Add LTX-2 third-party license notices for legal compliance (#1226)

LTX-2 (`ltx-core`, `ltx-pipelines`, `ltx-trainer`) is a third-party
dependency developed and provided by Lightricks. It is governed by the
[LTX Community License
Agreement](https://github.com/Lightricks/LTX-2/blob/main/LICENSE),
**not** the Apache 2.0 license that covers NVIDIA Model Optimizer. Per
legal guidance, all integration points must clearly surface this to
users.

- Add `[!WARNING]` license notice blocks at the top of all LTX-2-related
READMEs (`examples/diffusers`, `examples/diffusers/distillation`,
`examples/windows/diffusers/qad_example`)
- Add `warnings.warn(UserWarning)` at every LTX package import site in
Python files, covering both top-level and lazy imports:
  - `examples/diffusers/distillation/distillation_trainer.py`
  - `examples/diffusers/quantization/calibration.py`
  - `examples/diffusers/quantization/pipeline_manager.py`
-
`examples/windows/diffusers/qad_example/sample_example_qad_diffusers.py`
  - `modelopt/torch/export/diffusers_utils.py`
  - `modelopt/torch/quantization/plugins/diffusion/ltx2.py`
- Add license notice comment to `requirements.txt` files that list LTX
packages, so the obligation is visible at install time
- Update `.github/CODEOWNERS` so all `requirements*.txt` files (covering
variants like `requirements-dev.txt`) are owned by
`@NVIDIA/modelopt-setup-codeowners` regardless of location, via a
last-match-wins rule

**Design notes:**
- For library files (`diffusers_utils.py`, `ltx2.py`), the warning is
placed at the lazy import site inside functions — it fires only when
LTX-2 code paths are actually invoked, not at module import time, to
avoid polluting non-LTX users
- For example entry-point scripts that are LTX-2-only, the warning fires
at module load time (after all imports, to satisfy ruff E402)

- [ ] Confirm `pre-commit run --all-files` passes (ruff, mypy,
markdownlint, bandit all clean)
- [ ] Verify warning appears at runtime when running an LTX-2
quantization or distillation example
- [ ] Confirm non-LTX code paths (FLUX, SDXL, SD3) do not emit the
warning

🤖 Generated with [Claude Code](https://claude.com/claude-code)

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

* **Documentation**
* Added third-party license notices across documentation and
requirements files clarifying LTX-2 packages are governed by the LTX
Community License Agreement rather than NVIDIA Model Optimizer's Apache
2.0 license.

* **Chores**
  * Updated code ownership configuration for requirements files.
* Added runtime warnings to notify when LTX-2 dependencies are accessed.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>

Author: kevalmorabia97

.github/CODEOWNERS

@@ -54,3 +54,6 @@ modelopt/torch/utils @NVIDIA/modelopt-torch-utils-codeowners
 /examples/vlm_ptq @NVIDIA/modelopt-examples-vlm-codeowners
 /examples/vllm_serve @NVIDIA/modelopt-examples-llm_ptq-codeowners
 /examples/windows @NVIDIA/modelopt-windows-codeowners
+
+# Requirements files are owned by the setup team regardless of location
+requirements*.txt @NVIDIA/modelopt-setup-codeowners

examples/diffusers/README.md

@@ -118,6 +118,20 @@ python quantize.py \
 
 #### [LTX-2](https://github.com/Lightricks/LTX-2) FP4
 
+> [!WARNING]
+> **Third-Party License Notice — LTX-2**
+>
+> LTX-2 is a third-party model and set of packages developed and provided by Lightricks. LTX-2
+> is **not** covered by the Apache 2.0 license that governs NVIDIA Model Optimizer.
+>
+> By installing and using LTX-2 packages (`ltx-core`, `ltx-pipelines`, `ltx-trainer`) with
+> NVIDIA Model Optimizer, you **must** comply with the
+> [LTX Community License Agreement](https://github.com/Lightricks/LTX-2/blob/main/LICENSE).
+>
+> Any derivative models or fine-tuned weights produced from LTX-2 using NVIDIA Model Optimizer
+> (including quantized or distilled checkpoints) remain subject to the LTX Community License
+> Agreement and are **not** covered by Apache 2.0.
+
 This example produces three outputs: a PyTorch checkpoint (`--quantized-torch-ckpt-save-path`), a Hugging Face checkpoint (`--hf-ckpt-dir`), and a ComfyUI-compatible merged safetensor (`--extra-param merged_base_safetensor_path`).
 
 ```sh

examples/diffusers/distillation/README.md

@@ -1,5 +1,19 @@
 # LTX-2 Distillation Training with ModelOpt
 
+> [!WARNING]
+> **Third-Party License Notice — LTX-2**
+>
+> LTX-2 is a third-party model and set of packages developed and provided by Lightricks. LTX-2
+> is **not** covered by the Apache 2.0 license that governs NVIDIA Model Optimizer.
+>
+> By installing and using LTX-2 packages (`ltx-core`, `ltx-pipelines`, `ltx-trainer`) with
+> NVIDIA Model Optimizer, you **must** comply with the
+> [LTX Community License Agreement](https://github.com/Lightricks/LTX-2/blob/main/LICENSE).
+>
+> Any derivative models or fine-tuned weights produced from LTX-2 using NVIDIA Model Optimizer
+> (including quantized or distilled checkpoints) remain subject to the LTX Community License
+> Agreement and are **not** covered by Apache 2.0.
+
 Knowledge distillation for LTX-2 DiT models using NVIDIA ModelOpt. A frozen **teacher** guides a trainable **student** through a combined loss:
 
 ```text

examples/diffusers/distillation/distillation_trainer.py

@@ -35,6 +35,7 @@
 import json
 import os
 import time
+import warnings
 from pathlib import Path
 from typing import Literal
 
@@ -51,6 +52,17 @@
 import modelopt.torch.opt as mto
 import modelopt.torch.quantization as mtq
 
+warnings.warn(
+    "LTX-2 packages (ltx-core, ltx-pipelines, ltx-trainer) are provided by Lightricks and are "
+    "NOT covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply "
+    "with the LTX Community License Agreement when installing and using LTX-2 with NVIDIA Model "
+    "Optimizer. Any derivative models or fine-tuned weights from LTX-2 (including quantized or "
+    "distilled checkpoints) remain subject to the LTX Community License Agreement, not Apache "
+    "2.0. See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE",
+    UserWarning,
+    stacklevel=1,
+)
+
 # Custom quantization configs. Checked before mtq built-in configs.
 # Add your own configs here; they take precedence over mtq.* attributes.
 CUSTOM_QUANT_CONFIGS: dict[str, dict] = {

examples/diffusers/distillation/requirements.txt

@@ -1,3 +1,9 @@
+# Third-Party License Notice: LTX-2 packages below are provided by Lightricks and are NOT
+# covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply with
+# the LTX Community License Agreement when installing and using LTX-2 with NVIDIA Model
+# Optimizer. Any derivative models or fine-tuned weights produced from LTX-2 (including
+# quantized or distilled checkpoints) remain subject to the LTX Community License Agreement,
+# not Apache 2.0. See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE
 ltx-core @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-core
 ltx-pipelines @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-pipelines
 ltx-trainer @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-trainer

examples/diffusers/quantization/calibration.py

@@ -14,6 +14,7 @@
 # limitations under the License.
 
 import logging
+import warnings
 from pathlib import Path
 from typing import Any
 
@@ -120,6 +121,16 @@ def _run_wan_video_calibration(
         self.pipe(prompt=prompt_batch, **kwargs).frames
 
     def _run_ltx2_calibration(self, prompt_batch: list[str], extra_args: dict[str, Any]) -> None:
+        warnings.warn(
+            "LTX-2 packages (ltx-core, ltx-pipelines, ltx-trainer) are provided by Lightricks and are NOT "
+            "covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply "
+            "with the LTX Community License Agreement when installing and using LTX-2 with NVIDIA "
+            "Model Optimizer. Any derivative models or fine-tuned weights from LTX-2 remain "
+            "subject to the LTX Community License Agreement, not Apache 2.0. "
+            "See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE",
+            UserWarning,
+            stacklevel=2,
+        )
         from ltx_core.model.video_vae import TilingConfig
         from ltx_pipelines.utils.constants import (
             DEFAULT_AUDIO_GUIDER_PARAMS,

examples/diffusers/quantization/pipeline_manager.py

@@ -14,6 +14,7 @@
 # limitations under the License.
 
 import logging
+import warnings
 from collections.abc import Iterator
 from typing import Any
 
@@ -227,6 +228,16 @@ def _create_ltx2_pipeline(self) -> Any:
         if not gemma_root:
             raise ValueError("Missing required extra_param: gemma_root.")
 
+        warnings.warn(
+            "LTX-2 packages (ltx-core, ltx-pipelines, ltx-trainer) are provided by Lightricks and are NOT "
+            "covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply "
+            "with the LTX Community License Agreement when installing and using LTX-2 with NVIDIA "
+            "Model Optimizer. Any derivative models or fine-tuned weights from LTX-2 remain "
+            "subject to the LTX Community License Agreement, not Apache 2.0. "
+            "See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE",
+            UserWarning,
+            stacklevel=2,
+        )
         from ltx_core.loader import LTXV_LORA_COMFY_RENAMING_MAP, LoraPathStrengthAndSDOps
         from ltx_core.quantization import QuantizationPolicy
         from ltx_pipelines.ti2vid_two_stages import TI2VidTwoStagesPipeline

examples/windows/torch_onnx/diffusers/qad_example/README.md

@@ -1,5 +1,19 @@
 # LTX-2 QAD Example (Quantization-Aware Distillation)
 
+> [!WARNING]
+> **Third-Party License Notice — LTX-2**
+>
+> LTX-2 is a third-party model and set of packages developed and provided by Lightricks. LTX-2
+> is **not** covered by the Apache 2.0 license that governs NVIDIA Model Optimizer.
+>
+> By installing and using LTX-2 packages (`ltx-core`, `ltx-pipelines`, `ltx-trainer`) with
+> NVIDIA Model Optimizer, you **must** comply with the
+> [LTX Community License Agreement](https://github.com/Lightricks/LTX-2/blob/main/LICENSE).
+>
+> Any derivative models or fine-tuned weights produced from LTX-2 using NVIDIA Model Optimizer
+> (including quantized or distilled checkpoints) remain subject to the LTX Community License
+> Agreement and are **not** covered by Apache 2.0.
+
 **Note:** This is a **sample script for illustrating the QAD pipeline**. It has been verified to run on a **Linux RTX 5090** system, but runs into **OOM (Out of Memory)** on that configuration.
 
 This example demonstrates **Quantization-Aware Distillation (QAD)** for [LTX-2](https://github.com/Lightricks/LTX-2) using the native LTX training loop and [NVIDIA ModelOpt](https://github.com/NVIDIA/Model-Optimizer). It combines:

examples/windows/torch_onnx/diffusers/qad_example/requirements.txt

@@ -1,12 +1,11 @@
-accelerate
 # LTX-2 packages (from Lightricks/LTX-2 monorepo)
+# Third-Party License Notice: LTX-2 packages below are provided by Lightricks and are NOT
+# covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply with
+# the LTX Community License Agreement when installing and using LTX-2 with NVIDIA Model
+# Optimizer. Any derivative models or fine-tuned weights produced from LTX-2 (including
+# quantized or distilled checkpoints) remain subject to the LTX Community License Agreement,
+# not Apache 2.0. See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE
 ltx-core @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-core
 ltx-pipelines @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-pipelines
 ltx-trainer @ git+https://github.com/Lightricks/LTX-2.git#subdirectory=packages/ltx-trainer
-
-# NVIDIA ModelOpt (quantization & distillation)
-nvidia-modelopt
-pyyaml
-safetensors
-
-torch>=2.0
+nvidia-modelopt[hf]

examples/windows/torch_onnx/diffusers/qad_example/sample_example_qad_diffusers.py

@@ -44,6 +44,7 @@
 import os
 import struct
 import sys
+import warnings
 from collections import Counter
 from pathlib import Path
 
@@ -66,6 +67,17 @@
 from modelopt.torch.distill.distillation_model import DistillationModel
 from modelopt.torch.quantization.config import NVFP4_DEFAULT_CFG
 
+warnings.warn(
+    "LTX-2 packages (ltx-core, ltx-pipelines, ltx-trainer) are provided by Lightricks and are "
+    "NOT covered by the Apache 2.0 license governing NVIDIA Model Optimizer. You MUST comply "
+    "with the LTX Community License Agreement when installing and using LTX-2 with NVIDIA Model "
+    "Optimizer. Any derivative models or fine-tuned weights from LTX-2 (including quantized or "
+    "distilled checkpoints) remain subject to the LTX Community License Agreement, not Apache "
+    "2.0. See: https://github.com/Lightricks/LTX-2/blob/main/LICENSE",
+    UserWarning,
+    stacklevel=1,
+)
+
 logger = logging.getLogger(__name__)
 
 # ─── Constants ────────────────────────────────────────────────────────────────