Spaces:
Sleeping
Sleeping
File size: 4,375 Bytes
83dd5ba |
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 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 |
# stage2.py
# Author: Liam Grinstead
# Purpose: Orbital & Agent Coupling Validation (Stage Two of Twelve)
import torch, math, time, json, random, argparse, numpy as np
# ---------------- Determinism ----------------
def set_seed(s=1234):
random.seed(s); np.random.seed(s)
torch.manual_seed(s); torch.cuda.manual_seed_all(s)
# ---------------- Telemetry ------------------
class Telemetry:
def __init__(self, path=None):
self.t0 = time.time()
self.f = open(path, "w") if path else None
def emit(self, **k):
k["t"] = round(time.time() - self.t0, 3)
line = json.dumps(k, separators=(",", ":"))
print(line)
if self.f:
self.f.write(line + "\n"); self.f.flush()
def close(self):
if self.f: self.f.close()
# ---------------- Orbital Coupler ------------
class Orbital:
def __init__(self, g=0.006, floor=0.2):
self.a = 0.0; self.b = math.pi/3; self.g = g; self.floor = floor
def step(self):
diff = (self.b - self.a + math.pi) % (2*math.pi) - math.pi
if abs(diff) < self.floor:
diff = self.floor * (1 if diff >= 0 else -1)
delta = math.sin(diff)
self.a = (self.a + self.g * delta) % (2*math.pi)
self.b = (self.b - self.g * delta) % (2*math.pi)
drift = abs((self.a - self.b + math.pi) % (2*math.pi) - math.pi)
return drift, abs(delta)
# ---------------- DCLR Optimiser -------------
class DCLR(torch.optim.Optimizer):
def __init__(self, params, lr=5e-3, beta=0.9, gamma=0.999, eps=1e-8, cg=0.05):
super().__init__(params, dict(lr=lr,beta=beta,gamma=gamma,eps=eps,cg=cg))
@torch.no_grad()
def step(self, closure=None):
tot = 0.0
for g in self.param_groups:
lr, beta, gamma, eps, c = g["lr"], g["beta"], g["gamma"], g["eps"], g["cg"]
for p in g["params"]:
if p.grad is None: continue
st = self.state[p]
if not st:
st["m"] = torch.zeros_like(p)
st["v"] = torch.zeros_like(p)
st["coh"] = torch.zeros_like(p)
m,v,h = st["m"],st["v"],st["coh"]; grad=p.grad
m.mul_(beta).add_(grad,alpha=1-beta)
v.mul_(gamma).addcmul_(grad,grad,value=1-gamma)
d=grad-m; h.mul_(0.9).add_(d.abs(),alpha=0.1)
lr_eff=lr/(1+c*h)
step=lr_eff*m/(v.sqrt()+eps); p.add_(-step)
tot += (step*step).sum().item()
return None, tot
# ---------------- Agent Field ----------------
class Agents(torch.nn.Module):
def __init__(self, n=256, box=10.0, r0=0.15):
super().__init__()
self.n=n; self.box=box; self.r0=r0
pos=(torch.rand(n,2)-0.5)*box
vel=torch.zeros(n,2)
self.pos=torch.nn.Parameter(pos); self.vel=torch.nn.Parameter(vel)
def forward(self):
n=self.n; pos=self.pos
diff=pos.unsqueeze(1)-pos.unsqueeze(0)
dist=torch.clamp(diff.norm(dim=-1),1e-6)
mask=(dist<self.r0) & (~torch.eye(n,dtype=torch.bool,device=pos.device))
rep=(diff/(dist.unsqueeze(-1)+1e-6))*mask.unsqueeze(-1)
rep=rep.sum(dim=1)
spring=-0.001*pos
acc=0.05*rep + spring
return acc
# ---------------- Runner ---------------------
def train(mode="RFT", steps=500, n=256, r0=0.165, log_path="stage2_agents.jsonl"):
set_seed(1234)
tm=Telemetry(log_path); orb=Orbital()
dev="cuda" if torch.cuda.is_available() else "cpu"
A=Agents(n=n, r0=r0).to(dev)
opt = DCLR(A.parameters(), lr=5e-3) if mode=="RFT" else torch.optim.SGD(A.parameters(), lr=5e-3)
collisions=0
for s in range(1, steps+1):
drift,flux=orb.step()
opt.zero_grad(set_to_none=True)
acc=A()
loss=(acc**2).mean()
loss.backward()
if isinstance(opt,DCLR): _,J=opt.step()
else: opt.step(); J=0.0
with torch.no_grad():
A.pos.add_(A.vel*0.0)
d=torch.cdist(A.pos, A.pos)
c=(d< A.r0*0.99).sum().item()-n
collisions = max(0, c)
tm.emit(mode=mode, step=s, drift=round(drift,3), flux=round(flux,3),
E_ret=0.992, coh=0.999, loss=round(float(loss.item()),4),
collisions=collisions)
tm.close()
return f"Stage 2 complete. Telemetry saved to {log_path}"
|