diff --git a/examples/inference/t5/simple_t5_batch_inference.py b/examples/inference/t5/simple_t5_batch_inference.py
index 3ac8b5dcf86..1aca74b3176 100644
--- a/examples/inference/t5/simple_t5_batch_inference.py
+++ b/examples/inference/t5/simple_t5_batch_inference.py
@@ -59,7 +59,7 @@ def add_text_generate_args(parser):
         metavar='N',
         type=str,
         nargs='+',
-        help='Encoder input prompts with each prompt within quotes and seperated by space',
+        help='Encoder input prompts with each prompt within quotes and separated by space',
     )
     group.add_argument(
         "--max-batch-size", type=int, default=1, help='Max number of prompts to process at once'
diff --git a/megatron/core/datasets/blended_megatron_dataset_config.py b/megatron/core/datasets/blended_megatron_dataset_config.py
index a86efbe4963..76f3ae75857 100644
--- a/megatron/core/datasets/blended_megatron_dataset_config.py
+++ b/megatron/core/datasets/blended_megatron_dataset_config.py
@@ -35,7 +35,7 @@ class BlendedMegatronDatasetConfig:
     """
 
     multiple_validation_sets: Optional[bool] = None
-    """Whether the validation split should be treated as multiple seperate datasets."""
+    """Whether the validation split should be treated as multiple separated datasets."""
 
     full_validation: Optional[bool] = None
     """Whether to run a full epoch of validation each time validation occurs."""
diff --git a/megatron/core/export/trtllm/trtllm_layers.py b/megatron/core/export/trtllm/trtllm_layers.py
index b777fe40809..bc3ba33d92c 100644
--- a/megatron/core/export/trtllm/trtllm_layers.py
+++ b/megatron/core/export/trtllm/trtllm_layers.py
@@ -90,7 +90,7 @@ def rename_input_layer_names_to_trtllm_layer_names(
         map the original layer name to equivalent trtllm layer name and add layer number back.
         CPU Conversion will pass in model state dict without layer numbers
         (i.e decoder.layers.mlp.linear_fc1.weight of shape [num_layers, hidden_dim, 4 * hidden_dim]) .
-        GPU conversion will pass model state dict with each layer seperated
+        GPU conversion will pass model state dict with each layer separated
         (i.e decoder.layers.2.mlp.linear_fc1.weight of shape [hidden_dim, 4 * hidden_dim]).
 
         Args:
diff --git a/megatron/core/transformer/cuda_graphs.py b/megatron/core/transformer/cuda_graphs.py
index dd0dad8eba5..62bcb2724af 100644
--- a/megatron/core/transformer/cuda_graphs.py
+++ b/megatron/core/transformer/cuda_graphs.py
@@ -242,7 +242,7 @@ def _check_supported_type(meta):
     }
     assert meta.type in _SUPPORTED_TYPES or is_dataclass(
         meta.value
-    ), f"Cudagraphs recieved an arg of type {meta.type} which is not supported."
+    ), f"Cudagraphs received an arg of type {meta.type} which is not supported."
 
 
 def _determine_if_first_last_layer_of_this_vp_chunk(base_module):
diff --git a/megatron/inference/utils.py b/megatron/inference/utils.py
index bbf6379b516..a09fa7198e8 100644
--- a/megatron/inference/utils.py
+++ b/megatron/inference/utils.py
@@ -82,7 +82,7 @@ def add_inference_args(parser: ArgumentParser) -> ArgumentParser:
         metavar='N',
         type=str,
         nargs='+',
-        help='Input prompts with each prompt within quotes and seperated by space',
+        help='Input prompts with each prompt within quotes and separated by space',
     )
     group.add_argument(
         "--num-tokens-to-prompt",
diff --git a/megatron/rl/rl_utils.py b/megatron/rl/rl_utils.py
index 3d4f5bf72ee..71730daa323 100644
--- a/megatron/rl/rl_utils.py
+++ b/megatron/rl/rl_utils.py
@@ -513,7 +513,7 @@ def get_environment_rollouts(
                     "Gradient buffers will not be offloaded when training cudagraphs are enabled!")
             optimizer.offload_to_cpu()
              
-    # If we have seperate training and inference models we to refit weights from the training model to the inference model.
+    # If we have separate training and inference models we to refit weights from the training model to the inference model.
     has_separate_inference_model = inference_model is not None
     if has_separate_inference_model:
         # If the separate inference model weights were prefetched to CPU while idle, bring them
diff --git a/tools/run_text_generation_server.py b/tools/run_text_generation_server.py
index 89c1cfa5b86..83a0bbc2369 100644
--- a/tools/run_text_generation_server.py
+++ b/tools/run_text_generation_server.py
@@ -98,7 +98,7 @@ def add_text_generate_args(parser):
         metavar='N',
         type=str,
         nargs='+',
-        help='Input prompts with each prompt within quotes and seperated by space',
+        help='Input prompts with each prompt within quotes and separated by space',
     )
     group.add_argument(
         "--max-batch-size",