diff --git a/unsloth/__init__.py b/unsloth/__init__.py
index 0105199fba..298ed13399 100644
--- a/unsloth/__init__.py
+++ b/unsloth/__init__.py
@@ -17,17 +17,20 @@ import importlib
 import sys
 from packaging.version import Version
 
-# Define a list of modules to check
-MODULES_TO_CHECK = ["bitsandbytes"]
+# # Define a list of modules to check
+# MODULES_TO_CHECK = ["bitsandbytes"]
 
-# Check if any of the modules in the list have been imported
-for module in MODULES_TO_CHECK:
-    if module in sys.modules:
-        raise ImportError(f"Unsloth: Please import Unsloth before {module}.")
-    pass
-pass
+# # Check if any of the modules in the list have been imported
+# for module in MODULES_TO_CHECK:
+#     if module in sys.modules:
+#         raise ImportError(f"Unsloth: Please import Unsloth before {module}.")
+#     pass
+# pass
 
-# Currently only supports 1 GPU, or else seg faults will occur.    
+# Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so
+# enabling it will require much more work, so we have to prioritize. Please understand!
+# We do have a beta version, which you can contact us about!
+# Thank you for your understanding and we appreciate it immensely!
 if "CUDA_VISIBLE_DEVICES" in os.environ:
     os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
     devices = os.environ["CUDA_VISIBLE_DEVICES"]
@@ -36,6 +39,10 @@ if "CUDA_VISIBLE_DEVICES" in os.environ:
         first_id = devices.split(",")[0]
         warnings.warn(
             f"Unsloth: 'CUDA_VISIBLE_DEVICES' is currently {devices} \n"\
+            "Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so "\
+            "enabling it will require much more work, so we have to prioritize. Please understand!"\
+            "We do have a beta version, which you can contact us about!\n"\
+            "Thank you for your understanding and we appreciate it immensely!\n\n"\
             "Multiple CUDA devices detected but we require a single device.\n"\
             f"We will override CUDA_VISIBLE_DEVICES to first device: {first_id}."
         )
diff --git a/unsloth/models/llama.py b/unsloth/models/llama.py
index 2d8e6a0748..2368a37672 100644
--- a/unsloth/models/llama.py
+++ b/unsloth/models/llama.py
@@ -1165,10 +1165,10 @@ class FastLlamaModel:
                 inner_training_loop = Trainer._original_training_loop
         except:
             raise RuntimeError(
-                "Our OSS was designed for people with few GPU resources to level the playing field.\n"
-                "The OSS Apache 2 license only supports one GPU - please obtain a commercial license.\n"
-                "We're a 2 person team, so we still have to fund our development costs - thanks!\n"
-                "If you don't, please consider at least sponsoring us through Ko-fi! Appreciate it!",
+                'Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so '\
+                'enabling it will require much more work, so we have to prioritize. Please understand!\n'\
+                'We do have a separate beta version, which you can contact us about!\n'\
+                'Thank you for your understanding and we appreciate it immensely!'
             )
         pass
 
@@ -1201,7 +1201,10 @@ class FastLlamaModel:
         output = re.findall(rb'([\\d]{1,})[\\s]{1,}M', output)
         output = sum(int(x.decode('utf-8'))/1024 > 4 for x in output)
         if output > 1: raise RuntimeError(
-            'Error: More than 1 GPUs have a lot of VRAM usage. Please obtain a commercial license.')
+            'Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so '\\
+            'enabling it will require much more work, so we have to prioritize. Please understand!\\n'\\
+            'We do have a separate beta version, which you can contact us about!\\n'\\
+            'Thank you for your understanding and we appreciate it immensely!')
         for _ in range(3):
             gc.collect()
             torch.cuda.empty_cache()"""
@@ -1214,10 +1217,10 @@ class FastLlamaModel:
             args.gradient_accumulation_steps // self._train_batch_size
         if n_total_devices > 1:
             logger.warning_once(
-                "* Our OSS was designed for people with few GPU resources to level the playing field.\\n"
-                "* The OSS Apache 2 license only supports one GPU - please obtain a commercial license.\\n"
-                "* We're a 2 person team, so we still have to fund our development costs - thanks!\\n"
-                "* If you don't, please consider at least sponsoring us through Ko-fi! Appreciate it!",
+                '* Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so ' \\
+                '* enabling it will require much more work, so we have to prioritize. Please understand!\\n' \\
+                '* We do have a separate beta version, which you can contact us about!\\n'\\
+                '* Thank you for your understanding and we appreciate it immensely!'
             )
         debug_info ="""
         debug_info = debug_info.split('\n')
@@ -1244,10 +1247,10 @@ class FastLlamaModel:
         n_total_devices = total_batches // ga // bsz
         if n_total_devices > 1:
             logger.warning_once(
-                "* Our OSS was designed for people with few GPU resources to level the playing field.\\n"
-                "* The OSS Apache 2 license only supports one GPU - please obtain a commercial license.\\n"
-                "* We're a 2 person team, so we still have to fund our development costs - thanks!\\n"
-                "* If you don't, please consider at least sponsoring us through Ko-fi! Appreciate it!",
+                '* Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so ' \\
+                '* enabling it will require much more work, so we have to prioritize. Please understand!\\n' \\
+                '* We do have a separate beta version, which you can contact us about!\\n'\\
+                '* Thank you for your understanding and we appreciate it immensely!'
             )
             divisor = n_total_devices / 1
             bsz = self._train_batch_size = max(int(bsz / divisor), 1)
@@ -1273,10 +1276,10 @@ class FastLlamaModel:
         )
         if "n_total_devices >" not in inner_training_loop:
             raise RuntimeError(
-                "Our OSS was designed for people with few GPU resources to level the playing field.\n"
-                "The OSS Apache 2 license only supports one GPU - please obtain a commercial license.\n"
-                "We're a 2 person team, so we still have to fund our development costs - thanks!\n"
-                "If you don't, please consider at least sponsoring us through Ko-fi! Appreciate it!",
+                'Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so '\
+                'enabling it will require much more work, so we have to prioritize. Please understand!\n'\
+                'We do have a separate beta version, which you can contact us about!\n'\
+                'Thank you for your understanding and we appreciate it immensely!'
             )
         pass
         inner_training_loop = inner_training_loop.replace(
@@ -1783,10 +1786,10 @@ class FastLlamaModel:
         from transformers.trainer import Trainer 
         if Trainer._inner_training_loop.__name__ != "_fast_inner_training_loop":
             raise RuntimeError(
-                "Our OSS was designed for people with few GPU resources to level the playing field.\n"
-                "The OSS Apache 2 license only supports one GPU - please obtain a commercial license.\n"
-                "We're a 2 person team, so we still have to fund our development costs - thanks!\n"
-                "If you don't, please consider at least sponsoring us through Ko-fi! Appreciate it!",
+                'Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so '\
+                'enabling it will require much more work, so we have to prioritize. Please understand!\n'\
+                'We do have a separate beta version, which you can contact us about!\n'\
+                'Thank you for your understanding and we appreciate it immensely!'
             )
         pass
 
diff --git a/unsloth/tokenizer_utils.py b/unsloth/tokenizer_utils.py
index fe2dc06c44..50b09275aa 100644
--- a/unsloth/tokenizer_utils.py
+++ b/unsloth/tokenizer_utils.py
@@ -954,7 +954,7 @@ def patch_sft_trainer_tokenizer():
     "\n"\
     "if self._inner_training_loop.__name__ != '_fast_inner_training_loop':\n"\
     "    raise RuntimeError(\n"\
-    "       'Do not edit specific areas of the Unsloth codebase or you will get CUDA segfaults.'\n"\
+    "       'Please do not edit specific areas of the Unsloth codebase or you will get CUDA segfaults.'\n"\
     "    )\n"\
     "pass\n"\
     "n_devices = torch.cuda.device_count()\n"\
@@ -964,7 +964,10 @@ def patch_sft_trainer_tokenizer():
     "output = re.findall(rb'([\\d]{1,})[\\s]{1,}M', output)\n"\
     "output = sum(int(x.decode('utf-8'))/1024 > 4 for x in output)\n"\
     "if output > 1: raise RuntimeError(\n"\
-    "    'Error: More than 1 GPUs have a lot of VRAM usage. Please obtain a commercial license.')\n"\
+    "    'Unsloth currently does not work on multi GPU setups - sadly we are a 2 brother team so '\\\n"\
+    "    'enabling it will require much more work, so we have to prioritize. Please understand!\\n'\\\n"\
+    "    'We do have a separate beta version, which you can contact us about!\\n'\\\n"\
+    "    'Thank you for your understanding and we appreciate it immensely!')\n"\
     "for _ in range(3):\n"\
     "    gc.collect()\n"\
     "    torch.cuda.empty_cache()\n"\