File size: 2,731 Bytes
672ecc3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 | import copy
import hashlib
import json
import math
from collections import defaultdict
from pathlib import Path
import torch
from torch import nn
from torch.nn import functional as F
from world import D
REVISION = 'world-update-v3-routed-suite-listwise-20261006-r2'
PUBLIC = 'per_public_test'
GLOBAL = 'whole_training_suite'
class SemanticExecutionWorld(nn.Module):
def __init__(self, public_world):
super().__init__()
if public_world.mode != 'opf':
raise ValueError('Extension requires the native OPF world')
self.public_world = public_world
self.suite_readout = copy.deepcopy(public_world.readout)
old = self.suite_readout[-1]
final = nn.Linear(old.in_features, 7, device=old.weight.device, dtype=old.weight.dtype)
with torch.no_grad():
final.weight[:6].copy_(old.weight)
final.bias[:6].copy_(old.bias)
final.weight[6].zero_()
final.bias[6].zero_()
self.suite_readout[-1] = final
@torch.no_grad()
def score_suite(self, contexts):
self.eval()
encoded = self.public_world.encode(contexts)
(latent, _) = self.public_world.predict_latent(encoded)
output = self.suite_readout(latent)
probabilities = F.softmax(output[:, :5], dim=-1)
partial = torch.sigmoid(output[:, 6])
return dict(scope=GLOBAL, full_suite_pass_probability=probabilities[:, 0], predicted_pass_fraction=partial, utility=probabilities[:, 0] + 0.25 * partial, failure_status_probabilities=probabilities, predicted_mean_normalized_seconds=F.softplus(output[:, 5]), calibration_verified=False, semantic_correctness_guaranteed=False)
@torch.no_grad()
def score_public(self, contexts):
return self.public_world.score(contexts)
@classmethod
def load(cls, path, base_path, device='cpu'):
from artifacts import checkpoint_manifest
from world import ExecutionWorld
(manifest, identity) = checkpoint_manifest(path)
if manifest['mode'] != 'semantic_opf':
raise ValueError('Unexpected model architecture')
world = ExecutionWorld.load(Path(path) / 'public_world', base_path, device=device)
if manifest['arithmetic_precision'] == 'float32':
world.float()
elif manifest['arithmetic_precision'] != 'native_bfloat16_backbone_float32_domain':
raise ValueError('Unknown model precision')
model = cls(world)
state = torch.load(Path(path) / 'suite_readout.pt', map_location=device, weights_only=True)
model.suite_readout.load_state_dict(state, strict=True)
model.identity = identity
return model.to(device).requires_grad_(False).eval()
|