{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8208469055374593, "eval_steps": 42, "global_step": 252, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003257328990228013, "eval_loss": 6.045982360839844, "eval_runtime": 22.8171, "eval_samples_per_second": 22.658, "eval_steps_per_second": 5.697, "step": 1 }, { "epoch": 0.009771986970684038, "grad_norm": 7.921604633331299, "learning_rate": 3e-05, "loss": 5.6539, "step": 3 }, { "epoch": 0.019543973941368076, "grad_norm": 8.00938606262207, "learning_rate": 6e-05, "loss": 4.9183, "step": 6 }, { "epoch": 0.029315960912052116, "grad_norm": 7.276914596557617, "learning_rate": 9e-05, "loss": 3.1081, "step": 9 }, { "epoch": 0.03908794788273615, "grad_norm": 4.77871036529541, "learning_rate": 9.999588943391597e-05, "loss": 2.5988, "step": 12 }, { "epoch": 0.048859934853420196, "grad_norm": 4.304369926452637, "learning_rate": 9.99743108100344e-05, "loss": 2.3816, "step": 15 }, { "epoch": 0.05863192182410423, "grad_norm": 4.264029502868652, "learning_rate": 9.993424445916923e-05, "loss": 2.302, "step": 18 }, { "epoch": 0.06840390879478828, "grad_norm": 3.8603391647338867, "learning_rate": 9.987570520365104e-05, "loss": 2.205, "step": 21 }, { "epoch": 0.0781758957654723, "grad_norm": 3.673312187194824, "learning_rate": 9.979871469976196e-05, "loss": 1.9937, "step": 24 }, { "epoch": 0.08794788273615635, "grad_norm": 3.670795202255249, "learning_rate": 9.970330142972401e-05, "loss": 1.9628, "step": 27 }, { "epoch": 0.09771986970684039, "grad_norm": 4.024224281311035, "learning_rate": 9.95895006911623e-05, "loss": 1.9128, "step": 30 }, { "epoch": 0.10749185667752444, "grad_norm": 3.246854782104492, "learning_rate": 9.945735458404681e-05, "loss": 1.8845, "step": 33 }, { "epoch": 0.11726384364820847, "grad_norm": 3.4581379890441895, "learning_rate": 9.930691199511775e-05, "loss": 1.7264, "step": 36 }, { "epoch": 0.1270358306188925, "grad_norm": 3.0552656650543213, "learning_rate": 9.91382285798002e-05, "loss": 1.6868, "step": 39 }, { "epoch": 0.13680781758957655, "grad_norm": 2.96836519241333, "learning_rate": 9.895136674161465e-05, "loss": 1.7595, "step": 42 }, { "epoch": 0.13680781758957655, "eval_loss": 1.7990862131118774, "eval_runtime": 22.1331, "eval_samples_per_second": 23.359, "eval_steps_per_second": 5.874, "step": 42 }, { "epoch": 0.1465798045602606, "grad_norm": 2.9569833278656006, "learning_rate": 9.874639560909117e-05, "loss": 1.6317, "step": 45 }, { "epoch": 0.1563517915309446, "grad_norm": 2.9722719192504883, "learning_rate": 9.852339101019574e-05, "loss": 1.4734, "step": 48 }, { "epoch": 0.16612377850162866, "grad_norm": 3.663701295852661, "learning_rate": 9.828243544427796e-05, "loss": 1.7673, "step": 51 }, { "epoch": 0.1758957654723127, "grad_norm": 2.7686305046081543, "learning_rate": 9.802361805155097e-05, "loss": 1.5428, "step": 54 }, { "epoch": 0.18566775244299674, "grad_norm": 2.781644582748413, "learning_rate": 9.774703458011453e-05, "loss": 1.7697, "step": 57 }, { "epoch": 0.19543973941368079, "grad_norm": 2.7254538536071777, "learning_rate": 9.745278735053343e-05, "loss": 1.5856, "step": 60 }, { "epoch": 0.20521172638436483, "grad_norm": 2.544527769088745, "learning_rate": 9.714098521798465e-05, "loss": 1.5917, "step": 63 }, { "epoch": 0.21498371335504887, "grad_norm": 3.226173162460327, "learning_rate": 9.681174353198687e-05, "loss": 1.5432, "step": 66 }, { "epoch": 0.2247557003257329, "grad_norm": 2.7061498165130615, "learning_rate": 9.64651840937276e-05, "loss": 1.6754, "step": 69 }, { "epoch": 0.23452768729641693, "grad_norm": 2.9291999340057373, "learning_rate": 9.610143511100354e-05, "loss": 1.5638, "step": 72 }, { "epoch": 0.24429967426710097, "grad_norm": 2.4796836376190186, "learning_rate": 9.572063115079063e-05, "loss": 1.474, "step": 75 }, { "epoch": 0.254071661237785, "grad_norm": 2.7676734924316406, "learning_rate": 9.53229130894619e-05, "loss": 1.5904, "step": 78 }, { "epoch": 0.26384364820846906, "grad_norm": 2.8210036754608154, "learning_rate": 9.490842806067095e-05, "loss": 1.6858, "step": 81 }, { "epoch": 0.2736156351791531, "grad_norm": 2.67210054397583, "learning_rate": 9.44773294009206e-05, "loss": 1.5434, "step": 84 }, { "epoch": 0.2736156351791531, "eval_loss": 1.6323200464248657, "eval_runtime": 22.1692, "eval_samples_per_second": 23.321, "eval_steps_per_second": 5.864, "step": 84 }, { "epoch": 0.28338762214983715, "grad_norm": 2.819925308227539, "learning_rate": 9.40297765928369e-05, "loss": 1.5062, "step": 87 }, { "epoch": 0.2931596091205212, "grad_norm": 2.894965410232544, "learning_rate": 9.356593520616948e-05, "loss": 1.5876, "step": 90 }, { "epoch": 0.30293159609120524, "grad_norm": 2.7666544914245605, "learning_rate": 9.308597683653975e-05, "loss": 1.4585, "step": 93 }, { "epoch": 0.3127035830618892, "grad_norm": 2.8919670581817627, "learning_rate": 9.259007904196023e-05, "loss": 1.5846, "step": 96 }, { "epoch": 0.32247557003257327, "grad_norm": 2.660968542098999, "learning_rate": 9.207842527714767e-05, "loss": 1.5343, "step": 99 }, { "epoch": 0.3322475570032573, "grad_norm": 2.652714967727661, "learning_rate": 9.155120482565521e-05, "loss": 1.591, "step": 102 }, { "epoch": 0.34201954397394135, "grad_norm": 2.7949202060699463, "learning_rate": 9.10086127298478e-05, "loss": 1.4614, "step": 105 }, { "epoch": 0.3517915309446254, "grad_norm": 2.527437686920166, "learning_rate": 9.045084971874738e-05, "loss": 1.5641, "step": 108 }, { "epoch": 0.36156351791530944, "grad_norm": 2.352187395095825, "learning_rate": 8.987812213377424e-05, "loss": 1.4186, "step": 111 }, { "epoch": 0.3713355048859935, "grad_norm": 2.577479124069214, "learning_rate": 8.929064185241213e-05, "loss": 1.4607, "step": 114 }, { "epoch": 0.3811074918566775, "grad_norm": 2.7002623081207275, "learning_rate": 8.868862620982534e-05, "loss": 1.6182, "step": 117 }, { "epoch": 0.39087947882736157, "grad_norm": 2.521603584289551, "learning_rate": 8.807229791845673e-05, "loss": 1.4524, "step": 120 }, { "epoch": 0.4006514657980456, "grad_norm": 2.723996877670288, "learning_rate": 8.744188498563641e-05, "loss": 1.2303, "step": 123 }, { "epoch": 0.41042345276872966, "grad_norm": 2.888556957244873, "learning_rate": 8.679762062923175e-05, "loss": 1.376, "step": 126 }, { "epoch": 0.41042345276872966, "eval_loss": 1.5374845266342163, "eval_runtime": 22.1691, "eval_samples_per_second": 23.321, "eval_steps_per_second": 5.864, "step": 126 }, { "epoch": 0.4201954397394137, "grad_norm": 2.6070010662078857, "learning_rate": 8.613974319136958e-05, "loss": 1.313, "step": 129 }, { "epoch": 0.42996742671009774, "grad_norm": 2.7287192344665527, "learning_rate": 8.54684960502629e-05, "loss": 1.4142, "step": 132 }, { "epoch": 0.43973941368078173, "grad_norm": 2.9047019481658936, "learning_rate": 8.478412753017433e-05, "loss": 1.4128, "step": 135 }, { "epoch": 0.4495114006514658, "grad_norm": 3.378373622894287, "learning_rate": 8.408689080954998e-05, "loss": 1.5116, "step": 138 }, { "epoch": 0.4592833876221498, "grad_norm": 2.828160285949707, "learning_rate": 8.33770438273574e-05, "loss": 1.3085, "step": 141 }, { "epoch": 0.46905537459283386, "grad_norm": 2.577202081680298, "learning_rate": 8.265484918766243e-05, "loss": 1.3892, "step": 144 }, { "epoch": 0.4788273615635179, "grad_norm": 2.659550428390503, "learning_rate": 8.192057406248028e-05, "loss": 1.4546, "step": 147 }, { "epoch": 0.48859934853420195, "grad_norm": 3.5885229110717773, "learning_rate": 8.117449009293668e-05, "loss": 1.2797, "step": 150 }, { "epoch": 0.498371335504886, "grad_norm": 2.8732728958129883, "learning_rate": 8.041687328877567e-05, "loss": 1.405, "step": 153 }, { "epoch": 0.50814332247557, "grad_norm": 2.7093594074249268, "learning_rate": 7.964800392625129e-05, "loss": 1.4728, "step": 156 }, { "epoch": 0.5179153094462541, "grad_norm": 2.460803747177124, "learning_rate": 7.886816644444098e-05, "loss": 1.5428, "step": 159 }, { "epoch": 0.5276872964169381, "grad_norm": 2.5754201412200928, "learning_rate": 7.807764934001874e-05, "loss": 1.332, "step": 162 }, { "epoch": 0.5374592833876222, "grad_norm": 2.630692481994629, "learning_rate": 7.727674506052743e-05, "loss": 1.4519, "step": 165 }, { "epoch": 0.5472312703583062, "grad_norm": 3.039989948272705, "learning_rate": 7.646574989618938e-05, "loss": 1.4669, "step": 168 }, { "epoch": 0.5472312703583062, "eval_loss": 1.471885085105896, "eval_runtime": 22.1606, "eval_samples_per_second": 23.33, "eval_steps_per_second": 5.866, "step": 168 }, { "epoch": 0.5570032573289903, "grad_norm": 2.361525297164917, "learning_rate": 7.564496387029532e-05, "loss": 1.3612, "step": 171 }, { "epoch": 0.5667752442996743, "grad_norm": 2.969438076019287, "learning_rate": 7.481469062821252e-05, "loss": 1.3524, "step": 174 }, { "epoch": 0.5765472312703583, "grad_norm": 2.654615879058838, "learning_rate": 7.39752373250527e-05, "loss": 1.3684, "step": 177 }, { "epoch": 0.5863192182410424, "grad_norm": 2.500892162322998, "learning_rate": 7.312691451204178e-05, "loss": 1.2688, "step": 180 }, { "epoch": 0.5960912052117264, "grad_norm": 2.7902750968933105, "learning_rate": 7.227003602163295e-05, "loss": 1.213, "step": 183 }, { "epoch": 0.6058631921824105, "grad_norm": 2.477710008621216, "learning_rate": 7.14049188514063e-05, "loss": 1.4632, "step": 186 }, { "epoch": 0.6156351791530945, "grad_norm": 2.8017044067382812, "learning_rate": 7.05318830467969e-05, "loss": 1.4655, "step": 189 }, { "epoch": 0.6254071661237784, "grad_norm": 2.2910585403442383, "learning_rate": 6.965125158269619e-05, "loss": 1.3622, "step": 192 }, { "epoch": 0.6351791530944625, "grad_norm": 2.3485262393951416, "learning_rate": 6.876335024396872e-05, "loss": 1.3191, "step": 195 }, { "epoch": 0.6449511400651465, "grad_norm": 2.221445083618164, "learning_rate": 6.786850750493006e-05, "loss": 1.2862, "step": 198 }, { "epoch": 0.6547231270358306, "grad_norm": 2.2322473526000977, "learning_rate": 6.696705440782938e-05, "loss": 1.2873, "step": 201 }, { "epoch": 0.6644951140065146, "grad_norm": 2.134915828704834, "learning_rate": 6.605932444038229e-05, "loss": 1.2457, "step": 204 }, { "epoch": 0.6742671009771987, "grad_norm": 2.2600533962249756, "learning_rate": 6.514565341239861e-05, "loss": 1.3544, "step": 207 }, { "epoch": 0.6840390879478827, "grad_norm": 2.104802370071411, "learning_rate": 6.422637933155162e-05, "loss": 1.2662, "step": 210 }, { "epoch": 0.6840390879478827, "eval_loss": 1.3924102783203125, "eval_runtime": 22.1429, "eval_samples_per_second": 23.348, "eval_steps_per_second": 5.871, "step": 210 }, { "epoch": 0.6938110749185668, "grad_norm": 2.490628719329834, "learning_rate": 6.330184227833376e-05, "loss": 1.4418, "step": 213 }, { "epoch": 0.7035830618892508, "grad_norm": 2.3577187061309814, "learning_rate": 6.237238428024572e-05, "loss": 1.2653, "step": 216 }, { "epoch": 0.7133550488599348, "grad_norm": 2.6168723106384277, "learning_rate": 6.143834918526527e-05, "loss": 1.4725, "step": 219 }, { "epoch": 0.7231270358306189, "grad_norm": 2.343191146850586, "learning_rate": 6.0500082534642464e-05, "loss": 1.3543, "step": 222 }, { "epoch": 0.7328990228013029, "grad_norm": 2.482631206512451, "learning_rate": 5.955793143506863e-05, "loss": 1.3523, "step": 225 }, { "epoch": 0.742671009771987, "grad_norm": 2.1662986278533936, "learning_rate": 5.861224443026595e-05, "loss": 1.2848, "step": 228 }, { "epoch": 0.752442996742671, "grad_norm": 2.496674060821533, "learning_rate": 5.766337137204579e-05, "loss": 1.3919, "step": 231 }, { "epoch": 0.762214983713355, "grad_norm": 2.5266311168670654, "learning_rate": 5.6711663290882776e-05, "loss": 1.4481, "step": 234 }, { "epoch": 0.7719869706840391, "grad_norm": 2.154391288757324, "learning_rate": 5.575747226605298e-05, "loss": 1.1874, "step": 237 }, { "epoch": 0.7817589576547231, "grad_norm": 2.2282350063323975, "learning_rate": 5.480115129538409e-05, "loss": 1.2684, "step": 240 }, { "epoch": 0.7915309446254072, "grad_norm": 2.377781391143799, "learning_rate": 5.384305416466584e-05, "loss": 1.2627, "step": 243 }, { "epoch": 0.8013029315960912, "grad_norm": 2.505831480026245, "learning_rate": 5.288353531676873e-05, "loss": 1.2982, "step": 246 }, { "epoch": 0.8110749185667753, "grad_norm": 2.424248456954956, "learning_rate": 5.192294972051992e-05, "loss": 1.2661, "step": 249 }, { "epoch": 0.8208469055374593, "grad_norm": 2.3504350185394287, "learning_rate": 5.0961652739384356e-05, "loss": 1.3146, "step": 252 }, { "epoch": 0.8208469055374593, "eval_loss": 1.3527233600616455, "eval_runtime": 22.1769, "eval_samples_per_second": 23.313, "eval_steps_per_second": 5.862, "step": 252 } ], "logging_steps": 3, "max_steps": 500, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 42, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.9007271092197786e+17, "train_batch_size": 4, "trial_name": null, "trial_params": null }