MCPcopy Create free account
hub / github.com/ReinFlow/ReinFlow / p_losses

Method p_losses

model/diffusion/diffusion.py:349–374  ·  view source on GitHub ↗

If predicting epsilon: E_{t, x0, ε} [||ε - ε_θ(√α̅ₜx0 + √(1-α̅ₜ)ε, t)||² Args: x_start: (batch_size, horizon_steps, action_dim) cond: dict with keys as step and value as observation t: batch of integers

(
        self,
        x_start,
        cond: dict,
        t,
    )

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 1

lossMethod · 0.95

Calls 1

q_sampleMethod · 0.95

Tested by

no test coverage detected