Qwen3-VL mode working; /unload; normal model loading times

.local/share/pytorch_pod/python-apps/ai-model.py

@@ -18,6 +18,7 @@ from transformers import (
     AutoTokenizer,
     AutoProcessor,
     AutoModel,
+    AutoModelForVision2Seq,
 )
 from transformers.utils.import_utils import is_flash_attn_2_available
 
@@ -177,6 +178,10 @@ class EmbeddingResponse(BaseModel):
     usage: Usage
 
 
+class PreloadRequest(BaseModel):
+    model: str
+
+
 # -----------------------------------------------------------------------------
 # Helpers
 # -----------------------------------------------------------------------------
@@ -248,27 +253,28 @@ def _load_model_locked(model_id: str):
         # Load Generation Model
         # Check if it is a VL model
         if "VL" in model_id:
-            # Attempt to load as VL
+            # Use AutoModelForVision2Seq for VL models
-            # Using AutoModelForVision2Seq or AutoModelForCausalLM
+            # The configuration class Qwen3VLConfig requires Vision2Seq or AutoModel
-            # depending on the specific model support in transformers
             try:
-                from transformers import Qwen2VLForConditionalGeneration
+                print(f"Loading {model_id} with AutoModelForVision2Seq...")
-
+                model = AutoModelForVision2Seq.from_pretrained(
-                model_class = Qwen2VLForConditionalGeneration
-            except ImportError:
-                # Fallback to AutoModel if specific class not available
-                model_class = AutoModelForCausalLM
-
-            # Note: We use AutoModelForCausalLM for broad compatibility.
-            # Qwen2-VL requires Qwen2VLForConditionalGeneration for vision.
-            # We will try AutoModelForCausalLM first.
-
-            model = AutoModelForCausalLM.from_pretrained(
                     model_id,
                     torch_dtype=dtype,
                     device_map=device_map,
                     attn_implementation=attn_impl,
-                trust_remote_code=True,  # Often needed for new architectures
+                    trust_remote_code=True,
+                    low_cpu_mem_usage=True,
+                ).eval()
+            except Exception as e:
+                print(f"Vision2Seq failed: {e}. Fallback to AutoModel...")
+                # Fallback to generic AutoModel if Vision2Seq fails
+                model = AutoModel.from_pretrained(
+                    model_id,
+                    torch_dtype=dtype,
+                    device_map=device_map,
+                    attn_implementation=attn_impl,
+                    trust_remote_code=True,
+                    low_cpu_mem_usage=True,
                 ).eval()
 
             # Processor/Tokenizer
@@ -284,12 +290,14 @@ def _load_model_locked(model_id: str):
             _loaded_model_type = "generation"
         else:
             # Standard Text Model (GPT-OSS)
+            print(f"Loading {model_id} with AutoModelForCausalLM...")
             model = AutoModelForCausalLM.from_pretrained(
                 model_id,
                 torch_dtype=dtype,
                 device_map=device_map,
                 attn_implementation=attn_impl,
                 trust_remote_code=True,
+                low_cpu_mem_usage=True,
             ).eval()
             processor = AutoTokenizer.from_pretrained(
                 model_id, trust_remote_code=True
@@ -367,6 +375,41 @@ def _extract_embeddings(outputs) -> torch.Tensor:
 # -----------------------------------------------------------------------------
 
 
+@app.post("/preload")
+def preload_model(request: PreloadRequest):
+    model_id = request.model.strip()
+    if model_id not in ALLOWED_MODEL_IDS:
+        raise HTTPException(
+            status_code=400,
+            detail=f"Model {model_id} not in allowed models.",
+        )
+
+    with _model_lock:
+        try:
+            _ensure_model_loaded(model_id)
+        except Exception as e:
+            raise HTTPException(
+                status_code=500, detail=f"Failed to load model: {e}"
+            )
+
+    return {
+        "status": "ok",
+        "loaded_model_id": _loaded_model_id,
+        "vram_bytes": _current_vram_info(),
+    }
+
+
+@app.post("/unload")
+def unload_model():
+    with _model_lock:
+        stats = _unload_model_locked()
+    return {
+        "status": "ok",
+        "vram_bytes": _current_vram_info(),
+        "stats": stats,
+    }
+
+
 @app.get("/health")
 def health():
     cuda_ok = bool(torch.cuda.is_available())