Update app.py

app.py

@@ -521,11 +521,13 @@ def start_training(
     config_path = f"/tmp/tmp_configs/{uuid.uuid4()}_{slugged_lora_name}.yaml"
     with open(config_path, "w") as f:
         yaml.dump(config, f)
-
+   
+    
+    print(config_path)
     # Simulate training
-    job = get_job(config_path)
-    job.run()
-    job.cleanup()
+    # job = get_job(config_path)
+    # job.run()
+    # job.cleanup()
     print(f"[INFO] Starting training with config: {config_path}")
     print(json.dumps(config, indent=2))
     return f"Training started successfully with config: {config_path}"
@@ -632,77 +634,3 @@ def auto_run_lora_from_repo():
 
 
 
-
-def _run_lora_from_repo():
-    try:
-        # Set HF cache path if not already set
-        os.environ["HF_HOME"] = "/tmp/hf_cache"
-        os.makedirs("/tmp/hf_cache", exist_ok=True)
-
-        # Create temporary directory to hold downloaded files
-        local_dir = Path(f"/tmp/{LORA_NAME}-{uuid.uuid4()}")
-        os.makedirs(local_dir, exist_ok=True)
-
-        # Download snapshot from model repo using allow_patterns
-        snapshot_path = snapshot_download(
-            repo_id=REPO_ID,
-            repo_type="model",
-            local_dir=local_dir,
-            local_dir_use_symlinks=False,
-            allow_patterns=[f"{FOLDER_IN_REPO}/*"],  # only that folder
-        )
-
-        # Target subfolder inside the snapshot
-        image_dir = Path(snapshot_path) / FOLDER_IN_REPO
-
-        # Collect all image files (recursively)
-        image_paths = list(image_dir.rglob("*.jpg")) + \
-                      list(image_dir.rglob("*.jpeg")) + \
-                      list(image_dir.rglob("*.png"))
-
-        if not image_paths:
-            return JSONResponse(
-                status_code=400,
-                content={"error": "No images found in the HF repo folder."}
-            )
-
-        # Create auto captions
-        captions = [
-            f"Autogenerated caption for {img.stem} in the {CONCEPT_SENTENCE} [trigger]"
-            for img in image_paths
-        ]
-
-        # Prepare dataset
-        dataset_path = create_dataset(image_paths, *captions)
-
-        # Start training
-        result = start_training(
-            lora_name=LORA_NAME,
-            concept_sentence=CONCEPT_SENTENCE,
-            steps=1000,
-            lr=4e-4,
-            rank=16,
-            model_to_train="dev",
-            low_vram=True,
-            dataset_folder=dataset_path,
-            sample_1=f"A stylized portrait using {CONCEPT_SENTENCE}",
-            sample_2=f"A cat in the {CONCEPT_SENTENCE}",
-            sample_3=f"A selfie processed in {CONCEPT_SENTENCE}",
-            use_more_advanced_options=True,
-            more_advanced_options="""
-training:
-  seed: 42
-  precision: bf16
-  batch_size: 2
-augmentation:
-  flip: true
-  color_jitter: true
-"""
-        )
-
-        return {"message": result}
-
-    except PermissionError as pe:
-        return JSONResponse(status_code=500, content={"error": f"Permission denied: {pe}"})
-    except Exception as e:
-        return JSONResponse(status_code=500, content={"error": str(e)})