Upload folder using huggingface_hub
Browse files- config.json +30 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- trainer_state.json +510 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bert-base-uncased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0"
|
| 13 |
+
},
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"label2id": {
|
| 17 |
+
"LABEL_0": 0
|
| 18 |
+
},
|
| 19 |
+
"layer_norm_eps": 1e-12,
|
| 20 |
+
"max_position_embeddings": 512,
|
| 21 |
+
"model_type": "bert",
|
| 22 |
+
"num_attention_heads": 12,
|
| 23 |
+
"num_hidden_layers": 12,
|
| 24 |
+
"pad_token_id": 0,
|
| 25 |
+
"position_embedding_type": "absolute",
|
| 26 |
+
"transformers_version": "4.6.0.dev0",
|
| 27 |
+
"type_vocab_size": 2,
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 30522
|
| 30 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f33afe3b1d6cda469e758a7a94a438e0c364345deeb750320779499a50ec93f7
|
| 3 |
+
size 875990225
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af05cbe48f08400e7d12faabe4d1a6205d33142c1b598ceee71996fc8acbb27e
|
| 3 |
+
size 438021385
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:665ccd583a8492bb1b433db6ea2c5a85fb0ab94a21940a2445d098659b89e706
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,510 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.06226002424955368,
|
| 3 |
+
"best_model_checkpoint": "training_output/reranker_focal_answer/checkpoint-26182",
|
| 4 |
+
"epoch": 26.0,
|
| 5 |
+
"global_step": 26182,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.5,
|
| 12 |
+
"learning_rate": 4.950347567030785e-05,
|
| 13 |
+
"loss": 0.0273,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.99,
|
| 18 |
+
"learning_rate": 4.900695134061569e-05,
|
| 19 |
+
"loss": 0.013,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.0,
|
| 24 |
+
"eval_loss": 0.06812480837106705,
|
| 25 |
+
"eval_runtime": 56.2831,
|
| 26 |
+
"eval_samples_per_second": 556.508,
|
| 27 |
+
"step": 1007
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 1.49,
|
| 31 |
+
"learning_rate": 4.8510427010923536e-05,
|
| 32 |
+
"loss": 0.0109,
|
| 33 |
+
"step": 1500
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.99,
|
| 37 |
+
"learning_rate": 4.801390268123138e-05,
|
| 38 |
+
"loss": 0.0106,
|
| 39 |
+
"step": 2000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 2.0,
|
| 43 |
+
"eval_loss": 0.07386605441570282,
|
| 44 |
+
"eval_runtime": 56.2854,
|
| 45 |
+
"eval_samples_per_second": 556.486,
|
| 46 |
+
"step": 2014
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 2.48,
|
| 50 |
+
"learning_rate": 4.7517378351539225e-05,
|
| 51 |
+
"loss": 0.0095,
|
| 52 |
+
"step": 2500
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 2.98,
|
| 56 |
+
"learning_rate": 4.702085402184707e-05,
|
| 57 |
+
"loss": 0.0095,
|
| 58 |
+
"step": 3000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 3.0,
|
| 62 |
+
"eval_loss": 0.06689883023500443,
|
| 63 |
+
"eval_runtime": 56.2997,
|
| 64 |
+
"eval_samples_per_second": 556.344,
|
| 65 |
+
"step": 3021
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 3.48,
|
| 69 |
+
"learning_rate": 4.6524329692154915e-05,
|
| 70 |
+
"loss": 0.0083,
|
| 71 |
+
"step": 3500
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 3.97,
|
| 75 |
+
"learning_rate": 4.602780536246276e-05,
|
| 76 |
+
"loss": 0.0085,
|
| 77 |
+
"step": 4000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 4.0,
|
| 81 |
+
"eval_loss": 0.06948165595531464,
|
| 82 |
+
"eval_runtime": 56.3068,
|
| 83 |
+
"eval_samples_per_second": 556.274,
|
| 84 |
+
"step": 4028
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 4.47,
|
| 88 |
+
"learning_rate": 4.5531281032770604e-05,
|
| 89 |
+
"loss": 0.0072,
|
| 90 |
+
"step": 4500
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 4.97,
|
| 94 |
+
"learning_rate": 4.503475670307845e-05,
|
| 95 |
+
"loss": 0.0073,
|
| 96 |
+
"step": 5000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 5.0,
|
| 100 |
+
"eval_loss": 0.06620071828365326,
|
| 101 |
+
"eval_runtime": 56.3226,
|
| 102 |
+
"eval_samples_per_second": 556.118,
|
| 103 |
+
"step": 5035
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 5.46,
|
| 107 |
+
"learning_rate": 4.453823237338629e-05,
|
| 108 |
+
"loss": 0.0062,
|
| 109 |
+
"step": 5500
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 5.96,
|
| 113 |
+
"learning_rate": 4.404170804369414e-05,
|
| 114 |
+
"loss": 0.0064,
|
| 115 |
+
"step": 6000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 6.0,
|
| 119 |
+
"eval_loss": 0.06671101599931717,
|
| 120 |
+
"eval_runtime": 56.334,
|
| 121 |
+
"eval_samples_per_second": 556.005,
|
| 122 |
+
"step": 6042
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 6.45,
|
| 126 |
+
"learning_rate": 4.354518371400199e-05,
|
| 127 |
+
"loss": 0.0054,
|
| 128 |
+
"step": 6500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 6.95,
|
| 132 |
+
"learning_rate": 4.3048659384309834e-05,
|
| 133 |
+
"loss": 0.0056,
|
| 134 |
+
"step": 7000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 7.0,
|
| 138 |
+
"eval_loss": 0.06806950271129608,
|
| 139 |
+
"eval_runtime": 56.3386,
|
| 140 |
+
"eval_samples_per_second": 555.96,
|
| 141 |
+
"step": 7049
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 7.45,
|
| 145 |
+
"learning_rate": 4.255213505461768e-05,
|
| 146 |
+
"loss": 0.0049,
|
| 147 |
+
"step": 7500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 7.94,
|
| 151 |
+
"learning_rate": 4.205561072492552e-05,
|
| 152 |
+
"loss": 0.0049,
|
| 153 |
+
"step": 8000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 8.0,
|
| 157 |
+
"eval_loss": 0.06494650989770889,
|
| 158 |
+
"eval_runtime": 56.3526,
|
| 159 |
+
"eval_samples_per_second": 555.821,
|
| 160 |
+
"step": 8056
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 8.44,
|
| 164 |
+
"learning_rate": 4.155908639523337e-05,
|
| 165 |
+
"loss": 0.0043,
|
| 166 |
+
"step": 8500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 8.94,
|
| 170 |
+
"learning_rate": 4.106256206554122e-05,
|
| 171 |
+
"loss": 0.0044,
|
| 172 |
+
"step": 9000
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 9.0,
|
| 176 |
+
"eval_loss": 0.06944319605827332,
|
| 177 |
+
"eval_runtime": 56.3534,
|
| 178 |
+
"eval_samples_per_second": 555.814,
|
| 179 |
+
"step": 9063
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 9.43,
|
| 183 |
+
"learning_rate": 4.0566037735849064e-05,
|
| 184 |
+
"loss": 0.0038,
|
| 185 |
+
"step": 9500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 9.93,
|
| 189 |
+
"learning_rate": 4.006951340615691e-05,
|
| 190 |
+
"loss": 0.0039,
|
| 191 |
+
"step": 10000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 10.0,
|
| 195 |
+
"eval_loss": 0.06448082625865936,
|
| 196 |
+
"eval_runtime": 56.3612,
|
| 197 |
+
"eval_samples_per_second": 555.737,
|
| 198 |
+
"step": 10070
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 10.43,
|
| 202 |
+
"learning_rate": 3.957298907646475e-05,
|
| 203 |
+
"loss": 0.0035,
|
| 204 |
+
"step": 10500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 10.92,
|
| 208 |
+
"learning_rate": 3.90764647467726e-05,
|
| 209 |
+
"loss": 0.0035,
|
| 210 |
+
"step": 11000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 11.0,
|
| 214 |
+
"eval_loss": 0.06930028647184372,
|
| 215 |
+
"eval_runtime": 56.3501,
|
| 216 |
+
"eval_samples_per_second": 555.846,
|
| 217 |
+
"step": 11077
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 11.42,
|
| 221 |
+
"learning_rate": 3.857994041708044e-05,
|
| 222 |
+
"loss": 0.0031,
|
| 223 |
+
"step": 11500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 11.92,
|
| 227 |
+
"learning_rate": 3.8083416087388287e-05,
|
| 228 |
+
"loss": 0.0032,
|
| 229 |
+
"step": 12000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 12.0,
|
| 233 |
+
"eval_loss": 0.06855376809835434,
|
| 234 |
+
"eval_runtime": 56.3685,
|
| 235 |
+
"eval_samples_per_second": 555.665,
|
| 236 |
+
"step": 12084
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 12.41,
|
| 240 |
+
"learning_rate": 3.758689175769613e-05,
|
| 241 |
+
"loss": 0.0028,
|
| 242 |
+
"step": 12500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 12.91,
|
| 246 |
+
"learning_rate": 3.7090367428003976e-05,
|
| 247 |
+
"loss": 0.003,
|
| 248 |
+
"step": 13000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 13.0,
|
| 252 |
+
"eval_loss": 0.06912727653980255,
|
| 253 |
+
"eval_runtime": 56.3545,
|
| 254 |
+
"eval_samples_per_second": 555.803,
|
| 255 |
+
"step": 13091
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 13.41,
|
| 259 |
+
"learning_rate": 3.659384309831182e-05,
|
| 260 |
+
"loss": 0.0027,
|
| 261 |
+
"step": 13500
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 13.9,
|
| 265 |
+
"learning_rate": 3.6097318768619665e-05,
|
| 266 |
+
"loss": 0.0027,
|
| 267 |
+
"step": 14000
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 14.0,
|
| 271 |
+
"eval_loss": 0.06820587813854218,
|
| 272 |
+
"eval_runtime": 56.3582,
|
| 273 |
+
"eval_samples_per_second": 555.767,
|
| 274 |
+
"step": 14098
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 14.4,
|
| 278 |
+
"learning_rate": 3.560079443892751e-05,
|
| 279 |
+
"loss": 0.0025,
|
| 280 |
+
"step": 14500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 14.9,
|
| 284 |
+
"learning_rate": 3.5104270109235354e-05,
|
| 285 |
+
"loss": 0.0025,
|
| 286 |
+
"step": 15000
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 15.0,
|
| 290 |
+
"eval_loss": 0.06939652562141418,
|
| 291 |
+
"eval_runtime": 56.3589,
|
| 292 |
+
"eval_samples_per_second": 555.76,
|
| 293 |
+
"step": 15105
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 15.39,
|
| 297 |
+
"learning_rate": 3.46077457795432e-05,
|
| 298 |
+
"loss": 0.0023,
|
| 299 |
+
"step": 15500
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 15.89,
|
| 303 |
+
"learning_rate": 3.4111221449851043e-05,
|
| 304 |
+
"loss": 0.0023,
|
| 305 |
+
"step": 16000
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 16.0,
|
| 309 |
+
"eval_loss": 0.06309447437524796,
|
| 310 |
+
"eval_runtime": 56.3564,
|
| 311 |
+
"eval_samples_per_second": 555.784,
|
| 312 |
+
"step": 16112
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 16.39,
|
| 316 |
+
"learning_rate": 3.361469712015889e-05,
|
| 317 |
+
"loss": 0.0022,
|
| 318 |
+
"step": 16500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 16.88,
|
| 322 |
+
"learning_rate": 3.311817279046673e-05,
|
| 323 |
+
"loss": 0.0023,
|
| 324 |
+
"step": 17000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 17.0,
|
| 328 |
+
"eval_loss": 0.06566380709409714,
|
| 329 |
+
"eval_runtime": 56.4037,
|
| 330 |
+
"eval_samples_per_second": 555.318,
|
| 331 |
+
"step": 17119
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 17.38,
|
| 335 |
+
"learning_rate": 3.262164846077458e-05,
|
| 336 |
+
"loss": 0.0021,
|
| 337 |
+
"step": 17500
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 17.87,
|
| 341 |
+
"learning_rate": 3.212512413108242e-05,
|
| 342 |
+
"loss": 0.002,
|
| 343 |
+
"step": 18000
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 18.0,
|
| 347 |
+
"eval_loss": 0.06591986119747162,
|
| 348 |
+
"eval_runtime": 56.3845,
|
| 349 |
+
"eval_samples_per_second": 555.507,
|
| 350 |
+
"step": 18126
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 18.37,
|
| 354 |
+
"learning_rate": 3.1628599801390267e-05,
|
| 355 |
+
"loss": 0.002,
|
| 356 |
+
"step": 18500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 18.87,
|
| 360 |
+
"learning_rate": 3.113207547169811e-05,
|
| 361 |
+
"loss": 0.002,
|
| 362 |
+
"step": 19000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 19.0,
|
| 366 |
+
"eval_loss": 0.0679154321551323,
|
| 367 |
+
"eval_runtime": 56.3981,
|
| 368 |
+
"eval_samples_per_second": 555.373,
|
| 369 |
+
"step": 19133
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 19.36,
|
| 373 |
+
"learning_rate": 3.0635551142005956e-05,
|
| 374 |
+
"loss": 0.0018,
|
| 375 |
+
"step": 19500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 19.86,
|
| 379 |
+
"learning_rate": 3.0139026812313804e-05,
|
| 380 |
+
"loss": 0.0019,
|
| 381 |
+
"step": 20000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 20.0,
|
| 385 |
+
"eval_loss": 0.06612461805343628,
|
| 386 |
+
"eval_runtime": 56.3877,
|
| 387 |
+
"eval_samples_per_second": 555.476,
|
| 388 |
+
"step": 20140
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 20.36,
|
| 392 |
+
"learning_rate": 2.964250248262165e-05,
|
| 393 |
+
"loss": 0.0019,
|
| 394 |
+
"step": 20500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 20.85,
|
| 398 |
+
"learning_rate": 2.9145978152929493e-05,
|
| 399 |
+
"loss": 0.0017,
|
| 400 |
+
"step": 21000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 21.0,
|
| 404 |
+
"eval_loss": 0.06515755504369736,
|
| 405 |
+
"eval_runtime": 56.3934,
|
| 406 |
+
"eval_samples_per_second": 555.42,
|
| 407 |
+
"step": 21147
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 21.35,
|
| 411 |
+
"learning_rate": 2.8649453823237338e-05,
|
| 412 |
+
"loss": 0.0017,
|
| 413 |
+
"step": 21500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 21.85,
|
| 417 |
+
"learning_rate": 2.8152929493545182e-05,
|
| 418 |
+
"loss": 0.0018,
|
| 419 |
+
"step": 22000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 22.0,
|
| 423 |
+
"eval_loss": 0.062369268387556076,
|
| 424 |
+
"eval_runtime": 56.3907,
|
| 425 |
+
"eval_samples_per_second": 555.446,
|
| 426 |
+
"step": 22154
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 22.34,
|
| 430 |
+
"learning_rate": 2.7656405163853027e-05,
|
| 431 |
+
"loss": 0.0017,
|
| 432 |
+
"step": 22500
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 22.84,
|
| 436 |
+
"learning_rate": 2.7159880834160878e-05,
|
| 437 |
+
"loss": 0.0017,
|
| 438 |
+
"step": 23000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 23.0,
|
| 442 |
+
"eval_loss": 0.064911849796772,
|
| 443 |
+
"eval_runtime": 56.3943,
|
| 444 |
+
"eval_samples_per_second": 555.411,
|
| 445 |
+
"step": 23161
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 23.34,
|
| 449 |
+
"learning_rate": 2.6663356504468723e-05,
|
| 450 |
+
"loss": 0.0017,
|
| 451 |
+
"step": 23500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 23.83,
|
| 455 |
+
"learning_rate": 2.6166832174776567e-05,
|
| 456 |
+
"loss": 0.0016,
|
| 457 |
+
"step": 24000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 24.0,
|
| 461 |
+
"eval_loss": 0.06586817651987076,
|
| 462 |
+
"eval_runtime": 56.3912,
|
| 463 |
+
"eval_samples_per_second": 555.441,
|
| 464 |
+
"step": 24168
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 24.33,
|
| 468 |
+
"learning_rate": 2.5670307845084412e-05,
|
| 469 |
+
"loss": 0.0016,
|
| 470 |
+
"step": 24500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 24.83,
|
| 474 |
+
"learning_rate": 2.5173783515392257e-05,
|
| 475 |
+
"loss": 0.0015,
|
| 476 |
+
"step": 25000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 25.0,
|
| 480 |
+
"eval_loss": 0.06666699051856995,
|
| 481 |
+
"eval_runtime": 56.3936,
|
| 482 |
+
"eval_samples_per_second": 555.418,
|
| 483 |
+
"step": 25175
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 25.32,
|
| 487 |
+
"learning_rate": 2.46772591857001e-05,
|
| 488 |
+
"loss": 0.0015,
|
| 489 |
+
"step": 25500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 25.82,
|
| 493 |
+
"learning_rate": 2.4180734856007946e-05,
|
| 494 |
+
"loss": 0.0016,
|
| 495 |
+
"step": 26000
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 26.0,
|
| 499 |
+
"eval_loss": 0.06226002424955368,
|
| 500 |
+
"eval_runtime": 56.4017,
|
| 501 |
+
"eval_samples_per_second": 555.338,
|
| 502 |
+
"step": 26182
|
| 503 |
+
}
|
| 504 |
+
],
|
| 505 |
+
"max_steps": 50350,
|
| 506 |
+
"num_train_epochs": 50,
|
| 507 |
+
"total_flos": 1.4079942379067136e+17,
|
| 508 |
+
"trial_name": null,
|
| 509 |
+
"trial_params": null
|
| 510 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f5f4eaa6716b95dd342e62939c6f1cb03103ed92637fb120d53f34759682a1b
|
| 3 |
+
size 2351
|