Upload folder using huggingface_hub
Browse files- config.json +26 -0
- generation_config.json +6 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +316 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/storage2/fs1/btc/Active/yeli/xiaoxiao.zhou/tokenize/tokenizers/DNAbert2_theirBPE/pretrain/models/model_2/checkpoint-150000",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 3072,
|
| 13 |
+
"layer_norm_eps": 1e-12,
|
| 14 |
+
"max_length": 512,
|
| 15 |
+
"max_position_embeddings": 512,
|
| 16 |
+
"model_type": "bert",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 12,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.29.2",
|
| 23 |
+
"type_vocab_size": 2,
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 4096
|
| 26 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"max_length": 512,
|
| 4 |
+
"pad_token_id": 0,
|
| 5 |
+
"transformers_version": "4.29.2"
|
| 6 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d18ceeab6f3859f90a0b69e4d319a2f4a11186f1f83a81893bef54a51fea39a
|
| 3 |
+
size 713660538
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f955d3c0433c470a0bc280e8b61264489e5209007c9d6dd388b1ebe23f70f4b
|
| 3 |
+
size 356844726
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6d6c569bea0b6e042e097ec2a4b6cfec780f8e0a2a2181b3d4babd06fb07ebf
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d7ab81f6f167a77d305fd2072c3f0653a829da163e38e23511eabde99bbdc11
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.824392126592049,
|
| 5 |
+
"global_step": 50000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.1,
|
| 12 |
+
"learning_rate": 7.142857142857143e-06,
|
| 13 |
+
"loss": 5.3189,
|
| 14 |
+
"step": 1000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.19,
|
| 18 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 19 |
+
"loss": 5.3266,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.29,
|
| 24 |
+
"learning_rate": 2.1428571428571428e-05,
|
| 25 |
+
"loss": 5.3215,
|
| 26 |
+
"step": 3000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.39,
|
| 30 |
+
"learning_rate": 2.857142857142857e-05,
|
| 31 |
+
"loss": 5.3156,
|
| 32 |
+
"step": 4000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.48,
|
| 36 |
+
"learning_rate": 3.571428571428572e-05,
|
| 37 |
+
"loss": 5.3094,
|
| 38 |
+
"step": 5000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.58,
|
| 42 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 43 |
+
"loss": 5.294,
|
| 44 |
+
"step": 6000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.68,
|
| 48 |
+
"learning_rate": 5e-05,
|
| 49 |
+
"loss": 5.2852,
|
| 50 |
+
"step": 7000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.77,
|
| 54 |
+
"learning_rate": 4.883720930232558e-05,
|
| 55 |
+
"loss": 5.2715,
|
| 56 |
+
"step": 8000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.87,
|
| 60 |
+
"learning_rate": 4.7674418604651164e-05,
|
| 61 |
+
"loss": 5.2477,
|
| 62 |
+
"step": 9000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.96,
|
| 66 |
+
"learning_rate": 4.651162790697675e-05,
|
| 67 |
+
"loss": 5.2361,
|
| 68 |
+
"step": 10000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.06,
|
| 72 |
+
"learning_rate": 4.5348837209302326e-05,
|
| 73 |
+
"loss": 5.2234,
|
| 74 |
+
"step": 11000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.16,
|
| 78 |
+
"learning_rate": 4.418604651162791e-05,
|
| 79 |
+
"loss": 5.2037,
|
| 80 |
+
"step": 12000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.25,
|
| 84 |
+
"learning_rate": 4.302325581395349e-05,
|
| 85 |
+
"loss": 5.1853,
|
| 86 |
+
"step": 13000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.35,
|
| 90 |
+
"learning_rate": 4.186046511627907e-05,
|
| 91 |
+
"loss": 5.1751,
|
| 92 |
+
"step": 14000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.45,
|
| 96 |
+
"learning_rate": 4.0697674418604655e-05,
|
| 97 |
+
"loss": 5.1663,
|
| 98 |
+
"step": 15000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.54,
|
| 102 |
+
"learning_rate": 3.953488372093023e-05,
|
| 103 |
+
"loss": 5.1615,
|
| 104 |
+
"step": 16000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.64,
|
| 108 |
+
"learning_rate": 3.837209302325582e-05,
|
| 109 |
+
"loss": 5.1547,
|
| 110 |
+
"step": 17000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.74,
|
| 114 |
+
"learning_rate": 3.7209302325581394e-05,
|
| 115 |
+
"loss": 5.1431,
|
| 116 |
+
"step": 18000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.83,
|
| 120 |
+
"learning_rate": 3.604651162790698e-05,
|
| 121 |
+
"loss": 5.1293,
|
| 122 |
+
"step": 19000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.93,
|
| 126 |
+
"learning_rate": 3.488372093023256e-05,
|
| 127 |
+
"loss": 5.1278,
|
| 128 |
+
"step": 20000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.03,
|
| 132 |
+
"learning_rate": 3.372093023255814e-05,
|
| 133 |
+
"loss": 5.1174,
|
| 134 |
+
"step": 21000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 2.12,
|
| 138 |
+
"learning_rate": 3.2558139534883724e-05,
|
| 139 |
+
"loss": 5.1033,
|
| 140 |
+
"step": 22000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.22,
|
| 144 |
+
"learning_rate": 3.13953488372093e-05,
|
| 145 |
+
"loss": 5.1019,
|
| 146 |
+
"step": 23000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.32,
|
| 150 |
+
"learning_rate": 3.0232558139534883e-05,
|
| 151 |
+
"loss": 5.096,
|
| 152 |
+
"step": 24000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.41,
|
| 156 |
+
"learning_rate": 2.9069767441860467e-05,
|
| 157 |
+
"loss": 5.0878,
|
| 158 |
+
"step": 25000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.51,
|
| 162 |
+
"learning_rate": 2.7906976744186048e-05,
|
| 163 |
+
"loss": 5.0847,
|
| 164 |
+
"step": 26000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.61,
|
| 168 |
+
"learning_rate": 2.674418604651163e-05,
|
| 169 |
+
"loss": 5.0731,
|
| 170 |
+
"step": 27000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.7,
|
| 174 |
+
"learning_rate": 2.5581395348837212e-05,
|
| 175 |
+
"loss": 5.0672,
|
| 176 |
+
"step": 28000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.8,
|
| 180 |
+
"learning_rate": 2.441860465116279e-05,
|
| 181 |
+
"loss": 5.0717,
|
| 182 |
+
"step": 29000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 2.89,
|
| 186 |
+
"learning_rate": 2.3255813953488374e-05,
|
| 187 |
+
"loss": 5.0631,
|
| 188 |
+
"step": 30000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.99,
|
| 192 |
+
"learning_rate": 2.2093023255813955e-05,
|
| 193 |
+
"loss": 5.0626,
|
| 194 |
+
"step": 31000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.09,
|
| 198 |
+
"learning_rate": 2.0930232558139536e-05,
|
| 199 |
+
"loss": 5.06,
|
| 200 |
+
"step": 32000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.18,
|
| 204 |
+
"learning_rate": 1.9767441860465116e-05,
|
| 205 |
+
"loss": 5.0461,
|
| 206 |
+
"step": 33000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.28,
|
| 210 |
+
"learning_rate": 1.8604651162790697e-05,
|
| 211 |
+
"loss": 5.039,
|
| 212 |
+
"step": 34000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.38,
|
| 216 |
+
"learning_rate": 1.744186046511628e-05,
|
| 217 |
+
"loss": 5.0416,
|
| 218 |
+
"step": 35000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.47,
|
| 222 |
+
"learning_rate": 1.6279069767441862e-05,
|
| 223 |
+
"loss": 5.0437,
|
| 224 |
+
"step": 36000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.57,
|
| 228 |
+
"learning_rate": 1.5116279069767441e-05,
|
| 229 |
+
"loss": 5.03,
|
| 230 |
+
"step": 37000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 3.67,
|
| 234 |
+
"learning_rate": 1.3953488372093024e-05,
|
| 235 |
+
"loss": 5.027,
|
| 236 |
+
"step": 38000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 3.76,
|
| 240 |
+
"learning_rate": 1.2790697674418606e-05,
|
| 241 |
+
"loss": 5.0289,
|
| 242 |
+
"step": 39000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 3.86,
|
| 246 |
+
"learning_rate": 1.1627906976744187e-05,
|
| 247 |
+
"loss": 5.0246,
|
| 248 |
+
"step": 40000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 3.96,
|
| 252 |
+
"learning_rate": 1.0465116279069768e-05,
|
| 253 |
+
"loss": 5.0209,
|
| 254 |
+
"step": 41000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.05,
|
| 258 |
+
"learning_rate": 9.302325581395349e-06,
|
| 259 |
+
"loss": 5.0167,
|
| 260 |
+
"step": 42000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 4.15,
|
| 264 |
+
"learning_rate": 8.139534883720931e-06,
|
| 265 |
+
"loss": 5.0108,
|
| 266 |
+
"step": 43000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 4.25,
|
| 270 |
+
"learning_rate": 6.976744186046512e-06,
|
| 271 |
+
"loss": 5.0121,
|
| 272 |
+
"step": 44000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 4.34,
|
| 276 |
+
"learning_rate": 5.8139534883720935e-06,
|
| 277 |
+
"loss": 5.0122,
|
| 278 |
+
"step": 45000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.44,
|
| 282 |
+
"learning_rate": 4.651162790697674e-06,
|
| 283 |
+
"loss": 5.0222,
|
| 284 |
+
"step": 46000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 4.53,
|
| 288 |
+
"learning_rate": 3.488372093023256e-06,
|
| 289 |
+
"loss": 5.0162,
|
| 290 |
+
"step": 47000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.63,
|
| 294 |
+
"learning_rate": 2.325581395348837e-06,
|
| 295 |
+
"loss": 5.0045,
|
| 296 |
+
"step": 48000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.73,
|
| 300 |
+
"learning_rate": 1.1627906976744186e-06,
|
| 301 |
+
"loss": 5.0047,
|
| 302 |
+
"step": 49000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 4.82,
|
| 306 |
+
"learning_rate": 0.0,
|
| 307 |
+
"loss": 5.005,
|
| 308 |
+
"step": 50000
|
| 309 |
+
}
|
| 310 |
+
],
|
| 311 |
+
"max_steps": 50000,
|
| 312 |
+
"num_train_epochs": 5,
|
| 313 |
+
"total_flos": 1.2629229170809897e+18,
|
| 314 |
+
"trial_name": null,
|
| 315 |
+
"trial_params": null
|
| 316 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7a0b2d34455400cc6664215841506159df7ebf59680d22509badf6d4fae42c2
|
| 3 |
+
size 4600
|