.053068823360637174, | |
| "grad_norm": 0.4655727744102478, | |
| "learning_rate": 9.997351058548742e-05, | |
| "loss": 2.4552, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.053500277209097634, | |
| "grad_norm": 0.47453373670578003, | |
| "learning_rate": 9.997115521822125e-05, | |
| "loss": 2.4435, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.0539317310575581, | |
| "grad_norm": 0.4548121392726898, | |
| "learning_rate": 9.996869957657866e-05, | |
| "loss": 2.3989, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.05436318490601857, | |
| "grad_norm": 0.4468235373497009, | |
| "learning_rate": 9.996614366548727e-05, | |
| "loss": 2.3674, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.05479463875447903, | |
| "grad_norm": 0.4576756954193115, | |
| "learning_rate": 9.996348749007589e-05, | |
| "loss": 2.3689, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.055226092602939494, | |
| "grad_norm": 0.4505023956298828, | |
| "learning_rate": 9.996073105567449e-05, | |
| "loss": 2.4009, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.05565754645139996, | |
| "grad_norm": 0.4463840425014496, | |
| "learning_rate": 9.99578743678143e-05, | |
| "loss": 2.3846, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.05608900029986043, | |
| "grad_norm": 0.44599780440330505, | |
| "learning_rate": 9.995491743222763e-05, | |
| "loss": 2.4243, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.05652045414832089, | |
| "grad_norm": 0.44260427355766296, | |
| "learning_rate": 9.995186025484803e-05, | |
| "loss": 2.366, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.056951907996781355, | |
| "grad_norm": 0.4404599368572235, | |
| "learning_rate": 9.994870284181017e-05, | |
| "loss": 2.3635, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.05738336184524182, | |
| "grad_norm": 0.43710505962371826, | |
| "learning_rate": 9.994544519944984e-05, | |
| "loss": 2.3563, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.05781481569370228, | |
| "grad_norm": 0.42433786392211914, | |
| "learning_rate": 9.9942087334304e-05, | |
| "loss": 2.3752, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.05824626954216275, | |
| "grad_norm": 0.4220280945301056, | |
| "learning_rate": 9.993862925311067e-05, | |
| "loss": 2.394, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.058677723390623215, | |
| "grad_norm": 0.4359777271747589, | |
| "learning_rate": 9.9935070962809e-05, | |
| "loss": 2.3852, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.059109177239083675, | |
| "grad_norm": 0.4458872973918915, | |
| "learning_rate": 9.993141247053923e-05, | |
| "loss": 2.3866, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.05954063108754414, | |
| "grad_norm": 0.44301262497901917, | |
| "learning_rate": 9.992765378364263e-05, | |
| "loss": 2.3885, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.05997208493600461, | |
| "grad_norm": 0.4382476508617401, | |
| "learning_rate": 9.992379490966156e-05, | |
| "loss": 2.3537, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.060403538784465076, | |
| "grad_norm": 0.435742050409317, | |
| "learning_rate": 9.991983585633941e-05, | |
| "loss": 2.3634, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.060403538784465076, | |
| "eval_loss": 2.374647617340088, | |
| "eval_runtime": 2732.6747, | |
| "eval_samples_per_second": 27.675, | |
| "eval_steps_per_second": 13.838, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.060834992632925536, | |
| "grad_norm": 0.46364474296569824, | |
| "learning_rate": 9.991577663162062e-05, | |
| "loss": 2.3667, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.061266446481386, | |
| "grad_norm": 0.43860843777656555, | |
| "learning_rate": 9.991161724365057e-05, | |
| "loss": 2.3723, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.06169790032984647, | |
| "grad_norm": 0.42228421568870544, | |
| "learning_rate": 9.990735770077571e-05, | |
| "loss": 2.3963, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.06212935417830693, | |
| "grad_norm": 0.41791942715644836, | |
| "learning_rate": 9.990299801154342e-05, | |
| "loss": 2.3626, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.0625608080267674, | |
| "grad_norm": 0.4274737238883972, | |
| "learning_rate": 9.989853818470206e-05, | |
| "loss": 2.3904, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.06299226187522786, | |
| "grad_norm": 0.4228742718696594, | |
| "learning_rate": 9.989397822920088e-05, | |
| "loss": 2.3824, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.06342371572368832, | |
| "grad_norm": 0.4259440004825592, | |
| "learning_rate": 9.988931815419013e-05, | |
| "loss": 2.3413, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.06385516957214879, | |
| "grad_norm": 0.40943819284439087, | |
| "learning_rate": 9.988455796902094e-05, | |
| "loss": 2.3766, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.06428662342060926, | |
| "grad_norm": 0.4232550859451294, | |
| "learning_rate": 9.987969768324528e-05, | |
| "loss": 2.3293, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.06471807726906972, | |
| "grad_norm": 0.41405293345451355, | |
| "learning_rate": 9.987473730661602e-05, | |
| "loss": 2.3586, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.06514953111753019, | |
| "grad_norm": 0.4351080060005188, | |
| "learning_rate": 9.986967684908688e-05, | |
| "loss": 2.3015, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.06558098496599064, | |
| "grad_norm": 0.4127940237522125, | |
| "learning_rate": 9.986451632081242e-05, | |
| "loss": 2.3478, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.06601243881445111, | |
| "grad_norm": 0.4163227379322052, | |
| "learning_rate": 9.985925573214798e-05, | |
| "loss": 2.3294, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.06644389266291158, | |
| "grad_norm": 0.4200190603733063, | |
| "learning_rate": 9.985389509364968e-05, | |
| "loss": 2.3291, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.06687534651137204, | |
| "grad_norm": 0.4104471206665039, | |
| "learning_rate": 9.984843441607443e-05, | |
| "loss": 2.3712, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.06730680035983251, | |
| "grad_norm": 0.4211075007915497, | |
| "learning_rate": 9.984287371037988e-05, | |
| "loss": 2.2668, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.06773825420829298, | |
| "grad_norm": 0.42335718870162964, | |
| "learning_rate": 9.98372129877244e-05, | |
| "loss": 2.329, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.06816970805675344, | |
| "grad_norm": 0.4369964301586151, | |
| "learning_rate": 9.983145225946702e-05, | |
| "loss": 2.3559, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.0686011619052139, | |
| "grad_norm": 0.4052545130252838, | |
| "learning_rate": 9.982559153716749e-05, | |
| "loss": 2.3116, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.06903261575367436, | |
| "grad_norm": 0.4120001196861267, | |
| "learning_rate": 9.981963083258622e-05, | |
| "loss": 2.3408, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.06903261575367436, | |
| "eval_loss": 2.3286283016204834, | |
| "eval_runtime": 2709.736, | |
| "eval_samples_per_second": 27.91, | |
| "eval_steps_per_second": 13.955, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.06946406960213483, | |
| "grad_norm": 0.4071601629257202, | |
| "learning_rate": 9.98135701576842e-05, | |
| "loss": 2.3101, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.0698955234505953, | |
| "grad_norm": 0.4207800626754761, | |
| "learning_rate": 9.980740952462309e-05, | |
| "loss": 2.3192, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.07032697729905577, | |
| "grad_norm": 0.43044227361679077, | |
| "learning_rate": 9.980114894576504e-05, | |
| "loss": 2.3054, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.07075843114751623, | |
| "grad_norm": 0.3968140184879303, | |
| "learning_rate": 9.979478843367287e-05, | |
| "loss": 2.2884, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.0711898849959767, | |
| "grad_norm": 0.39424625039100647, | |
| "learning_rate": 9.978832800110983e-05, | |
| "loss": 2.336, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.07162133884443715, | |
| "grad_norm": 0.40985167026519775, | |
| "learning_rate": 9.978176766103974e-05, | |
| "loss": 2.3301, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.07205279269289762, | |
| "grad_norm": 0.3952298164367676, | |
| "learning_rate": 9.977510742662685e-05, | |
| "loss": 2.3406, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.07248424654135809, | |
| "grad_norm": 0.4053609073162079, | |
| "learning_rate": 9.976834731123591e-05, | |
| "loss": 2.2933, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.07291570038981855, | |
| "grad_norm": 0.4119470417499542, | |
| "learning_rate": 9.976148732843207e-05, | |
| "loss": 2.2927, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.07334715423827902, | |
| "grad_norm": 0.4225163459777832, | |
| "learning_rate": 9.975452749198086e-05, | |
| "loss": 2.3243, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.07377860808673949, | |
| "grad_norm": 0.410567969083786, | |
| "learning_rate": 9.974746781584823e-05, | |
| "loss": 2.2995, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.07421006193519995, | |
| "grad_norm": 0.3905579447746277, | |
| "learning_rate": 9.974030831420043e-05, | |
| "loss": 2.2875, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.0746415157836604, | |
| "grad_norm": 0.4086805284023285, | |
| "learning_rate": 9.973304900140404e-05, | |
| "loss": 2.3132, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.07507296963212087, | |
| "grad_norm": 0.4010264575481415, | |
| "learning_rate": 9.972568989202591e-05, | |
| "loss": 2.2886, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.07550442348058134, | |
| "grad_norm": 0.41073790192604065, | |
| "learning_rate": 9.971823100083319e-05, | |
| "loss": 2.3267, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.0759358773290418, | |
| "grad_norm": 0.39557185769081116, | |
| "learning_rate": 9.971067234279322e-05, | |
| "loss": 2.3053, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.07636733117750227, | |
| "grad_norm": 0.39963647723197937, | |
| "learning_rate": 9.970301393307353e-05, | |
| "loss": 2.3193, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.07679878502596274, | |
| "grad_norm": 0.38299834728240967, | |
| "learning_rate": 9.969525578704186e-05, | |
| "loss": 2.2845, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.0772302388744232, | |
| "grad_norm": 0.3826388418674469, | |
| "learning_rate": 9.968739792026602e-05, | |
| "loss": 2.2879, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.07766169272288366, | |
| "grad_norm": 0.39117395877838135, | |
| "learning_rate": 9.967944034851398e-05, | |
| "loss": 2.2922, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.07766169272288366, | |
| "eval_loss": 2.290743589401245, | |
| "eval_runtime": 2787.9306, | |
| "eval_samples_per_second": 27.127, | |
| "eval_steps_per_second": 13.563, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.07809314657134413, | |
| "grad_norm": 0.4105989933013916, | |
| "learning_rate": 9.967138308775374e-05, | |
| "loss": 2.3296, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.0785246004198046, | |
| "grad_norm": 0.3829767107963562, | |
| "learning_rate": 9.966322615415341e-05, | |
| "loss": 2.2909, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.07895605426826506, | |
| "grad_norm": 0.38462504744529724, | |
| "learning_rate": 9.965496956408102e-05, | |
| "loss": 2.3014, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.07938750811672553, | |
| "grad_norm": 0.3984980583190918, | |
| "learning_rate": 9.964661333410464e-05, | |
| "loss": 2.2933, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.079818961965186, | |
| "grad_norm": 0.41218262910842896, | |
| "learning_rate": 9.963815748099225e-05, | |
| "loss": 2.3321, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.08025041581364645, | |
| "grad_norm": 0.39859455823898315, | |
| "learning_rate": 9.962960202171174e-05, | |
| "loss": 2.2745, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.08068186966210691, | |
| "grad_norm": 0.3806203305721283, | |
| "learning_rate": 9.962094697343089e-05, | |
| "loss": 2.2691, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.08111332351056738, | |
| "grad_norm": 0.3818638324737549, | |
| "learning_rate": 9.961219235351729e-05, | |
| "loss": 2.3255, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.08154477735902785, | |
| "grad_norm": 0.390399307012558, | |
| "learning_rate": 9.96033381795384e-05, | |
| "loss": 2.2906, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.08197623120748831, | |
| "grad_norm": 0.37718382477760315, | |
| "learning_rate": 9.959438446926135e-05, | |
| "loss": 2.2903, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.08240768505594878, | |
| "grad_norm": 0.3930068016052246, | |
| "learning_rate": 9.95853312406531e-05, | |
| "loss": 2.2832, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.08283913890440925, | |
| "grad_norm": 0.3778354227542877, | |
| "learning_rate": 9.957617851188026e-05, | |
| "loss": 2.2438, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.0832705927528697, | |
| "grad_norm": 0.39089974761009216, | |
| "learning_rate": 9.956692630130911e-05, | |
| "loss": 2.3055, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.08370204660133017, | |
| "grad_norm": 0.3801150619983673, | |
| "learning_rate": 9.955757462750554e-05, | |
| "loss": 2.2887, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.08413350044979064, | |
| "grad_norm": 0.37772729992866516, | |
| "learning_rate": 9.954812350923504e-05, | |
| "loss": 2.2617, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.0845649542982511, | |
| "grad_norm": 0.4012361168861389, | |
| "learning_rate": 9.953857296546267e-05, | |
| "loss": 2.2583, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.08499640814671157, | |
| "grad_norm": 0.37904253602027893, | |
| "learning_rate": 9.952892301535295e-05, | |
| "loss": 2.2582, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.08542786199517204, | |
| "grad_norm": 0.38929301500320435, | |
| "learning_rate": 9.951917367826995e-05, | |
| "loss": 2.2581, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.0858593158436325, | |
| "grad_norm": 0.37145188450813293, | |
| "learning_rate": 9.95093249737771e-05, | |
| "loss": 2.2422, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.08629076969209296, | |
| "grad_norm": 0.3830346465110779, | |
| "learning_rate": 9.949937692163726e-05, | |
| "loss": 2.2527, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.08629076969209296, | |
| "eval_loss": 2.2615647315979004, | |
| "eval_runtime": 2822.4337, | |
| "eval_samples_per_second": 26.795, | |
| "eval_steps_per_second": 13.398, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.08672222354055342, | |
| "grad_norm": 0.37199145555496216, | |
| "learning_rate": 9.948932954181264e-05, | |
| "loss": 2.2656, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.08715367738901389, | |
| "grad_norm": 0.385570764541626, | |
| "learning_rate": 9.947918285446475e-05, | |
| "loss": 2.2667, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.08758513123747436, | |
| "grad_norm": 0.38618603348731995, | |
| "learning_rate": 9.946893687995444e-05, | |
| "loss": 2.2459, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.08801658508593482, | |
| "grad_norm": 0.37942782044410706, | |
| "learning_rate": 9.94585916388417e-05, | |
| "loss": 2.2473, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.08844803893439529, | |
| "grad_norm": 0.37803226709365845, | |
| "learning_rate": 9.944814715188578e-05, | |
| "loss": 2.2623, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.08887949278285574, | |
| "grad_norm": 0.37103068828582764, | |
| "learning_rate": 9.943760344004505e-05, | |
| "loss": 2.2822, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.08931094663131621, | |
| "grad_norm": 0.3779771327972412, | |
| "learning_rate": 9.942696052447699e-05, | |
| "loss": 2.2473, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.08974240047977668, | |
| "grad_norm": 0.38384005427360535, | |
| "learning_rate": 9.94162184265382e-05, | |
| "loss": 2.2615, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.09017385432823714, | |
| "grad_norm": 0.36723431944847107, | |
| "learning_rate": 9.940537716778417e-05, | |
| "loss": 2.2656, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.09060530817669761, | |
| "grad_norm": 0.39431023597717285, | |
| "learning_rate": 9.939443676996955e-05, | |
| "loss": 2.2645, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.09103676202515808, | |
| "grad_norm": 0.3861570954322815, | |
| "learning_rate": 9.938339725504778e-05, | |
| "loss": 2.2158, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.09146821587361854, | |
| "grad_norm": 0.38060426712036133, | |
| "learning_rate": 9.937225864517128e-05, | |
| "loss": 2.2341, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.091899669722079, | |
| "grad_norm": 0.3649984300136566, | |
| "learning_rate": 9.936102096269126e-05, | |
| "loss": 2.2333, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.09233112357053946, | |
| "grad_norm": 0.37162846326828003, | |
| "learning_rate": 9.93496842301578e-05, | |
| "loss": 2.262, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.09276257741899993, | |
| "grad_norm": 0.3689543604850769, | |
| "learning_rate": 9.933824847031967e-05, | |
| "loss": 2.2342, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.0931940312674604, | |
| "grad_norm": 0.3751949071884155, | |
| "learning_rate": 9.932671370612437e-05, | |
| "loss": 2.2452, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.09362548511592086, | |
| "grad_norm": 0.3793388605117798, | |
| "learning_rate": 9.931507996071812e-05, | |
| "loss": 2.216, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.09405693896438133, | |
| "grad_norm": 0.37986430525779724, | |
| "learning_rate": 9.930334725744568e-05, | |
| "loss": 2.2414, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.0944883928128418, | |
| "grad_norm": 0.37280675768852234, | |
| "learning_rate": 9.929151561985044e-05, | |
| "loss": 2.2244, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.09491984666130225, | |
| "grad_norm": 0.39744243025779724, | |
| "learning_rate": 9.927958507167429e-05, | |
| "loss": 2.2515, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.09491984666130225, | |
| "eval_loss": 2.235520362854004, | |
| "eval_runtime": 2908.2681, | |
| "eval_samples_per_second": 26.004, | |
| "eval_steps_per_second": 13.002, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.09535130050976272, | |
| "grad_norm": 0.3687732517719269, | |
| "learning_rate": 9.926755563685763e-05, | |
| "loss": 2.2522, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.09578275435822318, | |
| "grad_norm": 0.36527740955352783, | |
| "learning_rate": 9.925542733953923e-05, | |
| "loss": 2.2731, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.09621420820668365, | |
| "grad_norm": 0.3645661175251007, | |
| "learning_rate": 9.924320020405632e-05, | |
| "loss": 2.2084, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.09664566205514412, | |
| "grad_norm": 0.36427396535873413, | |
| "learning_rate": 9.92308742549444e-05, | |
| "loss": 2.2074, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.09707711590360459, | |
| "grad_norm": 0.35815566778182983, | |
| "learning_rate": 9.921844951693726e-05, | |
| "loss": 2.212, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.09750856975206505, | |
| "grad_norm": 0.3574189841747284, | |
| "learning_rate": 9.920592601496696e-05, | |
| "loss": 2.2262, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.0979400236005255, | |
| "grad_norm": 0.3735974133014679, | |
| "learning_rate": 9.919330377416373e-05, | |
| "loss": 2.23, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 0.09837147744898597, | |
| "grad_norm": 0.38068506121635437, | |
| "learning_rate": 9.918058281985592e-05, | |
| "loss": 2.1978, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.09880293129744644, | |
| "grad_norm": 0.36570560932159424, | |
| "learning_rate": 9.916776317756996e-05, | |
| "loss": 2.2228, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 0.0992343851459069, | |
| "grad_norm": 0.3642862141132355, | |
| "learning_rate": 9.915484487303035e-05, | |
| "loss": 2.2383, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.09966583899436737, | |
| "grad_norm": 0.36596134305000305, | |
| "learning_rate": 9.914182793215951e-05, | |
| "loss": 2.2302, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 0.10009729284282784, | |
| "grad_norm": 0.36998990178108215, | |
| "learning_rate": 9.912871238107786e-05, | |
| "loss": 2.2384, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.1005287466912883, | |
| "grad_norm": 0.36373093724250793, | |
| "learning_rate": 9.911549824610362e-05, | |
| "loss": 2.1848, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 0.10096020053974876, | |
| "grad_norm": 0.36697497963905334, | |
| "learning_rate": 9.910218555375289e-05, | |
| "loss": 2.2126, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.10139165438820923, | |
| "grad_norm": 0.36996662616729736, | |
| "learning_rate": 9.908877433073953e-05, | |
| "loss": 2.2276, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 0.1018231082366697, | |
| "grad_norm": 0.36146995425224304, | |
| "learning_rate": 9.907526460397507e-05, | |
| "loss": 2.2043, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.10225456208513016, | |
| "grad_norm": 0.36770108342170715, | |
| "learning_rate": 9.906165640056877e-05, | |
| "loss": 2.2223, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 0.10268601593359063, | |
| "grad_norm": 0.3661516606807709, | |
| "learning_rate": 9.904794974782745e-05, | |
| "loss": 2.1971, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.1031174697820511, | |
| "grad_norm": 0.36954957246780396, | |
| "learning_rate": 9.90341446732555e-05, | |
| "loss": 2.2063, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 0.10354892363051155, | |
| "grad_norm": 0.3615468740463257, | |
| "learning_rate": 9.902024120455482e-05, | |
| "loss": 2.2116, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.10354892363051155, | |
| "eval_loss": 2.214376926422119, | |
| "eval_runtime": 2985.5808, | |
| "eval_samples_per_second": 25.331, | |
| "eval_steps_per_second": 12.666, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.10398037747897201, | |
| "grad_norm": 0.35979902744293213, | |
| "learning_rate": 9.900623936962471e-05, | |
| "loss": 2.1718, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 0.10441183132743248, | |
| "grad_norm": 0.37038519978523254, | |
| "learning_rate": 9.899213919656191e-05, | |
| "loss": 2.2364, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.10484328517589295, | |
| "grad_norm": 0.349633127450943, | |
| "learning_rate": 9.897794071366045e-05, | |
| "loss": 2.2055, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 0.10527473902435341, | |
| "grad_norm": 0.365032821893692, | |
| "learning_rate": 9.896364394941166e-05, | |
| "loss": 2.2218, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.10570619287281388, | |
| "grad_norm": 0.3733338415622711, | |
| "learning_rate": 9.894924893250406e-05, | |
| "loss": 2.2368, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 0.10613764672127435, | |
| "grad_norm": 0.3741872310638428, | |
| "learning_rate": 9.893475569182335e-05, | |
| "loss": 2.2012, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.1065691005697348, | |
| "grad_norm": 0.35796815156936646, | |
| "learning_rate": 9.892016425645232e-05, | |
| "loss": 2.2186, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 0.10700055441819527, | |
| "grad_norm": 0.35955625772476196, | |
| "learning_rate": 9.89054746556708e-05, | |
| "loss": 2.2257, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.10743200826665573, | |
| "grad_norm": 0.36324945092201233, | |
| "learning_rate": 9.889068691895558e-05, | |
| "loss": 2.2069, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 0.1078634621151162, | |
| "grad_norm": 0.3579570949077606, | |
| "learning_rate": 9.887580107598044e-05, | |
| "loss": 2.2368, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.10829491596357667, | |
| "grad_norm": 0.3664620518684387, | |
| "learning_rate": 9.886081715661599e-05, | |
| "loss": 2.2064, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 0.10872636981203714, | |
| "grad_norm": 0.3607580065727234, | |
| "learning_rate": 9.884573519092959e-05, | |
| "loss": 2.2008, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 0.1091578236604976, | |
| "grad_norm": 0.3472786247730255, | |
| "learning_rate": 9.883055520918544e-05, | |
| "loss": 2.1839, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 0.10958927750895806, | |
| "grad_norm": 0.34541594982147217, | |
| "learning_rate": 9.881527724184436e-05, | |
| "loss": 2.2265, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 0.11002073135741852, | |
| "grad_norm": 0.360601544380188, | |
| "learning_rate": 9.87999013195638e-05, | |
| "loss": 2.2101, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 0.11045218520587899, | |
| "grad_norm": 0.3593735992908478, | |
| "learning_rate": 9.87844274731978e-05, | |
| "loss": 2.2014, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 0.11088363905433946, | |
| "grad_norm": 0.3615984320640564, | |
| "learning_rate": 9.876885573379687e-05, | |
| "loss": 2.2163, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 0.11131509290279992, | |
| "grad_norm": 0.3485114574432373, | |
| "learning_rate": 9.875318613260794e-05, | |
| "loss": 2.2154, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 0.11174654675126039, | |
| "grad_norm": 0.3505745232105255, | |
| "learning_rate": 9.873741870107438e-05, | |
| "loss": 2.1613, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 0.11217800059972086, | |
| "grad_norm": 0.3599075376987457, | |
| "learning_rate": 9.872155347083582e-05, | |
| "loss": 2.1743, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.11217800059972086, | |
| "eval_loss": 2.195662021636963, | |
| "eval_runtime": 2862.6512, | |
| "eval_samples_per_second": 26.419, | |
| "eval_steps_per_second": 13.209, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.11260945444818131, | |
| "grad_norm": 0.35765302181243896, | |
| "learning_rate": 9.870559047372811e-05, | |
| "loss": 2.2033, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 0.11304090829664178, | |
| "grad_norm": 0.3540419340133667, | |
| "learning_rate": 9.868952974178339e-05, | |
| "loss": 2.153, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 0.11347236214510224, | |
| "grad_norm": 0.3633355498313904, | |
| "learning_rate": 9.867337130722978e-05, | |
| "loss": 2.1969, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 0.11390381599356271, | |
| "grad_norm": 0.35366296768188477, | |
| "learning_rate": 9.865711520249156e-05, | |
| "loss": 2.1863, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.11433526984202318, | |
| "grad_norm": 0.3593711853027344, | |
| "learning_rate": 9.864076146018897e-05, | |
| "loss": 2.1632, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 0.11476672369048364, | |
| "grad_norm": 0.3670215904712677, | |
| "learning_rate": 9.862431011313815e-05, | |
| "loss": 2.1884, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 0.1151981775389441, | |
| "grad_norm": 0.3451824486255646, | |
| "learning_rate": 9.860776119435114e-05, | |
| "loss": 2.1869, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 0.11562963138740456, | |
| "grad_norm": 0.36133304238319397, | |
| "learning_rate": 9.85911147370357e-05, | |
| "loss": 2.1586, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 0.11606108523586503, | |
| "grad_norm": 0.35326626896858215, | |
| "learning_rate": 9.85743707745954e-05, | |
| "loss": 2.2124, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 0.1164925390843255, | |
| "grad_norm": 0.3657006025314331, | |
| "learning_rate": 9.855752934062943e-05, | |
| "loss": 2.1942, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.11692399293278596, | |
| "grad_norm": 0.3521730601787567, | |
| "learning_rate": 9.854059046893257e-05, | |
| "loss": 2.1945, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 0.11735544678124643, | |
| "grad_norm": 0.3588502109050751, | |
| "learning_rate": 9.852355419349511e-05, | |
| "loss": 2.2061, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 0.1177869006297069, | |
| "grad_norm": 0.36119502782821655, | |
| "learning_rate": 9.850642054850284e-05, | |
| "loss": 2.1927, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 0.11821835447816735, | |
| "grad_norm": 0.35996031761169434, | |
| "learning_rate": 9.848918956833686e-05, | |
| "loss": 2.1971, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 0.11864980832662782, | |
| "grad_norm": 0.34750646352767944, | |
| "learning_rate": 9.847186128757366e-05, | |
| "loss": 2.1965, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 0.11908126217508828, | |
| "grad_norm": 0.3519524335861206, | |
| "learning_rate": 9.845443574098496e-05, | |
| "loss": 2.1557, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 0.11951271602354875, | |
| "grad_norm": 0.35633113980293274, | |
| "learning_rate": 9.843691296353765e-05, | |
| "loss": 2.1867, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 0.11994416987200922, | |
| "grad_norm": 0.37132638692855835, | |
| "learning_rate": 9.84192929903937e-05, | |
| "loss": 2.1871, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 0.12037562372046968, | |
| "grad_norm": 0.3432469367980957, | |
| "learning_rate": 9.840157585691014e-05, | |
| "loss": 2.1724, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 0.12080707756893015, | |
| "grad_norm": 0.34275302290916443, | |
| "learning_rate": 9.838376159863902e-05, | |
| "loss": 2.1507, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.12080707756893015, | |
| "eval_loss": 2.179051637649536, | |
| "eval_runtime": 2834.8461, | |
| "eval_samples_per_second": 26.678, | |
| "eval_steps_per_second": 13.339, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.1212385314173906, | |
| "grad_norm": 0.34372302889823914, | |
| "learning_rate": 9.836585025132719e-05, | |
| "loss": 2.1675, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 0.12166998526585107, | |
| "grad_norm": 0.34981855750083923, | |
| "learning_rate": 9.834784185091638e-05, | |
| "loss": 2.2123, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 0.12210143911431154, | |
| "grad_norm": 0.3538338840007782, | |
| "learning_rate": 9.832973643354307e-05, | |
| "loss": 2.1979, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 0.122532892962772, | |
| "grad_norm": 0.3431382477283478, | |
| "learning_rate": 9.831153403553838e-05, | |
| "loss": 2.2037, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 0.12296434681123247, | |
| "grad_norm": 0.35622259974479675, | |
| "learning_rate": 9.829323469342808e-05, | |
| "loss": 2.159, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 0.12339580065969294, | |
| "grad_norm": 0.35167625546455383, | |
| "learning_rate": 9.827483844393249e-05, | |
| "loss": 2.1838, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 0.1238272545081534, | |
| "grad_norm": 0.3605762720108032, | |
| "learning_rate": 9.825634532396628e-05, | |
| "loss": 2.1947, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 0.12425870835661386, | |
| "grad_norm": 0.36447280645370483, | |
| "learning_rate": 9.823775537063867e-05, | |
| "loss": 2.1854, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 0.12469016220507433, | |
| "grad_norm": 0.3465559780597687, | |
| "learning_rate": 9.821906862125301e-05, | |
| "loss": 2.1467, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 0.1251216160535348, | |
| "grad_norm": 0.34778252243995667, | |
| "learning_rate": 9.820028511330703e-05, | |
| "loss": 2.1531, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.12555306990199525, | |
| "grad_norm": 0.34816524386405945, | |
| "learning_rate": 9.818140488449257e-05, | |
| "loss": 2.168, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 0.1259845237504557, | |
| "grad_norm": 0.3492778241634369, | |
| "learning_rate": 9.816242797269553e-05, | |
| "loss": 2.1599, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 0.12641597759891618, | |
| "grad_norm": 0.3482818901538849, | |
| "learning_rate": 9.814335441599583e-05, | |
| "loss": 2.1515, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 0.12684743144737665, | |
| "grad_norm": 0.3657529950141907, | |
| "learning_rate": 9.812418425266733e-05, | |
| "loss": 2.1656, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 0.1272788852958371, | |
| "grad_norm": 0.3417568802833557, | |
| "learning_rate": 9.810491752117776e-05, | |
| "loss": 2.156, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 0.12771033914429758, | |
| "grad_norm": 0.35225942730903625, | |
| "learning_rate": 9.808555426018863e-05, | |
| "loss": 2.1734, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 0.12814179299275805, | |
| "grad_norm": 0.3499196767807007, | |
| "learning_rate": 9.806609450855511e-05, | |
| "loss": 2.1257, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 0.1285732468412185, | |
| "grad_norm": 0.3475897014141083, | |
| "learning_rate": 9.804653830532599e-05, | |
| "loss": 2.1718, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 0.12900470068967898, | |
| "grad_norm": 0.3465918302536011, | |
| "learning_rate": 9.80268856897437e-05, | |
| "loss": 2.1536, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 0.12943615453813945, | |
| "grad_norm": 0.3532946705818176, | |
| "learning_rate": 9.800713670124399e-05, | |
| "loss": 2.1877, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.12943615453813945, | |
| "eval_loss": 2.16441011428833, | |
| "eval_runtime": 2772.359, | |
| "eval_samples_per_second": 27.279, | |
| "eval_steps_per_second": 13.64, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.12986760838659991, | |
| "grad_norm": 0.3611823618412018, | |
| "learning_rate": 9.798729137945612e-05, | |
| "loss": 2.1417, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 0.13029906223506038, | |
| "grad_norm": 0.35656318068504333, | |
| "learning_rate": 9.79673497642026e-05, | |
| "loss": 2.1631, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 0.13073051608352085, | |
| "grad_norm": 0.3490156829357147, | |
| "learning_rate": 9.794731189549918e-05, | |
| "loss": 2.1591, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 0.1311619699319813, | |
| "grad_norm": 0.34025728702545166, | |
| "learning_rate": 9.792717781355473e-05, | |
| "loss": 2.1742, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 0.13159342378044175, | |
| "grad_norm": 0.33115556836128235, | |
| "learning_rate": 9.790694755877123e-05, | |
| "loss": 2.1254, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 0.13202487762890222, | |
| "grad_norm": 0.3736386001110077, | |
| "learning_rate": 9.788662117174362e-05, | |
| "loss": 2.1415, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 0.1324563314773627, | |
| "grad_norm": 0.3499143719673157, | |
| "learning_rate": 9.786619869325972e-05, | |
| "loss": 2.1464, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 0.13288778532582315, | |
| "grad_norm": 0.33793404698371887, | |
| "learning_rate": 9.784568016430024e-05, | |
| "loss": 2.1691, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 0.13331923917428362, | |
| "grad_norm": 0.3370670974254608, | |
| "learning_rate": 9.782506562603857e-05, | |
| "loss": 2.178, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 0.1337506930227441, | |
| "grad_norm": 0.34184974431991577, | |
| "learning_rate": 9.780435511984073e-05, | |
| "loss": 2.1577, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.13418214687120456, | |
| "grad_norm": 0.36029738187789917, | |
| "learning_rate": 9.778354868726544e-05, | |
| "loss": 2.1941, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 0.13461360071966502, | |
| "grad_norm": 0.342674195766449, | |
| "learning_rate": 9.776264637006376e-05, | |
| "loss": 2.1351, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 0.1350450545681255, | |
| "grad_norm": 0.35412395000457764, | |
| "learning_rate": 9.774164821017924e-05, | |
| "loss": 2.1266, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 0.13547650841658596, | |
| "grad_norm": 0.3524283766746521, | |
| "learning_rate": 9.772055424974775e-05, | |
| "loss": 2.1708, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 0.13590796226504642, | |
| "grad_norm": 0.33916088938713074, | |
| "learning_rate": 9.769936453109738e-05, | |
| "loss": 2.1399, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 0.1363394161135069, | |
| "grad_norm": 0.3382519483566284, | |
| "learning_rate": 9.767807909674838e-05, | |
| "loss": 2.1421, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 0.13677086996196736, | |
| "grad_norm": 0.3560401499271393, | |
| "learning_rate": 9.765669798941304e-05, | |
| "loss": 2.1618, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 0.1372023238104278, | |
| "grad_norm": 0.34324225783348083, | |
| "learning_rate": 9.763522125199567e-05, | |
| "loss": 2.1811, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 0.13763377765888826, | |
| "grad_norm": 0.3366933763027191, | |
| "learning_rate": 9.761364892759246e-05, | |
| "loss": 2.1512, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 0.13806523150734873, | |
| "grad_norm": 0.3429983854293823, | |
| "learning_rate": 9.759198105949141e-05, | |
| "loss": 2.155, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.13806523150734873, | |
| "eval_loss": 2.151211977005005, | |
| "eval_runtime": 2914.79, | |
| "eval_samples_per_second": 25.946, | |
| "eval_steps_per_second": 12.973, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.1384966853558092, | |
| "grad_norm": 0.349167138338089, | |
| "learning_rate": 9.757021769117225e-05, | |
| "loss": 2.1388, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 0.13892813920426966, | |
| "grad_norm": 0.3502015471458435, | |
| "learning_rate": 9.754835886630634e-05, | |
| "loss": 2.1562, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 0.13935959305273013, | |
| "grad_norm": 0.3419562876224518, | |
| "learning_rate": 9.752640462875655e-05, | |
| "loss": 2.1739, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 0.1397910469011906, | |
| "grad_norm": 0.3341248631477356, | |
| "learning_rate": 9.750435502257729e-05, | |
| "loss": 2.1397, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 0.14022250074965106, | |
| "grad_norm": 0.3449030816555023, | |
| "learning_rate": 9.748221009201429e-05, | |
| "loss": 2.1816, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 0.14065395459811153, | |
| "grad_norm": 0.33365193009376526, | |
| "learning_rate": 9.745996988150453e-05, | |
| "loss": 2.1348, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 0.141085408446572, | |
| "grad_norm": 0.34570497274398804, | |
| "learning_rate": 9.743763443567624e-05, | |
| "loss": 2.1359, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 0.14151686229503246, | |
| "grad_norm": 0.34485459327697754, | |
| "learning_rate": 9.741520379934876e-05, | |
| "loss": 2.1368, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 0.14194831614349293, | |
| "grad_norm": 0.3529675304889679, | |
| "learning_rate": 9.739267801753238e-05, | |
| "loss": 2.1268, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 0.1423797699919534, | |
| "grad_norm": 0.34084296226501465, | |
| "learning_rate": 9.737005713542837e-05, | |
| "loss": 2.1367, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.14281122384041384, | |
| "grad_norm": 0.33580780029296875, | |
| "learning_rate": 9.734734119842882e-05, | |
| "loss": 2.154, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 0.1432426776888743, | |
| "grad_norm": 0.3536083996295929, | |
| "learning_rate": 9.732453025211654e-05, | |
| "loss": 2.1375, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 0.14367413153733477, | |
| "grad_norm": 0.34920045733451843, | |
| "learning_rate": 9.730162434226502e-05, | |
| "loss": 2.1264, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 0.14410558538579524, | |
| "grad_norm": 0.34508806467056274, | |
| "learning_rate": 9.727862351483827e-05, | |
| "loss": 2.1525, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 0.1445370392342557, | |
| "grad_norm": 0.3422926664352417, | |
| "learning_rate": 9.72555278159908e-05, | |
| "loss": 2.1303, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 0.14496849308271617, | |
| "grad_norm": 0.3599984645843506, | |
| "learning_rate": 9.723233729206749e-05, | |
| "loss": 2.0945, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 0.14539994693117664, | |
| "grad_norm": 0.33985811471939087, | |
| "learning_rate": 9.720905198960348e-05, | |
| "loss": 2.1518, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 0.1458314007796371, | |
| "grad_norm": 0.3494996726512909, | |
| "learning_rate": 9.718567195532413e-05, | |
| "loss": 2.1526, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 0.14626285462809757, | |
| "grad_norm": 0.34371593594551086, | |
| "learning_rate": 9.716219723614486e-05, | |
| "loss": 2.1285, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 0.14669430847655804, | |
| "grad_norm": 0.34658172726631165, | |
| "learning_rate": 9.713862787917109e-05, | |
| "loss": 2.1572, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.14669430847655804, | |
| "eval_loss": 2.1390128135681152, | |
| "eval_runtime": 3020.0517, | |
| "eval_samples_per_second": 25.042, | |
| "eval_steps_per_second": 12.521, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.1471257623250185, | |
| "grad_norm": 0.3454444408416748, | |
| "learning_rate": 9.711496393169817e-05, | |
| "loss": 2.1443, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 0.14755721617347897, | |
| "grad_norm": 0.3368396759033203, | |
| "learning_rate": 9.709120544121123e-05, | |
| "loss": 2.1328, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 0.14798867002193944, | |
| "grad_norm": 0.35348692536354065, | |
| "learning_rate": 9.706735245538515e-05, | |
| "loss": 2.16, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 0.1484201238703999, | |
| "grad_norm": 0.35535895824432373, | |
| "learning_rate": 9.70434050220844e-05, | |
| "loss": 2.1613, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 0.14885157771886035, | |
| "grad_norm": 0.3357240557670593, | |
| "learning_rate": 9.701936318936297e-05, | |
| "loss": 2.1302, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 0.1492830315673208, | |
| "grad_norm": 0.33816105127334595, | |
| "learning_rate": 9.69952270054643e-05, | |
| "loss": 2.1315, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 0.14971448541578128, | |
| "grad_norm": 0.35641545057296753, | |
| "learning_rate": 9.697099651882116e-05, | |
| "loss": 2.121, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 0.15014593926424175, | |
| "grad_norm": 0.33782002329826355, | |
| "learning_rate": 9.694667177805552e-05, | |
| "loss": 2.1376, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 0.1505773931127022, | |
| "grad_norm": 0.33257773518562317, | |
| "learning_rate": 9.692225283197849e-05, | |
| "loss": 2.1159, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 0.15100884696116268, | |
| "grad_norm": 0.3327760100364685, | |
| "learning_rate": 9.689773972959026e-05, | |
| "loss": 2.1422, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.15144030080962315, | |
| "grad_norm": 0.33078470826148987, | |
| "learning_rate": 9.68731325200799e-05, | |
| "loss": 2.1279, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 0.1518717546580836, | |
| "grad_norm": 0.35639920830726624, | |
| "learning_rate": 9.684843125282538e-05, | |
| "loss": 2.1226, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 0.15230320850654408, | |
| "grad_norm": 0.33569735288619995, | |
| "learning_rate": 9.682363597739334e-05, | |
| "loss": 2.1263, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 0.15273466235500455, | |
| "grad_norm": 0.35261237621307373, | |
| "learning_rate": 9.679874674353913e-05, | |
| "loss": 2.1135, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 0.153166116203465, | |
| "grad_norm": 0.3444455564022064, | |
| "learning_rate": 9.67737636012066e-05, | |
| "loss": 2.0787, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 0.15359757005192548, | |
| "grad_norm": 0.32956936955451965, | |
| "learning_rate": 9.674868660052805e-05, | |
| "loss": 2.1283, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 0.15402902390038595, | |
| "grad_norm": 0.3369726836681366, | |
| "learning_rate": 9.672351579182413e-05, | |
| "loss": 2.1337, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 0.1544604777488464, | |
| "grad_norm": 0.3625190854072571, | |
| "learning_rate": 9.66982512256037e-05, | |
| "loss": 2.1165, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 0.15489193159730685, | |
| "grad_norm": 0.3413086235523224, | |
| "learning_rate": 9.667289295256379e-05, | |
| "loss": 2.1242, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 0.15532338544576732, | |
| "grad_norm": 0.34174641966819763, | |
| "learning_rate": 9.664744102358947e-05, | |
| "loss": 2.1325, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.15532338544576732, | |
| "eval_loss": 2.128389835357666, | |
| "eval_runtime": 2954.9582, | |
| "eval_samples_per_second": 25.594, | |
| "eval_steps_per_second": 12.797, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.1557548392942278, | |
| "grad_norm": 0.3447898328304291, | |
| "learning_rate": 9.662189548975373e-05, | |
| "loss": 2.1418, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 0.15618629314268825, | |
| "grad_norm": 0.3458497226238251, | |
| "learning_rate": 9.659625640231735e-05, | |
| "loss": 2.1415, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 0.15661774699114872, | |
| "grad_norm": 0.3496282994747162, | |
| "learning_rate": 9.657052381272893e-05, | |
| "loss": 2.1318, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 0.1570492008396092, | |
| "grad_norm": 0.3375896215438843, | |
| "learning_rate": 9.654469777262462e-05, | |
| "loss": 2.1342, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 0.15748065468806965, | |
| "grad_norm": 0.3378855288028717, | |
| "learning_rate": 9.651877833382815e-05, | |
| "loss": 2.126, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 0.15791210853653012, | |
| "grad_norm": 0.3357175588607788, | |
| "learning_rate": 9.649276554835058e-05, | |
| "loss": 2.14, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 0.1583435623849906, | |
| "grad_norm": 0.3392798602581024, | |
| "learning_rate": 9.64666594683904e-05, | |
| "loss": 2.1071, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 0.15877501623345106, | |
| "grad_norm": 0.33450230956077576, | |
| "learning_rate": 9.644046014633319e-05, | |
| "loss": 2.1219, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 0.15920647008191152, | |
| "grad_norm": 0.33854368329048157, | |
| "learning_rate": 9.641416763475174e-05, | |
| "loss": 2.1206, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 0.159637923930372, | |
| "grad_norm": 0.3468214273452759, | |
| "learning_rate": 9.638778198640576e-05, | |
| "loss": 2.1227, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.16006937777883246, | |
| "grad_norm": 0.3371368944644928, | |
| "learning_rate": 9.636130325424192e-05, | |
| "loss": 2.1259, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 0.1605008316272929, | |
| "grad_norm": 0.3576333224773407, | |
| "learning_rate": 9.633473149139361e-05, | |
| "loss": 2.1255, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 0.16093228547575336, | |
| "grad_norm": 0.34628191590309143, | |
| "learning_rate": 9.630806675118093e-05, | |
| "loss": 2.1616, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 0.16136373932421383, | |
| "grad_norm": 0.3629646599292755, | |
| "learning_rate": 9.628130908711056e-05, | |
| "loss": 2.1303, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 0.1617951931726743, | |
| "grad_norm": 0.3423612117767334, | |
| "learning_rate": 9.625445855287562e-05, | |
| "loss": 2.1291, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 0.16222664702113476, | |
| "grad_norm": 0.3248550295829773, | |
| "learning_rate": 9.622751520235564e-05, | |
| "loss": 2.1231, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 0.16265810086959523, | |
| "grad_norm": 0.3261340260505676, | |
| "learning_rate": 9.620047908961633e-05, | |
| "loss": 2.1206, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 0.1630895547180557, | |
| "grad_norm": 0.33387574553489685, | |
| "learning_rate": 9.61733502689096e-05, | |
| "loss": 2.1371, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 0.16352100856651616, | |
| "grad_norm": 0.3315752446651459, | |
| "learning_rate": 9.614612879467336e-05, | |
| "loss": 2.1134, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 0.16395246241497663, | |
| "grad_norm": 0.340299129486084, | |
| "learning_rate": 9.611881472153141e-05, | |
| "loss": 2.0927, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.16395246241497663, | |
| "eval_loss": 2.118671417236328, | |
| "eval_runtime": 3059.0098, | |
| "eval_samples_per_second": 24.723, | |
| "eval_steps_per_second": 12.362, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.1643839162634371, | |
| "grad_norm": 0.3347771465778351, | |
| "learning_rate": 9.609140810429347e-05, | |
| "loss": 2.072, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 0.16481537011189756, | |
| "grad_norm": 0.3293216824531555, | |
| "learning_rate": 9.606390899795483e-05, | |
| "loss": 2.0933, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 0.16524682396035803, | |
| "grad_norm": 0.3466013967990875, | |
| "learning_rate": 9.60363174576965e-05, | |
| "loss": 2.0871, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 0.1656782778088185, | |
| "grad_norm": 0.34989696741104126, | |
| "learning_rate": 9.600863353888485e-05, | |
| "loss": 2.1029, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 0.16610973165727894, | |
| "grad_norm": 0.3422117829322815, | |
| "learning_rate": 9.598085729707174e-05, | |
| "loss": 2.1076, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 0.1665411855057394, | |
| "grad_norm": 0.32813316583633423, | |
| "learning_rate": 9.59529887879942e-05, | |
| "loss": 2.089, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 0.16697263935419987, | |
| "grad_norm": 0.3397763967514038, | |
| "learning_rate": 9.592502806757442e-05, | |
| "loss": 2.0949, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 0.16740409320266034, | |
| "grad_norm": 0.3325379490852356, | |
| "learning_rate": 9.589697519191967e-05, | |
| "loss": 2.1023, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 0.1678355470511208, | |
| "grad_norm": 0.3426936864852905, | |
| "learning_rate": 9.586883021732213e-05, | |
| "loss": 2.1364, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 0.16826700089958127, | |
| "grad_norm": 0.3614189922809601, | |
| "learning_rate": 9.584059320025876e-05, | |
| "loss": 2.13, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.16869845474804174, | |
| "grad_norm": 0.3224940001964569, | |
| "learning_rate": 9.581226419739122e-05, | |
| "loss": 2.094, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 0.1691299085965022, | |
| "grad_norm": 0.34935611486434937, | |
| "learning_rate": 9.578384326556577e-05, | |
| "loss": 2.1284, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 0.16956136244496267, | |
| "grad_norm": 0.33929896354675293, | |
| "learning_rate": 9.575533046181315e-05, | |
| "loss": 2.1261, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 0.16999281629342314, | |
| "grad_norm": 0.31563815474510193, | |
| "learning_rate": 9.572672584334845e-05, | |
| "loss": 2.1374, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 0.1704242701418836, | |
| "grad_norm": 0.33976227045059204, | |
| "learning_rate": 9.569802946757097e-05, | |
| "loss": 2.0851, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 0.17085572399034407, | |
| "grad_norm": 0.35277485847473145, | |
| "learning_rate": 9.566924139206413e-05, | |
| "loss": 2.1082, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 0.17128717783880454, | |
| "grad_norm": 0.34546715021133423, | |
| "learning_rate": 9.564036167459544e-05, | |
| "loss": 2.1219, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 0.171718631687265, | |
| "grad_norm": 0.32824838161468506, | |
| "learning_rate": 9.56113903731162e-05, | |
| "loss": 2.1053, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 0.17215008553572544, | |
| "grad_norm": 0.35051465034484863, | |
| "learning_rate": 9.558232754576154e-05, | |
| "loss": 2.0876, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 0.1725815393841859, | |
| "grad_norm": 0.3445969223976135, | |
| "learning_rate": 9.555317325085024e-05, | |
| "loss": 2.116, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.1725815393841859, | |
| "eval_loss": 2.1093990802764893, | |
| "eval_runtime": 2936.2197, | |
| "eval_samples_per_second": 25.757, | |
| "eval_steps_per_second": 12.878, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.17301299323264638, | |
| "grad_norm": 0.35169127583503723, | |
| "learning_rate": 9.552392754688465e-05, | |
| "loss": 2.075, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 0.17344444708110685, | |
| "grad_norm": 0.34275829792022705, | |
| "learning_rate": 9.549459049255049e-05, | |
| "loss": 2.1033, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 0.1738759009295673, | |
| "grad_norm": 0.3412991464138031, | |
| "learning_rate": 9.546516214671684e-05, | |
| "loss": 2.1315, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 0.17430735477802778, | |
| "grad_norm": 0.32284075021743774, | |
| "learning_rate": 9.543564256843593e-05, | |
| "loss": 2.121, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 0.17473880862648825, | |
| "grad_norm": 0.33116912841796875, | |
| "learning_rate": 9.540603181694307e-05, | |
| "loss": 2.0749, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 0.1751702624749487, | |
| "grad_norm": 0.34137117862701416, | |
| "learning_rate": 9.537632995165658e-05, | |
| "loss": 2.1103, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 0.17560171632340918, | |
| "grad_norm": 0.3352222442626953, | |
| "learning_rate": 9.534653703217751e-05, | |
| "loss": 2.0891, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 0.17603317017186965, | |
| "grad_norm": 0.3311406373977661, | |
| "learning_rate": 9.53166531182897e-05, | |
| "loss": 2.1171, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 0.1764646240203301, | |
| "grad_norm": 0.33530110120773315, | |
| "learning_rate": 9.528667826995955e-05, | |
| "loss": 2.1015, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 0.17689607786879058, | |
| "grad_norm": 0.32751473784446716, | |
| "learning_rate": 9.525661254733596e-05, | |
| "loss": 2.0727, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.17732753171725105, | |
| "grad_norm": 0.3411867320537567, | |
| "learning_rate": 9.522645601075016e-05, | |
| "loss": 2.1006, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 0.17775898556571149, | |
| "grad_norm": 0.3302497863769531, | |
| "learning_rate": 9.519620872071561e-05, | |
| "loss": 2.1083, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 0.17819043941417195, | |
| "grad_norm": 0.3380363881587982, | |
| "learning_rate": 9.516587073792792e-05, | |
| "loss": 2.115, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 0.17862189326263242, | |
| "grad_norm": 0.3405621349811554, | |
| "learning_rate": 9.513544212326461e-05, | |
| "loss": 2.1424, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 0.1790533471110929, | |
| "grad_norm": 0.3401827812194824, | |
| "learning_rate": 9.510492293778513e-05, | |
| "loss": 2.0871, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 0.17948480095955335, | |
| "grad_norm": 0.34303560853004456, | |
| "learning_rate": 9.507431324273067e-05, | |
| "loss": 2.1215, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 0.17991625480801382, | |
| "grad_norm": 0.34137266874313354, | |
| "learning_rate": 9.504361309952401e-05, | |
| "loss": 2.1171, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 0.1803477086564743, | |
| "grad_norm": 0.32752570509910583, | |
| "learning_rate": 9.501282256976945e-05, | |
| "loss": 2.1038, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 0.18077916250493475, | |
| "grad_norm": 0.33904024958610535, | |
| "learning_rate": 9.498194171525267e-05, | |
| "loss": 2.1143, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 0.18121061635339522, | |
| "grad_norm": 0.3322898745536804, | |
| "learning_rate": 9.495097059794057e-05, | |
| "loss": 2.1101, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.18121061635339522, | |
| "eval_loss": 2.1013271808624268, | |
| "eval_runtime": 2858.1247, | |
| "eval_samples_per_second": 26.461, | |
| "eval_steps_per_second": 13.23, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.1816420702018557, | |
| "grad_norm": 0.3301975131034851, | |
| "learning_rate": 9.491990927998117e-05, | |
| "loss": 2.0876, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 0.18207352405031615, | |
| "grad_norm": 0.3501399755477905, | |
| "learning_rate": 9.488875782370356e-05, | |
| "loss": 2.1131, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 0.18250497789877662, | |
| "grad_norm": 0.33008062839508057, | |
| "learning_rate": 9.485751629161763e-05, | |
| "loss": 2.1086, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 0.1829364317472371, | |
| "grad_norm": 0.3411719501018524, | |
| "learning_rate": 9.482618474641407e-05, | |
| "loss": 2.0703, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 0.18336788559569756, | |
| "grad_norm": 0.33804383873939514, | |
| "learning_rate": 9.479476325096414e-05, | |
| "loss": 2.1167, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 0.183799339444158, | |
| "grad_norm": 0.34236985445022583, | |
| "learning_rate": 9.476325186831967e-05, | |
| "loss": 2.0685, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 0.18423079329261846, | |
| "grad_norm": 0.3437114953994751, | |
| "learning_rate": 9.47316506617128e-05, | |
| "loss": 2.0929, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 0.18466224714107893, | |
| "grad_norm": 0.33738139271736145, | |
| "learning_rate": 9.469995969455593e-05, | |
| "loss": 2.0855, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 0.1850937009895394, | |
| "grad_norm": 0.3287017345428467, | |
| "learning_rate": 9.46681790304416e-05, | |
| "loss": 2.1089, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 0.18552515483799986, | |
| "grad_norm": 0.3263178765773773, | |
| "learning_rate": 9.463630873314233e-05, | |
| "loss": 2.0775, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.18595660868646033, | |
| "grad_norm": 0.3348768353462219, | |
| "learning_rate": 9.460434886661049e-05, | |
| "loss": 2.1013, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 0.1863880625349208, | |
| "grad_norm": 0.3428316116333008, | |
| "learning_rate": 9.457229949497822e-05, | |
| "loss": 2.0863, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 0.18681951638338126, | |
| "grad_norm": 0.3480794429779053, | |
| "learning_rate": 9.454016068255717e-05, | |
| "loss": 2.1077, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 0.18725097023184173, | |
| "grad_norm": 0.3354591131210327, | |
| "learning_rate": 9.450793249383858e-05, | |
| "loss": 2.0962, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 0.1876824240803022, | |
| "grad_norm": 0.33805784583091736, | |
| "learning_rate": 9.4475614993493e-05, | |
| "loss": 2.1287, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 0.18811387792876266, | |
| "grad_norm": 0.32974615693092346, | |
| "learning_rate": 9.444320824637014e-05, | |
| "loss": 2.0657, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 0.18854533177722313, | |
| "grad_norm": 0.33754146099090576, | |
| "learning_rate": 9.441071231749889e-05, | |
| "loss": 2.0863, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 0.1889767856256836, | |
| "grad_norm": 0.3371632993221283, | |
| "learning_rate": 9.437812727208702e-05, | |
| "loss": 2.1099, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 0.18940823947414406, | |
| "grad_norm": 0.3343505561351776, | |
| "learning_rate": 9.434545317552115e-05, | |
| "loss": 2.0833, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 0.1898396933226045, | |
| "grad_norm": 0.33940064907073975, | |
| "learning_rate": 9.431269009336661e-05, | |
| "loss": 2.1066, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.1898396933226045, | |
| "eval_loss": 2.093938112258911, | |
| "eval_runtime": 2820.9238, | |
| "eval_samples_per_second": 26.81, | |
| "eval_steps_per_second": 13.405, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.19027114717106497, | |
| "grad_norm": 0.32727915048599243, | |
| "learning_rate": 9.427983809136727e-05, | |
| "loss": 2.0494, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 0.19070260101952544, | |
| "grad_norm": 0.3239200711250305, | |
| "learning_rate": 9.424689723544545e-05, | |
| "loss": 2.0716, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 0.1911340548679859, | |
| "grad_norm": 0.3169387876987457, | |
| "learning_rate": 9.421386759170177e-05, | |
| "loss": 2.0903, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 0.19156550871644637, | |
| "grad_norm": 0.3533649146556854, | |
| "learning_rate": 9.418074922641498e-05, | |
| "loss": 2.1177, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 0.19199696256490684, | |
| "grad_norm": 0.3385666608810425, | |
| "learning_rate": 9.41475422060419e-05, | |
| "loss": 2.0788, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 0.1924284164133673, | |
| "grad_norm": 0.3421185612678528, | |
| "learning_rate": 9.411424659721722e-05, | |
| "loss": 2.1098, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 0.19285987026182777, | |
| "grad_norm": 0.3351210653781891, | |
| "learning_rate": 9.408086246675345e-05, | |
| "loss": 2.0829, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 0.19329132411028824, | |
| "grad_norm": 0.32542884349823, | |
| "learning_rate": 9.404738988164065e-05, | |
| "loss": 2.056, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 0.1937227779587487, | |
| "grad_norm": 0.32976922392845154, | |
| "learning_rate": 9.401382890904645e-05, | |
| "loss": 2.1095, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 0.19415423180720917, | |
| "grad_norm": 0.3366084098815918, | |
| "learning_rate": 9.39801796163158e-05, | |
| "loss": 2.051, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.19458568565566964, | |
| "grad_norm": 0.32825997471809387, | |
| "learning_rate": 9.394644207097089e-05, | |
| "loss": 2.0851, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 0.1950171395041301, | |
| "grad_norm": 0.33241936564445496, | |
| "learning_rate": 9.391261634071102e-05, | |
| "loss": 2.1218, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 0.19544859335259054, | |
| "grad_norm": 0.33546289801597595, | |
| "learning_rate": 9.38787024934124e-05, | |
| "loss": 2.0866, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 0.195880047201051, | |
| "grad_norm": 0.3313877284526825, | |
| "learning_rate": 9.384470059712811e-05, | |
| "loss": 2.0722, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 0.19631150104951148, | |
| "grad_norm": 0.3336232900619507, | |
| "learning_rate": 9.381061072008786e-05, | |
| "loss": 2.0831, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 0.19674295489797194, | |
| "grad_norm": 0.3445178270339966, | |
| "learning_rate": 9.377643293069798e-05, | |
| "loss": 2.1004, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 0.1971744087464324, | |
| "grad_norm": 0.33248141407966614, | |
| "learning_rate": 9.374216729754113e-05, | |
| "loss": 2.071, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 0.19760586259489288, | |
| "grad_norm": 0.3260776996612549, | |
| "learning_rate": 9.370781388937631e-05, | |
| "loss": 2.1116, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 0.19803731644335335, | |
| "grad_norm": 0.3381562829017639, | |
| "learning_rate": 9.367337277513858e-05, | |
| "loss": 2.0624, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 0.1984687702918138, | |
| "grad_norm": 0.3296578824520111, | |
| "learning_rate": 9.363884402393905e-05, | |
| "loss": 2.0885, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.1984687702918138, | |
| "eval_loss": 2.086437463760376, | |
| "eval_runtime": 2791.6728, | |
| "eval_samples_per_second": 27.091, | |
| "eval_steps_per_second": 13.545, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.19890022414027428, | |
| "grad_norm": 0.3351718485355377, | |
| "learning_rate": 9.360422770506469e-05, | |
| "loss": 2.1152, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 0.19933167798873475, | |
| "grad_norm": 0.344475120306015, | |
| "learning_rate": 9.356952388797815e-05, | |
| "loss": 2.0992, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 0.1997631318371952, | |
| "grad_norm": 0.32130861282348633, | |
| "learning_rate": 9.353473264231768e-05, | |
| "loss": 2.1008, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 0.20019458568565568, | |
| "grad_norm": 0.3410738408565521, | |
| "learning_rate": 9.349985403789696e-05, | |
| "loss": 2.0973, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 0.20062603953411615, | |
| "grad_norm": 0.3305729627609253, | |
| "learning_rate": 9.346488814470498e-05, | |
| "loss": 2.0821, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 0.2010574933825766, | |
| "grad_norm": 0.3449229598045349, | |
| "learning_rate": 9.342983503290587e-05, | |
| "loss": 2.1085, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 0.20148894723103705, | |
| "grad_norm": 0.3349471092224121, | |
| "learning_rate": 9.339469477283882e-05, | |
| "loss": 2.0927, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 0.20192040107949752, | |
| "grad_norm": 0.35080793499946594, | |
| "learning_rate": 9.335946743501783e-05, | |
| "loss": 2.083, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 0.20235185492795799, | |
| "grad_norm": 0.33375632762908936, | |
| "learning_rate": 9.332415309013165e-05, | |
| "loss": 2.1197, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 0.20278330877641845, | |
| "grad_norm": 0.3398565351963043, | |
| "learning_rate": 9.328875180904368e-05, | |
| "loss": 2.0895, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.20321476262487892, | |
| "grad_norm": 0.3695748746395111, | |
| "learning_rate": 9.32532636627917e-05, | |
| "loss": 2.0526, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 0.2036462164733394, | |
| "grad_norm": 0.3366464078426361, | |
| "learning_rate": 9.321768872258784e-05, | |
| "loss": 2.0781, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 0.20407767032179985, | |
| "grad_norm": 0.32691192626953125, | |
| "learning_rate": 9.31820270598184e-05, | |
| "loss": 2.0826, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 0.20450912417026032, | |
| "grad_norm": 0.33307206630706787, | |
| "learning_rate": 9.314627874604364e-05, | |
| "loss": 2.0769, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 0.2049405780187208, | |
| "grad_norm": 0.32544660568237305, | |
| "learning_rate": 9.311044385299778e-05, | |
| "loss": 2.0789, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 0.20537203186718125, | |
| "grad_norm": 0.3445671796798706, | |
| "learning_rate": 9.307452245258872e-05, | |
| "loss": 2.063, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 0.20580348571564172, | |
| "grad_norm": 0.3327982723712921, | |
| "learning_rate": 9.303851461689794e-05, | |
| "loss": 2.0549, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 0.2062349395641022, | |
| "grad_norm": 0.34181317687034607, | |
| "learning_rate": 9.300242041818044e-05, | |
| "loss": 2.1197, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 0.20666639341256265, | |
| "grad_norm": 0.3366542160511017, | |
| "learning_rate": 9.296623992886442e-05, | |
| "loss": 2.0681, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 0.2070978472610231, | |
| "grad_norm": 0.33302879333496094, | |
| "learning_rate": 9.292997322155129e-05, | |
| "loss": 2.096, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.2070978472610231, | |
| "eval_loss": 2.080108404159546, | |
| "eval_runtime": 2808.0771, | |
| "eval_samples_per_second": 26.932, | |
| "eval_steps_per_second": 13.466, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.20752930110948356, | |
| "grad_norm": 0.32614579796791077, | |
| "learning_rate": 9.28936203690155e-05, | |
| "loss": 2.0692, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 0.20796075495794403, | |
| "grad_norm": 0.33064261078834534, | |
| "learning_rate": 9.285718144420429e-05, | |
| "loss": 2.0901, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 0.2083922088064045, | |
| "grad_norm": 0.31495994329452515, | |
| "learning_rate": 9.282065652023767e-05, | |
| "loss": 2.0594, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 0.20882366265486496, | |
| "grad_norm": 0.3352917730808258, | |
| "learning_rate": 9.278404567040817e-05, | |
| "loss": 2.0911, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 0.20925511650332543, | |
| "grad_norm": 0.34422972798347473, | |
| "learning_rate": 9.274734896818082e-05, | |
| "loss": 2.0875, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 0.2096865703517859, | |
| "grad_norm": 0.3241231143474579, | |
| "learning_rate": 9.271056648719285e-05, | |
| "loss": 2.0378, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 0.21011802420024636, | |
| "grad_norm": 0.33844032883644104, | |
| "learning_rate": 9.267369830125367e-05, | |
| "loss": 2.0835, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 0.21054947804870683, | |
| "grad_norm": 0.32184383273124695, | |
| "learning_rate": 9.263674448434461e-05, | |
| "loss": 2.0463, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 0.2109809318971673, | |
| "grad_norm": 0.3213908076286316, | |
| "learning_rate": 9.259970511061891e-05, | |
| "loss": 2.052, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 0.21141238574562776, | |
| "grad_norm": 0.335366815328598, | |
| "learning_rate": 9.256258025440143e-05, | |
| "loss": 2.102, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.21184383959408823, | |
| "grad_norm": 0.3458082377910614, | |
| "learning_rate": 9.252536999018856e-05, | |
| "loss": 2.0397, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 0.2122752934425487, | |
| "grad_norm": 0.3523118197917938, | |
| "learning_rate": 9.248807439264813e-05, | |
| "loss": 2.0426, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 0.21270674729100916, | |
| "grad_norm": 0.34614822268486023, | |
| "learning_rate": 9.245069353661916e-05, | |
| "loss": 2.0637, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 0.2131382011394696, | |
| "grad_norm": 0.3328052759170532, | |
| "learning_rate": 9.241322749711176e-05, | |
| "loss": 2.0866, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 0.21356965498793007, | |
| "grad_norm": 0.3421047031879425, | |
| "learning_rate": 9.237567634930699e-05, | |
| "loss": 2.0718, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 0.21400110883639054, | |
| "grad_norm": 0.3291900157928467, | |
| "learning_rate": 9.233804016855665e-05, | |
| "loss": 2.0594, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 0.214432562684851, | |
| "grad_norm": 0.3402421176433563, | |
| "learning_rate": 9.230031903038323e-05, | |
| "loss": 2.0633, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 0.21486401653331147, | |
| "grad_norm": 0.33524367213249207, | |
| "learning_rate": 9.226251301047966e-05, | |
| "loss": 2.0842, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 0.21529547038177194, | |
| "grad_norm": 0.3257879614830017, | |
| "learning_rate": 9.222462218470918e-05, | |
| "loss": 2.0567, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 0.2157269242302324, | |
| "grad_norm": 0.3245730698108673, | |
| "learning_rate": 9.218664662910528e-05, | |
| "loss": 2.0792, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.2157269242302324, | |
| "eval_loss": 2.073448419570923, | |
| "eval_runtime": 2834.2159, | |
| "eval_samples_per_second": 26.684, | |
| "eval_steps_per_second": 13.342, | |
| "step": 25000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 115888, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.113474465792e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |