Upload training_log.json with huggingface_hub
Browse files- training_log.json +667 -0
training_log.json
ADDED
|
@@ -0,0 +1,667 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 246,
|
| 3 |
+
"best_metric": 0.244215190410614,
|
| 4 |
+
"best_model_checkpoint": "C:\\Users\\Devansh Tyagi\\Desktop\\Projects\\Code-Autopsy\\checkpoints\\checkpoint-246",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 20,
|
| 7 |
+
"global_step": 246,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 1.2720158830285073,
|
| 14 |
+
"epoch": 0.06144393241167435,
|
| 15 |
+
"grad_norm": 0.2734375,
|
| 16 |
+
"learning_rate": 6.153846153846155e-05,
|
| 17 |
+
"loss": 2.1623348236083983,
|
| 18 |
+
"mean_token_accuracy": 0.6029356420040131,
|
| 19 |
+
"num_tokens": 7832.0,
|
| 20 |
+
"step": 5
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 1.2995335638523102,
|
| 24 |
+
"epoch": 0.1228878648233487,
|
| 25 |
+
"grad_norm": 0.435546875,
|
| 26 |
+
"learning_rate": 0.00013846153846153847,
|
| 27 |
+
"loss": 2.1435720443725588,
|
| 28 |
+
"mean_token_accuracy": 0.6072336934506893,
|
| 29 |
+
"num_tokens": 15169.0,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 1.3567075744271277,
|
| 34 |
+
"epoch": 0.18433179723502305,
|
| 35 |
+
"grad_norm": 0.5703125,
|
| 36 |
+
"learning_rate": 0.00019999091026277928,
|
| 37 |
+
"loss": 1.8734064102172852,
|
| 38 |
+
"mean_token_accuracy": 0.6238010875880718,
|
| 39 |
+
"num_tokens": 22884.0,
|
| 40 |
+
"step": 15
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 1.5021331459283829,
|
| 44 |
+
"epoch": 0.2457757296466974,
|
| 45 |
+
"grad_norm": 0.59765625,
|
| 46 |
+
"learning_rate": 0.0001996729429353878,
|
| 47 |
+
"loss": 1.6023683547973633,
|
| 48 |
+
"mean_token_accuracy": 0.6496468104422093,
|
| 49 |
+
"num_tokens": 30047.0,
|
| 50 |
+
"step": 20
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.2457757296466974,
|
| 54 |
+
"eval_entropy": 1.4432531578900063,
|
| 55 |
+
"eval_loss": 1.3970026969909668,
|
| 56 |
+
"eval_mean_token_accuracy": 0.6886225922466958,
|
| 57 |
+
"eval_num_tokens": 30047.0,
|
| 58 |
+
"eval_runtime": 98.8045,
|
| 59 |
+
"eval_samples_per_second": 0.739,
|
| 60 |
+
"eval_steps_per_second": 0.739,
|
| 61 |
+
"step": 20
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"entropy": 1.3366006165742874,
|
| 65 |
+
"epoch": 0.30721966205837176,
|
| 66 |
+
"grad_norm": 0.6171875,
|
| 67 |
+
"learning_rate": 0.00019890213986358148,
|
| 68 |
+
"loss": 1.2767066955566406,
|
| 69 |
+
"mean_token_accuracy": 0.7060731440782547,
|
| 70 |
+
"num_tokens": 37561.0,
|
| 71 |
+
"step": 25
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 0.8943048655986786,
|
| 75 |
+
"epoch": 0.3686635944700461,
|
| 76 |
+
"grad_norm": 0.69921875,
|
| 77 |
+
"learning_rate": 0.00019768200297248193,
|
| 78 |
+
"loss": 0.9884555816650391,
|
| 79 |
+
"mean_token_accuracy": 0.772250434756279,
|
| 80 |
+
"num_tokens": 45476.0,
|
| 81 |
+
"step": 30
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 0.6866158157587051,
|
| 85 |
+
"epoch": 0.43010752688172044,
|
| 86 |
+
"grad_norm": 0.734375,
|
| 87 |
+
"learning_rate": 0.00019601807560796713,
|
| 88 |
+
"loss": 0.7711129665374756,
|
| 89 |
+
"mean_token_accuracy": 0.8339575499296188,
|
| 90 |
+
"num_tokens": 53130.0,
|
| 91 |
+
"step": 35
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 0.6559636496007443,
|
| 95 |
+
"epoch": 0.4915514592933948,
|
| 96 |
+
"grad_norm": 0.86328125,
|
| 97 |
+
"learning_rate": 0.00019391791735205182,
|
| 98 |
+
"loss": 0.6632838726043702,
|
| 99 |
+
"mean_token_accuracy": 0.8613109961152077,
|
| 100 |
+
"num_tokens": 60631.0,
|
| 101 |
+
"step": 40
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4915514592933948,
|
| 105 |
+
"eval_entropy": 0.5521646117510861,
|
| 106 |
+
"eval_loss": 0.630871057510376,
|
| 107 |
+
"eval_mean_token_accuracy": 0.8730851166868863,
|
| 108 |
+
"eval_num_tokens": 60631.0,
|
| 109 |
+
"eval_runtime": 97.1898,
|
| 110 |
+
"eval_samples_per_second": 0.751,
|
| 111 |
+
"eval_steps_per_second": 0.751,
|
| 112 |
+
"step": 40
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"entropy": 0.5553950414061546,
|
| 116 |
+
"epoch": 0.5529953917050692,
|
| 117 |
+
"grad_norm": 0.68359375,
|
| 118 |
+
"learning_rate": 0.00019139106967807062,
|
| 119 |
+
"loss": 0.6109470844268798,
|
| 120 |
+
"mean_token_accuracy": 0.8708647087216377,
|
| 121 |
+
"num_tokens": 68147.0,
|
| 122 |
+
"step": 45
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"entropy": 0.5107979021966458,
|
| 126 |
+
"epoch": 0.6144393241167435,
|
| 127 |
+
"grad_norm": 0.498046875,
|
| 128 |
+
"learning_rate": 0.0001884490126017005,
|
| 129 |
+
"loss": 0.5680758953094482,
|
| 130 |
+
"mean_token_accuracy": 0.8782492533326149,
|
| 131 |
+
"num_tokens": 75622.0,
|
| 132 |
+
"step": 50
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.5033965408802032,
|
| 136 |
+
"epoch": 0.6758832565284179,
|
| 137 |
+
"grad_norm": 0.50390625,
|
| 138 |
+
"learning_rate": 0.00018510511252476587,
|
| 139 |
+
"loss": 0.5488744735717773,
|
| 140 |
+
"mean_token_accuracy": 0.8770601689815521,
|
| 141 |
+
"num_tokens": 83564.0,
|
| 142 |
+
"step": 55
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.4694289192557335,
|
| 146 |
+
"epoch": 0.7373271889400922,
|
| 147 |
+
"grad_norm": 0.4375,
|
| 148 |
+
"learning_rate": 0.00018137456150878303,
|
| 149 |
+
"loss": 0.4926635265350342,
|
| 150 |
+
"mean_token_accuracy": 0.8881596863269806,
|
| 151 |
+
"num_tokens": 91409.0,
|
| 152 |
+
"step": 60
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.7373271889400922,
|
| 156 |
+
"eval_entropy": 0.4363853490515931,
|
| 157 |
+
"eval_loss": 0.46880096197128296,
|
| 158 |
+
"eval_mean_token_accuracy": 0.8960385273580682,
|
| 159 |
+
"eval_num_tokens": 91409.0,
|
| 160 |
+
"eval_runtime": 95.7743,
|
| 161 |
+
"eval_samples_per_second": 0.762,
|
| 162 |
+
"eval_steps_per_second": 0.762,
|
| 163 |
+
"step": 60
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"entropy": 0.43721060529351236,
|
| 167 |
+
"epoch": 0.7987711213517665,
|
| 168 |
+
"grad_norm": 0.60546875,
|
| 169 |
+
"learning_rate": 0.00017727430825413792,
|
| 170 |
+
"loss": 0.4627737522125244,
|
| 171 |
+
"mean_token_accuracy": 0.9004349738359452,
|
| 172 |
+
"num_tokens": 98711.0,
|
| 173 |
+
"step": 65
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"entropy": 0.4222001314163208,
|
| 177 |
+
"epoch": 0.8602150537634409,
|
| 178 |
+
"grad_norm": 0.66015625,
|
| 179 |
+
"learning_rate": 0.00017282298109847345,
|
| 180 |
+
"loss": 0.47431230545043945,
|
| 181 |
+
"mean_token_accuracy": 0.8917310431599617,
|
| 182 |
+
"num_tokens": 105785.0,
|
| 183 |
+
"step": 70
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"entropy": 0.4233152411878109,
|
| 187 |
+
"epoch": 0.9216589861751152,
|
| 188 |
+
"grad_norm": 0.443359375,
|
| 189 |
+
"learning_rate": 0.00016804080338412108,
|
| 190 |
+
"loss": 0.4605244159698486,
|
| 191 |
+
"mean_token_accuracy": 0.902459979057312,
|
| 192 |
+
"num_tokens": 112840.0,
|
| 193 |
+
"step": 75
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 0.4323478534817696,
|
| 197 |
+
"epoch": 0.9831029185867896,
|
| 198 |
+
"grad_norm": 0.4453125,
|
| 199 |
+
"learning_rate": 0.00016294950157908132,
|
| 200 |
+
"loss": 0.44331941604614256,
|
| 201 |
+
"mean_token_accuracy": 0.897197200357914,
|
| 202 |
+
"num_tokens": 120233.0,
|
| 203 |
+
"step": 80
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.9831029185867896,
|
| 207 |
+
"eval_entropy": 0.4033231739312002,
|
| 208 |
+
"eval_loss": 0.39226189255714417,
|
| 209 |
+
"eval_mean_token_accuracy": 0.9040106412482588,
|
| 210 |
+
"eval_num_tokens": 120233.0,
|
| 211 |
+
"eval_runtime": 99.948,
|
| 212 |
+
"eval_samples_per_second": 0.73,
|
| 213 |
+
"eval_steps_per_second": 0.73,
|
| 214 |
+
"step": 80
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"entropy": 0.4201068043708801,
|
| 218 |
+
"epoch": 1.0368663594470047,
|
| 219 |
+
"grad_norm": 0.5,
|
| 220 |
+
"learning_rate": 0.00015757220656897896,
|
| 221 |
+
"loss": 0.41355881690979,
|
| 222 |
+
"mean_token_accuracy": 0.9027445759092059,
|
| 223 |
+
"num_tokens": 126530.0,
|
| 224 |
+
"step": 85
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"entropy": 0.3881114762276411,
|
| 228 |
+
"epoch": 1.098310291858679,
|
| 229 |
+
"grad_norm": 0.4609375,
|
| 230 |
+
"learning_rate": 0.00015193334856844528,
|
| 231 |
+
"loss": 0.3866158723831177,
|
| 232 |
+
"mean_token_accuracy": 0.9028143763542176,
|
| 233 |
+
"num_tokens": 134195.0,
|
| 234 |
+
"step": 90
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"entropy": 0.37071702964603903,
|
| 238 |
+
"epoch": 1.1597542242703534,
|
| 239 |
+
"grad_norm": 0.58203125,
|
| 240 |
+
"learning_rate": 0.00014605854612936728,
|
| 241 |
+
"loss": 0.38562917709350586,
|
| 242 |
+
"mean_token_accuracy": 0.9037379220128059,
|
| 243 |
+
"num_tokens": 142095.0,
|
| 244 |
+
"step": 95
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"entropy": 0.39508153647184374,
|
| 248 |
+
"epoch": 1.2211981566820276,
|
| 249 |
+
"grad_norm": 0.498046875,
|
| 250 |
+
"learning_rate": 0.00013997448975026382,
|
| 251 |
+
"loss": 0.37968151569366454,
|
| 252 |
+
"mean_token_accuracy": 0.902898819744587,
|
| 253 |
+
"num_tokens": 150281.0,
|
| 254 |
+
"step": 100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.2211981566820276,
|
| 258 |
+
"eval_entropy": 0.3728377917041517,
|
| 259 |
+
"eval_loss": 0.32667699456214905,
|
| 260 |
+
"eval_mean_token_accuracy": 0.9168023311928527,
|
| 261 |
+
"eval_num_tokens": 150281.0,
|
| 262 |
+
"eval_runtime": 100.3144,
|
| 263 |
+
"eval_samples_per_second": 0.728,
|
| 264 |
+
"eval_steps_per_second": 0.728,
|
| 265 |
+
"step": 100
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"entropy": 0.3990515094250441,
|
| 269 |
+
"epoch": 1.282642089093702,
|
| 270 |
+
"grad_norm": 0.498046875,
|
| 271 |
+
"learning_rate": 0.00013370882061557635,
|
| 272 |
+
"loss": 0.35355989933013915,
|
| 273 |
+
"mean_token_accuracy": 0.9085300818085671,
|
| 274 |
+
"num_tokens": 158311.0,
|
| 275 |
+
"step": 105
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"entropy": 0.34259250313043593,
|
| 279 |
+
"epoch": 1.3440860215053765,
|
| 280 |
+
"grad_norm": 0.515625,
|
| 281 |
+
"learning_rate": 0.0001272900050157875,
|
| 282 |
+
"loss": 0.3187845706939697,
|
| 283 |
+
"mean_token_accuracy": 0.9141712740063668,
|
| 284 |
+
"num_tokens": 165531.0,
|
| 285 |
+
"step": 110
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"entropy": 0.3513215810060501,
|
| 289 |
+
"epoch": 1.4055299539170507,
|
| 290 |
+
"grad_norm": 0.447265625,
|
| 291 |
+
"learning_rate": 0.00012074720501890484,
|
| 292 |
+
"loss": 0.34308681488037107,
|
| 293 |
+
"mean_token_accuracy": 0.9139169871807098,
|
| 294 |
+
"num_tokens": 172730.0,
|
| 295 |
+
"step": 115
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"entropy": 0.3432418659329414,
|
| 299 |
+
"epoch": 1.466973886328725,
|
| 300 |
+
"grad_norm": 0.4609375,
|
| 301 |
+
"learning_rate": 0.00011411014598087644,
|
| 302 |
+
"loss": 0.3543131113052368,
|
| 303 |
+
"mean_token_accuracy": 0.9060193613171578,
|
| 304 |
+
"num_tokens": 180674.0,
|
| 305 |
+
"step": 120
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.466973886328725,
|
| 309 |
+
"eval_entropy": 0.3156628967964486,
|
| 310 |
+
"eval_loss": 0.28925180435180664,
|
| 311 |
+
"eval_mean_token_accuracy": 0.9195440672848323,
|
| 312 |
+
"eval_num_tokens": 180674.0,
|
| 313 |
+
"eval_runtime": 100.8258,
|
| 314 |
+
"eval_samples_per_second": 0.724,
|
| 315 |
+
"eval_steps_per_second": 0.724,
|
| 316 |
+
"step": 120
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"entropy": 0.328844403848052,
|
| 320 |
+
"epoch": 1.5284178187403994,
|
| 321 |
+
"grad_norm": 0.6484375,
|
| 322 |
+
"learning_rate": 0.0001074089814968676,
|
| 323 |
+
"loss": 0.3134729862213135,
|
| 324 |
+
"mean_token_accuracy": 0.9172038078308106,
|
| 325 |
+
"num_tokens": 187909.0,
|
| 326 |
+
"step": 125
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"entropy": 0.3347533904016018,
|
| 330 |
+
"epoch": 1.5898617511520738,
|
| 331 |
+
"grad_norm": 0.46484375,
|
| 332 |
+
"learning_rate": 0.0001006741564069543,
|
| 333 |
+
"loss": 0.32318768501281736,
|
| 334 |
+
"mean_token_accuracy": 0.915935255587101,
|
| 335 |
+
"num_tokens": 195418.0,
|
| 336 |
+
"step": 130
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"entropy": 0.31085881143808364,
|
| 340 |
+
"epoch": 1.651305683563748,
|
| 341 |
+
"grad_norm": 0.6953125,
|
| 342 |
+
"learning_rate": 9.393626847862763e-05,
|
| 343 |
+
"loss": 0.28849263191223146,
|
| 344 |
+
"mean_token_accuracy": 0.9287568554282188,
|
| 345 |
+
"num_tokens": 202302.0,
|
| 346 |
+
"step": 135
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"entropy": 0.36759571749716996,
|
| 350 |
+
"epoch": 1.7127496159754223,
|
| 351 |
+
"grad_norm": 0.5546875,
|
| 352 |
+
"learning_rate": 8.722592939451625e-05,
|
| 353 |
+
"loss": 0.3258425951004028,
|
| 354 |
+
"mean_token_accuracy": 0.9135418727993965,
|
| 355 |
+
"num_tokens": 210159.0,
|
| 356 |
+
"step": 140
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.7127496159754223,
|
| 360 |
+
"eval_entropy": 0.2999503336948891,
|
| 361 |
+
"eval_loss": 0.2642817199230194,
|
| 362 |
+
"eval_mean_token_accuracy": 0.9280164380596109,
|
| 363 |
+
"eval_num_tokens": 210159.0,
|
| 364 |
+
"eval_runtime": 95.0047,
|
| 365 |
+
"eval_samples_per_second": 0.768,
|
| 366 |
+
"eval_steps_per_second": 0.768,
|
| 367 |
+
"step": 140
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"entropy": 0.2923012483865023,
|
| 371 |
+
"epoch": 1.7741935483870968,
|
| 372 |
+
"grad_norm": 0.470703125,
|
| 373 |
+
"learning_rate": 8.057362567688942e-05,
|
| 374 |
+
"loss": 0.28111426830291747,
|
| 375 |
+
"mean_token_accuracy": 0.9294544145464897,
|
| 376 |
+
"num_tokens": 217668.0,
|
| 377 |
+
"step": 145
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"entropy": 0.3098321039229631,
|
| 381 |
+
"epoch": 1.8356374807987712,
|
| 382 |
+
"grad_norm": 0.400390625,
|
| 383 |
+
"learning_rate": 7.400958018079008e-05,
|
| 384 |
+
"loss": 0.2959801197052002,
|
| 385 |
+
"mean_token_accuracy": 0.9246162533760071,
|
| 386 |
+
"num_tokens": 224914.0,
|
| 387 |
+
"step": 150
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"entropy": 0.30831411490216853,
|
| 391 |
+
"epoch": 1.8970814132104454,
|
| 392 |
+
"grad_norm": 0.435546875,
|
| 393 |
+
"learning_rate": 6.756361478506578e-05,
|
| 394 |
+
"loss": 0.3086764097213745,
|
| 395 |
+
"mean_token_accuracy": 0.9224985137581825,
|
| 396 |
+
"num_tokens": 232318.0,
|
| 397 |
+
"step": 155
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"entropy": 0.29551686625927687,
|
| 401 |
+
"epoch": 1.9585253456221197,
|
| 402 |
+
"grad_norm": 0.58203125,
|
| 403 |
+
"learning_rate": 6.1265014905121e-05,
|
| 404 |
+
"loss": 0.2773712635040283,
|
| 405 |
+
"mean_token_accuracy": 0.9252645134925842,
|
| 406 |
+
"num_tokens": 239672.0,
|
| 407 |
+
"step": 160
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.9585253456221197,
|
| 411 |
+
"eval_entropy": 0.26235738804895586,
|
| 412 |
+
"eval_loss": 0.25179651379585266,
|
| 413 |
+
"eval_mean_token_accuracy": 0.9296977715949489,
|
| 414 |
+
"eval_num_tokens": 239672.0,
|
| 415 |
+
"eval_runtime": 95.912,
|
| 416 |
+
"eval_samples_per_second": 0.761,
|
| 417 |
+
"eval_steps_per_second": 0.761,
|
| 418 |
+
"step": 160
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"entropy": 0.26158358571784834,
|
| 422 |
+
"epoch": 2.0122887864823347,
|
| 423 |
+
"grad_norm": 0.4140625,
|
| 424 |
+
"learning_rate": 5.5142396442938795e-05,
|
| 425 |
+
"loss": 0.24724166393280028,
|
| 426 |
+
"mean_token_accuracy": 0.9323797225952148,
|
| 427 |
+
"num_tokens": 246006.0,
|
| 428 |
+
"step": 165
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"entropy": 0.2724688397720456,
|
| 432 |
+
"epoch": 2.0737327188940093,
|
| 433 |
+
"grad_norm": 0.61328125,
|
| 434 |
+
"learning_rate": 4.9223575778847085e-05,
|
| 435 |
+
"loss": 0.2526975393295288,
|
| 436 |
+
"mean_token_accuracy": 0.9308534324169159,
|
| 437 |
+
"num_tokens": 253437.0,
|
| 438 |
+
"step": 170
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"entropy": 0.2712310256436467,
|
| 442 |
+
"epoch": 2.1351766513056836,
|
| 443 |
+
"grad_norm": 0.439453125,
|
| 444 |
+
"learning_rate": 4.353544339568448e-05,
|
| 445 |
+
"loss": 0.2677891254425049,
|
| 446 |
+
"mean_token_accuracy": 0.9274151250720024,
|
| 447 |
+
"num_tokens": 260911.0,
|
| 448 |
+
"step": 175
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"entropy": 0.2781375008635223,
|
| 452 |
+
"epoch": 2.196620583717358,
|
| 453 |
+
"grad_norm": 0.55078125,
|
| 454 |
+
"learning_rate": 3.8103841709519084e-05,
|
| 455 |
+
"loss": 0.29916017055511473,
|
| 456 |
+
"mean_token_accuracy": 0.9209732711315155,
|
| 457 |
+
"num_tokens": 268425.0,
|
| 458 |
+
"step": 180
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.196620583717358,
|
| 462 |
+
"eval_entropy": 0.2463044097570524,
|
| 463 |
+
"eval_loss": 0.24747803807258606,
|
| 464 |
+
"eval_mean_token_accuracy": 0.930726853135514,
|
| 465 |
+
"eval_num_tokens": 268425.0,
|
| 466 |
+
"eval_runtime": 95.1097,
|
| 467 |
+
"eval_samples_per_second": 0.768,
|
| 468 |
+
"eval_steps_per_second": 0.768,
|
| 469 |
+
"step": 180
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"entropy": 0.2690023283474147,
|
| 473 |
+
"epoch": 2.258064516129032,
|
| 474 |
+
"grad_norm": 0.4765625,
|
| 475 |
+
"learning_rate": 3.29534476619609e-05,
|
| 476 |
+
"loss": 0.27658329010009763,
|
| 477 |
+
"mean_token_accuracy": 0.9289590641856194,
|
| 478 |
+
"num_tokens": 275782.0,
|
| 479 |
+
"step": 185
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"entropy": 0.2697915196418762,
|
| 483 |
+
"epoch": 2.3195084485407067,
|
| 484 |
+
"grad_norm": 0.44921875,
|
| 485 |
+
"learning_rate": 2.8107660607477328e-05,
|
| 486 |
+
"loss": 0.28579936027526853,
|
| 487 |
+
"mean_token_accuracy": 0.9249747917056084,
|
| 488 |
+
"num_tokens": 283440.0,
|
| 489 |
+
"step": 190
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"entropy": 0.28678075782954693,
|
| 493 |
+
"epoch": 2.380952380952381,
|
| 494 |
+
"grad_norm": 0.5078125,
|
| 495 |
+
"learning_rate": 2.3588496005063287e-05,
|
| 496 |
+
"loss": 0.300011134147644,
|
| 497 |
+
"mean_token_accuracy": 0.9201881185173988,
|
| 498 |
+
"num_tokens": 290391.0,
|
| 499 |
+
"step": 195
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"entropy": 0.2683371202088892,
|
| 503 |
+
"epoch": 2.442396313364055,
|
| 504 |
+
"grad_norm": 0.49609375,
|
| 505 |
+
"learning_rate": 1.9416485397247795e-05,
|
| 506 |
+
"loss": 0.26392982006072996,
|
| 507 |
+
"mean_token_accuracy": 0.9297228127717971,
|
| 508 |
+
"num_tokens": 298116.0,
|
| 509 |
+
"step": 200
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.442396313364055,
|
| 513 |
+
"eval_entropy": 0.24947702108997188,
|
| 514 |
+
"eval_loss": 0.24490928649902344,
|
| 515 |
+
"eval_mean_token_accuracy": 0.931492513989749,
|
| 516 |
+
"eval_num_tokens": 298116.0,
|
| 517 |
+
"eval_runtime": 96.5333,
|
| 518 |
+
"eval_samples_per_second": 0.756,
|
| 519 |
+
"eval_steps_per_second": 0.756,
|
| 520 |
+
"step": 200
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"entropy": 0.27692094454541805,
|
| 524 |
+
"epoch": 2.50384024577573,
|
| 525 |
+
"grad_norm": 0.396484375,
|
| 526 |
+
"learning_rate": 1.5610583130854128e-05,
|
| 527 |
+
"loss": 0.2804937601089478,
|
| 528 |
+
"mean_token_accuracy": 0.922454084455967,
|
| 529 |
+
"num_tokens": 305768.0,
|
| 530 |
+
"step": 205
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"entropy": 0.29743164833635094,
|
| 534 |
+
"epoch": 2.565284178187404,
|
| 535 |
+
"grad_norm": 0.396484375,
|
| 536 |
+
"learning_rate": 1.2188080243301437e-05,
|
| 537 |
+
"loss": 0.29574849605560305,
|
| 538 |
+
"mean_token_accuracy": 0.9198502600193024,
|
| 539 |
+
"num_tokens": 313660.0,
|
| 540 |
+
"step": 210
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"entropy": 0.2640182925388217,
|
| 544 |
+
"epoch": 2.6267281105990783,
|
| 545 |
+
"grad_norm": 0.49609375,
|
| 546 |
+
"learning_rate": 9.164525905680709e-06,
|
| 547 |
+
"loss": 0.25743927955627444,
|
| 548 |
+
"mean_token_accuracy": 0.9311310544610023,
|
| 549 |
+
"num_tokens": 320841.0,
|
| 550 |
+
"step": 215
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"entropy": 0.27618018090724944,
|
| 554 |
+
"epoch": 2.688172043010753,
|
| 555 |
+
"grad_norm": 0.51953125,
|
| 556 |
+
"learning_rate": 6.553656779506468e-06,
|
| 557 |
+
"loss": 0.2789269685745239,
|
| 558 |
+
"mean_token_accuracy": 0.9260447949171067,
|
| 559 |
+
"num_tokens": 328470.0,
|
| 560 |
+
"step": 220
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.688172043010753,
|
| 564 |
+
"eval_entropy": 0.24862186620904975,
|
| 565 |
+
"eval_loss": 0.24429786205291748,
|
| 566 |
+
"eval_mean_token_accuracy": 0.9314827543415435,
|
| 567 |
+
"eval_num_tokens": 328470.0,
|
| 568 |
+
"eval_runtime": 93.481,
|
| 569 |
+
"eval_samples_per_second": 0.781,
|
| 570 |
+
"eval_steps_per_second": 0.781,
|
| 571 |
+
"step": 220
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"entropy": 0.2760592779144645,
|
| 575 |
+
"epoch": 2.749615975422427,
|
| 576 |
+
"grad_norm": 0.40625,
|
| 577 |
+
"learning_rate": 4.367334608091389e-06,
|
| 578 |
+
"loss": 0.29435703754425047,
|
| 579 |
+
"mean_token_accuracy": 0.9220241814851761,
|
| 580 |
+
"num_tokens": 336107.0,
|
| 581 |
+
"step": 225
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"entropy": 0.2630997773259878,
|
| 585 |
+
"epoch": 2.8110599078341014,
|
| 586 |
+
"grad_norm": 0.466796875,
|
| 587 |
+
"learning_rate": 2.6154923260801934e-06,
|
| 588 |
+
"loss": 0.2583890914916992,
|
| 589 |
+
"mean_token_accuracy": 0.9310009226202964,
|
| 590 |
+
"num_tokens": 343384.0,
|
| 591 |
+
"step": 230
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"entropy": 0.28990690847858785,
|
| 595 |
+
"epoch": 2.8725038402457757,
|
| 596 |
+
"grad_norm": 0.390625,
|
| 597 |
+
"learning_rate": 1.3060889319784885e-06,
|
| 598 |
+
"loss": 0.2960776090621948,
|
| 599 |
+
"mean_token_accuracy": 0.9223004102706909,
|
| 600 |
+
"num_tokens": 351180.0,
|
| 601 |
+
"step": 235
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"entropy": 0.2609599939547479,
|
| 605 |
+
"epoch": 2.93394777265745,
|
| 606 |
+
"grad_norm": 0.388671875,
|
| 607 |
+
"learning_rate": 4.4507332870059594e-07,
|
| 608 |
+
"loss": 0.25511951446533204,
|
| 609 |
+
"mean_token_accuracy": 0.9273923814296723,
|
| 610 |
+
"num_tokens": 359106.0,
|
| 611 |
+
"step": 240
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.93394777265745,
|
| 615 |
+
"eval_entropy": 0.24859930197261784,
|
| 616 |
+
"eval_loss": 0.24424654245376587,
|
| 617 |
+
"eval_mean_token_accuracy": 0.9321725441984934,
|
| 618 |
+
"eval_num_tokens": 359106.0,
|
| 619 |
+
"eval_runtime": 96.2762,
|
| 620 |
+
"eval_samples_per_second": 0.758,
|
| 621 |
+
"eval_steps_per_second": 0.758,
|
| 622 |
+
"step": 240
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"entropy": 0.2684542080387473,
|
| 626 |
+
"epoch": 2.9953917050691246,
|
| 627 |
+
"grad_norm": 0.390625,
|
| 628 |
+
"learning_rate": 3.635729641654484e-08,
|
| 629 |
+
"loss": 0.2732724666595459,
|
| 630 |
+
"mean_token_accuracy": 0.9262633189558983,
|
| 631 |
+
"num_tokens": 366418.0,
|
| 632 |
+
"step": 245
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.0,
|
| 636 |
+
"eval_entropy": 0.24880487002330284,
|
| 637 |
+
"eval_loss": 0.244215190410614,
|
| 638 |
+
"eval_mean_token_accuracy": 0.9320176945973749,
|
| 639 |
+
"eval_num_tokens": 366954.0,
|
| 640 |
+
"eval_runtime": 94.6685,
|
| 641 |
+
"eval_samples_per_second": 0.771,
|
| 642 |
+
"eval_steps_per_second": 0.771,
|
| 643 |
+
"step": 246
|
| 644 |
+
}
|
| 645 |
+
],
|
| 646 |
+
"logging_steps": 5,
|
| 647 |
+
"max_steps": 246,
|
| 648 |
+
"num_input_tokens_seen": 0,
|
| 649 |
+
"num_train_epochs": 3,
|
| 650 |
+
"save_steps": 20,
|
| 651 |
+
"stateful_callbacks": {
|
| 652 |
+
"TrainerControl": {
|
| 653 |
+
"args": {
|
| 654 |
+
"should_epoch_stop": false,
|
| 655 |
+
"should_evaluate": false,
|
| 656 |
+
"should_log": false,
|
| 657 |
+
"should_save": true,
|
| 658 |
+
"should_training_stop": true
|
| 659 |
+
},
|
| 660 |
+
"attributes": {}
|
| 661 |
+
}
|
| 662 |
+
},
|
| 663 |
+
"total_flos": 1.5578662344763392e+16,
|
| 664 |
+
"train_batch_size": 1,
|
| 665 |
+
"trial_name": null,
|
| 666 |
+
"trial_params": null
|
| 667 |
+
}
|