From 6c3e0a7b7e4b2a1b6488705f874857197b8244f8 Mon Sep 17 00:00:00 2001
From: Matthew Ding <matthew.ding@databricks.com>
Date: Sun, 22 Dec 2024 22:22:03 -0800
Subject: [PATCH] add debugging statement

---
 llmfoundry/data/finetuning/dataloader.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/llmfoundry/data/finetuning/dataloader.py b/llmfoundry/data/finetuning/dataloader.py
index fce694f160..74e2ff1185 100644
--- a/llmfoundry/data/finetuning/dataloader.py
+++ b/llmfoundry/data/finetuning/dataloader.py
@@ -179,6 +179,8 @@ def build_finetuning_dataloader(
         padding/waste rates for different `cfg.dataset.packing_ratio` choices,
         given a starting workload YAML.
     """
+    print(f"🚨🚨🚨 build_finetuning_dataloader was called with the following arguments: tokenizer={tokenizer}, device_batch_size={device_batch_size}, dataset={dataset}, num_workers={num_workers}, drop_last={drop_last}, pin_memory={pin_memory}, prefetch_factor={prefetch_factor}, persistent_workers={persistent_workers}, timeout={timeout}")
+
     dataset_cfg = dataset
     is_streaming = (
         dataset_cfg.get('remote') is not None or