timesteps, set_timesteps(sigmas=..)

hlky · Nov 23, 2024 · 4e9119f · 4e9119f
1 parent 98a52db
commit 4e9119f
Showing 1 changed file with 17 additions and 3 deletions.
diff --git a/src/diffusers/schedulers/scheduling_euler_discrete.py b/src/diffusers/schedulers/scheduling_euler_discrete.py
@@ -251,12 +251,14 @@ def __init__(
             0 + timestep_offset, num_train_timesteps - 1 + timestep_offset, num_train_timesteps, dtype=float
         )[::-1].copy()
         timesteps = torch.from_numpy(timesteps).to(dtype=torch.float32)
+        print(timesteps)
 
         if use_flow_match:
             sigmas = timesteps / num_train_timesteps
             if not use_dynamic_shifting:
                 # when use_dynamic_shifting is True, we apply the timestep shifting on the fly based on the image resolution
                 sigmas = shift * sigmas / (1 + (shift - 1) * sigmas)
+            print(sigmas)
         else:
             sigmas = (((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5).flip(0)
 
@@ -268,6 +270,7 @@ def __init__(
             self.timesteps = torch.Tensor([0.25 * sigma.log() for sigma in sigmas])
         elif use_flow_match:
             self.timesteps = sigmas * num_train_timesteps
+            print(self.timesteps)
         else:
             self.timesteps = timesteps
 
@@ -407,6 +410,19 @@ def set_timesteps(
             log_sigmas = np.log(np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5))
             sigmas = np.array(sigmas).astype(np.float32)
             timesteps = np.array([self._sigma_to_t(sigma, log_sigmas) for sigma in sigmas[:-1]])
+        elif sigmas is not None and self.config.use_flow_match:
+            sigmas = np.array(sigmas).astype(np.float32)
+            timesteps = sigmas * self.config.num_train_timesteps
+
+            if self.config.use_flow_match:
+                if self.config.use_dynamic_shifting:
+                    sigmas = self.time_shift(mu, 1.0, sigmas)
+                else:
+                    sigmas = self.config.shift * sigmas / (1 + (self.config.shift - 1) * sigmas)
+
+            if self.config.invert_sigmas:
+                sigmas = 1.0 - sigmas
+            timesteps = sigmas * self.config.num_train_timesteps
         elif sigmas is None:
             if timesteps is not None:
                 timesteps = np.array(timesteps).astype(np.float32)
@@ -482,7 +498,7 @@ def set_timesteps(
 
             if self.config.invert_sigmas:
                 sigmas = 1.0 - sigmas
-                timesteps = sigmas * self.config.num_train_timesteps
+            timesteps = sigmas * self.config.num_train_timesteps
 
             if self.config.final_sigmas_type == "sigma_min":
                 sigma_last = ((1 - self.alphas_cumprod[0]) / self.alphas_cumprod[0]) ** 0.5
@@ -502,8 +518,6 @@ def set_timesteps(
         # TODO: Support the full EDM scalings for all prediction types and timestep types
         if self.config.timestep_type == "continuous" and self.config.prediction_type == "v_prediction":
             self.timesteps = torch.Tensor([0.25 * sigma.log() for sigma in sigmas[:-1]]).to(device=device)
-        elif self.config.use_flow_match:
-            self.timesteps = sigmas[:-1] * self.config.num_train_timesteps
         else:
             self.timesteps = torch.from_numpy(timesteps.astype(np.float32)).to(device=device)