Upload 6 files
Browse files- model/generation_config.json +7 -0
- model/model.safetensors +3 -0
- model/rng_state.pth +3 -0
- model/scheduler.pt +3 -0
- model/trainer_state.json +318 -0
- model/training_args.bin +3 -0
model/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.47.0"
|
| 7 |
+
}
|
model/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de52eaba2eee8309b67c8c4e2f976b32617abd28a08fa8761b00e1214711700c
|
| 3 |
+
size 242041896
|
model/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96d2f6417dab4121134db6c9a0a52d5ca647f6116469c0d746f9c1e2ffc1ab10
|
| 3 |
+
size 14244
|
model/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abe219c40d13942f8616d969af97cde95d76ae15a5c271bc32d6bd60e10cc69a
|
| 3 |
+
size 1064
|
model/trainer_state.json
ADDED
|
@@ -0,0 +1,318 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 1966,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1017293997965412,
|
| 13 |
+
"grad_norm": 1.07757568359375,
|
| 14 |
+
"learning_rate": 4.985827402694507e-05,
|
| 15 |
+
"loss": 1.4448,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1017293997965412,
|
| 20 |
+
"eval_loss": 0.5420752167701721,
|
| 21 |
+
"eval_runtime": 81.2287,
|
| 22 |
+
"eval_samples_per_second": 193.478,
|
| 23 |
+
"eval_steps_per_second": 24.191,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2034587995930824,
|
| 28 |
+
"grad_norm": 0.9271150827407837,
|
| 29 |
+
"learning_rate": 4.943470300789534e-05,
|
| 30 |
+
"loss": 0.6104,
|
| 31 |
+
"step": 200
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.2034587995930824,
|
| 35 |
+
"eval_loss": 0.39222481846809387,
|
| 36 |
+
"eval_runtime": 80.7623,
|
| 37 |
+
"eval_samples_per_second": 194.596,
|
| 38 |
+
"eval_steps_per_second": 24.331,
|
| 39 |
+
"step": 200
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.3051881993896236,
|
| 43 |
+
"grad_norm": 0.8333455324172974,
|
| 44 |
+
"learning_rate": 4.873408942403743e-05,
|
| 45 |
+
"loss": 0.4651,
|
| 46 |
+
"step": 300
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.3051881993896236,
|
| 50 |
+
"eval_loss": 0.32220107316970825,
|
| 51 |
+
"eval_runtime": 80.7592,
|
| 52 |
+
"eval_samples_per_second": 194.603,
|
| 53 |
+
"eval_steps_per_second": 24.332,
|
| 54 |
+
"step": 300
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.4069175991861648,
|
| 58 |
+
"grad_norm": 0.8338972330093384,
|
| 59 |
+
"learning_rate": 4.7764376886723956e-05,
|
| 60 |
+
"loss": 0.4106,
|
| 61 |
+
"step": 400
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.4069175991861648,
|
| 65 |
+
"eval_loss": 0.2739136815071106,
|
| 66 |
+
"eval_runtime": 80.8876,
|
| 67 |
+
"eval_samples_per_second": 194.294,
|
| 68 |
+
"eval_steps_per_second": 24.293,
|
| 69 |
+
"step": 400
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.508646998982706,
|
| 73 |
+
"grad_norm": 0.6324864029884338,
|
| 74 |
+
"learning_rate": 4.6536560072189665e-05,
|
| 75 |
+
"loss": 0.3699,
|
| 76 |
+
"step": 500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.508646998982706,
|
| 80 |
+
"eval_loss": 0.2465265691280365,
|
| 81 |
+
"eval_runtime": 80.8298,
|
| 82 |
+
"eval_samples_per_second": 194.433,
|
| 83 |
+
"eval_steps_per_second": 24.31,
|
| 84 |
+
"step": 500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.6103763987792472,
|
| 88 |
+
"grad_norm": 0.7570372223854065,
|
| 89 |
+
"learning_rate": 4.506456006305641e-05,
|
| 90 |
+
"loss": 0.3256,
|
| 91 |
+
"step": 600
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.6103763987792472,
|
| 95 |
+
"eval_loss": 0.22816891968250275,
|
| 96 |
+
"eval_runtime": 80.8701,
|
| 97 |
+
"eval_samples_per_second": 194.336,
|
| 98 |
+
"eval_steps_per_second": 24.298,
|
| 99 |
+
"step": 600
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.7121057985757884,
|
| 103 |
+
"grad_norm": 0.6379811763763428,
|
| 104 |
+
"learning_rate": 4.336506651001469e-05,
|
| 105 |
+
"loss": 0.3194,
|
| 106 |
+
"step": 700
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.7121057985757884,
|
| 110 |
+
"eval_loss": 0.21390707790851593,
|
| 111 |
+
"eval_runtime": 80.9268,
|
| 112 |
+
"eval_samples_per_second": 194.2,
|
| 113 |
+
"eval_steps_per_second": 24.281,
|
| 114 |
+
"step": 700
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.8138351983723296,
|
| 118 |
+
"grad_norm": 0.8042012453079224,
|
| 119 |
+
"learning_rate": 4.145734840326494e-05,
|
| 120 |
+
"loss": 0.3032,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.8138351983723296,
|
| 125 |
+
"eval_loss": 0.20113198459148407,
|
| 126 |
+
"eval_runtime": 80.8778,
|
| 127 |
+
"eval_samples_per_second": 194.318,
|
| 128 |
+
"eval_steps_per_second": 24.296,
|
| 129 |
+
"step": 800
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.9155645981688708,
|
| 133 |
+
"grad_norm": 0.721193253993988,
|
| 134 |
+
"learning_rate": 3.936303559920664e-05,
|
| 135 |
+
"loss": 0.2787,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.9155645981688708,
|
| 140 |
+
"eval_loss": 0.19362396001815796,
|
| 141 |
+
"eval_runtime": 80.735,
|
| 142 |
+
"eval_samples_per_second": 194.662,
|
| 143 |
+
"eval_steps_per_second": 24.339,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.017293997965412,
|
| 148 |
+
"grad_norm": 0.5696636438369751,
|
| 149 |
+
"learning_rate": 3.710587357944272e-05,
|
| 150 |
+
"loss": 0.2672,
|
| 151 |
+
"step": 1000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.017293997965412,
|
| 155 |
+
"eval_loss": 0.18811464309692383,
|
| 156 |
+
"eval_runtime": 80.9806,
|
| 157 |
+
"eval_samples_per_second": 194.071,
|
| 158 |
+
"eval_steps_per_second": 24.265,
|
| 159 |
+
"step": 1000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.1190233977619533,
|
| 163 |
+
"grad_norm": 0.5195199847221375,
|
| 164 |
+
"learning_rate": 3.471145422266069e-05,
|
| 165 |
+
"loss": 0.2665,
|
| 166 |
+
"step": 1100
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.1190233977619533,
|
| 170 |
+
"eval_loss": 0.18336237967014313,
|
| 171 |
+
"eval_runtime": 81.0197,
|
| 172 |
+
"eval_samples_per_second": 193.978,
|
| 173 |
+
"eval_steps_per_second": 24.253,
|
| 174 |
+
"step": 1100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.2207527975584944,
|
| 178 |
+
"grad_norm": 0.5171718001365662,
|
| 179 |
+
"learning_rate": 3.220692564191985e-05,
|
| 180 |
+
"loss": 0.2597,
|
| 181 |
+
"step": 1200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.2207527975584944,
|
| 185 |
+
"eval_loss": 0.17948384582996368,
|
| 186 |
+
"eval_runtime": 80.9251,
|
| 187 |
+
"eval_samples_per_second": 194.204,
|
| 188 |
+
"eval_steps_per_second": 24.282,
|
| 189 |
+
"step": 1200
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.3224821973550356,
|
| 193 |
+
"grad_norm": 0.6012802124023438,
|
| 194 |
+
"learning_rate": 2.9620684377232173e-05,
|
| 195 |
+
"loss": 0.2457,
|
| 196 |
+
"step": 1300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.3224821973550356,
|
| 200 |
+
"eval_loss": 0.17490074038505554,
|
| 201 |
+
"eval_runtime": 80.6967,
|
| 202 |
+
"eval_samples_per_second": 194.754,
|
| 203 |
+
"eval_steps_per_second": 24.35,
|
| 204 |
+
"step": 1300
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.4242115971515767,
|
| 208 |
+
"grad_norm": 0.8013952374458313,
|
| 209 |
+
"learning_rate": 2.6982053433381048e-05,
|
| 210 |
+
"loss": 0.2446,
|
| 211 |
+
"step": 1400
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.4242115971515767,
|
| 215 |
+
"eval_loss": 0.17289136350154877,
|
| 216 |
+
"eval_runtime": 80.8694,
|
| 217 |
+
"eval_samples_per_second": 194.338,
|
| 218 |
+
"eval_steps_per_second": 24.298,
|
| 219 |
+
"step": 1400
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.5259409969481181,
|
| 223 |
+
"grad_norm": 0.5540690422058105,
|
| 224 |
+
"learning_rate": 2.4320949813410496e-05,
|
| 225 |
+
"loss": 0.2417,
|
| 226 |
+
"step": 1500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.5259409969481181,
|
| 230 |
+
"eval_loss": 0.1698693186044693,
|
| 231 |
+
"eval_runtime": 80.754,
|
| 232 |
+
"eval_samples_per_second": 194.616,
|
| 233 |
+
"eval_steps_per_second": 24.333,
|
| 234 |
+
"step": 1500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.627670396744659,
|
| 238 |
+
"grad_norm": 0.6591632962226868,
|
| 239 |
+
"learning_rate": 2.166754531731575e-05,
|
| 240 |
+
"loss": 0.2312,
|
| 241 |
+
"step": 1600
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.627670396744659,
|
| 245 |
+
"eval_loss": 0.16674023866653442,
|
| 246 |
+
"eval_runtime": 80.7008,
|
| 247 |
+
"eval_samples_per_second": 194.744,
|
| 248 |
+
"eval_steps_per_second": 24.349,
|
| 249 |
+
"step": 1600
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.7293997965412005,
|
| 253 |
+
"grad_norm": 1.1154977083206177,
|
| 254 |
+
"learning_rate": 1.90519244518262e-05,
|
| 255 |
+
"loss": 0.2362,
|
| 256 |
+
"step": 1700
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.7293997965412005,
|
| 260 |
+
"eval_loss": 0.16516855359077454,
|
| 261 |
+
"eval_runtime": 80.7504,
|
| 262 |
+
"eval_samples_per_second": 194.624,
|
| 263 |
+
"eval_steps_per_second": 24.334,
|
| 264 |
+
"step": 1700
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.8311291963377416,
|
| 268 |
+
"grad_norm": 0.6043060421943665,
|
| 269 |
+
"learning_rate": 1.6503743329926174e-05,
|
| 270 |
+
"loss": 0.2308,
|
| 271 |
+
"step": 1800
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.8311291963377416,
|
| 275 |
+
"eval_loss": 0.16346405446529388,
|
| 276 |
+
"eval_runtime": 80.6619,
|
| 277 |
+
"eval_samples_per_second": 194.838,
|
| 278 |
+
"eval_steps_per_second": 24.361,
|
| 279 |
+
"step": 1800
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.9328585961342828,
|
| 283 |
+
"grad_norm": 0.6331253051757812,
|
| 284 |
+
"learning_rate": 1.4051893427537416e-05,
|
| 285 |
+
"loss": 0.2278,
|
| 286 |
+
"step": 1900
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.9328585961342828,
|
| 290 |
+
"eval_loss": 0.16204476356506348,
|
| 291 |
+
"eval_runtime": 80.7035,
|
| 292 |
+
"eval_samples_per_second": 194.738,
|
| 293 |
+
"eval_steps_per_second": 24.348,
|
| 294 |
+
"step": 1900
|
| 295 |
+
}
|
| 296 |
+
],
|
| 297 |
+
"logging_steps": 100,
|
| 298 |
+
"max_steps": 2949,
|
| 299 |
+
"num_input_tokens_seen": 0,
|
| 300 |
+
"num_train_epochs": 3,
|
| 301 |
+
"save_steps": 500,
|
| 302 |
+
"stateful_callbacks": {
|
| 303 |
+
"TrainerControl": {
|
| 304 |
+
"args": {
|
| 305 |
+
"should_epoch_stop": false,
|
| 306 |
+
"should_evaluate": false,
|
| 307 |
+
"should_log": false,
|
| 308 |
+
"should_save": true,
|
| 309 |
+
"should_training_stop": false
|
| 310 |
+
},
|
| 311 |
+
"attributes": {}
|
| 312 |
+
}
|
| 313 |
+
},
|
| 314 |
+
"total_flos": 8507720982331392.0,
|
| 315 |
+
"train_batch_size": 64,
|
| 316 |
+
"trial_name": null,
|
| 317 |
+
"trial_params": null
|
| 318 |
+
}
|
model/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cd7d483f5b00f9b5d354c02efd4836ba470c8d241b371b36d550ab140ed4cb8
|
| 3 |
+
size 5304
|