Training in progress, step 9500, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2384234968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ca7c4e3749f06ebc1778c062b3d70c4f488a26b411ddd7c3d301ae4023802d1
|
| 3 |
size 2384234968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4768662910
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d45b61751a61e4b6f882922592537bf8e092f455741fa220a9008ab320f07ad
|
| 3 |
size 4768662910
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a81369904e00a468d2ec4beb1dd4e8f30c6191c2e29c4144f662ff07eadf5eab
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -8108,6 +8108,456 @@
|
|
| 8108 |
"mean_token_accuracy": 0.9694716207683086,
|
| 8109 |
"num_tokens": 36864000.0,
|
| 8110 |
"step": 9000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8111 |
}
|
| 8112 |
],
|
| 8113 |
"logging_steps": 10,
|
|
@@ -8127,7 +8577,7 @@
|
|
| 8127 |
"attributes": {}
|
| 8128 |
}
|
| 8129 |
},
|
| 8130 |
-
"total_flos":
|
| 8131 |
"train_batch_size": 1,
|
| 8132 |
"trial_name": null,
|
| 8133 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9182736455463728,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 9500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 8108 |
"mean_token_accuracy": 0.9694716207683086,
|
| 8109 |
"num_tokens": 36864000.0,
|
| 8110 |
"step": 9000
|
| 8111 |
+
},
|
| 8112 |
+
{
|
| 8113 |
+
"epoch": 0.8709100575129284,
|
| 8114 |
+
"grad_norm": 0.7295970320701599,
|
| 8115 |
+
"learning_rate": 7.175080558539206e-06,
|
| 8116 |
+
"loss": 0.1497,
|
| 8117 |
+
"mean_token_accuracy": 0.9693737745285034,
|
| 8118 |
+
"num_tokens": 36904960.0,
|
| 8119 |
+
"step": 9010
|
| 8120 |
+
},
|
| 8121 |
+
{
|
| 8122 |
+
"epoch": 0.8718766613503456,
|
| 8123 |
+
"grad_norm": 0.7322930097579956,
|
| 8124 |
+
"learning_rate": 7.121374865735768e-06,
|
| 8125 |
+
"loss": 0.144,
|
| 8126 |
+
"mean_token_accuracy": 0.9701565556228161,
|
| 8127 |
+
"num_tokens": 36945920.0,
|
| 8128 |
+
"step": 9020
|
| 8129 |
+
},
|
| 8130 |
+
{
|
| 8131 |
+
"epoch": 0.8728432651877628,
|
| 8132 |
+
"grad_norm": 0.7817639112472534,
|
| 8133 |
+
"learning_rate": 7.067669172932331e-06,
|
| 8134 |
+
"loss": 0.1716,
|
| 8135 |
+
"mean_token_accuracy": 0.9648238711059094,
|
| 8136 |
+
"num_tokens": 36986880.0,
|
| 8137 |
+
"step": 9030
|
| 8138 |
+
},
|
| 8139 |
+
{
|
| 8140 |
+
"epoch": 0.87380986902518,
|
| 8141 |
+
"grad_norm": 0.6489200592041016,
|
| 8142 |
+
"learning_rate": 7.013963480128894e-06,
|
| 8143 |
+
"loss": 0.1481,
|
| 8144 |
+
"mean_token_accuracy": 0.9706457898020744,
|
| 8145 |
+
"num_tokens": 37027840.0,
|
| 8146 |
+
"step": 9040
|
| 8147 |
+
},
|
| 8148 |
+
{
|
| 8149 |
+
"epoch": 0.8747764728625973,
|
| 8150 |
+
"grad_norm": 0.8958516716957092,
|
| 8151 |
+
"learning_rate": 6.960257787325457e-06,
|
| 8152 |
+
"loss": 0.1468,
|
| 8153 |
+
"mean_token_accuracy": 0.9693003885447979,
|
| 8154 |
+
"num_tokens": 37068800.0,
|
| 8155 |
+
"step": 9050
|
| 8156 |
+
},
|
| 8157 |
+
{
|
| 8158 |
+
"epoch": 0.8757430767000145,
|
| 8159 |
+
"grad_norm": 0.7400612235069275,
|
| 8160 |
+
"learning_rate": 6.906552094522021e-06,
|
| 8161 |
+
"loss": 0.1493,
|
| 8162 |
+
"mean_token_accuracy": 0.9704745575785637,
|
| 8163 |
+
"num_tokens": 37109760.0,
|
| 8164 |
+
"step": 9060
|
| 8165 |
+
},
|
| 8166 |
+
{
|
| 8167 |
+
"epoch": 0.8767096805374317,
|
| 8168 |
+
"grad_norm": 0.6856247186660767,
|
| 8169 |
+
"learning_rate": 6.852846401718582e-06,
|
| 8170 |
+
"loss": 0.1293,
|
| 8171 |
+
"mean_token_accuracy": 0.9728718161582947,
|
| 8172 |
+
"num_tokens": 37150720.0,
|
| 8173 |
+
"step": 9070
|
| 8174 |
+
},
|
| 8175 |
+
{
|
| 8176 |
+
"epoch": 0.877676284374849,
|
| 8177 |
+
"grad_norm": 0.7188942432403564,
|
| 8178 |
+
"learning_rate": 6.799140708915146e-06,
|
| 8179 |
+
"loss": 0.1254,
|
| 8180 |
+
"mean_token_accuracy": 0.9737769059836865,
|
| 8181 |
+
"num_tokens": 37191680.0,
|
| 8182 |
+
"step": 9080
|
| 8183 |
+
},
|
| 8184 |
+
{
|
| 8185 |
+
"epoch": 0.8786428882122662,
|
| 8186 |
+
"grad_norm": 0.5900483727455139,
|
| 8187 |
+
"learning_rate": 6.745435016111707e-06,
|
| 8188 |
+
"loss": 0.1473,
|
| 8189 |
+
"mean_token_accuracy": 0.9698874719440937,
|
| 8190 |
+
"num_tokens": 37232640.0,
|
| 8191 |
+
"step": 9090
|
| 8192 |
+
},
|
| 8193 |
+
{
|
| 8194 |
+
"epoch": 0.8796094920496834,
|
| 8195 |
+
"grad_norm": 0.7418830990791321,
|
| 8196 |
+
"learning_rate": 6.691729323308271e-06,
|
| 8197 |
+
"loss": 0.1365,
|
| 8198 |
+
"mean_token_accuracy": 0.9710371784865857,
|
| 8199 |
+
"num_tokens": 37273600.0,
|
| 8200 |
+
"step": 9100
|
| 8201 |
+
},
|
| 8202 |
+
{
|
| 8203 |
+
"epoch": 0.8805760958871006,
|
| 8204 |
+
"grad_norm": 0.6702271699905396,
|
| 8205 |
+
"learning_rate": 6.6380236305048335e-06,
|
| 8206 |
+
"loss": 0.1411,
|
| 8207 |
+
"mean_token_accuracy": 0.969960855692625,
|
| 8208 |
+
"num_tokens": 37314560.0,
|
| 8209 |
+
"step": 9110
|
| 8210 |
+
},
|
| 8211 |
+
{
|
| 8212 |
+
"epoch": 0.8815426997245179,
|
| 8213 |
+
"grad_norm": 0.8139868378639221,
|
| 8214 |
+
"learning_rate": 6.584317937701397e-06,
|
| 8215 |
+
"loss": 0.1456,
|
| 8216 |
+
"mean_token_accuracy": 0.969985319674015,
|
| 8217 |
+
"num_tokens": 37355520.0,
|
| 8218 |
+
"step": 9120
|
| 8219 |
+
},
|
| 8220 |
+
{
|
| 8221 |
+
"epoch": 0.8825093035619351,
|
| 8222 |
+
"grad_norm": 0.6658422350883484,
|
| 8223 |
+
"learning_rate": 6.53061224489796e-06,
|
| 8224 |
+
"loss": 0.1475,
|
| 8225 |
+
"mean_token_accuracy": 0.9693982362747192,
|
| 8226 |
+
"num_tokens": 37396480.0,
|
| 8227 |
+
"step": 9130
|
| 8228 |
+
},
|
| 8229 |
+
{
|
| 8230 |
+
"epoch": 0.8834759073993523,
|
| 8231 |
+
"grad_norm": 0.702655017375946,
|
| 8232 |
+
"learning_rate": 6.476906552094522e-06,
|
| 8233 |
+
"loss": 0.1379,
|
| 8234 |
+
"mean_token_accuracy": 0.9712817974388599,
|
| 8235 |
+
"num_tokens": 37437440.0,
|
| 8236 |
+
"step": 9140
|
| 8237 |
+
},
|
| 8238 |
+
{
|
| 8239 |
+
"epoch": 0.8844425112367696,
|
| 8240 |
+
"grad_norm": 0.7942471504211426,
|
| 8241 |
+
"learning_rate": 6.423200859291086e-06,
|
| 8242 |
+
"loss": 0.1531,
|
| 8243 |
+
"mean_token_accuracy": 0.9690068490803242,
|
| 8244 |
+
"num_tokens": 37478400.0,
|
| 8245 |
+
"step": 9150
|
| 8246 |
+
},
|
| 8247 |
+
{
|
| 8248 |
+
"epoch": 0.8854091150741868,
|
| 8249 |
+
"grad_norm": 0.7765222787857056,
|
| 8250 |
+
"learning_rate": 6.369495166487647e-06,
|
| 8251 |
+
"loss": 0.1528,
|
| 8252 |
+
"mean_token_accuracy": 0.9692270055413246,
|
| 8253 |
+
"num_tokens": 37519360.0,
|
| 8254 |
+
"step": 9160
|
| 8255 |
+
},
|
| 8256 |
+
{
|
| 8257 |
+
"epoch": 0.886375718911604,
|
| 8258 |
+
"grad_norm": 0.5786271095275879,
|
| 8259 |
+
"learning_rate": 6.315789473684211e-06,
|
| 8260 |
+
"loss": 0.1372,
|
| 8261 |
+
"mean_token_accuracy": 0.9717954978346824,
|
| 8262 |
+
"num_tokens": 37560320.0,
|
| 8263 |
+
"step": 9170
|
| 8264 |
+
},
|
| 8265 |
+
{
|
| 8266 |
+
"epoch": 0.8873423227490214,
|
| 8267 |
+
"grad_norm": 0.6450340151786804,
|
| 8268 |
+
"learning_rate": 6.262083780880773e-06,
|
| 8269 |
+
"loss": 0.1624,
|
| 8270 |
+
"mean_token_accuracy": 0.9683219164609909,
|
| 8271 |
+
"num_tokens": 37601280.0,
|
| 8272 |
+
"step": 9180
|
| 8273 |
+
},
|
| 8274 |
+
{
|
| 8275 |
+
"epoch": 0.8883089265864386,
|
| 8276 |
+
"grad_norm": 0.6178423166275024,
|
| 8277 |
+
"learning_rate": 6.2083780880773365e-06,
|
| 8278 |
+
"loss": 0.1253,
|
| 8279 |
+
"mean_token_accuracy": 0.973556748777628,
|
| 8280 |
+
"num_tokens": 37642240.0,
|
| 8281 |
+
"step": 9190
|
| 8282 |
+
},
|
| 8283 |
+
{
|
| 8284 |
+
"epoch": 0.8892755304238558,
|
| 8285 |
+
"grad_norm": 0.7988136410713196,
|
| 8286 |
+
"learning_rate": 6.1546723952739e-06,
|
| 8287 |
+
"loss": 0.1451,
|
| 8288 |
+
"mean_token_accuracy": 0.969740703701973,
|
| 8289 |
+
"num_tokens": 37683200.0,
|
| 8290 |
+
"step": 9200
|
| 8291 |
+
},
|
| 8292 |
+
{
|
| 8293 |
+
"epoch": 0.890242134261273,
|
| 8294 |
+
"grad_norm": 0.6491620540618896,
|
| 8295 |
+
"learning_rate": 6.100966702470462e-06,
|
| 8296 |
+
"loss": 0.1374,
|
| 8297 |
+
"mean_token_accuracy": 0.9728228926658631,
|
| 8298 |
+
"num_tokens": 37724160.0,
|
| 8299 |
+
"step": 9210
|
| 8300 |
+
},
|
| 8301 |
+
{
|
| 8302 |
+
"epoch": 0.8912087380986903,
|
| 8303 |
+
"grad_norm": 0.9206412434577942,
|
| 8304 |
+
"learning_rate": 6.047261009667025e-06,
|
| 8305 |
+
"loss": 0.1457,
|
| 8306 |
+
"mean_token_accuracy": 0.9691780783236027,
|
| 8307 |
+
"num_tokens": 37765120.0,
|
| 8308 |
+
"step": 9220
|
| 8309 |
+
},
|
| 8310 |
+
{
|
| 8311 |
+
"epoch": 0.8921753419361075,
|
| 8312 |
+
"grad_norm": 0.672639787197113,
|
| 8313 |
+
"learning_rate": 5.993555316863588e-06,
|
| 8314 |
+
"loss": 0.1409,
|
| 8315 |
+
"mean_token_accuracy": 0.9709882564842701,
|
| 8316 |
+
"num_tokens": 37806080.0,
|
| 8317 |
+
"step": 9230
|
| 8318 |
+
},
|
| 8319 |
+
{
|
| 8320 |
+
"epoch": 0.8931419457735247,
|
| 8321 |
+
"grad_norm": 0.626698911190033,
|
| 8322 |
+
"learning_rate": 5.939849624060151e-06,
|
| 8323 |
+
"loss": 0.1344,
|
| 8324 |
+
"mean_token_accuracy": 0.9726027339696884,
|
| 8325 |
+
"num_tokens": 37847040.0,
|
| 8326 |
+
"step": 9240
|
| 8327 |
+
},
|
| 8328 |
+
{
|
| 8329 |
+
"epoch": 0.894108549610942,
|
| 8330 |
+
"grad_norm": 0.7072364091873169,
|
| 8331 |
+
"learning_rate": 5.886143931256713e-06,
|
| 8332 |
+
"loss": 0.1351,
|
| 8333 |
+
"mean_token_accuracy": 0.9722113452851773,
|
| 8334 |
+
"num_tokens": 37888000.0,
|
| 8335 |
+
"step": 9250
|
| 8336 |
+
},
|
| 8337 |
+
{
|
| 8338 |
+
"epoch": 0.8950751534483592,
|
| 8339 |
+
"grad_norm": 0.6620608568191528,
|
| 8340 |
+
"learning_rate": 5.832438238453276e-06,
|
| 8341 |
+
"loss": 0.131,
|
| 8342 |
+
"mean_token_accuracy": 0.9728228934109211,
|
| 8343 |
+
"num_tokens": 37928960.0,
|
| 8344 |
+
"step": 9260
|
| 8345 |
+
},
|
| 8346 |
+
{
|
| 8347 |
+
"epoch": 0.8960417572857764,
|
| 8348 |
+
"grad_norm": 0.649089515209198,
|
| 8349 |
+
"learning_rate": 5.7787325456498395e-06,
|
| 8350 |
+
"loss": 0.1534,
|
| 8351 |
+
"mean_token_accuracy": 0.9675146721303463,
|
| 8352 |
+
"num_tokens": 37969920.0,
|
| 8353 |
+
"step": 9270
|
| 8354 |
+
},
|
| 8355 |
+
{
|
| 8356 |
+
"epoch": 0.8970083611231937,
|
| 8357 |
+
"grad_norm": 0.8602608442306519,
|
| 8358 |
+
"learning_rate": 5.725026852846402e-06,
|
| 8359 |
+
"loss": 0.1635,
|
| 8360 |
+
"mean_token_accuracy": 0.9658512689173222,
|
| 8361 |
+
"num_tokens": 38010880.0,
|
| 8362 |
+
"step": 9280
|
| 8363 |
+
},
|
| 8364 |
+
{
|
| 8365 |
+
"epoch": 0.8979749649606109,
|
| 8366 |
+
"grad_norm": 0.6629733443260193,
|
| 8367 |
+
"learning_rate": 5.671321160042965e-06,
|
| 8368 |
+
"loss": 0.1449,
|
| 8369 |
+
"mean_token_accuracy": 0.9702054776251317,
|
| 8370 |
+
"num_tokens": 38051840.0,
|
| 8371 |
+
"step": 9290
|
| 8372 |
+
},
|
| 8373 |
+
{
|
| 8374 |
+
"epoch": 0.8989415687980281,
|
| 8375 |
+
"grad_norm": 0.6702824831008911,
|
| 8376 |
+
"learning_rate": 5.617615467239528e-06,
|
| 8377 |
+
"loss": 0.1383,
|
| 8378 |
+
"mean_token_accuracy": 0.9710861049592495,
|
| 8379 |
+
"num_tokens": 38092800.0,
|
| 8380 |
+
"step": 9300
|
| 8381 |
+
},
|
| 8382 |
+
{
|
| 8383 |
+
"epoch": 0.8999081726354453,
|
| 8384 |
+
"grad_norm": 0.633313000202179,
|
| 8385 |
+
"learning_rate": 5.563909774436091e-06,
|
| 8386 |
+
"loss": 0.1405,
|
| 8387 |
+
"mean_token_accuracy": 0.9705724023282528,
|
| 8388 |
+
"num_tokens": 38133760.0,
|
| 8389 |
+
"step": 9310
|
| 8390 |
+
},
|
| 8391 |
+
{
|
| 8392 |
+
"epoch": 0.9008747764728626,
|
| 8393 |
+
"grad_norm": 0.701628565788269,
|
| 8394 |
+
"learning_rate": 5.510204081632653e-06,
|
| 8395 |
+
"loss": 0.1523,
|
| 8396 |
+
"mean_token_accuracy": 0.9679794482886791,
|
| 8397 |
+
"num_tokens": 38174720.0,
|
| 8398 |
+
"step": 9320
|
| 8399 |
+
},
|
| 8400 |
+
{
|
| 8401 |
+
"epoch": 0.9018413803102798,
|
| 8402 |
+
"grad_norm": 0.6978937387466431,
|
| 8403 |
+
"learning_rate": 5.456498388829216e-06,
|
| 8404 |
+
"loss": 0.1476,
|
| 8405 |
+
"mean_token_accuracy": 0.9705234795808793,
|
| 8406 |
+
"num_tokens": 38215680.0,
|
| 8407 |
+
"step": 9330
|
| 8408 |
+
},
|
| 8409 |
+
{
|
| 8410 |
+
"epoch": 0.902807984147697,
|
| 8411 |
+
"grad_norm": 0.6952319741249084,
|
| 8412 |
+
"learning_rate": 5.4027926960257785e-06,
|
| 8413 |
+
"loss": 0.1294,
|
| 8414 |
+
"mean_token_accuracy": 0.9723825819790364,
|
| 8415 |
+
"num_tokens": 38256640.0,
|
| 8416 |
+
"step": 9340
|
| 8417 |
+
},
|
| 8418 |
+
{
|
| 8419 |
+
"epoch": 0.9037745879851143,
|
| 8420 |
+
"grad_norm": 0.7691527605056763,
|
| 8421 |
+
"learning_rate": 5.349087003222342e-06,
|
| 8422 |
+
"loss": 0.1352,
|
| 8423 |
+
"mean_token_accuracy": 0.9732876695692539,
|
| 8424 |
+
"num_tokens": 38297600.0,
|
| 8425 |
+
"step": 9350
|
| 8426 |
+
},
|
| 8427 |
+
{
|
| 8428 |
+
"epoch": 0.9047411918225315,
|
| 8429 |
+
"grad_norm": 0.8168832659721375,
|
| 8430 |
+
"learning_rate": 5.295381310418905e-06,
|
| 8431 |
+
"loss": 0.1408,
|
| 8432 |
+
"mean_token_accuracy": 0.9707191728055478,
|
| 8433 |
+
"num_tokens": 38338560.0,
|
| 8434 |
+
"step": 9360
|
| 8435 |
+
},
|
| 8436 |
+
{
|
| 8437 |
+
"epoch": 0.9057077956599487,
|
| 8438 |
+
"grad_norm": 0.7141739130020142,
|
| 8439 |
+
"learning_rate": 5.241675617615467e-06,
|
| 8440 |
+
"loss": 0.1324,
|
| 8441 |
+
"mean_token_accuracy": 0.9724315024912358,
|
| 8442 |
+
"num_tokens": 38379520.0,
|
| 8443 |
+
"step": 9370
|
| 8444 |
+
},
|
| 8445 |
+
{
|
| 8446 |
+
"epoch": 0.9066743994973661,
|
| 8447 |
+
"grad_norm": 0.7228880524635315,
|
| 8448 |
+
"learning_rate": 5.187969924812031e-06,
|
| 8449 |
+
"loss": 0.1449,
|
| 8450 |
+
"mean_token_accuracy": 0.9706213280558587,
|
| 8451 |
+
"num_tokens": 38420480.0,
|
| 8452 |
+
"step": 9380
|
| 8453 |
+
},
|
| 8454 |
+
{
|
| 8455 |
+
"epoch": 0.9076410033347833,
|
| 8456 |
+
"grad_norm": 0.6438316702842712,
|
| 8457 |
+
"learning_rate": 5.134264232008593e-06,
|
| 8458 |
+
"loss": 0.1572,
|
| 8459 |
+
"mean_token_accuracy": 0.9682240657508373,
|
| 8460 |
+
"num_tokens": 38461440.0,
|
| 8461 |
+
"step": 9390
|
| 8462 |
+
},
|
| 8463 |
+
{
|
| 8464 |
+
"epoch": 0.9086076071722005,
|
| 8465 |
+
"grad_norm": 0.608272910118103,
|
| 8466 |
+
"learning_rate": 5.080558539205156e-06,
|
| 8467 |
+
"loss": 0.1341,
|
| 8468 |
+
"mean_token_accuracy": 0.9718688815832138,
|
| 8469 |
+
"num_tokens": 38502400.0,
|
| 8470 |
+
"step": 9400
|
| 8471 |
+
},
|
| 8472 |
+
{
|
| 8473 |
+
"epoch": 0.9095742110096177,
|
| 8474 |
+
"grad_norm": 0.5616933107376099,
|
| 8475 |
+
"learning_rate": 5.0268528464017184e-06,
|
| 8476 |
+
"loss": 0.1406,
|
| 8477 |
+
"mean_token_accuracy": 0.9704990208148956,
|
| 8478 |
+
"num_tokens": 38543360.0,
|
| 8479 |
+
"step": 9410
|
| 8480 |
+
},
|
| 8481 |
+
{
|
| 8482 |
+
"epoch": 0.910540814847035,
|
| 8483 |
+
"grad_norm": 0.758497416973114,
|
| 8484 |
+
"learning_rate": 4.9731471535982815e-06,
|
| 8485 |
+
"loss": 0.1427,
|
| 8486 |
+
"mean_token_accuracy": 0.9709637947380543,
|
| 8487 |
+
"num_tokens": 38584320.0,
|
| 8488 |
+
"step": 9420
|
| 8489 |
+
},
|
| 8490 |
+
{
|
| 8491 |
+
"epoch": 0.9115074186844522,
|
| 8492 |
+
"grad_norm": 0.6793957948684692,
|
| 8493 |
+
"learning_rate": 4.919441460794845e-06,
|
| 8494 |
+
"loss": 0.1438,
|
| 8495 |
+
"mean_token_accuracy": 0.9701320923864841,
|
| 8496 |
+
"num_tokens": 38625280.0,
|
| 8497 |
+
"step": 9430
|
| 8498 |
+
},
|
| 8499 |
+
{
|
| 8500 |
+
"epoch": 0.9124740225218694,
|
| 8501 |
+
"grad_norm": 0.7057655453681946,
|
| 8502 |
+
"learning_rate": 4.865735767991407e-06,
|
| 8503 |
+
"loss": 0.1398,
|
| 8504 |
+
"mean_token_accuracy": 0.9713307209312916,
|
| 8505 |
+
"num_tokens": 38666240.0,
|
| 8506 |
+
"step": 9440
|
| 8507 |
+
},
|
| 8508 |
+
{
|
| 8509 |
+
"epoch": 0.9134406263592867,
|
| 8510 |
+
"grad_norm": 0.7207940220832825,
|
| 8511 |
+
"learning_rate": 4.812030075187971e-06,
|
| 8512 |
+
"loss": 0.1577,
|
| 8513 |
+
"mean_token_accuracy": 0.9684442207217216,
|
| 8514 |
+
"num_tokens": 38707200.0,
|
| 8515 |
+
"step": 9450
|
| 8516 |
+
},
|
| 8517 |
+
{
|
| 8518 |
+
"epoch": 0.9144072301967039,
|
| 8519 |
+
"grad_norm": 0.625789999961853,
|
| 8520 |
+
"learning_rate": 4.758324382384533e-06,
|
| 8521 |
+
"loss": 0.1324,
|
| 8522 |
+
"mean_token_accuracy": 0.9726516611874103,
|
| 8523 |
+
"num_tokens": 38748160.0,
|
| 8524 |
+
"step": 9460
|
| 8525 |
+
},
|
| 8526 |
+
{
|
| 8527 |
+
"epoch": 0.9153738340341211,
|
| 8528 |
+
"grad_norm": 0.7996525764465332,
|
| 8529 |
+
"learning_rate": 4.704618689581096e-06,
|
| 8530 |
+
"loss": 0.1499,
|
| 8531 |
+
"mean_token_accuracy": 0.9679549895226955,
|
| 8532 |
+
"num_tokens": 38789120.0,
|
| 8533 |
+
"step": 9470
|
| 8534 |
+
},
|
| 8535 |
+
{
|
| 8536 |
+
"epoch": 0.9163404378715384,
|
| 8537 |
+
"grad_norm": 0.6713771820068359,
|
| 8538 |
+
"learning_rate": 4.650912996777658e-06,
|
| 8539 |
+
"loss": 0.152,
|
| 8540 |
+
"mean_token_accuracy": 0.9697651579976082,
|
| 8541 |
+
"num_tokens": 38830080.0,
|
| 8542 |
+
"step": 9480
|
| 8543 |
+
},
|
| 8544 |
+
{
|
| 8545 |
+
"epoch": 0.9173070417089556,
|
| 8546 |
+
"grad_norm": 0.9124431014060974,
|
| 8547 |
+
"learning_rate": 4.5972073039742214e-06,
|
| 8548 |
+
"loss": 0.133,
|
| 8549 |
+
"mean_token_accuracy": 0.9723336569964885,
|
| 8550 |
+
"num_tokens": 38871040.0,
|
| 8551 |
+
"step": 9490
|
| 8552 |
+
},
|
| 8553 |
+
{
|
| 8554 |
+
"epoch": 0.9182736455463728,
|
| 8555 |
+
"grad_norm": 0.6981998085975647,
|
| 8556 |
+
"learning_rate": 4.5435016111707845e-06,
|
| 8557 |
+
"loss": 0.1483,
|
| 8558 |
+
"mean_token_accuracy": 0.9684931464493275,
|
| 8559 |
+
"num_tokens": 38912000.0,
|
| 8560 |
+
"step": 9500
|
| 8561 |
}
|
| 8562 |
],
|
| 8563 |
"logging_steps": 10,
|
|
|
|
| 8577 |
"attributes": {}
|
| 8578 |
}
|
| 8579 |
},
|
| 8580 |
+
"total_flos": 1.02836817887232e+17,
|
| 8581 |
"train_batch_size": 1,
|
| 8582 |
"trial_name": null,
|
| 8583 |
"trial_params": null
|