Add NaN debug prints to identify source of NaN loss

scripts/training/train_flux_lora.py

@@ -255,6 +255,13 @@ def main():
                 ).input_ids.to(encode_device)
                 encoder_hidden_states = text_encoder_2(text_ids_2)[0]
 
+            # Debug NaN on first step
+            if global_step == resume_step:
+                print(f"    DEBUG latents: min={latents.min():.4f} max={latents.max():.4f} nan={latents.isnan().any()}")
+                print(f"    DEBUG pooled: min={pooled_prompt_embeds.min():.4f} max={pooled_prompt_embeds.max():.4f} nan={pooled_prompt_embeds.isnan().any()}")
+                print(f"    DEBUG hidden: min={encoder_hidden_states.min():.4f} max={encoder_hidden_states.max():.4f} nan={encoder_hidden_states.isnan().any()}")
+                print(f"    DEBUG vae_shift={vae_shift} vae_scale={vae_scale}")
+
             # Move to train device
             latents = latents.to(train_device)
             pooled_prompt_embeds = pooled_prompt_embeds.to(train_device)
@@ -291,6 +298,10 @@ def main():
                 return_dict=False,
             )[0]
 
+            # Debug NaN on first few steps
+            if global_step < resume_step + 3:
+                print(f"    DEBUG step {global_step}: pred nan={model_pred.isnan().any()} target nan={target.isnan().any()} pred range=[{model_pred.min():.4f}, {model_pred.max():.4f}]", flush=True)
+
             loss = F.mse_loss(model_pred.float(), target.float())
             loss = loss / args.gradient_accumulation
             loss.backward()