Training in progress, step 160000, checkpoint
Browse files
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:d3fd70c2ee9103dacc2e8460e11a544a0da08573c15a524763aa473249698e80
|
| 3 |
size 14244
|
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": 3.
|
| 6 |
"eval_steps": 5000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2348,6 +2348,162 @@
|
|
| 2348 |
"eval_samples_per_second": 440.711,
|
| 2349 |
"eval_steps_per_second": 13.772,
|
| 2350 |
"step": 150000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2351 |
}
|
| 2352 |
],
|
| 2353 |
"logging_steps": 500,
|
|
@@ -2367,7 +2523,7 @@
|
|
| 2367 |
"attributes": {}
|
| 2368 |
}
|
| 2369 |
},
|
| 2370 |
-
"total_flos": 3.
|
| 2371 |
"train_batch_size": 32,
|
| 2372 |
"trial_name": null,
|
| 2373 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.972983710766786,
|
| 6 |
"eval_steps": 5000,
|
| 7 |
+
"global_step": 160000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2348 |
"eval_samples_per_second": 440.711,
|
| 2349 |
"eval_steps_per_second": 13.772,
|
| 2350 |
"step": 150000
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 3.737087802940008,
|
| 2354 |
+
"grad_norm": NaN,
|
| 2355 |
+
"learning_rate": 0.0004900742091849483,
|
| 2356 |
+
"loss": 0.0,
|
| 2357 |
+
"step": 150500
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 3.749503377036154,
|
| 2361 |
+
"grad_norm": NaN,
|
| 2362 |
+
"learning_rate": 0.0004900742091849483,
|
| 2363 |
+
"loss": 0.0,
|
| 2364 |
+
"step": 151000
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 3.7619189511323006,
|
| 2368 |
+
"grad_norm": NaN,
|
| 2369 |
+
"learning_rate": 0.0004900742091849483,
|
| 2370 |
+
"loss": 0.0,
|
| 2371 |
+
"step": 151500
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 3.7743345252284466,
|
| 2375 |
+
"grad_norm": NaN,
|
| 2376 |
+
"learning_rate": 0.0004900742091849483,
|
| 2377 |
+
"loss": 0.0,
|
| 2378 |
+
"step": 152000
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 3.7867500993245926,
|
| 2382 |
+
"grad_norm": NaN,
|
| 2383 |
+
"learning_rate": 0.0004900742091849483,
|
| 2384 |
+
"loss": 0.0,
|
| 2385 |
+
"step": 152500
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 3.799165673420739,
|
| 2389 |
+
"grad_norm": NaN,
|
| 2390 |
+
"learning_rate": 0.0004900742091849483,
|
| 2391 |
+
"loss": 0.0,
|
| 2392 |
+
"step": 153000
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 3.811581247516885,
|
| 2396 |
+
"grad_norm": NaN,
|
| 2397 |
+
"learning_rate": 0.0004900742091849483,
|
| 2398 |
+
"loss": 0.0,
|
| 2399 |
+
"step": 153500
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 3.823996821613031,
|
| 2403 |
+
"grad_norm": NaN,
|
| 2404 |
+
"learning_rate": 0.0004900742091849483,
|
| 2405 |
+
"loss": 0.0,
|
| 2406 |
+
"step": 154000
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 3.8364123957091776,
|
| 2410 |
+
"grad_norm": NaN,
|
| 2411 |
+
"learning_rate": 0.0004900742091849483,
|
| 2412 |
+
"loss": 0.0,
|
| 2413 |
+
"step": 154500
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 3.848827969805324,
|
| 2417 |
+
"grad_norm": NaN,
|
| 2418 |
+
"learning_rate": 0.0004900742091849483,
|
| 2419 |
+
"loss": 0.0,
|
| 2420 |
+
"step": 155000
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 3.848827969805324,
|
| 2424 |
+
"eval_loss": NaN,
|
| 2425 |
+
"eval_runtime": 2929.1046,
|
| 2426 |
+
"eval_samples_per_second": 439.961,
|
| 2427 |
+
"eval_steps_per_second": 13.749,
|
| 2428 |
+
"step": 155000
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 3.86124354390147,
|
| 2432 |
+
"grad_norm": NaN,
|
| 2433 |
+
"learning_rate": 0.0004900742091849483,
|
| 2434 |
+
"loss": 0.0,
|
| 2435 |
+
"step": 155500
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 3.873659117997616,
|
| 2439 |
+
"grad_norm": NaN,
|
| 2440 |
+
"learning_rate": 0.0004900742091849483,
|
| 2441 |
+
"loss": 0.0,
|
| 2442 |
+
"step": 156000
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 3.8860746920937626,
|
| 2446 |
+
"grad_norm": NaN,
|
| 2447 |
+
"learning_rate": 0.0004900742091849483,
|
| 2448 |
+
"loss": 0.0,
|
| 2449 |
+
"step": 156500
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 3.8984902661899086,
|
| 2453 |
+
"grad_norm": NaN,
|
| 2454 |
+
"learning_rate": 0.0004900742091849483,
|
| 2455 |
+
"loss": 0.0,
|
| 2456 |
+
"step": 157000
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 3.9109058402860546,
|
| 2460 |
+
"grad_norm": NaN,
|
| 2461 |
+
"learning_rate": 0.0004900742091849483,
|
| 2462 |
+
"loss": 0.0,
|
| 2463 |
+
"step": 157500
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 3.923321414382201,
|
| 2467 |
+
"grad_norm": NaN,
|
| 2468 |
+
"learning_rate": 0.0004900742091849483,
|
| 2469 |
+
"loss": 0.0,
|
| 2470 |
+
"step": 158000
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 3.935736988478347,
|
| 2474 |
+
"grad_norm": NaN,
|
| 2475 |
+
"learning_rate": 0.0004900742091849483,
|
| 2476 |
+
"loss": 0.0,
|
| 2477 |
+
"step": 158500
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 3.9481525625744935,
|
| 2481 |
+
"grad_norm": NaN,
|
| 2482 |
+
"learning_rate": 0.0004900742091849483,
|
| 2483 |
+
"loss": 0.0,
|
| 2484 |
+
"step": 159000
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 3.9605681366706396,
|
| 2488 |
+
"grad_norm": NaN,
|
| 2489 |
+
"learning_rate": 0.0004900742091849483,
|
| 2490 |
+
"loss": 0.0,
|
| 2491 |
+
"step": 159500
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 3.972983710766786,
|
| 2495 |
+
"grad_norm": NaN,
|
| 2496 |
+
"learning_rate": 0.0004900742091849483,
|
| 2497 |
+
"loss": 0.0,
|
| 2498 |
+
"step": 160000
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 3.972983710766786,
|
| 2502 |
+
"eval_loss": NaN,
|
| 2503 |
+
"eval_runtime": 2925.5422,
|
| 2504 |
+
"eval_samples_per_second": 440.496,
|
| 2505 |
+
"eval_steps_per_second": 13.766,
|
| 2506 |
+
"step": 160000
|
| 2507 |
}
|
| 2508 |
],
|
| 2509 |
"logging_steps": 500,
|
|
|
|
| 2523 |
"attributes": {}
|
| 2524 |
}
|
| 2525 |
},
|
| 2526 |
+
"total_flos": 3.3978863868862464e+17,
|
| 2527 |
"train_batch_size": 32,
|
| 2528 |
"trial_name": null,
|
| 2529 |
"trial_params": null
|