enable studio for intel GPU

studio/backend/core/inference/inference.py

@@ -1642,8 +1642,31 @@ def _generate_dac(
             + text
             + "<|text_end|>\n<|audio_start|><|global_features_start|>\n"
         )
+        import contextlib
+
         with torch.inference_mode():
-            with torch.amp.autocast("cuda", dtype = model.dtype):
+            # Derive the autocast device from the loaded model, not from the
+            # global backend: a CPU-fallback DAC on an XPU/CUDA host must not
+            # open a GPU autocast context around CPU tensors.
+            device_type = (
+                model.device.type
+                if hasattr(model.device, "type")
+                else str(model.device).split(":", 1)[0]
+            )
+            # Clamp to autocast-supported backends so exotic devices
+            # (e.g. "meta" during accelerate offloaded loading) do not raise.
+            # MPS is autocast-supported since torch 2.3, keep it in the set.
+            if device_type not in ("cuda", "xpu", "mps", "cpu"):
+                device_type = "cpu"
+            # CPU and XPU autocast only accept bfloat16/float16. For a
+            # float32 model, skip autocast entirely to avoid raising or
+            # producing a warning on every generate call.
+            autocast_dtype_supported = model.dtype in (torch.bfloat16, torch.float16)
+            if device_type in ("cpu", "xpu") and not autocast_dtype_supported:
+                autocast_ctx = contextlib.nullcontext()
+            else:
+                autocast_ctx = torch.amp.autocast(device_type, dtype = model.dtype)
+            with autocast_ctx:
                 inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)
                 generated = model.generate(
                     **inputs,

studio/backend/core/inference/llama_cpp.py

@@ -1438,8 +1438,9 @@ def unload_model(self) -> bool:
                 LlamaCppBackend._codec_mgr = None
                 import torch
 
-                if torch.cuda.is_available():
-                    torch.cuda.empty_cache()
+                from utils.hardware import clear_gpu_cache
+
+                clear_gpu_cache()
             return True
 
     def _kill_process(self):
@@ -3016,7 +3017,9 @@ def init_audio_codec(self, audio_type: str) -> None:
         if LlamaCppBackend._codec_mgr is None:
             LlamaCppBackend._codec_mgr = AudioCodecManager()
 
-        device = "cuda" if torch.cuda.is_available() else "cpu"
+        from utils.hardware import get_torch_device_str
+
+        device = get_torch_device_str()
         model_repo_path = None
 
         # BiCodec needs a repo with BiCodec/ weights — download canonical SparkTTS
@@ -3090,7 +3093,9 @@ def generate_audio_response(
 
         import torch
 
-        device = "cuda" if torch.cuda.is_available() else "cpu"
+        from utils.hardware import get_torch_device_str
+
+        device = get_torch_device_str()
         return LlamaCppBackend._codec_mgr.decode(
             audio_type, device, token_ids = token_ids, text = data.get("content", "")
         )

studio/backend/core/training/trainer.py

@@ -1532,7 +1532,9 @@ def _preprocess_snac_dataset(self, dataset, custom_format_mapping = None):
 
         SNAC_MODEL_NAME = "hubertsiuzdak/snac_24khz"
         SNAC_SAMPLE_RATE = 24000
-        device = "cuda" if torch.cuda.is_available() else "cpu"
+        from utils.hardware import get_torch_device_str
+
+        device = get_torch_device_str()
         max_length = self.max_seq_length or 2048
         tokenizer = self.tokenizer
 
@@ -1708,7 +1710,9 @@ def _preprocess_snac_dataset(self, dataset, custom_format_mapping = None):
         import gc
 
         gc.collect()
-        torch.cuda.empty_cache()
+        from utils.hardware import clear_gpu_cache
+
+        clear_gpu_cache()
         self._cuda_audio_used = True
 
         if not processed_examples:
@@ -1736,7 +1740,9 @@ def _preprocess_bicodec_dataset(self, dataset, custom_format_mapping = None):
 
         import subprocess
 
-        device = "cuda" if torch.cuda.is_available() else "cpu"
+        from utils.hardware import get_torch_device_str
+
+        device = get_torch_device_str()
 
         # The sparktts Python package lives in the SparkAudio/Spark-TTS GitHub repo,
         # NOT in the unsloth/Spark-TTS-0.5B HF model repo. Clone it if needed.
@@ -1936,7 +1942,9 @@ def extract_wav2vec2_features(wavs: torch.Tensor) -> torch.Tensor:
         import gc
 
         gc.collect()
-        torch.cuda.empty_cache()
+        from utils.hardware import clear_gpu_cache
+
+        clear_gpu_cache()
         self._cuda_audio_used = True
 
         if not processed_examples:
@@ -1971,7 +1979,9 @@ def _preprocess_dac_dataset(self, dataset, custom_format_mapping = None):
         from datasets import Dataset as HFDataset
         from utils.paths import ensure_dir, tmp_root
 
-        device = "cuda" if torch.cuda.is_available() else "cpu"
+        from utils.hardware import get_torch_device_str
+
+        device = get_torch_device_str()
 
         # Clone OuteTTS repo (same as audio_codecs._load_dac)
         import subprocess
@@ -2149,7 +2159,9 @@ def _preprocess_dac_dataset(self, dataset, custom_format_mapping = None):
         import gc
 
         gc.collect()
-        torch.cuda.empty_cache()
+        from utils.hardware import clear_gpu_cache
+
+        clear_gpu_cache()
         self._cuda_audio_used = True
 
         if not processed_examples:

studio/backend/utils/hardware/__init__.py

@@ -28,6 +28,7 @@
     estimate_required_model_memory_gb,
     auto_select_gpu_ids,
     prepare_gpu_selection,
+    get_torch_device_str,
     safe_num_proc,
     safe_thread_num_proc,
     dataset_map_num_proc,
@@ -68,6 +69,7 @@
     "estimate_required_model_memory_gb",
     "auto_select_gpu_ids",
     "prepare_gpu_selection",
+    "get_torch_device_str",
     "safe_num_proc",
     "safe_thread_num_proc",
     "dataset_map_num_proc",