| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9956709956709957, | |
| "eval_steps": 500, | |
| "global_step": 115, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1e-05, | |
| "loss": 2.0642, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0704, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0758, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4e-05, | |
| "loss": 2.0031, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-05, | |
| "loss": 2.0466, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 6e-05, | |
| "loss": 2.027, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 7e-05, | |
| "loss": 2.0737, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 8e-05, | |
| "loss": 2.1039, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9e-05, | |
| "loss": 2.0212, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0001, | |
| "loss": 2.047, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.999780139628657e-05, | |
| "loss": 2.0435, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.99912057785006e-05, | |
| "loss": 2.0305, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.998021372668808e-05, | |
| "loss": 1.9602, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.996482620753565e-05, | |
| "loss": 2.0138, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.994504457428558e-05, | |
| "loss": 2.065, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.992087056661677e-05, | |
| "loss": 2.0051, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 9.989230631049171e-05, | |
| "loss": 2.008, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.985935431796962e-05, | |
| "loss": 1.9578, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.982201748698542e-05, | |
| "loss": 1.9591, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.978029910109491e-05, | |
| "loss": 1.9495, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.973420282918601e-05, | |
| "loss": 1.9121, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.968373272515612e-05, | |
| "loss": 1.9639, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.962889322755555e-05, | |
| "loss": 1.9431, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 9.956968915919725e-05, | |
| "loss": 1.9283, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.950612572673255e-05, | |
| "loss": 1.9112, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.943820852019344e-05, | |
| "loss": 1.9942, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.936594351250082e-05, | |
| "loss": 1.9536, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.928933705893924e-05, | |
| "loss": 1.9364, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.920839589659803e-05, | |
| "loss": 1.9485, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.91231271437788e-05, | |
| "loss": 1.9743, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.903353829936943e-05, | |
| "loss": 1.9492, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.893963724218455e-05, | |
| "loss": 1.9612, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.884143223027266e-05, | |
| "loss": 1.8852, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.873893190018995e-05, | |
| "loss": 1.9104, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 9.863214526624065e-05, | |
| "loss": 1.9762, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 9.852108171968436e-05, | |
| "loss": 1.9602, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 9.840575102791013e-05, | |
| "loss": 1.9283, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.828616333357743e-05, | |
| "loss": 1.8863, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 9.816232915372423e-05, | |
| "loss": 1.9776, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.8034259378842e-05, | |
| "loss": 1.8874, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.790196527191811e-05, | |
| "loss": 1.8779, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 9.776545846744509e-05, | |
| "loss": 1.9399, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 9.762475097039767e-05, | |
| "loss": 1.8758, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.747985515517683e-05, | |
| "loss": 1.903, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.733078376452171e-05, | |
| "loss": 1.9438, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 9.717754990838881e-05, | |
| "loss": 1.885, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.702016706279913e-05, | |
| "loss": 1.9278, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 9.685864906865303e-05, | |
| "loss": 1.9612, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 9.669301013051297e-05, | |
| "loss": 1.9352, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 9.652326481535435e-05, | |
| "loss": 1.9198, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 9.634942805128433e-05, | |
| "loss": 1.9552, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 9.617151512622917e-05, | |
| "loss": 1.9767, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 9.598954168658955e-05, | |
| "loss": 1.9933, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.580352373586467e-05, | |
| "loss": 1.9751, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.561347763324484e-05, | |
| "loss": 1.8376, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.541942009217273e-05, | |
| "loss": 1.9485, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 9.522136817887353e-05, | |
| "loss": 1.885, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.501933931085416e-05, | |
| "loss": 1.9241, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 9.481335125537138e-05, | |
| "loss": 1.8931, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 9.460342212786932e-05, | |
| "loss": 1.9544, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 9.43895703903864e-05, | |
| "loss": 1.8637, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 9.417181484993154e-05, | |
| "loss": 1.964, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.395017465683036e-05, | |
| "loss": 1.8605, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.372466930304091e-05, | |
| "loss": 1.9388, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 9.349531862043952e-05, | |
| "loss": 1.9195, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.32621427790767e-05, | |
| "loss": 1.9328, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 9.302516228540327e-05, | |
| "loss": 1.8901, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 9.278439798046697e-05, | |
| "loss": 1.9296, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 9.253987103807958e-05, | |
| "loss": 1.9282, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 9.229160296295488e-05, | |
| "loss": 1.9448, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 9.203961558881731e-05, | |
| "loss": 1.9211, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 9.178393107648193e-05, | |
| "loss": 1.8658, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 9.15245719119055e-05, | |
| "loss": 1.9358, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 9.126156090420888e-05, | |
| "loss": 1.9446, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 9.099492118367123e-05, | |
| "loss": 1.8403, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 9.072467619969572e-05, | |
| "loss": 1.9857, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.878, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 9.017346582226289e-05, | |
| "loss": 1.9484, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 8.98925489045329e-05, | |
| "loss": 1.8543, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 8.960812367055646e-05, | |
| "loss": 1.8613, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.93202151338687e-05, | |
| "loss": 1.887, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 8.902884861434065e-05, | |
| "loss": 1.8897, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 8.873404973595285e-05, | |
| "loss": 1.93, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.843584442454158e-05, | |
| "loss": 1.951, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.81342589055191e-05, | |
| "loss": 1.9591, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.782931970156707e-05, | |
| "loss": 1.8649, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.752105363030414e-05, | |
| "loss": 1.9752, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.720948780192746e-05, | |
| "loss": 1.9189, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 8.689464961682852e-05, | |
| "loss": 1.9188, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 8.657656676318346e-05, | |
| "loss": 1.9376, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 8.625526721451798e-05, | |
| "loss": 1.8662, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 8.593077922724733e-05, | |
| "loss": 1.8113, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.560313133819125e-05, | |
| "loss": 1.8575, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.527235236206436e-05, | |
| "loss": 1.8991, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.493847138894209e-05, | |
| "loss": 1.8475, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.46015177817023e-05, | |
| "loss": 1.8795, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.426152117344313e-05, | |
| "loss": 1.8596, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.391851146487675e-05, | |
| "loss": 1.9481, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.357251882169994e-05, | |
| "loss": 1.9213, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.322357367194109e-05, | |
| "loss": 1.9243, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.28717067032843e-05, | |
| "loss": 1.8764, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 8.251694886037052e-05, | |
| "loss": 1.9728, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 8.215933134207618e-05, | |
| "loss": 1.884, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 8.179888559876943e-05, | |
| "loss": 1.9589, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.143564332954425e-05, | |
| "loss": 2.017, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 8.106963647943274e-05, | |
| "loss": 1.9672, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.070089723659566e-05, | |
| "loss": 1.8711, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 8.032945802949179e-05, | |
| "loss": 1.9211, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 7.995535152402591e-05, | |
| "loss": 1.9019, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 7.957861062067614e-05, | |
| "loss": 1.8887, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 7.919926845160037e-05, | |
| "loss": 1.9923, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 7.881735837772274e-05, | |
| "loss": 1.8621, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 7.843291398579946e-05, | |
| "loss": 1.9584, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 7.804596908546529e-05, | |
| "loss": 1.943, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.765655770625997e-05, | |
| "loss": 1.8953, | |
| "step": 115 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 345, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 115, | |
| "total_flos": 2.6099686427118797e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |