Training in progress, step 450, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 912336848
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:afb10fcd25ab47f6824a47a0e69a659b482321bd359014b11cd02b39c18c712b
|
3 |
size 912336848
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 463916756
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f412242b8b4c1088fdef955be25e987f1a3ae214ed80585611bc33ab628f9141
|
3 |
size 463916756
|
last-checkpoint/rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14244
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f56a3eb79b75d9133a792f2e57fd736a8e6c8d874fe198547337ba33d890f211
|
3 |
size 14244
|
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:c9636ae38b683f4b5b714bdf172e563b0c593e0efe94f07eea78547963bfbfae
|
3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
4 |
-
"epoch": 0.
|
5 |
"eval_steps": 150,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -2131,6 +2131,1064 @@
|
|
2131 |
"eval_samples_per_second": 9.109,
|
2132 |
"eval_steps_per_second": 2.278,
|
2133 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2134 |
}
|
2135 |
],
|
2136 |
"logging_steps": 1,
|
@@ -2154,12 +3212,12 @@
|
|
2154 |
"should_evaluate": false,
|
2155 |
"should_log": false,
|
2156 |
"should_save": true,
|
2157 |
-
"should_training_stop":
|
2158 |
},
|
2159 |
"attributes": {}
|
2160 |
}
|
2161 |
},
|
2162 |
-
"total_flos":
|
2163 |
"train_batch_size": 8,
|
2164 |
"trial_name": null,
|
2165 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.8302884697914124,
|
3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-450",
|
4 |
+
"epoch": 0.08468197214904027,
|
5 |
"eval_steps": 150,
|
6 |
+
"global_step": 450,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
2131 |
"eval_samples_per_second": 9.109,
|
2132 |
"eval_steps_per_second": 2.278,
|
2133 |
"step": 300
|
2134 |
+
},
|
2135 |
+
{
|
2136 |
+
"epoch": 0.056642830259691385,
|
2137 |
+
"grad_norm": 1.5962673425674438,
|
2138 |
+
"learning_rate": 2.6813618894527138e-05,
|
2139 |
+
"loss": 3.5087,
|
2140 |
+
"step": 301
|
2141 |
+
},
|
2142 |
+
{
|
2143 |
+
"epoch": 0.05683101242002258,
|
2144 |
+
"grad_norm": 1.5255632400512695,
|
2145 |
+
"learning_rate": 2.6490591592961578e-05,
|
2146 |
+
"loss": 3.4712,
|
2147 |
+
"step": 302
|
2148 |
+
},
|
2149 |
+
{
|
2150 |
+
"epoch": 0.05701919458035378,
|
2151 |
+
"grad_norm": 1.542389154434204,
|
2152 |
+
"learning_rate": 2.6168819172567392e-05,
|
2153 |
+
"loss": 3.2852,
|
2154 |
+
"step": 303
|
2155 |
+
},
|
2156 |
+
{
|
2157 |
+
"epoch": 0.05720737674068498,
|
2158 |
+
"grad_norm": 1.664982557296753,
|
2159 |
+
"learning_rate": 2.5848318808857606e-05,
|
2160 |
+
"loss": 3.5038,
|
2161 |
+
"step": 304
|
2162 |
+
},
|
2163 |
+
{
|
2164 |
+
"epoch": 0.057395558901016185,
|
2165 |
+
"grad_norm": 1.4651890993118286,
|
2166 |
+
"learning_rate": 2.5529107609445733e-05,
|
2167 |
+
"loss": 3.1837,
|
2168 |
+
"step": 305
|
2169 |
+
},
|
2170 |
+
{
|
2171 |
+
"epoch": 0.057583741061347386,
|
2172 |
+
"grad_norm": 1.486304521560669,
|
2173 |
+
"learning_rate": 2.521120261313241e-05,
|
2174 |
+
"loss": 3.2311,
|
2175 |
+
"step": 306
|
2176 |
+
},
|
2177 |
+
{
|
2178 |
+
"epoch": 0.05777192322167859,
|
2179 |
+
"grad_norm": 1.4881970882415771,
|
2180 |
+
"learning_rate": 2.4894620788996037e-05,
|
2181 |
+
"loss": 3.419,
|
2182 |
+
"step": 307
|
2183 |
+
},
|
2184 |
+
{
|
2185 |
+
"epoch": 0.05796010538200978,
|
2186 |
+
"grad_norm": 1.4943639039993286,
|
2187 |
+
"learning_rate": 2.457937903548695e-05,
|
2188 |
+
"loss": 3.3085,
|
2189 |
+
"step": 308
|
2190 |
+
},
|
2191 |
+
{
|
2192 |
+
"epoch": 0.058148287542340985,
|
2193 |
+
"grad_norm": 1.5298432111740112,
|
2194 |
+
"learning_rate": 2.426549417952542e-05,
|
2195 |
+
"loss": 3.608,
|
2196 |
+
"step": 309
|
2197 |
+
},
|
2198 |
+
{
|
2199 |
+
"epoch": 0.058336469702672186,
|
2200 |
+
"grad_norm": 1.5149232149124146,
|
2201 |
+
"learning_rate": 2.3952982975603496e-05,
|
2202 |
+
"loss": 3.3398,
|
2203 |
+
"step": 310
|
2204 |
+
},
|
2205 |
+
{
|
2206 |
+
"epoch": 0.05852465186300339,
|
2207 |
+
"grad_norm": 1.492951512336731,
|
2208 |
+
"learning_rate": 2.3641862104890595e-05,
|
2209 |
+
"loss": 3.3529,
|
2210 |
+
"step": 311
|
2211 |
+
},
|
2212 |
+
{
|
2213 |
+
"epoch": 0.05871283402333459,
|
2214 |
+
"grad_norm": 1.448451280593872,
|
2215 |
+
"learning_rate": 2.3332148174343254e-05,
|
2216 |
+
"loss": 3.2021,
|
2217 |
+
"step": 312
|
2218 |
+
},
|
2219 |
+
{
|
2220 |
+
"epoch": 0.05890101618366579,
|
2221 |
+
"grad_norm": 1.4983866214752197,
|
2222 |
+
"learning_rate": 2.3023857715818532e-05,
|
2223 |
+
"loss": 3.3579,
|
2224 |
+
"step": 313
|
2225 |
+
},
|
2226 |
+
{
|
2227 |
+
"epoch": 0.059089198343996986,
|
2228 |
+
"grad_norm": 1.5230140686035156,
|
2229 |
+
"learning_rate": 2.2717007185191674e-05,
|
2230 |
+
"loss": 3.4931,
|
2231 |
+
"step": 314
|
2232 |
+
},
|
2233 |
+
{
|
2234 |
+
"epoch": 0.05927738050432819,
|
2235 |
+
"grad_norm": 1.5284234285354614,
|
2236 |
+
"learning_rate": 2.24116129614777e-05,
|
2237 |
+
"loss": 3.4331,
|
2238 |
+
"step": 315
|
2239 |
+
},
|
2240 |
+
{
|
2241 |
+
"epoch": 0.05946556266465939,
|
2242 |
+
"grad_norm": 1.4961637258529663,
|
2243 |
+
"learning_rate": 2.2107691345957133e-05,
|
2244 |
+
"loss": 3.3148,
|
2245 |
+
"step": 316
|
2246 |
+
},
|
2247 |
+
{
|
2248 |
+
"epoch": 0.05965374482499059,
|
2249 |
+
"grad_norm": 1.547951340675354,
|
2250 |
+
"learning_rate": 2.1805258561305862e-05,
|
2251 |
+
"loss": 3.3194,
|
2252 |
+
"step": 317
|
2253 |
+
},
|
2254 |
+
{
|
2255 |
+
"epoch": 0.05984192698532179,
|
2256 |
+
"grad_norm": 1.5121986865997314,
|
2257 |
+
"learning_rate": 2.1504330750729186e-05,
|
2258 |
+
"loss": 3.4504,
|
2259 |
+
"step": 318
|
2260 |
+
},
|
2261 |
+
{
|
2262 |
+
"epoch": 0.060030109145652995,
|
2263 |
+
"grad_norm": 1.51499605178833,
|
2264 |
+
"learning_rate": 2.120492397710022e-05,
|
2265 |
+
"loss": 3.4349,
|
2266 |
+
"step": 319
|
2267 |
+
},
|
2268 |
+
{
|
2269 |
+
"epoch": 0.06021829130598419,
|
2270 |
+
"grad_norm": 1.520011305809021,
|
2271 |
+
"learning_rate": 2.090705422210237e-05,
|
2272 |
+
"loss": 3.3619,
|
2273 |
+
"step": 320
|
2274 |
+
},
|
2275 |
+
{
|
2276 |
+
"epoch": 0.06040647346631539,
|
2277 |
+
"grad_norm": 1.5512540340423584,
|
2278 |
+
"learning_rate": 2.061073738537635e-05,
|
2279 |
+
"loss": 3.38,
|
2280 |
+
"step": 321
|
2281 |
+
},
|
2282 |
+
{
|
2283 |
+
"epoch": 0.06059465562664659,
|
2284 |
+
"grad_norm": 1.463521122932434,
|
2285 |
+
"learning_rate": 2.0315989283671473e-05,
|
2286 |
+
"loss": 3.1915,
|
2287 |
+
"step": 322
|
2288 |
+
},
|
2289 |
+
{
|
2290 |
+
"epoch": 0.060782837786977795,
|
2291 |
+
"grad_norm": 1.5051774978637695,
|
2292 |
+
"learning_rate": 2.0022825650001387e-05,
|
2293 |
+
"loss": 3.4725,
|
2294 |
+
"step": 323
|
2295 |
+
},
|
2296 |
+
{
|
2297 |
+
"epoch": 0.060971019947308996,
|
2298 |
+
"grad_norm": 1.4965063333511353,
|
2299 |
+
"learning_rate": 1.9731262132804274e-05,
|
2300 |
+
"loss": 3.3025,
|
2301 |
+
"step": 324
|
2302 |
+
},
|
2303 |
+
{
|
2304 |
+
"epoch": 0.0611592021076402,
|
2305 |
+
"grad_norm": 1.4896841049194336,
|
2306 |
+
"learning_rate": 1.9441314295107537e-05,
|
2307 |
+
"loss": 3.4046,
|
2308 |
+
"step": 325
|
2309 |
+
},
|
2310 |
+
{
|
2311 |
+
"epoch": 0.06134738426797139,
|
2312 |
+
"grad_norm": 1.4985949993133545,
|
2313 |
+
"learning_rate": 1.9152997613697183e-05,
|
2314 |
+
"loss": 3.4278,
|
2315 |
+
"step": 326
|
2316 |
+
},
|
2317 |
+
{
|
2318 |
+
"epoch": 0.061535566428302595,
|
2319 |
+
"grad_norm": 1.5722819566726685,
|
2320 |
+
"learning_rate": 1.8866327478291546e-05,
|
2321 |
+
"loss": 3.4424,
|
2322 |
+
"step": 327
|
2323 |
+
},
|
2324 |
+
{
|
2325 |
+
"epoch": 0.061723748588633796,
|
2326 |
+
"grad_norm": 1.457684874534607,
|
2327 |
+
"learning_rate": 1.8581319190720035e-05,
|
2328 |
+
"loss": 3.175,
|
2329 |
+
"step": 328
|
2330 |
+
},
|
2331 |
+
{
|
2332 |
+
"epoch": 0.061911930748965,
|
2333 |
+
"grad_norm": 1.501202940940857,
|
2334 |
+
"learning_rate": 1.8297987964106115e-05,
|
2335 |
+
"loss": 3.2706,
|
2336 |
+
"step": 329
|
2337 |
+
},
|
2338 |
+
{
|
2339 |
+
"epoch": 0.0621001129092962,
|
2340 |
+
"grad_norm": 1.4955599308013916,
|
2341 |
+
"learning_rate": 1.801634892205545e-05,
|
2342 |
+
"loss": 3.3183,
|
2343 |
+
"step": 330
|
2344 |
+
},
|
2345 |
+
{
|
2346 |
+
"epoch": 0.0622882950696274,
|
2347 |
+
"grad_norm": 1.52448308467865,
|
2348 |
+
"learning_rate": 1.7736417097848506e-05,
|
2349 |
+
"loss": 3.3863,
|
2350 |
+
"step": 331
|
2351 |
+
},
|
2352 |
+
{
|
2353 |
+
"epoch": 0.0624764772299586,
|
2354 |
+
"grad_norm": 1.5245345830917358,
|
2355 |
+
"learning_rate": 1.7458207433638223e-05,
|
2356 |
+
"loss": 3.4718,
|
2357 |
+
"step": 332
|
2358 |
+
},
|
2359 |
+
{
|
2360 |
+
"epoch": 0.0626646593902898,
|
2361 |
+
"grad_norm": 1.5418884754180908,
|
2362 |
+
"learning_rate": 1.718173477965236e-05,
|
2363 |
+
"loss": 3.3681,
|
2364 |
+
"step": 333
|
2365 |
+
},
|
2366 |
+
{
|
2367 |
+
"epoch": 0.062852841550621,
|
2368 |
+
"grad_norm": 1.487151861190796,
|
2369 |
+
"learning_rate": 1.6907013893400837e-05,
|
2370 |
+
"loss": 3.3089,
|
2371 |
+
"step": 334
|
2372 |
+
},
|
2373 |
+
{
|
2374 |
+
"epoch": 0.0630410237109522,
|
2375 |
+
"grad_norm": 1.4714837074279785,
|
2376 |
+
"learning_rate": 1.6634059438888033e-05,
|
2377 |
+
"loss": 3.181,
|
2378 |
+
"step": 335
|
2379 |
+
},
|
2380 |
+
{
|
2381 |
+
"epoch": 0.0632292058712834,
|
2382 |
+
"grad_norm": 1.5032023191452026,
|
2383 |
+
"learning_rate": 1.636288598583e-05,
|
2384 |
+
"loss": 3.3265,
|
2385 |
+
"step": 336
|
2386 |
+
},
|
2387 |
+
{
|
2388 |
+
"epoch": 0.0634173880316146,
|
2389 |
+
"grad_norm": 1.5267105102539062,
|
2390 |
+
"learning_rate": 1.6093508008876857e-05,
|
2391 |
+
"loss": 3.1744,
|
2392 |
+
"step": 337
|
2393 |
+
},
|
2394 |
+
{
|
2395 |
+
"epoch": 0.0636055701919458,
|
2396 |
+
"grad_norm": 1.517006516456604,
|
2397 |
+
"learning_rate": 1.5825939886840037e-05,
|
2398 |
+
"loss": 3.2575,
|
2399 |
+
"step": 338
|
2400 |
+
},
|
2401 |
+
{
|
2402 |
+
"epoch": 0.06379375235227701,
|
2403 |
+
"grad_norm": 1.4955118894577026,
|
2404 |
+
"learning_rate": 1.5560195901924894e-05,
|
2405 |
+
"loss": 3.2745,
|
2406 |
+
"step": 339
|
2407 |
+
},
|
2408 |
+
{
|
2409 |
+
"epoch": 0.06398193451260821,
|
2410 |
+
"grad_norm": 1.7153363227844238,
|
2411 |
+
"learning_rate": 1.5296290238968303e-05,
|
2412 |
+
"loss": 3.4614,
|
2413 |
+
"step": 340
|
2414 |
+
},
|
2415 |
+
{
|
2416 |
+
"epoch": 0.06417011667293941,
|
2417 |
+
"grad_norm": 1.5209360122680664,
|
2418 |
+
"learning_rate": 1.50342369846815e-05,
|
2419 |
+
"loss": 3.3276,
|
2420 |
+
"step": 341
|
2421 |
+
},
|
2422 |
+
{
|
2423 |
+
"epoch": 0.0643582988332706,
|
2424 |
+
"grad_norm": 1.5184190273284912,
|
2425 |
+
"learning_rate": 1.4774050126898164e-05,
|
2426 |
+
"loss": 3.3869,
|
2427 |
+
"step": 342
|
2428 |
+
},
|
2429 |
+
{
|
2430 |
+
"epoch": 0.0645464809936018,
|
2431 |
+
"grad_norm": 1.5385315418243408,
|
2432 |
+
"learning_rate": 1.451574355382776e-05,
|
2433 |
+
"loss": 3.4677,
|
2434 |
+
"step": 343
|
2435 |
+
},
|
2436 |
+
{
|
2437 |
+
"epoch": 0.064734663153933,
|
2438 |
+
"grad_norm": 1.4663329124450684,
|
2439 |
+
"learning_rate": 1.425933105331429e-05,
|
2440 |
+
"loss": 3.3069,
|
2441 |
+
"step": 344
|
2442 |
+
},
|
2443 |
+
{
|
2444 |
+
"epoch": 0.0649228453142642,
|
2445 |
+
"grad_norm": 1.5196608304977417,
|
2446 |
+
"learning_rate": 1.4004826312100216e-05,
|
2447 |
+
"loss": 3.3985,
|
2448 |
+
"step": 345
|
2449 |
+
},
|
2450 |
+
{
|
2451 |
+
"epoch": 0.0651110274745954,
|
2452 |
+
"grad_norm": 1.5887420177459717,
|
2453 |
+
"learning_rate": 1.3752242915095992e-05,
|
2454 |
+
"loss": 3.309,
|
2455 |
+
"step": 346
|
2456 |
+
},
|
2457 |
+
{
|
2458 |
+
"epoch": 0.06529920963492661,
|
2459 |
+
"grad_norm": 1.5067335367202759,
|
2460 |
+
"learning_rate": 1.3501594344654884e-05,
|
2461 |
+
"loss": 3.186,
|
2462 |
+
"step": 347
|
2463 |
+
},
|
2464 |
+
{
|
2465 |
+
"epoch": 0.06548739179525781,
|
2466 |
+
"grad_norm": 1.4992436170578003,
|
2467 |
+
"learning_rate": 1.3252893979853304e-05,
|
2468 |
+
"loss": 3.3428,
|
2469 |
+
"step": 348
|
2470 |
+
},
|
2471 |
+
{
|
2472 |
+
"epoch": 0.06567557395558901,
|
2473 |
+
"grad_norm": 1.6664701700210571,
|
2474 |
+
"learning_rate": 1.3006155095776707e-05,
|
2475 |
+
"loss": 3.4061,
|
2476 |
+
"step": 349
|
2477 |
+
},
|
2478 |
+
{
|
2479 |
+
"epoch": 0.06586375611592021,
|
2480 |
+
"grad_norm": 2.502802848815918,
|
2481 |
+
"learning_rate": 1.2761390862810907e-05,
|
2482 |
+
"loss": 3.0303,
|
2483 |
+
"step": 350
|
2484 |
+
},
|
2485 |
+
{
|
2486 |
+
"epoch": 0.06605193827625142,
|
2487 |
+
"grad_norm": 1.500586748123169,
|
2488 |
+
"learning_rate": 1.2518614345939212e-05,
|
2489 |
+
"loss": 3.4372,
|
2490 |
+
"step": 351
|
2491 |
+
},
|
2492 |
+
{
|
2493 |
+
"epoch": 0.06624012043658262,
|
2494 |
+
"grad_norm": 1.5125335454940796,
|
2495 |
+
"learning_rate": 1.227783850404487e-05,
|
2496 |
+
"loss": 3.3834,
|
2497 |
+
"step": 352
|
2498 |
+
},
|
2499 |
+
{
|
2500 |
+
"epoch": 0.06642830259691382,
|
2501 |
+
"grad_norm": 1.5365911722183228,
|
2502 |
+
"learning_rate": 1.2039076189219517e-05,
|
2503 |
+
"loss": 3.4872,
|
2504 |
+
"step": 353
|
2505 |
+
},
|
2506 |
+
{
|
2507 |
+
"epoch": 0.066616484757245,
|
2508 |
+
"grad_norm": 1.5221909284591675,
|
2509 |
+
"learning_rate": 1.1802340146077045e-05,
|
2510 |
+
"loss": 3.3993,
|
2511 |
+
"step": 354
|
2512 |
+
},
|
2513 |
+
{
|
2514 |
+
"epoch": 0.06680466691757621,
|
2515 |
+
"grad_norm": 1.5232350826263428,
|
2516 |
+
"learning_rate": 1.1567643011073392e-05,
|
2517 |
+
"loss": 3.5355,
|
2518 |
+
"step": 355
|
2519 |
+
},
|
2520 |
+
{
|
2521 |
+
"epoch": 0.06699284907790741,
|
2522 |
+
"grad_norm": 1.5335094928741455,
|
2523 |
+
"learning_rate": 1.1334997311832002e-05,
|
2524 |
+
"loss": 3.4007,
|
2525 |
+
"step": 356
|
2526 |
+
},
|
2527 |
+
{
|
2528 |
+
"epoch": 0.06718103123823861,
|
2529 |
+
"grad_norm": 1.4804530143737793,
|
2530 |
+
"learning_rate": 1.1104415466475087e-05,
|
2531 |
+
"loss": 3.3991,
|
2532 |
+
"step": 357
|
2533 |
+
},
|
2534 |
+
{
|
2535 |
+
"epoch": 0.06736921339856981,
|
2536 |
+
"grad_norm": 1.4774835109710693,
|
2537 |
+
"learning_rate": 1.0875909782960886e-05,
|
2538 |
+
"loss": 3.3081,
|
2539 |
+
"step": 358
|
2540 |
+
},
|
2541 |
+
{
|
2542 |
+
"epoch": 0.06755739555890101,
|
2543 |
+
"grad_norm": 1.507690191268921,
|
2544 |
+
"learning_rate": 1.0649492458426564e-05,
|
2545 |
+
"loss": 3.5331,
|
2546 |
+
"step": 359
|
2547 |
+
},
|
2548 |
+
{
|
2549 |
+
"epoch": 0.06774557771923222,
|
2550 |
+
"grad_norm": 1.4900294542312622,
|
2551 |
+
"learning_rate": 1.0425175578537299e-05,
|
2552 |
+
"loss": 3.4034,
|
2553 |
+
"step": 360
|
2554 |
+
},
|
2555 |
+
{
|
2556 |
+
"epoch": 0.06793375987956342,
|
2557 |
+
"grad_norm": 1.536511778831482,
|
2558 |
+
"learning_rate": 1.020297111684101e-05,
|
2559 |
+
"loss": 3.3285,
|
2560 |
+
"step": 361
|
2561 |
+
},
|
2562 |
+
{
|
2563 |
+
"epoch": 0.06812194203989462,
|
2564 |
+
"grad_norm": 1.4938386678695679,
|
2565 |
+
"learning_rate": 9.98289093412938e-06,
|
2566 |
+
"loss": 3.3401,
|
2567 |
+
"step": 362
|
2568 |
+
},
|
2569 |
+
{
|
2570 |
+
"epoch": 0.06831012420022582,
|
2571 |
+
"grad_norm": 1.498859167098999,
|
2572 |
+
"learning_rate": 9.764946777804646e-06,
|
2573 |
+
"loss": 3.4705,
|
2574 |
+
"step": 363
|
2575 |
+
},
|
2576 |
+
{
|
2577 |
+
"epoch": 0.06849830636055702,
|
2578 |
+
"grad_norm": 1.505251407623291,
|
2579 |
+
"learning_rate": 9.549150281252633e-06,
|
2580 |
+
"loss": 3.3769,
|
2581 |
+
"step": 364
|
2582 |
+
},
|
2583 |
+
{
|
2584 |
+
"epoch": 0.06868648852088823,
|
2585 |
+
"grad_norm": 1.5352188348770142,
|
2586 |
+
"learning_rate": 9.335512963221732e-06,
|
2587 |
+
"loss": 3.4495,
|
2588 |
+
"step": 365
|
2589 |
+
},
|
2590 |
+
{
|
2591 |
+
"epoch": 0.06887467068121943,
|
2592 |
+
"grad_norm": 1.5243737697601318,
|
2593 |
+
"learning_rate": 9.124046227208082e-06,
|
2594 |
+
"loss": 3.3756,
|
2595 |
+
"step": 366
|
2596 |
+
},
|
2597 |
+
{
|
2598 |
+
"epoch": 0.06906285284155061,
|
2599 |
+
"grad_norm": 1.4866875410079956,
|
2600 |
+
"learning_rate": 8.914761360846869e-06,
|
2601 |
+
"loss": 3.2548,
|
2602 |
+
"step": 367
|
2603 |
+
},
|
2604 |
+
{
|
2605 |
+
"epoch": 0.06925103500188182,
|
2606 |
+
"grad_norm": 1.70474374294281,
|
2607 |
+
"learning_rate": 8.707669535309793e-06,
|
2608 |
+
"loss": 3.3533,
|
2609 |
+
"step": 368
|
2610 |
+
},
|
2611 |
+
{
|
2612 |
+
"epoch": 0.06943921716221302,
|
2613 |
+
"grad_norm": 1.4734491109848022,
|
2614 |
+
"learning_rate": 8.502781804708826e-06,
|
2615 |
+
"loss": 3.3556,
|
2616 |
+
"step": 369
|
2617 |
+
},
|
2618 |
+
{
|
2619 |
+
"epoch": 0.06962739932254422,
|
2620 |
+
"grad_norm": 1.5178672075271606,
|
2621 |
+
"learning_rate": 8.30010910550611e-06,
|
2622 |
+
"loss": 3.4011,
|
2623 |
+
"step": 370
|
2624 |
+
},
|
2625 |
+
{
|
2626 |
+
"epoch": 0.06981558148287542,
|
2627 |
+
"grad_norm": 1.5426009893417358,
|
2628 |
+
"learning_rate": 8.09966225593024e-06,
|
2629 |
+
"loss": 3.3262,
|
2630 |
+
"step": 371
|
2631 |
+
},
|
2632 |
+
{
|
2633 |
+
"epoch": 0.07000376364320662,
|
2634 |
+
"grad_norm": 1.488108515739441,
|
2635 |
+
"learning_rate": 7.901451955398792e-06,
|
2636 |
+
"loss": 3.3038,
|
2637 |
+
"step": 372
|
2638 |
+
},
|
2639 |
+
{
|
2640 |
+
"epoch": 0.07019194580353783,
|
2641 |
+
"grad_norm": 1.5539402961730957,
|
2642 |
+
"learning_rate": 7.705488783947202e-06,
|
2643 |
+
"loss": 3.4263,
|
2644 |
+
"step": 373
|
2645 |
+
},
|
2646 |
+
{
|
2647 |
+
"epoch": 0.07038012796386903,
|
2648 |
+
"grad_norm": 1.5011879205703735,
|
2649 |
+
"learning_rate": 7.511783201664052e-06,
|
2650 |
+
"loss": 3.1954,
|
2651 |
+
"step": 374
|
2652 |
+
},
|
2653 |
+
{
|
2654 |
+
"epoch": 0.07056831012420023,
|
2655 |
+
"grad_norm": 1.5092988014221191,
|
2656 |
+
"learning_rate": 7.320345548132679e-06,
|
2657 |
+
"loss": 3.328,
|
2658 |
+
"step": 375
|
2659 |
+
},
|
2660 |
+
{
|
2661 |
+
"epoch": 0.07075649228453143,
|
2662 |
+
"grad_norm": 1.498982310295105,
|
2663 |
+
"learning_rate": 7.131186041879357e-06,
|
2664 |
+
"loss": 3.2402,
|
2665 |
+
"step": 376
|
2666 |
+
},
|
2667 |
+
{
|
2668 |
+
"epoch": 0.07094467444486263,
|
2669 |
+
"grad_norm": 1.5231132507324219,
|
2670 |
+
"learning_rate": 6.944314779827749e-06,
|
2671 |
+
"loss": 3.4769,
|
2672 |
+
"step": 377
|
2673 |
+
},
|
2674 |
+
{
|
2675 |
+
"epoch": 0.07113285660519383,
|
2676 |
+
"grad_norm": 1.5165536403656006,
|
2677 |
+
"learning_rate": 6.759741736760061e-06,
|
2678 |
+
"loss": 3.4978,
|
2679 |
+
"step": 378
|
2680 |
+
},
|
2681 |
+
{
|
2682 |
+
"epoch": 0.07132103876552502,
|
2683 |
+
"grad_norm": 1.531844973564148,
|
2684 |
+
"learning_rate": 6.577476764784546e-06,
|
2685 |
+
"loss": 3.3847,
|
2686 |
+
"step": 379
|
2687 |
+
},
|
2688 |
+
{
|
2689 |
+
"epoch": 0.07150922092585622,
|
2690 |
+
"grad_norm": 1.4768074750900269,
|
2691 |
+
"learning_rate": 6.397529592809614e-06,
|
2692 |
+
"loss": 3.1253,
|
2693 |
+
"step": 380
|
2694 |
+
},
|
2695 |
+
{
|
2696 |
+
"epoch": 0.07169740308618743,
|
2697 |
+
"grad_norm": 1.5122772455215454,
|
2698 |
+
"learning_rate": 6.219909826024589e-06,
|
2699 |
+
"loss": 3.3401,
|
2700 |
+
"step": 381
|
2701 |
+
},
|
2702 |
+
{
|
2703 |
+
"epoch": 0.07188558524651863,
|
2704 |
+
"grad_norm": 1.5160380601882935,
|
2705 |
+
"learning_rate": 6.0446269453868945e-06,
|
2706 |
+
"loss": 3.3382,
|
2707 |
+
"step": 382
|
2708 |
+
},
|
2709 |
+
{
|
2710 |
+
"epoch": 0.07207376740684983,
|
2711 |
+
"grad_norm": 1.5226458311080933,
|
2712 |
+
"learning_rate": 5.871690307116107e-06,
|
2713 |
+
"loss": 3.431,
|
2714 |
+
"step": 383
|
2715 |
+
},
|
2716 |
+
{
|
2717 |
+
"epoch": 0.07226194956718103,
|
2718 |
+
"grad_norm": 1.4903466701507568,
|
2719 |
+
"learning_rate": 5.701109142194422e-06,
|
2720 |
+
"loss": 3.2707,
|
2721 |
+
"step": 384
|
2722 |
+
},
|
2723 |
+
{
|
2724 |
+
"epoch": 0.07245013172751223,
|
2725 |
+
"grad_norm": 1.5170961618423462,
|
2726 |
+
"learning_rate": 5.532892555874059e-06,
|
2727 |
+
"loss": 3.3027,
|
2728 |
+
"step": 385
|
2729 |
+
},
|
2730 |
+
{
|
2731 |
+
"epoch": 0.07263831388784343,
|
2732 |
+
"grad_norm": 1.5108839273452759,
|
2733 |
+
"learning_rate": 5.3670495271910925e-06,
|
2734 |
+
"loss": 3.3152,
|
2735 |
+
"step": 386
|
2736 |
+
},
|
2737 |
+
{
|
2738 |
+
"epoch": 0.07282649604817464,
|
2739 |
+
"grad_norm": 1.51710045337677,
|
2740 |
+
"learning_rate": 5.203588908486279e-06,
|
2741 |
+
"loss": 3.2884,
|
2742 |
+
"step": 387
|
2743 |
+
},
|
2744 |
+
{
|
2745 |
+
"epoch": 0.07301467820850584,
|
2746 |
+
"grad_norm": 1.5793352127075195,
|
2747 |
+
"learning_rate": 5.042519424932513e-06,
|
2748 |
+
"loss": 3.5139,
|
2749 |
+
"step": 388
|
2750 |
+
},
|
2751 |
+
{
|
2752 |
+
"epoch": 0.07320286036883704,
|
2753 |
+
"grad_norm": 1.481472373008728,
|
2754 |
+
"learning_rate": 4.883849674069058e-06,
|
2755 |
+
"loss": 3.2427,
|
2756 |
+
"step": 389
|
2757 |
+
},
|
2758 |
+
{
|
2759 |
+
"epoch": 0.07339104252916824,
|
2760 |
+
"grad_norm": 1.5381510257720947,
|
2761 |
+
"learning_rate": 4.727588125342669e-06,
|
2762 |
+
"loss": 3.3485,
|
2763 |
+
"step": 390
|
2764 |
+
},
|
2765 |
+
{
|
2766 |
+
"epoch": 0.07357922468949944,
|
2767 |
+
"grad_norm": 1.508355975151062,
|
2768 |
+
"learning_rate": 4.573743119655516e-06,
|
2769 |
+
"loss": 3.295,
|
2770 |
+
"step": 391
|
2771 |
+
},
|
2772 |
+
{
|
2773 |
+
"epoch": 0.07376740684983063,
|
2774 |
+
"grad_norm": 1.527873158454895,
|
2775 |
+
"learning_rate": 4.422322868919937e-06,
|
2776 |
+
"loss": 3.2918,
|
2777 |
+
"step": 392
|
2778 |
+
},
|
2779 |
+
{
|
2780 |
+
"epoch": 0.07395558901016183,
|
2781 |
+
"grad_norm": 1.5130548477172852,
|
2782 |
+
"learning_rate": 4.273335455620097e-06,
|
2783 |
+
"loss": 3.3859,
|
2784 |
+
"step": 393
|
2785 |
+
},
|
2786 |
+
{
|
2787 |
+
"epoch": 0.07414377117049303,
|
2788 |
+
"grad_norm": 1.5391393899917603,
|
2789 |
+
"learning_rate": 4.126788832380629e-06,
|
2790 |
+
"loss": 3.3711,
|
2791 |
+
"step": 394
|
2792 |
+
},
|
2793 |
+
{
|
2794 |
+
"epoch": 0.07433195333082424,
|
2795 |
+
"grad_norm": 1.5213444232940674,
|
2796 |
+
"learning_rate": 3.982690821542035e-06,
|
2797 |
+
"loss": 3.4224,
|
2798 |
+
"step": 395
|
2799 |
+
},
|
2800 |
+
{
|
2801 |
+
"epoch": 0.07452013549115544,
|
2802 |
+
"grad_norm": 1.4749212265014648,
|
2803 |
+
"learning_rate": 3.8410491147432395e-06,
|
2804 |
+
"loss": 3.1882,
|
2805 |
+
"step": 396
|
2806 |
+
},
|
2807 |
+
{
|
2808 |
+
"epoch": 0.07470831765148664,
|
2809 |
+
"grad_norm": 1.4860862493515015,
|
2810 |
+
"learning_rate": 3.7018712725109926e-06,
|
2811 |
+
"loss": 3.2488,
|
2812 |
+
"step": 397
|
2813 |
+
},
|
2814 |
+
{
|
2815 |
+
"epoch": 0.07489649981181784,
|
2816 |
+
"grad_norm": 1.5500200986862183,
|
2817 |
+
"learning_rate": 3.5651647238562904e-06,
|
2818 |
+
"loss": 3.3369,
|
2819 |
+
"step": 398
|
2820 |
+
},
|
2821 |
+
{
|
2822 |
+
"epoch": 0.07508468197214904,
|
2823 |
+
"grad_norm": 1.8019529581069946,
|
2824 |
+
"learning_rate": 3.430936765877857e-06,
|
2825 |
+
"loss": 3.1557,
|
2826 |
+
"step": 399
|
2827 |
+
},
|
2828 |
+
{
|
2829 |
+
"epoch": 0.07527286413248024,
|
2830 |
+
"grad_norm": 2.5598955154418945,
|
2831 |
+
"learning_rate": 3.299194563372604e-06,
|
2832 |
+
"loss": 3.03,
|
2833 |
+
"step": 400
|
2834 |
+
},
|
2835 |
+
{
|
2836 |
+
"epoch": 0.07546104629281145,
|
2837 |
+
"grad_norm": 1.5204044580459595,
|
2838 |
+
"learning_rate": 3.1699451484532463e-06,
|
2839 |
+
"loss": 3.4419,
|
2840 |
+
"step": 401
|
2841 |
+
},
|
2842 |
+
{
|
2843 |
+
"epoch": 0.07564922845314265,
|
2844 |
+
"grad_norm": 1.4903305768966675,
|
2845 |
+
"learning_rate": 3.0431954201728784e-06,
|
2846 |
+
"loss": 3.3295,
|
2847 |
+
"step": 402
|
2848 |
+
},
|
2849 |
+
{
|
2850 |
+
"epoch": 0.07583741061347385,
|
2851 |
+
"grad_norm": 1.5254848003387451,
|
2852 |
+
"learning_rate": 2.9189521441567726e-06,
|
2853 |
+
"loss": 3.4107,
|
2854 |
+
"step": 403
|
2855 |
+
},
|
2856 |
+
{
|
2857 |
+
"epoch": 0.07602559277380504,
|
2858 |
+
"grad_norm": 1.5176067352294922,
|
2859 |
+
"learning_rate": 2.797221952241219e-06,
|
2860 |
+
"loss": 3.3387,
|
2861 |
+
"step": 404
|
2862 |
+
},
|
2863 |
+
{
|
2864 |
+
"epoch": 0.07621377493413624,
|
2865 |
+
"grad_norm": 1.495973825454712,
|
2866 |
+
"learning_rate": 2.6780113421195298e-06,
|
2867 |
+
"loss": 3.3161,
|
2868 |
+
"step": 405
|
2869 |
+
},
|
2870 |
+
{
|
2871 |
+
"epoch": 0.07640195709446744,
|
2872 |
+
"grad_norm": 1.5124212503433228,
|
2873 |
+
"learning_rate": 2.561326676995218e-06,
|
2874 |
+
"loss": 3.3538,
|
2875 |
+
"step": 406
|
2876 |
+
},
|
2877 |
+
{
|
2878 |
+
"epoch": 0.07659013925479864,
|
2879 |
+
"grad_norm": 1.4950138330459595,
|
2880 |
+
"learning_rate": 2.4471741852423237e-06,
|
2881 |
+
"loss": 3.363,
|
2882 |
+
"step": 407
|
2883 |
+
},
|
2884 |
+
{
|
2885 |
+
"epoch": 0.07677832141512984,
|
2886 |
+
"grad_norm": 1.5359971523284912,
|
2887 |
+
"learning_rate": 2.3355599600729915e-06,
|
2888 |
+
"loss": 3.4671,
|
2889 |
+
"step": 408
|
2890 |
+
},
|
2891 |
+
{
|
2892 |
+
"epoch": 0.07696650357546105,
|
2893 |
+
"grad_norm": 1.4483258724212646,
|
2894 |
+
"learning_rate": 2.2264899592121744e-06,
|
2895 |
+
"loss": 3.1561,
|
2896 |
+
"step": 409
|
2897 |
+
},
|
2898 |
+
{
|
2899 |
+
"epoch": 0.07715468573579225,
|
2900 |
+
"grad_norm": 1.483723521232605,
|
2901 |
+
"learning_rate": 2.1199700045797077e-06,
|
2902 |
+
"loss": 3.3924,
|
2903 |
+
"step": 410
|
2904 |
+
},
|
2905 |
+
{
|
2906 |
+
"epoch": 0.07734286789612345,
|
2907 |
+
"grad_norm": 1.4614593982696533,
|
2908 |
+
"learning_rate": 2.0160057819794466e-06,
|
2909 |
+
"loss": 3.2969,
|
2910 |
+
"step": 411
|
2911 |
+
},
|
2912 |
+
{
|
2913 |
+
"epoch": 0.07753105005645465,
|
2914 |
+
"grad_norm": 1.4894236326217651,
|
2915 |
+
"learning_rate": 1.9146028407958484e-06,
|
2916 |
+
"loss": 3.3376,
|
2917 |
+
"step": 412
|
2918 |
+
},
|
2919 |
+
{
|
2920 |
+
"epoch": 0.07771923221678585,
|
2921 |
+
"grad_norm": 1.501068115234375,
|
2922 |
+
"learning_rate": 1.8157665936977263e-06,
|
2923 |
+
"loss": 3.3462,
|
2924 |
+
"step": 413
|
2925 |
+
},
|
2926 |
+
{
|
2927 |
+
"epoch": 0.07790741437711705,
|
2928 |
+
"grad_norm": 1.4664418697357178,
|
2929 |
+
"learning_rate": 1.7195023163493252e-06,
|
2930 |
+
"loss": 3.1851,
|
2931 |
+
"step": 414
|
2932 |
+
},
|
2933 |
+
{
|
2934 |
+
"epoch": 0.07809559653744826,
|
2935 |
+
"grad_norm": 1.5121281147003174,
|
2936 |
+
"learning_rate": 1.6258151471287396e-06,
|
2937 |
+
"loss": 3.3562,
|
2938 |
+
"step": 415
|
2939 |
+
},
|
2940 |
+
{
|
2941 |
+
"epoch": 0.07828377869777944,
|
2942 |
+
"grad_norm": 1.5005747079849243,
|
2943 |
+
"learning_rate": 1.5347100868536246e-06,
|
2944 |
+
"loss": 3.276,
|
2945 |
+
"step": 416
|
2946 |
+
},
|
2947 |
+
{
|
2948 |
+
"epoch": 0.07847196085811065,
|
2949 |
+
"grad_norm": 1.5016510486602783,
|
2950 |
+
"learning_rate": 1.4461919985142735e-06,
|
2951 |
+
"loss": 3.1685,
|
2952 |
+
"step": 417
|
2953 |
+
},
|
2954 |
+
{
|
2955 |
+
"epoch": 0.07866014301844185,
|
2956 |
+
"grad_norm": 1.490427017211914,
|
2957 |
+
"learning_rate": 1.3602656070140275e-06,
|
2958 |
+
"loss": 3.3398,
|
2959 |
+
"step": 418
|
2960 |
+
},
|
2961 |
+
{
|
2962 |
+
"epoch": 0.07884832517877305,
|
2963 |
+
"grad_norm": 1.5081381797790527,
|
2964 |
+
"learning_rate": 1.27693549891707e-06,
|
2965 |
+
"loss": 3.378,
|
2966 |
+
"step": 419
|
2967 |
+
},
|
2968 |
+
{
|
2969 |
+
"epoch": 0.07903650733910425,
|
2970 |
+
"grad_norm": 1.5135114192962646,
|
2971 |
+
"learning_rate": 1.196206122203647e-06,
|
2972 |
+
"loss": 3.2584,
|
2973 |
+
"step": 420
|
2974 |
+
},
|
2975 |
+
{
|
2976 |
+
"epoch": 0.07922468949943545,
|
2977 |
+
"grad_norm": 1.503239631652832,
|
2978 |
+
"learning_rate": 1.1180817860325599e-06,
|
2979 |
+
"loss": 3.3742,
|
2980 |
+
"step": 421
|
2981 |
+
},
|
2982 |
+
{
|
2983 |
+
"epoch": 0.07941287165976665,
|
2984 |
+
"grad_norm": 1.5126641988754272,
|
2985 |
+
"learning_rate": 1.0425666605112517e-06,
|
2986 |
+
"loss": 3.4263,
|
2987 |
+
"step": 422
|
2988 |
+
},
|
2989 |
+
{
|
2990 |
+
"epoch": 0.07960105382009786,
|
2991 |
+
"grad_norm": 1.5308884382247925,
|
2992 |
+
"learning_rate": 9.696647764731337e-07,
|
2993 |
+
"loss": 3.3315,
|
2994 |
+
"step": 423
|
2995 |
+
},
|
2996 |
+
{
|
2997 |
+
"epoch": 0.07978923598042906,
|
2998 |
+
"grad_norm": 1.5186878442764282,
|
2999 |
+
"learning_rate": 8.993800252624862e-07,
|
3000 |
+
"loss": 3.3336,
|
3001 |
+
"step": 424
|
3002 |
+
},
|
3003 |
+
{
|
3004 |
+
"epoch": 0.07997741814076026,
|
3005 |
+
"grad_norm": 1.5928887128829956,
|
3006 |
+
"learning_rate": 8.317161585266964e-07,
|
3007 |
+
"loss": 3.641,
|
3008 |
+
"step": 425
|
3009 |
+
},
|
3010 |
+
{
|
3011 |
+
"epoch": 0.08016560030109146,
|
3012 |
+
"grad_norm": 1.5158437490463257,
|
3013 |
+
"learning_rate": 7.666767880160464e-07,
|
3014 |
+
"loss": 3.4028,
|
3015 |
+
"step": 426
|
3016 |
+
},
|
3017 |
+
{
|
3018 |
+
"epoch": 0.08035378246142266,
|
3019 |
+
"grad_norm": 1.5123188495635986,
|
3020 |
+
"learning_rate": 7.042653853909064e-07,
|
3021 |
+
"loss": 3.4125,
|
3022 |
+
"step": 427
|
3023 |
+
},
|
3024 |
+
{
|
3025 |
+
"epoch": 0.08054196462175386,
|
3026 |
+
"grad_norm": 1.4982678890228271,
|
3027 |
+
"learning_rate": 6.444852820364222e-07,
|
3028 |
+
"loss": 3.2116,
|
3029 |
+
"step": 428
|
3030 |
+
},
|
3031 |
+
{
|
3032 |
+
"epoch": 0.08073014678208505,
|
3033 |
+
"grad_norm": 1.5639753341674805,
|
3034 |
+
"learning_rate": 5.87339668884701e-07,
|
3035 |
+
"loss": 3.1289,
|
3036 |
+
"step": 429
|
3037 |
+
},
|
3038 |
+
{
|
3039 |
+
"epoch": 0.08091832894241625,
|
3040 |
+
"grad_norm": 1.572089672088623,
|
3041 |
+
"learning_rate": 5.328315962444874e-07,
|
3042 |
+
"loss": 3.3213,
|
3043 |
+
"step": 430
|
3044 |
+
},
|
3045 |
+
{
|
3046 |
+
"epoch": 0.08110651110274746,
|
3047 |
+
"grad_norm": 1.5132265090942383,
|
3048 |
+
"learning_rate": 4.809639736383431e-07,
|
3049 |
+
"loss": 3.3827,
|
3050 |
+
"step": 431
|
3051 |
+
},
|
3052 |
+
{
|
3053 |
+
"epoch": 0.08129469326307866,
|
3054 |
+
"grad_norm": 1.4733517169952393,
|
3055 |
+
"learning_rate": 4.317395696473214e-07,
|
3056 |
+
"loss": 3.2562,
|
3057 |
+
"step": 432
|
3058 |
+
},
|
3059 |
+
{
|
3060 |
+
"epoch": 0.08148287542340986,
|
3061 |
+
"grad_norm": 1.548933982849121,
|
3062 |
+
"learning_rate": 3.851610117632354e-07,
|
3063 |
+
"loss": 3.4471,
|
3064 |
+
"step": 433
|
3065 |
+
},
|
3066 |
+
{
|
3067 |
+
"epoch": 0.08167105758374106,
|
3068 |
+
"grad_norm": 1.4967180490493774,
|
3069 |
+
"learning_rate": 3.4123078624834216e-07,
|
3070 |
+
"loss": 3.0655,
|
3071 |
+
"step": 434
|
3072 |
+
},
|
3073 |
+
{
|
3074 |
+
"epoch": 0.08185923974407226,
|
3075 |
+
"grad_norm": 1.6242047548294067,
|
3076 |
+
"learning_rate": 2.9995123800270476e-07,
|
3077 |
+
"loss": 3.2837,
|
3078 |
+
"step": 435
|
3079 |
+
},
|
3080 |
+
{
|
3081 |
+
"epoch": 0.08204742190440346,
|
3082 |
+
"grad_norm": 1.4935169219970703,
|
3083 |
+
"learning_rate": 2.613245704389644e-07,
|
3084 |
+
"loss": 3.3263,
|
3085 |
+
"step": 436
|
3086 |
+
},
|
3087 |
+
{
|
3088 |
+
"epoch": 0.08223560406473467,
|
3089 |
+
"grad_norm": 1.4887654781341553,
|
3090 |
+
"learning_rate": 2.2535284536476242e-07,
|
3091 |
+
"loss": 3.1296,
|
3092 |
+
"step": 437
|
3093 |
+
},
|
3094 |
+
{
|
3095 |
+
"epoch": 0.08242378622506587,
|
3096 |
+
"grad_norm": 1.511015772819519,
|
3097 |
+
"learning_rate": 1.920379828726726e-07,
|
3098 |
+
"loss": 3.3212,
|
3099 |
+
"step": 438
|
3100 |
+
},
|
3101 |
+
{
|
3102 |
+
"epoch": 0.08261196838539707,
|
3103 |
+
"grad_norm": 1.5536808967590332,
|
3104 |
+
"learning_rate": 1.6138176123770554e-07,
|
3105 |
+
"loss": 3.5203,
|
3106 |
+
"step": 439
|
3107 |
+
},
|
3108 |
+
{
|
3109 |
+
"epoch": 0.08280015054572827,
|
3110 |
+
"grad_norm": 1.5585739612579346,
|
3111 |
+
"learning_rate": 1.333858168224178e-07,
|
3112 |
+
"loss": 3.2727,
|
3113 |
+
"step": 440
|
3114 |
+
},
|
3115 |
+
{
|
3116 |
+
"epoch": 0.08298833270605946,
|
3117 |
+
"grad_norm": 1.5475882291793823,
|
3118 |
+
"learning_rate": 1.0805164398952072e-07,
|
3119 |
+
"loss": 3.4168,
|
3120 |
+
"step": 441
|
3121 |
+
},
|
3122 |
+
{
|
3123 |
+
"epoch": 0.08317651486639066,
|
3124 |
+
"grad_norm": 1.5009870529174805,
|
3125 |
+
"learning_rate": 8.53805950221498e-08,
|
3126 |
+
"loss": 3.3267,
|
3127 |
+
"step": 442
|
3128 |
+
},
|
3129 |
+
{
|
3130 |
+
"epoch": 0.08336469702672186,
|
3131 |
+
"grad_norm": 1.5358216762542725,
|
3132 |
+
"learning_rate": 6.537388005167233e-08,
|
3133 |
+
"loss": 3.4914,
|
3134 |
+
"step": 443
|
3135 |
+
},
|
3136 |
+
{
|
3137 |
+
"epoch": 0.08355287918705306,
|
3138 |
+
"grad_norm": 1.5157291889190674,
|
3139 |
+
"learning_rate": 4.8032566993089225e-08,
|
3140 |
+
"loss": 3.3647,
|
3141 |
+
"step": 444
|
3142 |
+
},
|
3143 |
+
{
|
3144 |
+
"epoch": 0.08374106134738427,
|
3145 |
+
"grad_norm": 1.5044530630111694,
|
3146 |
+
"learning_rate": 3.3357581488030475e-08,
|
3147 |
+
"loss": 3.2696,
|
3148 |
+
"step": 445
|
3149 |
+
},
|
3150 |
+
{
|
3151 |
+
"epoch": 0.08392924350771547,
|
3152 |
+
"grad_norm": 1.5259320735931396,
|
3153 |
+
"learning_rate": 2.134970685536697e-08,
|
3154 |
+
"loss": 3.3305,
|
3155 |
+
"step": 446
|
3156 |
+
},
|
3157 |
+
{
|
3158 |
+
"epoch": 0.08411742566804667,
|
3159 |
+
"grad_norm": 1.4801864624023438,
|
3160 |
+
"learning_rate": 1.200958404936059e-08,
|
3161 |
+
"loss": 3.1884,
|
3162 |
+
"step": 447
|
3163 |
+
},
|
3164 |
+
{
|
3165 |
+
"epoch": 0.08430560782837787,
|
3166 |
+
"grad_norm": 1.4972193241119385,
|
3167 |
+
"learning_rate": 5.337711625497121e-09,
|
3168 |
+
"loss": 3.3929,
|
3169 |
+
"step": 448
|
3170 |
+
},
|
3171 |
+
{
|
3172 |
+
"epoch": 0.08449378998870907,
|
3173 |
+
"grad_norm": 1.7666032314300537,
|
3174 |
+
"learning_rate": 1.3344457138297906e-09,
|
3175 |
+
"loss": 3.4298,
|
3176 |
+
"step": 449
|
3177 |
+
},
|
3178 |
+
{
|
3179 |
+
"epoch": 0.08468197214904027,
|
3180 |
+
"grad_norm": 2.5446274280548096,
|
3181 |
+
"learning_rate": 0.0,
|
3182 |
+
"loss": 2.8081,
|
3183 |
+
"step": 450
|
3184 |
+
},
|
3185 |
+
{
|
3186 |
+
"epoch": 0.08468197214904027,
|
3187 |
+
"eval_loss": 0.8302884697914124,
|
3188 |
+
"eval_runtime": 982.1123,
|
3189 |
+
"eval_samples_per_second": 9.113,
|
3190 |
+
"eval_steps_per_second": 2.279,
|
3191 |
+
"step": 450
|
3192 |
}
|
3193 |
],
|
3194 |
"logging_steps": 1,
|
|
|
3212 |
"should_evaluate": false,
|
3213 |
"should_log": false,
|
3214 |
"should_save": true,
|
3215 |
+
"should_training_stop": true
|
3216 |
},
|
3217 |
"attributes": {}
|
3218 |
}
|
3219 |
},
|
3220 |
+
"total_flos": 1.0272928176694886e+18,
|
3221 |
"train_batch_size": 8,
|
3222 |
"trial_name": null,
|
3223 |
"trial_params": null
|