Update app.py

app.py

@@ -25,7 +25,7 @@ import spaces
 import gradio as gr
 
 import pk_workflow as pk
-from h3_dpmpp_2s_ancestral import use_dpmpp_2s_ancestral
+from h3_dpmpp_2s_ancestral import use_dpmpp_2s_ancestral, use_dpmpp_sde_gpu, use_seeds_2
 
 MODEL_REPO = os.environ.get("H3_MODEL_REPO", "MiniMaxAI/MiniMax-H3")
 CONDITIONER_SPACE = os.environ.get("H3_CONDITIONER", "dagloop5/qwen3vl-conditioner")
@@ -113,6 +113,8 @@ SAMPLERS = {
     "dpmpp_2m_sde_gpu": "dpmpp_2m_sde_gpu",
     "dpmpp_3m_sde_gpu": "dpmpp_3m_sde_gpu",
     "dpmpp_2s_ancestral": "dpmpp_2s_ancestral",
+    "dpmpp_sde_gpu": "dpmpp_sde_gpu",
+    "seeds_2": "seeds_2",
 }
 DEFAULT_SAMPLER = "euler"
 
@@ -562,18 +564,20 @@ def _generate(
     started = time.time()
     with pk.use_schedule(PIPE, steps, schedule, video_shift, audio_shift, sampler_name=sampler, seed=int(seed)):
         with use_dpmpp_2s_ancestral(PIPE, int(seed), enabled=(sampler == "dpmpp_2s_ancestral")):
-            state = PIPE(
-                prompt_embeds=prompt_embeds.to("cuda"),
-                text_token_tags=text_token_tags,
-                image=first_frame,
-                last_image=last_frame,
-                height=height,
-                width=width,
-                num_frames=num_frames,
-                num_inference_steps=requested_steps,
-                output_type="pt",
-                generator=torch.Generator("cpu").manual_seed(int(seed)),
-            )
+            with use_dpmpp_sde_gpu(PIPE, int(seed), enabled=(sampler == "dpmpp_sde_gpu")):
+                with use_seeds_2(PIPE, int(seed), enabled=(sampler == "seeds_2")):
+                    state = PIPE(
+                        prompt_embeds=prompt_embeds.to("cuda"),
+                        text_token_tags=text_token_tags,
+                        image=first_frame,
+                        last_image=last_frame,
+                        height=height,
+                        width=width,
+                        num_frames=num_frames,
+                        num_inference_steps=requested_steps,
+                        output_type="pt",
+                        generator=torch.Generator("cpu").manual_seed(int(seed)),
+                    )
             
     denoised = time.time() - started