Upload folder using huggingface_hub
Browse files
checkpoints/checkpoint-306/adapter.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:705b77f4f5fc3f2207f1ea7c7637f4091f6443783fc9b894788e32de66bc6a04
|
| 3 |
+
size 6439640
|
checkpoints/checkpoint-306/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ddd315c5e0840b81e818dd5ff765d8de5e5d04b7b6a9cdf05799e5bfbd7c0ce1
|
| 3 |
+
size 3304962
|
checkpoints/checkpoint-306/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:402fa8046bb84d6ccced5bc526160e4d722440748b38fcb0b6cdf48297e39aa6
|
| 3 |
+
size 14244
|
checkpoints/checkpoint-306/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cde0307485e9704120944cc03273844cb67ca42817c760970d1e4ff5341cf1d0
|
| 3 |
+
size 1064
|
checkpoints/checkpoint-306/trainer_state.json
ADDED
|
@@ -0,0 +1,280 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 306,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.3137254901960784,
|
| 13 |
+
"grad_norm": 21.246755599975586,
|
| 14 |
+
"learning_rate": 2.2580645161290322e-07,
|
| 15 |
+
"loss": 10.64,
|
| 16 |
+
"step": 16
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.6274509803921569,
|
| 20 |
+
"grad_norm": 22.127361297607422,
|
| 21 |
+
"learning_rate": 4.677419354838709e-07,
|
| 22 |
+
"loss": 10.6241,
|
| 23 |
+
"step": 32
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.9411764705882353,
|
| 27 |
+
"grad_norm": 20.265277862548828,
|
| 28 |
+
"learning_rate": 7.258064516129032e-07,
|
| 29 |
+
"loss": 10.5868,
|
| 30 |
+
"step": 48
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.0,
|
| 34 |
+
"eval_bleu": 0.011208247430062485,
|
| 35 |
+
"eval_cap_loss": 4.461696884211372,
|
| 36 |
+
"eval_con_loss": 2.0591132453843657,
|
| 37 |
+
"eval_loss": 6.349074924693388,
|
| 38 |
+
"step": 51
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.0,
|
| 42 |
+
"eval_bleu": 0.011208247430062485,
|
| 43 |
+
"eval_cap_loss": 4.461696884211372,
|
| 44 |
+
"eval_con_loss": 2.0591132453843657,
|
| 45 |
+
"eval_loss": 6.349074924693388,
|
| 46 |
+
"eval_runtime": 31.4,
|
| 47 |
+
"eval_samples_per_second": 25.764,
|
| 48 |
+
"eval_steps_per_second": 3.248,
|
| 49 |
+
"step": 51
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.2549019607843137,
|
| 53 |
+
"grad_norm": 20.630815505981445,
|
| 54 |
+
"learning_rate": 9.838709677419355e-07,
|
| 55 |
+
"loss": 10.3883,
|
| 56 |
+
"step": 64
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.5686274509803921,
|
| 60 |
+
"grad_norm": 20.142724990844727,
|
| 61 |
+
"learning_rate": 9.918989757867583e-07,
|
| 62 |
+
"loss": 10.4571,
|
| 63 |
+
"step": 80
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.8823529411764706,
|
| 67 |
+
"grad_norm": 17.457569122314453,
|
| 68 |
+
"learning_rate": 9.631619841257474e-07,
|
| 69 |
+
"loss": 10.3837,
|
| 70 |
+
"step": 96
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 2.0,
|
| 74 |
+
"eval_bleu": 0.01885383457992918,
|
| 75 |
+
"eval_cap_loss": 3.9823871734095553,
|
| 76 |
+
"eval_con_loss": 2.0591144795511283,
|
| 77 |
+
"eval_loss": 6.109422552819345,
|
| 78 |
+
"step": 102
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.0,
|
| 82 |
+
"eval_bleu": 0.01885383457992918,
|
| 83 |
+
"eval_cap_loss": 3.9823871734095553,
|
| 84 |
+
"eval_con_loss": 2.0591144795511283,
|
| 85 |
+
"eval_loss": 6.109422552819345,
|
| 86 |
+
"eval_runtime": 31.2828,
|
| 87 |
+
"eval_samples_per_second": 25.861,
|
| 88 |
+
"eval_steps_per_second": 3.261,
|
| 89 |
+
"step": 102
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.196078431372549,
|
| 93 |
+
"grad_norm": 14.928309440612793,
|
| 94 |
+
"learning_rate": 9.183978228987435e-07,
|
| 95 |
+
"loss": 10.2051,
|
| 96 |
+
"step": 112
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.5098039215686274,
|
| 100 |
+
"grad_norm": 15.862812995910645,
|
| 101 |
+
"learning_rate": 8.535533905932737e-07,
|
| 102 |
+
"loss": 10.256,
|
| 103 |
+
"step": 128
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 2.8235294117647056,
|
| 107 |
+
"grad_norm": 13.910375595092773,
|
| 108 |
+
"learning_rate": 7.737576698351878e-07,
|
| 109 |
+
"loss": 10.2118,
|
| 110 |
+
"step": 144
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.0,
|
| 114 |
+
"eval_bleu": 0.09867147788444039,
|
| 115 |
+
"eval_cap_loss": 3.6369867652070287,
|
| 116 |
+
"eval_con_loss": 2.059116457022873,
|
| 117 |
+
"eval_loss": 5.936726282624638,
|
| 118 |
+
"step": 153
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 3.0,
|
| 122 |
+
"eval_bleu": 0.09867147788444039,
|
| 123 |
+
"eval_cap_loss": 3.6369867652070287,
|
| 124 |
+
"eval_con_loss": 2.059116457022873,
|
| 125 |
+
"eval_loss": 5.936726282624638,
|
| 126 |
+
"eval_runtime": 31.5123,
|
| 127 |
+
"eval_samples_per_second": 25.673,
|
| 128 |
+
"eval_steps_per_second": 3.237,
|
| 129 |
+
"step": 153
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 3.1372549019607843,
|
| 133 |
+
"grad_norm": 13.83757209777832,
|
| 134 |
+
"learning_rate": 6.823851123199893e-07,
|
| 135 |
+
"loss": 10.0243,
|
| 136 |
+
"step": 160
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 3.450980392156863,
|
| 140 |
+
"grad_norm": 12.985786437988281,
|
| 141 |
+
"learning_rate": 5.832997383071659e-07,
|
| 142 |
+
"loss": 10.1282,
|
| 143 |
+
"step": 176
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 3.764705882352941,
|
| 147 |
+
"grad_norm": 13.04931926727295,
|
| 148 |
+
"learning_rate": 4.806917324978607e-07,
|
| 149 |
+
"loss": 10.0797,
|
| 150 |
+
"step": 192
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 4.0,
|
| 154 |
+
"eval_bleu": 0.12895394150561787,
|
| 155 |
+
"eval_cap_loss": 3.4135348516352035,
|
| 156 |
+
"eval_con_loss": 2.0591169128230975,
|
| 157 |
+
"eval_loss": 5.825001257307389,
|
| 158 |
+
"step": 204
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.0,
|
| 162 |
+
"eval_bleu": 0.12895394150561787,
|
| 163 |
+
"eval_cap_loss": 3.4135348516352035,
|
| 164 |
+
"eval_con_loss": 2.0591169128230975,
|
| 165 |
+
"eval_loss": 5.825001257307389,
|
| 166 |
+
"eval_runtime": 31.5247,
|
| 167 |
+
"eval_samples_per_second": 25.662,
|
| 168 |
+
"eval_steps_per_second": 3.236,
|
| 169 |
+
"step": 204
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.078431372549019,
|
| 173 |
+
"grad_norm": 13.757915496826172,
|
| 174 |
+
"learning_rate": 3.7890024686521205e-07,
|
| 175 |
+
"loss": 9.9267,
|
| 176 |
+
"step": 208
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.392156862745098,
|
| 180 |
+
"grad_norm": 14.317400932312012,
|
| 181 |
+
"learning_rate": 2.8222990386287614e-07,
|
| 182 |
+
"loss": 10.0335,
|
| 183 |
+
"step": 224
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.705882352941177,
|
| 187 |
+
"grad_norm": 11.822975158691406,
|
| 188 |
+
"learning_rate": 1.947687598396154e-07,
|
| 189 |
+
"loss": 10.0163,
|
| 190 |
+
"step": 240
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 5.0,
|
| 194 |
+
"eval_bleu": 0.13267291344127094,
|
| 195 |
+
"eval_cap_loss": 3.3223345349816715,
|
| 196 |
+
"eval_con_loss": 2.0591166206434663,
|
| 197 |
+
"eval_loss": 5.779400519296234,
|
| 198 |
+
"step": 255
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 5.0,
|
| 202 |
+
"eval_bleu": 0.13267291344127094,
|
| 203 |
+
"eval_cap_loss": 3.3223345349816715,
|
| 204 |
+
"eval_con_loss": 2.0591166206434663,
|
| 205 |
+
"eval_loss": 5.779400519296234,
|
| 206 |
+
"eval_runtime": 31.0795,
|
| 207 |
+
"eval_samples_per_second": 26.03,
|
| 208 |
+
"eval_steps_per_second": 3.282,
|
| 209 |
+
"step": 255
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 5.019607843137255,
|
| 213 |
+
"grad_norm": 14.300347328186035,
|
| 214 |
+
"learning_rate": 1.2021542673772584e-07,
|
| 215 |
+
"loss": 9.8869,
|
| 216 |
+
"step": 256
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.333333333333333,
|
| 220 |
+
"grad_norm": 11.969139099121094,
|
| 221 |
+
"learning_rate": 6.172266286180161e-08,
|
| 222 |
+
"loss": 10.0042,
|
| 223 |
+
"step": 272
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.647058823529412,
|
| 227 |
+
"grad_norm": 12.140946388244629,
|
| 228 |
+
"learning_rate": 2.1764047049913525e-08,
|
| 229 |
+
"loss": 10.0031,
|
| 230 |
+
"step": 288
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.96078431372549,
|
| 234 |
+
"grad_norm": 12.373481750488281,
|
| 235 |
+
"learning_rate": 2.029374413821949e-09,
|
| 236 |
+
"loss": 9.9969,
|
| 237 |
+
"step": 304
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.0,
|
| 241 |
+
"eval_bleu": 0.13520814810096016,
|
| 242 |
+
"eval_cap_loss": 3.3063340187072754,
|
| 243 |
+
"eval_con_loss": 2.059116517796236,
|
| 244 |
+
"eval_loss": 5.771400057802014,
|
| 245 |
+
"step": 306
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 6.0,
|
| 249 |
+
"eval_bleu": 0.13520814810096016,
|
| 250 |
+
"eval_cap_loss": 3.3063340187072754,
|
| 251 |
+
"eval_con_loss": 2.059116517796236,
|
| 252 |
+
"eval_loss": 5.771400057802014,
|
| 253 |
+
"eval_runtime": 31.1282,
|
| 254 |
+
"eval_samples_per_second": 25.989,
|
| 255 |
+
"eval_steps_per_second": 3.277,
|
| 256 |
+
"step": 306
|
| 257 |
+
}
|
| 258 |
+
],
|
| 259 |
+
"logging_steps": 16,
|
| 260 |
+
"max_steps": 306,
|
| 261 |
+
"num_input_tokens_seen": 0,
|
| 262 |
+
"num_train_epochs": 6,
|
| 263 |
+
"save_steps": 500,
|
| 264 |
+
"stateful_callbacks": {
|
| 265 |
+
"TrainerControl": {
|
| 266 |
+
"args": {
|
| 267 |
+
"should_epoch_stop": false,
|
| 268 |
+
"should_evaluate": false,
|
| 269 |
+
"should_log": false,
|
| 270 |
+
"should_save": true,
|
| 271 |
+
"should_training_stop": true
|
| 272 |
+
},
|
| 273 |
+
"attributes": {}
|
| 274 |
+
}
|
| 275 |
+
},
|
| 276 |
+
"total_flos": 0.0,
|
| 277 |
+
"train_batch_size": 64,
|
| 278 |
+
"trial_name": null,
|
| 279 |
+
"trial_params": null
|
| 280 |
+
}
|