diff options
Diffstat (limited to 'ep_run/casc_eq_train.py')
| -rw-r--r-- | ep_run/casc_eq_train.py | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/ep_run/casc_eq_train.py b/ep_run/casc_eq_train.py index 2e221f6..f360df1 100644 --- a/ep_run/casc_eq_train.py +++ b/ep_run/casc_eq_train.py @@ -17,7 +17,7 @@ ap.add_argument('--beta', type=float, default=0.003); ap.add_argument('--seed', ap.add_argument('--K', type=int, default=3) # fb (message-passing) rounds ap.add_argument('--geta', type=float, default=1.0) # fb mixing (1.0 = undamped) ap.add_argument('--save_every', type=int, default=1000); ap.add_argument('--log', type=int, default=100) -ap.add_argument('--wandb', default='ept-cascade') # ON BY DEFAULT (user directive 07-13); --wandb '' to disable +ap.add_argument('--wandb', default='auto') # ON BY DEFAULT; 'auto' = per-regime project (ept-fineweb-72m / ept-tinystories-42m); --wandb '' to disable ap.add_argument('--wandb_run', default='') ap.add_argument('--kmax', type=int, default=8) # adaptive fb rounds cap ap.add_argument('--noguard', action='store_true') # diagnosis: skip only non-finite grads @@ -526,6 +526,8 @@ if DDP: # belt & suspenders on top of identical init seeds: rank0's params are if RANK == 0: print(f'[ddp] world={WORLD} backend={args.ddp_backend} params broadcast; eff batch {args.B}x{WORLD}={args.B*WORLD}', flush=True) wb = None +if args.wandb == 'auto': + args.wandb = 'ept-fineweb-72m' if 'fineweb' in args.data else 'ept-tinystories-42m' if args.wandb and RANK == 0: try: import wandb as _w |
