Upload merge_ministral.py with huggingface_hub

merge_ministral.py

@@ -30,11 +30,13 @@ print(f"LoRA adapter: {LORA_ADAPTER}")
 print(f"Output repo: {OUTPUT_REPO}")
 
 print("\n=== Loading base model ===")
+# Load to CPU first to avoid offloading issues, then move to GPU for merge
 base = Mistral3ForConditionalGeneration.from_pretrained(
     BASE_MODEL,
     torch_dtype=torch.bfloat16,
-    device_map="auto",
-    trust_remote_code=True
+    device_map="cuda:0",  # Single GPU, no offloading
+    trust_remote_code=True,
+    low_cpu_mem_usage=True,
 )
 print(f"Base model loaded: {base.__class__.__name__}")
 
@@ -46,6 +48,10 @@ print("\n=== Merging weights ===")
 merged = model.merge_and_unload()
 print("Merge complete")
 
+print("\n=== Moving to CPU for save ===")
+merged = merged.cpu()
+torch.cuda.empty_cache()
+
 print("\n=== Saving merged model ===")
 merged.save_pretrained("./merged-model", safe_serialization=True)
 print("Model saved locally")