diff --git a/README.md b/README.md
index 593eff6c31..60013df938 100644
--- a/README.md
+++ b/README.md
@@ -487,6 +487,9 @@ datasets:
 dataset_prepared_path: data/last_run_prepared
 # push prepared dataset to hub
 push_dataset_to_hub: # repo path
+# The maximum number of processes to use while preprocessing your input dataset. This defaults to `os.cpu_count()`
+# if not set.
+dataset_processes: # defaults to os.cpu_count() if not set
 # push checkpoints to hub
 hub_model_id: # repo path to push finetuned model
 # how to push checkpoints to hub
diff --git a/src/axolotl/utils/config.py b/src/axolotl/utils/config.py
index 5a034ea0f2..a8c41d95b0 100644
--- a/src/axolotl/utils/config.py
+++ b/src/axolotl/utils/config.py
@@ -75,6 +75,8 @@ def normalize_config(cfg):
     else:
         cfg.torch_dtype = torch.float32
 
+    cfg.dataset_processes = cfg.dataset_processes or os.cpu_count()
+
     model_config = load_model_config(cfg)
     cfg.model_config_type = model_config.model_type
 
diff --git a/src/axolotl/utils/trainer.py b/src/axolotl/utils/trainer.py
index 3c75e4ec53..80ee5c8c61 100644
--- a/src/axolotl/utils/trainer.py
+++ b/src/axolotl/utils/trainer.py
@@ -400,19 +400,25 @@ def disable_datasets_caching():
 def process_datasets_for_packing(cfg, train_dataset, eval_dataset, tokenizer):
     drop_long = partial(drop_long_seq, sequence_len=cfg.sequence_len)
     with zero_first(is_main_process()):
-        train_dataset = train_dataset.filter(drop_long, num_proc=os.cpu_count())
+        train_dataset = train_dataset.filter(drop_long, num_proc=cfg.dataset_processes)
         if eval_dataset:
-            eval_dataset = eval_dataset.filter(drop_long, num_proc=os.cpu_count())
+            eval_dataset = eval_dataset.filter(
+                drop_long, num_proc=cfg.dataset_processes
+            )
 
         if cfg.group_by_length:
-            train_dataset = train_dataset.map(add_length, num_proc=os.cpu_count())
+            train_dataset = train_dataset.map(
+                add_length, num_proc=cfg.dataset_processes
+            )
 
         if cfg.sample_packing:
-            train_dataset = train_dataset.map(add_position_ids, num_proc=os.cpu_count())
+            train_dataset = train_dataset.map(
+                add_position_ids, num_proc=cfg.dataset_processes
+            )
             if cfg.eval_sample_packing is not False:
                 if eval_dataset:
                     eval_dataset = eval_dataset.map(
-                        add_position_ids, num_proc=os.cpu_count()
+                        add_position_ids, num_proc=cfg.dataset_processes
                     )
 
         # Phi doesn't want the attention_mask feature when training