diff --git a/.github/workflows/publish.yaml b/.github/workflows/publish.yaml
new file mode 100644
index 0000000000..94543c1a11
--- /dev/null
+++ b/.github/workflows/publish.yaml
@@ -0,0 +1,39 @@
+# To create a release, create a tag and push it to GitHub:
+#git tag -a "v0.0.1-beta" -m "beta version testing"
+#git push --tags
+# https://dev.to/iamtekson/publish-package-to-pypi-and-release-new-version-using-github-actions-108k
+name: Publish LitGPT to PyPI
+
+on:
+  push:
+    tags:
+      - "v*"
+jobs:
+  build-n-publish:
+    name: Build and publish to PyPI
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/litgpt
+    permissions:
+      id-token: write
+
+    steps:
+      - name: Checkout source
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.x"
+
+      - name: Build source and wheel distributions
+        run: |
+          python -m pip install --upgrade build twine
+          python -m build
+          twine check --strict dist/*
+      - name: Publish distribution to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          user: __token__
+          password: ${{ secrets.PYPI_API_TOKEN }}
diff --git a/pyproject.toml b/pyproject.toml
index 40029ae13b..63cc9d6d86 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -37,7 +37,7 @@ all = [
     "sentencepiece>=0.2.0",      # llama-based models
     "tokenizers>=0.15.2",        # pythia, falcon, redpajama
     "requests>=2.31.0",          # litgpt.data
-    "litdata>=0.2.2",            # litgpt.data
+    "litdata>=0.2.2,<0.2.6",     # litgpt.data
     "litserve>=0.1.0",           # litgpt.deploy
     "zstandard>=0.22.0",         # litgpt.data.prepare_slimpajama.py
     "pandas>=1.9.0",             # litgpt.data.prepare_starcoder.py
diff --git a/tests/test_lora.py b/tests/test_lora.py
index d283f1cf44..8f4edba90a 100644
--- a/tests/test_lora.py
+++ b/tests/test_lora.py
@@ -758,3 +758,13 @@ def test_lora_model_fsdp_init():
     model = fabric.setup(model)
     y = model(x)
     assert y.shape == torch.Size([2, 8, 512])
+
+    # verify that all the parameters, buffers and other attributes aren't on `meta` device
+    for m in model.modules():
+        for p_name, parameter in m.named_parameters():
+            assert not parameter.is_meta, f"Parameter `{p_name}` isn't materialized."
+        for b_name, buffer in m._buffers.items():
+            assert not buffer.is_meta, f"Buffer `{b_name}` isn't materialized."
+        for attr_name, attr_value in m.__dict__.items():
+            if isinstance(attr_value, torch.Tensor):
+                assert not attr_value.is_meta, f"Attribute `{attr_name}` isn't materialized."