Spaces:
Runtime error
Runtime error
Commit
•
89b5cec
1
Parent(s):
ba07099
use gpt2-medium
Browse files- app.py +1 -1
- model/gpt2-355M/config.json +38 -0
- model/gpt2-355M/pytorch_model.bin +3 -0
- model/gpt2-355M/trainer_state.json +256 -0
app.py
CHANGED
|
@@ -5,7 +5,7 @@ import torch
|
|
| 5 |
device = 'cuda' if torch.cuda.is_available() else 'cpu'
|
| 6 |
|
| 7 |
# load pretrained + finetuned GPT2
|
| 8 |
-
model = GPT2LMHeadModel.from_pretrained("./model/gpt2-
|
| 9 |
model = model.to(device)
|
| 10 |
|
| 11 |
# create tokenizer
|
|
|
|
| 5 |
device = 'cuda' if torch.cuda.is_available() else 'cpu'
|
| 6 |
|
| 7 |
# load pretrained + finetuned GPT2
|
| 8 |
+
model = GPT2LMHeadModel.from_pretrained("./model/gpt2-355M")
|
| 9 |
model = model.to(device)
|
| 10 |
|
| 11 |
# create tokenizer
|
model/gpt2-355M/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "gpt2-medium",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"gradient_checkpointing": false,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "gpt2",
|
| 15 |
+
"n_ctx": 1024,
|
| 16 |
+
"n_embd": 1024,
|
| 17 |
+
"n_head": 16,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 24,
|
| 20 |
+
"n_positions": 1024,
|
| 21 |
+
"n_special": 0,
|
| 22 |
+
"predict_special_tokens": true,
|
| 23 |
+
"resid_pdrop": 0.1,
|
| 24 |
+
"summary_activation": null,
|
| 25 |
+
"summary_first_dropout": 0.1,
|
| 26 |
+
"summary_proj_to_labels": true,
|
| 27 |
+
"summary_type": "cls_index",
|
| 28 |
+
"summary_use_proj": true,
|
| 29 |
+
"task_specific_params": {
|
| 30 |
+
"text-generation": {
|
| 31 |
+
"do_sample": true,
|
| 32 |
+
"max_length": 50
|
| 33 |
+
}
|
| 34 |
+
},
|
| 35 |
+
"transformers_version": "4.2.2",
|
| 36 |
+
"use_cache": true,
|
| 37 |
+
"vocab_size": 50257
|
| 38 |
+
}
|
model/gpt2-355M/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b0e684952adc84a05d296c526ead3c19827b804bb60ebaacc8857550efc3b9a
|
| 3 |
+
size 1444583837
|
model/gpt2-355M/trainer_state.json
ADDED
|
@@ -0,0 +1,256 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 14.662756598240469,
|
| 5 |
+
"global_step": 20000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.37,
|
| 12 |
+
"learning_rate": 5e-05,
|
| 13 |
+
"loss": 3.6227,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.73,
|
| 18 |
+
"learning_rate": 4.8747494989979966e-05,
|
| 19 |
+
"loss": 3.3016,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.1,
|
| 24 |
+
"learning_rate": 4.7494989979959916e-05,
|
| 25 |
+
"loss": 3.1452,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.47,
|
| 30 |
+
"learning_rate": 4.624248496993988e-05,
|
| 31 |
+
"loss": 2.932,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.83,
|
| 36 |
+
"learning_rate": 4.4989979959919844e-05,
|
| 37 |
+
"loss": 2.925,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.2,
|
| 42 |
+
"learning_rate": 4.37374749498998e-05,
|
| 43 |
+
"loss": 2.7771,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.57,
|
| 48 |
+
"learning_rate": 4.248496993987976e-05,
|
| 49 |
+
"loss": 2.6615,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 2.93,
|
| 54 |
+
"learning_rate": 4.123246492985972e-05,
|
| 55 |
+
"loss": 2.6681,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 3.3,
|
| 60 |
+
"learning_rate": 3.997995991983968e-05,
|
| 61 |
+
"loss": 2.4825,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 3.67,
|
| 66 |
+
"learning_rate": 3.872745490981964e-05,
|
| 67 |
+
"loss": 2.4556,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.03,
|
| 72 |
+
"learning_rate": 3.74749498997996e-05,
|
| 73 |
+
"loss": 2.4436,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 4.4,
|
| 78 |
+
"learning_rate": 3.6222444889779564e-05,
|
| 79 |
+
"loss": 2.2667,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 4.77,
|
| 84 |
+
"learning_rate": 3.496993987975952e-05,
|
| 85 |
+
"loss": 2.2716,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 5.13,
|
| 90 |
+
"learning_rate": 3.371743486973948e-05,
|
| 91 |
+
"loss": 2.2282,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 5.5,
|
| 96 |
+
"learning_rate": 3.246492985971944e-05,
|
| 97 |
+
"loss": 2.1086,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 5.87,
|
| 102 |
+
"learning_rate": 3.12124248496994e-05,
|
| 103 |
+
"loss": 2.1339,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 6.23,
|
| 108 |
+
"learning_rate": 2.9959919839679363e-05,
|
| 109 |
+
"loss": 2.0177,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 6.6,
|
| 114 |
+
"learning_rate": 2.870741482965932e-05,
|
| 115 |
+
"loss": 1.9853,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 6.96,
|
| 120 |
+
"learning_rate": 2.745490981963928e-05,
|
| 121 |
+
"loss": 1.9993,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 7.33,
|
| 126 |
+
"learning_rate": 2.620240480961924e-05,
|
| 127 |
+
"loss": 1.859,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 7.7,
|
| 132 |
+
"learning_rate": 2.49498997995992e-05,
|
| 133 |
+
"loss": 1.8696,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 8.06,
|
| 138 |
+
"learning_rate": 2.3697394789579158e-05,
|
| 139 |
+
"loss": 1.858,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 8.43,
|
| 144 |
+
"learning_rate": 2.244488977955912e-05,
|
| 145 |
+
"loss": 1.7593,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 8.8,
|
| 150 |
+
"learning_rate": 2.119238476953908e-05,
|
| 151 |
+
"loss": 1.7654,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.16,
|
| 156 |
+
"learning_rate": 1.993987975951904e-05,
|
| 157 |
+
"loss": 1.7326,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 9.53,
|
| 162 |
+
"learning_rate": 1.8687374749499e-05,
|
| 163 |
+
"loss": 1.6704,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 9.9,
|
| 168 |
+
"learning_rate": 1.743486973947896e-05,
|
| 169 |
+
"loss": 1.689,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 10.26,
|
| 174 |
+
"learning_rate": 1.6182364729458917e-05,
|
| 175 |
+
"loss": 1.6195,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 10.63,
|
| 180 |
+
"learning_rate": 1.492985971943888e-05,
|
| 181 |
+
"loss": 1.6111,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 11.0,
|
| 186 |
+
"learning_rate": 1.3677354709418838e-05,
|
| 187 |
+
"loss": 1.6198,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 11.36,
|
| 192 |
+
"learning_rate": 1.2424849699398798e-05,
|
| 193 |
+
"loss": 1.5393,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 11.73,
|
| 198 |
+
"learning_rate": 1.1172344689378759e-05,
|
| 199 |
+
"loss": 1.5502,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 12.1,
|
| 204 |
+
"learning_rate": 9.919839679358718e-06,
|
| 205 |
+
"loss": 1.5391,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 12.46,
|
| 210 |
+
"learning_rate": 8.667334669338678e-06,
|
| 211 |
+
"loss": 1.4915,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 12.83,
|
| 216 |
+
"learning_rate": 7.414829659318638e-06,
|
| 217 |
+
"loss": 1.507,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 13.2,
|
| 222 |
+
"learning_rate": 6.162324649298597e-06,
|
| 223 |
+
"loss": 1.4794,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 13.56,
|
| 228 |
+
"learning_rate": 4.9098196392785576e-06,
|
| 229 |
+
"loss": 1.4626,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 13.93,
|
| 234 |
+
"learning_rate": 3.6573146292585176e-06,
|
| 235 |
+
"loss": 1.464,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 14.3,
|
| 240 |
+
"learning_rate": 2.404809619238477e-06,
|
| 241 |
+
"loss": 1.4426,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 14.66,
|
| 246 |
+
"learning_rate": 1.152304609218437e-06,
|
| 247 |
+
"loss": 1.4393,
|
| 248 |
+
"step": 20000
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"max_steps": 20460,
|
| 252 |
+
"num_train_epochs": 15,
|
| 253 |
+
"total_flos": 43573965472923648,
|
| 254 |
+
"trial_name": null,
|
| 255 |
+
"trial_params": null
|
| 256 |
+
}
|