diff --git a/src/autotrain/__init__.py b/src/autotrain/__init__.py
index 5e21f2a6c0..15bf731d86 100644
--- a/src/autotrain/__init__.py
+++ b/src/autotrain/__init__.py
@@ -30,4 +30,4 @@
 warnings.filterwarnings("ignore", category=UserWarning, module="tensorflow")
 
 
-__version__ = "0.6.35.dev0"
+__version__ = "0.6.36.dev0"
diff --git a/src/autotrain/cli/run_llm.py b/src/autotrain/cli/run_llm.py
index 2198afc704..a1306b9900 100644
--- a/src/autotrain/cli/run_llm.py
+++ b/src/autotrain/cli/run_llm.py
@@ -429,10 +429,16 @@ def __init__(self, args):
                     break
                 print(f"Bot: {tgi.chat(prompt)}")
 
-        if not torch.cuda.is_available():
-            raise ValueError("No GPU found. Please install CUDA and try again.")
+        cuda_available = torch.cuda.is_available()
+        mps_available = torch.backends.mps.is_available()
 
-        self.num_gpus = torch.cuda.device_count()
+        if not cuda_available and not mps_available:
+            raise ValueError("No GPU/MPS device found. LLM training requires an accelerator")
+
+        if cuda_available:
+            self.num_gpus = torch.cuda.device_count()
+        elif mps_available:
+            self.num_gpus = 1
 
     def run(self):
         from autotrain.backend import EndpointsRunner, SpaceRunner
diff --git a/src/autotrain/trainers/clm/__main__.py b/src/autotrain/trainers/clm/__main__.py
index 406ed3b08d..72379d337f 100644
--- a/src/autotrain/trainers/clm/__main__.py
+++ b/src/autotrain/trainers/clm/__main__.py
@@ -134,8 +134,10 @@ def train(config):
                 bnb_4bit_compute_dtype=torch.float16,
                 bnb_4bit_use_double_quant=False,
             )
+            config.fp16 = True
         elif config.use_int8:
             bnb_config = BitsAndBytesConfig(load_in_8bit=config.use_int8)
+            config.fp16 = True
         else:
             bnb_config = None