Upload runs/latent_action_v16/latent_no_aux_s68/train.log with huggingface_hub
Browse files
runs/latent_action_v16/latent_no_aux_s68/train.log
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
`torch_dtype` is deprecated! Use `dtype` instead!
|
| 2 |
+
resumed from checkpoints/v15_teacher_rendered.pt
|
| 3 |
+
pre-packing dataset...
|
| 4 |
+
packed 56000 (skipped 0) trainable=648.4M
|
| 5 |
+
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
|
| 6 |
+
/usr/local/lib/python3.10/dist-packages/torch/utils/checkpoint.py:295: FutureWarning: `torch.cpu.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cpu', args...)` instead.
|
| 7 |
+
with torch.enable_grad(), device_autocast_ctx, torch.cpu.amp.autocast(**ctx.cpu_autocast_kwargs): # type: ignore[attr-defined]
|
| 8 |
+
step 50/1760 loss=0.6968 2.00s/step peak=9.9GB
|
| 9 |
+
step 100/1760 loss=0.6056 1.97s/step peak=10.1GB
|
| 10 |
+
step 150/1760 loss=0.5318 1.91s/step peak=10.1GB
|
| 11 |
+
step 200/1760 loss=0.5168 1.86s/step peak=10.1GB
|
| 12 |
+
step 250/1760 loss=0.5208 1.88s/step peak=10.1GB
|
| 13 |
+
step 300/1760 loss=0.4603 1.85s/step peak=10.1GB
|
| 14 |
+
step 350/1760 loss=0.5051 1.86s/step peak=10.1GB
|
| 15 |
+
step 400/1760 loss=0.4600 1.86s/step peak=10.1GB
|
| 16 |
+
saved runs/latent_action/latent_no_aux_s68/step400.pt
|
| 17 |
+
step 450/1760 loss=0.4918 1.85s/step peak=10.1GB
|
| 18 |
+
step 500/1760 loss=0.4826 1.85s/step peak=10.1GB
|
| 19 |
+
step 550/1760 loss=0.4811 1.85s/step peak=10.1GB
|
| 20 |
+
step 600/1760 loss=0.4749 1.84s/step peak=10.1GB
|
| 21 |
+
step 650/1760 loss=0.4722 1.83s/step peak=10.1GB
|
| 22 |
+
step 700/1760 loss=0.4721 1.82s/step peak=10.1GB
|
| 23 |
+
step 750/1760 loss=0.4463 1.81s/step peak=10.1GB
|
| 24 |
+
step 800/1760 loss=0.4703 1.81s/step peak=10.1GB
|
| 25 |
+
saved runs/latent_action/latent_no_aux_s68/step800.pt
|
| 26 |
+
step 850/1760 loss=0.4631 1.81s/step peak=10.1GB
|
| 27 |
+
step 900/1760 loss=0.4278 1.80s/step peak=10.1GB
|
| 28 |
+
step 950/1760 loss=0.4807 1.79s/step peak=10.1GB
|
| 29 |
+
step 1000/1760 loss=0.4690 1.79s/step peak=10.1GB
|
| 30 |
+
step 1050/1760 loss=0.4512 1.80s/step peak=10.1GB
|
| 31 |
+
step 1100/1760 loss=0.4349 1.80s/step peak=10.1GB
|
| 32 |
+
step 1150/1760 loss=0.4275 1.79s/step peak=10.1GB
|
| 33 |
+
step 1200/1760 loss=0.4576 1.79s/step peak=10.1GB
|
| 34 |
+
saved runs/latent_action/latent_no_aux_s68/step1200.pt
|
| 35 |
+
step 1250/1760 loss=0.4607 1.79s/step peak=10.1GB
|
| 36 |
+
step 1300/1760 loss=0.4543 1.79s/step peak=10.1GB
|
| 37 |
+
step 1350/1760 loss=0.4610 1.79s/step peak=10.1GB
|
| 38 |
+
step 1400/1760 loss=0.4085 1.79s/step peak=10.1GB
|
| 39 |
+
step 1450/1760 loss=0.4479 1.80s/step peak=10.1GB
|
| 40 |
+
step 1500/1760 loss=0.4674 1.79s/step peak=10.1GB
|
| 41 |
+
step 1550/1760 loss=0.4361 1.79s/step peak=10.1GB
|
| 42 |
+
step 1600/1760 loss=0.4162 1.79s/step peak=10.1GB
|
| 43 |
+
saved runs/latent_action/latent_no_aux_s68/step1600.pt
|
| 44 |
+
step 1650/1760 loss=0.4589 1.79s/step peak=10.1GB
|
| 45 |
+
step 1700/1760 loss=0.4222 1.79s/step peak=10.1GB
|
| 46 |
+
step 1750/1760 loss=0.4408 1.79s/step peak=10.1GB
|
| 47 |
+
saved runs/latent_action/latent_no_aux_s68/final.pt
|
| 48 |
+
done
|