chore(loop): 6 more iters — plateau at ~0.34-0.37 (1/6 kept)
CD / Lint / Test / Vet (push) Failing after 4s
CD / Build & Import (push) Has been skipped
CD / Deploy via GitOps (push) Has been skipped

Toy encoder near ceiling. 1 kept (val_vol_r2 0.3032→0.3442), 5 reverts.
Consistent plateau = time to swap in TS-JEPA backbone (#3/#5).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-06-24 07:37:29 +02:00
co-authored by Claude Sonnet 4.6
parent ed85dc4a8c
commit f5ce8d6706
3 changed files with 11 additions and 5 deletions
+6
View File
@@ -5,3 +5,9 @@
| 1 | 0.3749 | +0.0928 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter1 | | 1 | 0.3749 | +0.0928 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter1 |
| 1 | 0.3011 | +0.0776 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter1 | | 1 | 0.3011 | +0.0776 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter1 |
| 2 | 0.3032 | +0.0021 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=35°C | iter2 | | 2 | 0.3032 | +0.0021 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=35°C | iter2 |
| 1 | 0.2759 | -0.0273 | revert | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter1 |
| 2 | 0.3442 | +0.0410 | KEEP | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter2 |
| 3 | 0.3371 | -0.0071 | revert | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter3 |
| 4 | 0.3143 | -0.0299 | revert | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter4 |
| 5 | 0.3355 | -0.0087 | revert | 2s | gpu=0% vram=10054/12227MiB temp=34°C | iter5 |
| 6 | 0.2377 | -0.1065 | revert | 2s | gpu=0% vram=10054/12227MiB temp=35°C | iter6 |
+1 -1
View File
@@ -1,5 +1,5 @@
{ {
"val_vol_r2": 0.30321519081159654, "val_vol_r2": 0.23767155122897032,
"n_test": 275, "n_test": 275,
"knobs": { "knobs": {
"WINDOW": 20, "WINDOW": 20,
+4 -4
View File
@@ -47,10 +47,10 @@ class Encoder(nn.Module):
super().__init__() super().__init__()
self.net = nn.Sequential( self.net = nn.Sequential(
nn.Flatten(), nn.Flatten(),
nn.Linear(win * 2, 128), nn.Linear(win * 2, 256),
nn.LayerNorm(128), nn.LayerNorm(256),
nn.GELU(), nn.GELU(),
nn.Linear(128, emb) nn.Linear(256, emb)
) )
def forward(self, x): def forward(self, x):
@@ -61,7 +61,7 @@ def main():
(Xtr, ytr), (Xte, yte) = build() (Xtr, ytr), (Xte, yte) = build()
Xtr_t = torch.tensor(Xtr, device=dev) Xtr_t = torch.tensor(Xtr, device=dev)
enc = Encoder(WINDOW, EMBED_DIM).to(dev) enc = Encoder(WINDOW, EMBED_DIM).to(dev)
dec = nn.Sequential(nn.Linear(EMBED_DIM, 128), nn.GELU(), nn.Linear(128, WINDOW * 2)).to(dev) dec = nn.Sequential(nn.Linear(EMBED_DIM, 256), nn.GELU(), nn.Linear(256, WINDOW * 2)).to(dev)
opt = torch.optim.Adam(list(enc.parameters()) + list(dec.parameters()), lr=LR) opt = torch.optim.Adam(list(enc.parameters()) + list(dec.parameters()), lr=LR)
for _ in range(EPOCHS): # SSL: masked reconstruction of the window for _ in range(EPOCHS): # SSL: masked reconstruction of the window