Training in progress, step 1800, 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 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bee5c2b077b438ffc5771ab6d015bfbf3a620e6c2a965f0327b8cdd07148fb37
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b113d24576292e217d3db735b08200661f3e002afefb75f75c580548c83cb953
|
| 3 |
size 350603
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ec5b01c2a9dc2a433384ff06fff94d1c4ac5418ea070d618f11f44e79cf5478
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a41e0cd8d7ef42bfb03c5fe5412dd1bb05c5d842bf4a52234bf86b85a60a181
|
| 3 |
size 1465
|
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": 150,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2198,6 +2198,444 @@
|
|
| 2198 |
"eval_samples_per_second": 9.621,
|
| 2199 |
"eval_steps_per_second": 1.636,
|
| 2200 |
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2201 |
}
|
| 2202 |
],
|
| 2203 |
"logging_steps": 5,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.36,
|
| 6 |
"eval_steps": 150,
|
| 7 |
+
"global_step": 1800,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2198 |
"eval_samples_per_second": 9.621,
|
| 2199 |
"eval_steps_per_second": 1.636,
|
| 2200 |
"step": 1500
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.301,
|
| 2204 |
+
"grad_norm": 0.07124020904302597,
|
| 2205 |
+
"learning_rate": 3.3506656162420487e-06,
|
| 2206 |
+
"loss": 6.548468780517578,
|
| 2207 |
+
"step": 1505
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.302,
|
| 2211 |
+
"grad_norm": 0.08137217909097672,
|
| 2212 |
+
"learning_rate": 3.345780417082399e-06,
|
| 2213 |
+
"loss": 6.5169837951660154,
|
| 2214 |
+
"step": 1510
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.303,
|
| 2218 |
+
"grad_norm": 0.07528692483901978,
|
| 2219 |
+
"learning_rate": 3.3408805006837212e-06,
|
| 2220 |
+
"loss": 6.5109718322753904,
|
| 2221 |
+
"step": 1515
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.304,
|
| 2225 |
+
"grad_norm": 0.09413464367389679,
|
| 2226 |
+
"learning_rate": 3.335965920630716e-06,
|
| 2227 |
+
"loss": 6.50414810180664,
|
| 2228 |
+
"step": 1520
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.305,
|
| 2232 |
+
"grad_norm": 0.11192154139280319,
|
| 2233 |
+
"learning_rate": 3.3310367306684432e-06,
|
| 2234 |
+
"loss": 6.555927276611328,
|
| 2235 |
+
"step": 1525
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.306,
|
| 2239 |
+
"grad_norm": 0.09222781658172607,
|
| 2240 |
+
"learning_rate": 3.3260929847017352e-06,
|
| 2241 |
+
"loss": 6.508258056640625,
|
| 2242 |
+
"step": 1530
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.307,
|
| 2246 |
+
"grad_norm": 0.0601465106010437,
|
| 2247 |
+
"learning_rate": 3.3211347367946046e-06,
|
| 2248 |
+
"loss": 6.509712219238281,
|
| 2249 |
+
"step": 1535
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.308,
|
| 2253 |
+
"grad_norm": 0.07379843294620514,
|
| 2254 |
+
"learning_rate": 3.3161620411696573e-06,
|
| 2255 |
+
"loss": 6.513530731201172,
|
| 2256 |
+
"step": 1540
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.309,
|
| 2260 |
+
"grad_norm": 0.07197915762662888,
|
| 2261 |
+
"learning_rate": 3.311174952207496e-06,
|
| 2262 |
+
"loss": 6.489860534667969,
|
| 2263 |
+
"step": 1545
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.31,
|
| 2267 |
+
"grad_norm": 0.0705137625336647,
|
| 2268 |
+
"learning_rate": 3.3061735244461258e-06,
|
| 2269 |
+
"loss": 6.491265106201172,
|
| 2270 |
+
"step": 1550
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.311,
|
| 2274 |
+
"grad_norm": 0.062279585748910904,
|
| 2275 |
+
"learning_rate": 3.3011578125803596e-06,
|
| 2276 |
+
"loss": 6.476299285888672,
|
| 2277 |
+
"step": 1555
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.312,
|
| 2281 |
+
"grad_norm": 0.07427375763654709,
|
| 2282 |
+
"learning_rate": 3.2961278714612193e-06,
|
| 2283 |
+
"loss": 6.456025695800781,
|
| 2284 |
+
"step": 1560
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.313,
|
| 2288 |
+
"grad_norm": 0.07860921323299408,
|
| 2289 |
+
"learning_rate": 3.2910837560953345e-06,
|
| 2290 |
+
"loss": 6.443370056152344,
|
| 2291 |
+
"step": 1565
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.314,
|
| 2295 |
+
"grad_norm": 0.0653812438249588,
|
| 2296 |
+
"learning_rate": 3.2860255216443427e-06,
|
| 2297 |
+
"loss": 6.486808013916016,
|
| 2298 |
+
"step": 1570
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.315,
|
| 2302 |
+
"grad_norm": 0.07543846219778061,
|
| 2303 |
+
"learning_rate": 3.2809532234242863e-06,
|
| 2304 |
+
"loss": 6.4591796875,
|
| 2305 |
+
"step": 1575
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.316,
|
| 2309 |
+
"grad_norm": 0.06522826105356216,
|
| 2310 |
+
"learning_rate": 3.275866916905005e-06,
|
| 2311 |
+
"loss": 6.500060272216797,
|
| 2312 |
+
"step": 1580
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.317,
|
| 2316 |
+
"grad_norm": 0.05764804780483246,
|
| 2317 |
+
"learning_rate": 3.2707666577095326e-06,
|
| 2318 |
+
"loss": 6.507598876953125,
|
| 2319 |
+
"step": 1585
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.318,
|
| 2323 |
+
"grad_norm": 0.0758504718542099,
|
| 2324 |
+
"learning_rate": 3.265652501613486e-06,
|
| 2325 |
+
"loss": 6.476622009277344,
|
| 2326 |
+
"step": 1590
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.319,
|
| 2330 |
+
"grad_norm": 0.06581117957830429,
|
| 2331 |
+
"learning_rate": 3.2605245045444574e-06,
|
| 2332 |
+
"loss": 6.449163818359375,
|
| 2333 |
+
"step": 1595
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.32,
|
| 2337 |
+
"grad_norm": 0.2182990163564682,
|
| 2338 |
+
"learning_rate": 3.255382722581401e-06,
|
| 2339 |
+
"loss": 6.435308837890625,
|
| 2340 |
+
"step": 1600
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.321,
|
| 2344 |
+
"grad_norm": 0.07734459638595581,
|
| 2345 |
+
"learning_rate": 3.2502272119540204e-06,
|
| 2346 |
+
"loss": 6.482182312011719,
|
| 2347 |
+
"step": 1605
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.322,
|
| 2351 |
+
"grad_norm": 0.09094393253326416,
|
| 2352 |
+
"learning_rate": 3.245058029042154e-06,
|
| 2353 |
+
"loss": 6.544580078125,
|
| 2354 |
+
"step": 1610
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.323,
|
| 2358 |
+
"grad_norm": 0.08472870290279388,
|
| 2359 |
+
"learning_rate": 3.239875230375158e-06,
|
| 2360 |
+
"loss": 6.462835693359375,
|
| 2361 |
+
"step": 1615
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.324,
|
| 2365 |
+
"grad_norm": 0.08265180140733719,
|
| 2366 |
+
"learning_rate": 3.2346788726312887e-06,
|
| 2367 |
+
"loss": 6.495674133300781,
|
| 2368 |
+
"step": 1620
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.325,
|
| 2372 |
+
"grad_norm": 0.10989709943532944,
|
| 2373 |
+
"learning_rate": 3.2294690126370806e-06,
|
| 2374 |
+
"loss": 6.509843444824218,
|
| 2375 |
+
"step": 1625
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.326,
|
| 2379 |
+
"grad_norm": 0.11295666545629501,
|
| 2380 |
+
"learning_rate": 3.2242457073667287e-06,
|
| 2381 |
+
"loss": 6.541798400878906,
|
| 2382 |
+
"step": 1630
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.327,
|
| 2386 |
+
"grad_norm": 0.08597501367330551,
|
| 2387 |
+
"learning_rate": 3.2190090139414627e-06,
|
| 2388 |
+
"loss": 6.481275177001953,
|
| 2389 |
+
"step": 1635
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.328,
|
| 2393 |
+
"grad_norm": 0.06690355390310287,
|
| 2394 |
+
"learning_rate": 3.213758989628923e-06,
|
| 2395 |
+
"loss": 6.45312271118164,
|
| 2396 |
+
"step": 1640
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.329,
|
| 2400 |
+
"grad_norm": 0.0718226209282875,
|
| 2401 |
+
"learning_rate": 3.2084956918425335e-06,
|
| 2402 |
+
"loss": 6.4438018798828125,
|
| 2403 |
+
"step": 1645
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.33,
|
| 2407 |
+
"grad_norm": 0.11290576308965683,
|
| 2408 |
+
"learning_rate": 3.203219178140877e-06,
|
| 2409 |
+
"loss": 6.5095054626464846,
|
| 2410 |
+
"step": 1650
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.33,
|
| 2414 |
+
"eval_accuracy": 0.14094505494505494,
|
| 2415 |
+
"eval_loss": 6.380516529083252,
|
| 2416 |
+
"eval_runtime": 10.3519,
|
| 2417 |
+
"eval_samples_per_second": 9.66,
|
| 2418 |
+
"eval_steps_per_second": 1.642,
|
| 2419 |
+
"step": 1650
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.331,
|
| 2423 |
+
"grad_norm": 0.06508518010377884,
|
| 2424 |
+
"learning_rate": 3.1979295062270603e-06,
|
| 2425 |
+
"loss": 6.427039337158203,
|
| 2426 |
+
"step": 1655
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.332,
|
| 2430 |
+
"grad_norm": 0.06661232560873032,
|
| 2431 |
+
"learning_rate": 3.1926267339480895e-06,
|
| 2432 |
+
"loss": 6.569775390625,
|
| 2433 |
+
"step": 1660
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.333,
|
| 2437 |
+
"grad_norm": 0.07158288359642029,
|
| 2438 |
+
"learning_rate": 3.1873109192942307e-06,
|
| 2439 |
+
"loss": 6.505144500732422,
|
| 2440 |
+
"step": 1665
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 0.334,
|
| 2444 |
+
"grad_norm": 0.06938688457012177,
|
| 2445 |
+
"learning_rate": 3.1819821203983843e-06,
|
| 2446 |
+
"loss": 6.410604858398438,
|
| 2447 |
+
"step": 1670
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 0.335,
|
| 2451 |
+
"grad_norm": 0.08378569036722183,
|
| 2452 |
+
"learning_rate": 3.176640395535438e-06,
|
| 2453 |
+
"loss": 6.493407440185547,
|
| 2454 |
+
"step": 1675
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 0.336,
|
| 2458 |
+
"grad_norm": 0.08100678771734238,
|
| 2459 |
+
"learning_rate": 3.171285803121641e-06,
|
| 2460 |
+
"loss": 6.4384620666503904,
|
| 2461 |
+
"step": 1680
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.337,
|
| 2465 |
+
"grad_norm": 0.07643549144268036,
|
| 2466 |
+
"learning_rate": 3.165918401713958e-06,
|
| 2467 |
+
"loss": 6.493080139160156,
|
| 2468 |
+
"step": 1685
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.338,
|
| 2472 |
+
"grad_norm": 0.0757737085223198,
|
| 2473 |
+
"learning_rate": 3.16053825000943e-06,
|
| 2474 |
+
"loss": 6.489937591552734,
|
| 2475 |
+
"step": 1690
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 0.339,
|
| 2479 |
+
"grad_norm": 0.0830363929271698,
|
| 2480 |
+
"learning_rate": 3.155145406844535e-06,
|
| 2481 |
+
"loss": 6.443827819824219,
|
| 2482 |
+
"step": 1695
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 0.34,
|
| 2486 |
+
"grad_norm": 0.09771528095006943,
|
| 2487 |
+
"learning_rate": 3.1497399311945412e-06,
|
| 2488 |
+
"loss": 6.492597961425782,
|
| 2489 |
+
"step": 1700
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.341,
|
| 2493 |
+
"grad_norm": 0.10129442065954208,
|
| 2494 |
+
"learning_rate": 3.144321882172865e-06,
|
| 2495 |
+
"loss": 6.537218475341797,
|
| 2496 |
+
"step": 1705
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 0.342,
|
| 2500 |
+
"grad_norm": 0.06285049766302109,
|
| 2501 |
+
"learning_rate": 3.138891319030421e-06,
|
| 2502 |
+
"loss": 6.446696472167969,
|
| 2503 |
+
"step": 1710
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.343,
|
| 2507 |
+
"grad_norm": 0.07542657852172852,
|
| 2508 |
+
"learning_rate": 3.133448301154979e-06,
|
| 2509 |
+
"loss": 6.419207000732422,
|
| 2510 |
+
"step": 1715
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 0.344,
|
| 2514 |
+
"grad_norm": 0.07626835256814957,
|
| 2515 |
+
"learning_rate": 3.1279928880705104e-06,
|
| 2516 |
+
"loss": 6.4702308654785154,
|
| 2517 |
+
"step": 1720
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 0.345,
|
| 2521 |
+
"grad_norm": 0.08218275755643845,
|
| 2522 |
+
"learning_rate": 3.1225251394365383e-06,
|
| 2523 |
+
"loss": 6.416409301757812,
|
| 2524 |
+
"step": 1725
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.346,
|
| 2528 |
+
"grad_norm": 0.07094059139490128,
|
| 2529 |
+
"learning_rate": 3.1170451150474853e-06,
|
| 2530 |
+
"loss": 6.4315032958984375,
|
| 2531 |
+
"step": 1730
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.347,
|
| 2535 |
+
"grad_norm": 0.07319115102291107,
|
| 2536 |
+
"learning_rate": 3.1115528748320205e-06,
|
| 2537 |
+
"loss": 6.402560424804688,
|
| 2538 |
+
"step": 1735
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.348,
|
| 2542 |
+
"grad_norm": 0.10192760825157166,
|
| 2543 |
+
"learning_rate": 3.1060484788524034e-06,
|
| 2544 |
+
"loss": 6.432094573974609,
|
| 2545 |
+
"step": 1740
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.349,
|
| 2549 |
+
"grad_norm": 0.07734037935733795,
|
| 2550 |
+
"learning_rate": 3.100531987303826e-06,
|
| 2551 |
+
"loss": 6.432294464111328,
|
| 2552 |
+
"step": 1745
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 0.35,
|
| 2556 |
+
"grad_norm": 0.07318601757287979,
|
| 2557 |
+
"learning_rate": 3.0950034605137563e-06,
|
| 2558 |
+
"loss": 6.447303009033203,
|
| 2559 |
+
"step": 1750
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 0.351,
|
| 2563 |
+
"grad_norm": 0.06598563492298126,
|
| 2564 |
+
"learning_rate": 3.0894629589412776e-06,
|
| 2565 |
+
"loss": 6.5047248840332035,
|
| 2566 |
+
"step": 1755
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.352,
|
| 2570 |
+
"grad_norm": 0.10212764143943787,
|
| 2571 |
+
"learning_rate": 3.0839105431764273e-06,
|
| 2572 |
+
"loss": 6.521588134765625,
|
| 2573 |
+
"step": 1760
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.353,
|
| 2577 |
+
"grad_norm": 0.0771833136677742,
|
| 2578 |
+
"learning_rate": 3.0783462739395347e-06,
|
| 2579 |
+
"loss": 6.516807556152344,
|
| 2580 |
+
"step": 1765
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.354,
|
| 2584 |
+
"grad_norm": 0.10352446138858795,
|
| 2585 |
+
"learning_rate": 3.072770212080556e-06,
|
| 2586 |
+
"loss": 6.434762573242187,
|
| 2587 |
+
"step": 1770
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.355,
|
| 2591 |
+
"grad_norm": 0.08731052279472351,
|
| 2592 |
+
"learning_rate": 3.067182418578412e-06,
|
| 2593 |
+
"loss": 6.497015380859375,
|
| 2594 |
+
"step": 1775
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.356,
|
| 2598 |
+
"grad_norm": 0.09280135482549667,
|
| 2599 |
+
"learning_rate": 3.0615829545403154e-06,
|
| 2600 |
+
"loss": 6.473223876953125,
|
| 2601 |
+
"step": 1780
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.357,
|
| 2605 |
+
"grad_norm": 0.07267135381698608,
|
| 2606 |
+
"learning_rate": 3.055971881201109e-06,
|
| 2607 |
+
"loss": 6.508274841308594,
|
| 2608 |
+
"step": 1785
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.358,
|
| 2612 |
+
"grad_norm": 0.06757643818855286,
|
| 2613 |
+
"learning_rate": 3.0503492599225916e-06,
|
| 2614 |
+
"loss": 6.476979827880859,
|
| 2615 |
+
"step": 1790
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.359,
|
| 2619 |
+
"grad_norm": 0.06268265843391418,
|
| 2620 |
+
"learning_rate": 3.0447151521928475e-06,
|
| 2621 |
+
"loss": 6.417572784423828,
|
| 2622 |
+
"step": 1795
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.36,
|
| 2626 |
+
"grad_norm": 0.0907832533121109,
|
| 2627 |
+
"learning_rate": 3.0390696196255786e-06,
|
| 2628 |
+
"loss": 6.457040405273437,
|
| 2629 |
+
"step": 1800
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.36,
|
| 2633 |
+
"eval_accuracy": 0.1416092796092796,
|
| 2634 |
+
"eval_loss": 6.360020637512207,
|
| 2635 |
+
"eval_runtime": 10.3268,
|
| 2636 |
+
"eval_samples_per_second": 9.684,
|
| 2637 |
+
"eval_steps_per_second": 1.646,
|
| 2638 |
+
"step": 1800
|
| 2639 |
}
|
| 2640 |
],
|
| 2641 |
"logging_steps": 5,
|