MiniMax Music3 loading text encoder crash comfyui

[INFO] Found quantization metadata version 1
[INFO] CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
[INFO] Requested to load MiniMaxMusic3TEModel
Windows fatal exception: access violation

Stack (most recent call first):
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\model_patcher.py", line 1999 in load
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\model_patcher.py", line 2150 in partially_load
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\model_management.py", line 818 in model_use_more_vram
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\model_management.py", line 790 in model_load
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\model_management.py", line 1004 in load_models_gpu
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\sd.py", line 462 in load_model
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\sd.py", line 404 in encode_from_tokens
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy\sd.py", line 340 in encode_from_tokens_scheduled
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy_extras\nodes_minimax_music.py", line 37 in execute
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy_api\latest\_io.py", line 1990 in EXECUTE_NORMALIZED
File "M:\ComfyUI\313tor212cu130\ComfyUI\comfy_api\internal\__init__.py", line 149 in wrapped_func
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 306 in process_inputs
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 318 in _async_map_node_over_list
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 344 in get_output_data
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 545 in execute
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 789 in execute_async
File "asyncio\events.py", line 89 in _run
File "asyncio\base_events.py", line 2050 in _run_once
File "asyncio\base_events.py", line 683 in run_forever
File "asyncio\base_events.py", line 712 in run_until_complete
File "asyncio\runners.py", line 118 in run
File "asyncio\runners.py", line 195 in run
File "M:\ComfyUI\313tor212cu130\ComfyUI\execution.py", line 728 in execute
File "M:\ComfyUI\313tor212cu130\ComfyUI\main.py", line 372 in prompt_worker
File "threading.py", line 995 in run
File "threading.py", line 1044 in _bootstrap_inner
File "threading.py", line 1015 in _bootstrap

M:\ComfyUI\313tor212cu130>pause
Press any key to continue . . .

Start at comfy_extras/nodes_minimax_music.py line 37 and trace the MiniMaxMusic3TEModel load through comfy/model_patcher.py lines 1999 and 2150, as shown in the stack trace. Reproduce the Windows CUDA text-encoder load and determine what prevents the access violation; done means the model loads without the fatal crash.

Written by the indexing model from the issue text.