Check length of gpu_split in model_init

exllamav2/model_init.py

@@ -1,5 +1,6 @@
 
 import argparse, sys, os, glob, time
+import torch
 
 from exllamav2 import(
     ExLlamaV2,
@@ -167,6 +168,9 @@ def post_init_load(
     split = None
     if args.gpu_split and args.gpu_split != "auto":
         split = [float(alloc) for alloc in args.gpu_split.split(",")]
+        if len(split) > torch.cuda.device_count():
+            print(f" ## Error: Too many entries in gpu_split. {torch.cuda.device_count()} CUDA devices are available.")
+            sys.exit()
 
     if args.tensor_parallel:
         if args.gpu_split == "auto": split = None