| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.9069767441860463, |
| "eval_steps": 500, |
| "global_step": 84, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.03488372093023256, |
| "grad_norm": 34.417083740234375, |
| "learning_rate": 5.0000000000000004e-08, |
| "loss": 2.2576, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.06976744186046512, |
| "grad_norm": 34.94503402709961, |
| "learning_rate": 1.0000000000000001e-07, |
| "loss": 2.2928, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.10465116279069768, |
| "grad_norm": 34.11349105834961, |
| "learning_rate": 1.5000000000000002e-07, |
| "loss": 2.2273, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.13953488372093023, |
| "grad_norm": 35.753822326660156, |
| "learning_rate": 2.0000000000000002e-07, |
| "loss": 2.3202, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.1744186046511628, |
| "grad_norm": 34.57868957519531, |
| "learning_rate": 2.5000000000000004e-07, |
| "loss": 2.2515, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.20930232558139536, |
| "grad_norm": 34.45582580566406, |
| "learning_rate": 3.0000000000000004e-07, |
| "loss": 2.241, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.2441860465116279, |
| "grad_norm": 34.65391540527344, |
| "learning_rate": 3.5000000000000004e-07, |
| "loss": 2.2543, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.27906976744186046, |
| "grad_norm": 34.55806350708008, |
| "learning_rate": 4.0000000000000003e-07, |
| "loss": 2.2219, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.313953488372093, |
| "grad_norm": 34.35453414916992, |
| "learning_rate": 4.5000000000000003e-07, |
| "loss": 2.2105, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.3488372093023256, |
| "grad_norm": 34.049232482910156, |
| "learning_rate": 5.000000000000001e-07, |
| "loss": 2.1968, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.38372093023255816, |
| "grad_norm": 33.26045608520508, |
| "learning_rate": 5.5e-07, |
| "loss": 2.13, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.4186046511627907, |
| "grad_norm": 33.314796447753906, |
| "learning_rate": 6.000000000000001e-07, |
| "loss": 2.1306, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.45348837209302323, |
| "grad_norm": 32.00406265258789, |
| "learning_rate": 6.5e-07, |
| "loss": 2.0275, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.4883720930232558, |
| "grad_norm": 32.68654251098633, |
| "learning_rate": 7.000000000000001e-07, |
| "loss": 2.0442, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.5232558139534884, |
| "grad_norm": 31.643146514892578, |
| "learning_rate": 7.5e-07, |
| "loss": 1.9725, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.5581395348837209, |
| "grad_norm": 30.403486251831055, |
| "learning_rate": 8.000000000000001e-07, |
| "loss": 1.8724, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.5930232558139535, |
| "grad_norm": 29.716495513916016, |
| "learning_rate": 8.500000000000001e-07, |
| "loss": 1.8118, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.627906976744186, |
| "grad_norm": 29.22306251525879, |
| "learning_rate": 9.000000000000001e-07, |
| "loss": 1.7551, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.6627906976744186, |
| "grad_norm": 27.425785064697266, |
| "learning_rate": 9.500000000000001e-07, |
| "loss": 1.6087, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.6976744186046512, |
| "grad_norm": 26.653409957885742, |
| "learning_rate": 1.0000000000000002e-06, |
| "loss": 1.5191, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.7325581395348837, |
| "grad_norm": 25.145450592041016, |
| "learning_rate": 1.0500000000000001e-06, |
| "loss": 1.3758, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.7674418604651163, |
| "grad_norm": 24.684560775756836, |
| "learning_rate": 1.1e-06, |
| "loss": 1.2955, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.8023255813953488, |
| "grad_norm": 24.005306243896484, |
| "learning_rate": 1.1500000000000002e-06, |
| "loss": 1.1665, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.8372093023255814, |
| "grad_norm": 24.253219604492188, |
| "learning_rate": 1.2000000000000002e-06, |
| "loss": 1.0143, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.872093023255814, |
| "grad_norm": 24.97549057006836, |
| "learning_rate": 1.25e-06, |
| "loss": 0.875, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.9069767441860465, |
| "grad_norm": 22.349998474121094, |
| "learning_rate": 1.3e-06, |
| "loss": 0.6222, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.9418604651162791, |
| "grad_norm": 18.814115524291992, |
| "learning_rate": 1.3500000000000002e-06, |
| "loss": 0.4715, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.9767441860465116, |
| "grad_norm": 15.833977699279785, |
| "learning_rate": 1.4000000000000001e-06, |
| "loss": 0.374, |
| "step": 28 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 15.833977699279785, |
| "learning_rate": 1.45e-06, |
| "loss": 0.2818, |
| "step": 29 |
| }, |
| { |
| "epoch": 1.0348837209302326, |
| "grad_norm": 14.85798168182373, |
| "learning_rate": 1.5e-06, |
| "loss": 0.2559, |
| "step": 30 |
| }, |
| { |
| "epoch": 1.069767441860465, |
| "grad_norm": 6.723965167999268, |
| "learning_rate": 1.5500000000000002e-06, |
| "loss": 0.2073, |
| "step": 31 |
| }, |
| { |
| "epoch": 1.1046511627906976, |
| "grad_norm": 4.480190753936768, |
| "learning_rate": 1.6000000000000001e-06, |
| "loss": 0.1632, |
| "step": 32 |
| }, |
| { |
| "epoch": 1.1395348837209303, |
| "grad_norm": 3.228346824645996, |
| "learning_rate": 1.6500000000000003e-06, |
| "loss": 0.1432, |
| "step": 33 |
| }, |
| { |
| "epoch": 1.1744186046511629, |
| "grad_norm": 2.4646286964416504, |
| "learning_rate": 1.7000000000000002e-06, |
| "loss": 0.1166, |
| "step": 34 |
| }, |
| { |
| "epoch": 1.2093023255813953, |
| "grad_norm": 1.721303939819336, |
| "learning_rate": 1.75e-06, |
| "loss": 0.1034, |
| "step": 35 |
| }, |
| { |
| "epoch": 1.244186046511628, |
| "grad_norm": 1.6059235334396362, |
| "learning_rate": 1.8000000000000001e-06, |
| "loss": 0.0943, |
| "step": 36 |
| }, |
| { |
| "epoch": 1.2790697674418605, |
| "grad_norm": 1.390950322151184, |
| "learning_rate": 1.85e-06, |
| "loss": 0.0855, |
| "step": 37 |
| }, |
| { |
| "epoch": 1.3139534883720931, |
| "grad_norm": 1.2893034219741821, |
| "learning_rate": 1.9000000000000002e-06, |
| "loss": 0.0797, |
| "step": 38 |
| }, |
| { |
| "epoch": 1.3488372093023255, |
| "grad_norm": 1.0295542478561401, |
| "learning_rate": 1.9500000000000004e-06, |
| "loss": 0.072, |
| "step": 39 |
| }, |
| { |
| "epoch": 1.3837209302325582, |
| "grad_norm": 1.0248366594314575, |
| "learning_rate": 2.0000000000000003e-06, |
| "loss": 0.068, |
| "step": 40 |
| }, |
| { |
| "epoch": 1.4186046511627908, |
| "grad_norm": 0.7489507794380188, |
| "learning_rate": 2.05e-06, |
| "loss": 0.0613, |
| "step": 41 |
| }, |
| { |
| "epoch": 1.4534883720930232, |
| "grad_norm": 0.7149010300636292, |
| "learning_rate": 2.1000000000000002e-06, |
| "loss": 0.0596, |
| "step": 42 |
| }, |
| { |
| "epoch": 1.4883720930232558, |
| "grad_norm": 0.5656580328941345, |
| "learning_rate": 2.15e-06, |
| "loss": 0.0585, |
| "step": 43 |
| }, |
| { |
| "epoch": 1.5232558139534884, |
| "grad_norm": 0.6951081156730652, |
| "learning_rate": 2.2e-06, |
| "loss": 0.066, |
| "step": 44 |
| }, |
| { |
| "epoch": 1.558139534883721, |
| "grad_norm": 0.8682136535644531, |
| "learning_rate": 2.25e-06, |
| "loss": 0.0623, |
| "step": 45 |
| }, |
| { |
| "epoch": 1.5930232558139537, |
| "grad_norm": 0.6056260466575623, |
| "learning_rate": 2.3000000000000004e-06, |
| "loss": 0.054, |
| "step": 46 |
| }, |
| { |
| "epoch": 1.627906976744186, |
| "grad_norm": 0.35378050804138184, |
| "learning_rate": 2.35e-06, |
| "loss": 0.0626, |
| "step": 47 |
| }, |
| { |
| "epoch": 1.6627906976744184, |
| "grad_norm": 0.4003937244415283, |
| "learning_rate": 2.4000000000000003e-06, |
| "loss": 0.0524, |
| "step": 48 |
| }, |
| { |
| "epoch": 1.697674418604651, |
| "grad_norm": 0.527799665927887, |
| "learning_rate": 2.4500000000000003e-06, |
| "loss": 0.056, |
| "step": 49 |
| }, |
| { |
| "epoch": 1.7325581395348837, |
| "grad_norm": 0.5358547568321228, |
| "learning_rate": 2.5e-06, |
| "loss": 0.0527, |
| "step": 50 |
| }, |
| { |
| "epoch": 1.7674418604651163, |
| "grad_norm": 0.45014214515686035, |
| "learning_rate": 2.55e-06, |
| "loss": 0.0492, |
| "step": 51 |
| }, |
| { |
| "epoch": 1.802325581395349, |
| "grad_norm": 0.4785670340061188, |
| "learning_rate": 2.6e-06, |
| "loss": 0.0525, |
| "step": 52 |
| }, |
| { |
| "epoch": 1.8372093023255816, |
| "grad_norm": 0.4931303560733795, |
| "learning_rate": 2.6500000000000005e-06, |
| "loss": 0.0485, |
| "step": 53 |
| }, |
| { |
| "epoch": 1.872093023255814, |
| "grad_norm": 0.2308250367641449, |
| "learning_rate": 2.7000000000000004e-06, |
| "loss": 0.0499, |
| "step": 54 |
| }, |
| { |
| "epoch": 1.9069767441860463, |
| "grad_norm": 0.5777614712715149, |
| "learning_rate": 2.7500000000000004e-06, |
| "loss": 0.0527, |
| "step": 55 |
| }, |
| { |
| "epoch": 1.941860465116279, |
| "grad_norm": 0.32561731338500977, |
| "learning_rate": 2.8000000000000003e-06, |
| "loss": 0.052, |
| "step": 56 |
| }, |
| { |
| "epoch": 1.9767441860465116, |
| "grad_norm": 0.34972742199897766, |
| "learning_rate": 2.85e-06, |
| "loss": 0.0467, |
| "step": 57 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.43575090169906616, |
| "learning_rate": 2.9e-06, |
| "loss": 0.0441, |
| "step": 58 |
| }, |
| { |
| "epoch": 2.0348837209302326, |
| "grad_norm": 0.3564951419830322, |
| "learning_rate": 2.95e-06, |
| "loss": 0.0471, |
| "step": 59 |
| }, |
| { |
| "epoch": 2.0697674418604652, |
| "grad_norm": 0.3126629889011383, |
| "learning_rate": 3e-06, |
| "loss": 0.0494, |
| "step": 60 |
| }, |
| { |
| "epoch": 2.104651162790698, |
| "grad_norm": 0.4542100429534912, |
| "learning_rate": 3.05e-06, |
| "loss": 0.0456, |
| "step": 61 |
| }, |
| { |
| "epoch": 2.13953488372093, |
| "grad_norm": 0.4597831666469574, |
| "learning_rate": 3.1000000000000004e-06, |
| "loss": 0.049, |
| "step": 62 |
| }, |
| { |
| "epoch": 2.1744186046511627, |
| "grad_norm": 0.3821645677089691, |
| "learning_rate": 3.1500000000000003e-06, |
| "loss": 0.0461, |
| "step": 63 |
| }, |
| { |
| "epoch": 2.2093023255813953, |
| "grad_norm": 0.3955609202384949, |
| "learning_rate": 3.2000000000000003e-06, |
| "loss": 0.0472, |
| "step": 64 |
| }, |
| { |
| "epoch": 2.244186046511628, |
| "grad_norm": 0.33170193433761597, |
| "learning_rate": 3.2500000000000002e-06, |
| "loss": 0.0433, |
| "step": 65 |
| }, |
| { |
| "epoch": 2.2790697674418605, |
| "grad_norm": 0.389060378074646, |
| "learning_rate": 3.3000000000000006e-06, |
| "loss": 0.0438, |
| "step": 66 |
| }, |
| { |
| "epoch": 2.313953488372093, |
| "grad_norm": 0.31669771671295166, |
| "learning_rate": 3.3500000000000005e-06, |
| "loss": 0.0403, |
| "step": 67 |
| }, |
| { |
| "epoch": 2.3488372093023258, |
| "grad_norm": 0.36098915338516235, |
| "learning_rate": 3.4000000000000005e-06, |
| "loss": 0.0392, |
| "step": 68 |
| }, |
| { |
| "epoch": 2.383720930232558, |
| "grad_norm": 0.4905363619327545, |
| "learning_rate": 3.45e-06, |
| "loss": 0.0391, |
| "step": 69 |
| }, |
| { |
| "epoch": 2.4186046511627906, |
| "grad_norm": 0.32984688878059387, |
| "learning_rate": 3.5e-06, |
| "loss": 0.0432, |
| "step": 70 |
| }, |
| { |
| "epoch": 2.453488372093023, |
| "grad_norm": 0.3890271782875061, |
| "learning_rate": 3.5500000000000003e-06, |
| "loss": 0.0411, |
| "step": 71 |
| }, |
| { |
| "epoch": 2.488372093023256, |
| "grad_norm": 0.4199780821800232, |
| "learning_rate": 3.6000000000000003e-06, |
| "loss": 0.0404, |
| "step": 72 |
| }, |
| { |
| "epoch": 2.5232558139534884, |
| "grad_norm": 0.5035303831100464, |
| "learning_rate": 3.65e-06, |
| "loss": 0.0417, |
| "step": 73 |
| }, |
| { |
| "epoch": 2.558139534883721, |
| "grad_norm": 0.45307648181915283, |
| "learning_rate": 3.7e-06, |
| "loss": 0.0397, |
| "step": 74 |
| }, |
| { |
| "epoch": 2.5930232558139537, |
| "grad_norm": 0.36111533641815186, |
| "learning_rate": 3.7500000000000005e-06, |
| "loss": 0.0372, |
| "step": 75 |
| }, |
| { |
| "epoch": 2.6279069767441863, |
| "grad_norm": 0.4821806252002716, |
| "learning_rate": 3.8000000000000005e-06, |
| "loss": 0.0375, |
| "step": 76 |
| }, |
| { |
| "epoch": 2.6627906976744184, |
| "grad_norm": 0.4104216992855072, |
| "learning_rate": 3.85e-06, |
| "loss": 0.0363, |
| "step": 77 |
| }, |
| { |
| "epoch": 2.697674418604651, |
| "grad_norm": 0.4832363724708557, |
| "learning_rate": 3.900000000000001e-06, |
| "loss": 0.0406, |
| "step": 78 |
| }, |
| { |
| "epoch": 2.7325581395348837, |
| "grad_norm": 0.2809476852416992, |
| "learning_rate": 3.95e-06, |
| "loss": 0.037, |
| "step": 79 |
| }, |
| { |
| "epoch": 2.7674418604651163, |
| "grad_norm": 0.22463542222976685, |
| "learning_rate": 4.000000000000001e-06, |
| "loss": 0.0359, |
| "step": 80 |
| }, |
| { |
| "epoch": 2.802325581395349, |
| "grad_norm": 0.5055042505264282, |
| "learning_rate": 4.05e-06, |
| "loss": 0.0403, |
| "step": 81 |
| }, |
| { |
| "epoch": 2.8372093023255816, |
| "grad_norm": 0.26646772027015686, |
| "learning_rate": 4.1e-06, |
| "loss": 0.0346, |
| "step": 82 |
| }, |
| { |
| "epoch": 2.8720930232558137, |
| "grad_norm": 0.3047487437725067, |
| "learning_rate": 4.15e-06, |
| "loss": 0.0381, |
| "step": 83 |
| }, |
| { |
| "epoch": 2.9069767441860463, |
| "grad_norm": 0.24949510395526886, |
| "learning_rate": 4.2000000000000004e-06, |
| "loss": 0.0312, |
| "step": 84 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 168, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 6, |
| "save_steps": 28, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 7.14798164830113e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
