eval: load student after freeing teacher, a08 only

kaggle/train_voices.py

@@ -343,9 +343,6 @@ def stage_eval() -> None:
     mm = tts.model
     device = next(mm.parameters()).device
     teacher = mm.flowloss.cfm.model.float().eval()
-    student = copy.deepcopy(teacher)
-    student.load_state_dict(state)
-    student = student.to(device).eval()
 
     asr_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-small",
                         device=0, torch_dtype=torch.float16)
@@ -390,6 +387,14 @@ def stage_eval() -> None:
     del teacher
     torch.cuda.empty_cache()
 
+    # Pass 2: student only. Load it here, not alongside the teacher: two DiTs
+    # plus the 1.48 B backbone do not fit in 15 GB.
+    import copy as _copy
+
+    student = _copy.deepcopy(mm.flowloss.cfm.model.float())
+    student.load_state_dict(state)
+    student = student.to(device).eval()
+
     # Pass 2: student only, reusing the teacher audio for frame count.
     for v in sel:
         latent, spk = tts.prepare_prompt(str(BUNDLE / v["wav"]), use_spk_emb=True)