From a8749efde8bdb72bd60c310c500701565aa4c544 Mon Sep 17 00:00:00 2001
From: Kyle Sayers <kylesayrs@gmail.com>
Date: Wed, 28 Jan 2026 17:11:43 +0000
Subject: [PATCH 1/2] qwen3_vl_example

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
---
 .../multimodal_vision/qwen3_vl_example.py     | 124 ++++++++++++++++++
 1 file changed, 124 insertions(+)
 create mode 100644 examples/multimodal_vision/qwen3_vl_example.py

diff --git a/examples/multimodal_vision/qwen3_vl_example.py b/examples/multimodal_vision/qwen3_vl_example.py
new file mode 100644
index 0000000000..12bec4ba3a
--- /dev/null
+++ b/examples/multimodal_vision/qwen3_vl_example.py
@@ -0,0 +1,124 @@
+import base64
+from io import BytesIO
+
+import torch
+from datasets import load_dataset
+from qwen_vl_utils import process_vision_info
+from transformers import AutoProcessor, Qwen3VLForConditionalGeneration
+
+from llmcompressor import oneshot
+from llmcompressor.modifiers.awq import AWQModifier
+from llmcompressor.utils import dispatch_for_generation
+
+# Load model.
+model_id = "Qwen/Qwen3-VL-8B-Instruct"
+model = Qwen3VLForConditionalGeneration.from_pretrained(model_id, torch_dtype="auto")
+processor = AutoProcessor.from_pretrained(model_id)
+
+# Oneshot arguments
+DATASET_ID = "lmms-lab/flickr30k"
+DATASET_SPLIT = "test[:12]"
+NUM_CALIBRATION_SAMPLES = 12
+MAX_SEQUENCE_LENGTH = 2048
+
+# Load dataset and preprocess.
+ds = load_dataset(DATASET_ID, split=DATASET_SPLIT)
+ds = ds.shuffle(seed=42)
+
+
+# Apply chat template and tokenize inputs.
+def preprocess_and_tokenize(example):
+    # preprocess
+    buffered = BytesIO()
+    example["image"].save(buffered, format="PNG")
+    encoded_image = base64.b64encode(buffered.getvalue())
+    encoded_image_text = encoded_image.decode("utf-8")
+    base64_qwen = f"data:image;base64,{encoded_image_text}"
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {"type": "image", "image": base64_qwen},
+                {"type": "text", "text": "What does the image show?"},
+            ],
+        }
+    ]
+    text = processor.apply_chat_template(
+        messages, tokenize=False, add_generation_prompt=True
+    )
+    image_inputs, video_inputs = process_vision_info(messages)
+
+    # tokenize
+    return processor(
+        text=[text],
+        images=image_inputs,
+        videos=video_inputs,
+        padding=False,
+        max_length=MAX_SEQUENCE_LENGTH,
+        truncation=True,
+    )
+
+
+ds = ds.map(preprocess_and_tokenize, remove_columns=ds.column_names)
+
+
+# Define a oneshot data collator for multimodal inputs.
+def data_collator(batch):
+    assert len(batch) == 1
+    return {key: torch.tensor(value) for key, value in batch[0].items()}
+
+
+# Recipe
+recipe = AWQModifier(
+    scheme="W4A16",
+    ignore=["re:.*lm_head", "re:.*visual.*"],
+    duo_scaling=False,
+)
+
+# Perform oneshot
+oneshot(
+    model=model,
+    tokenizer=model_id,
+    dataset=ds,
+    recipe=recipe,
+    max_seq_length=MAX_SEQUENCE_LENGTH,
+    num_calibration_samples=NUM_CALIBRATION_SAMPLES,
+    data_collator=data_collator,
+    sequential_targets=["Qwen3VLTextDecoderLayer"],
+)
+
+# Confirm generations of the quantized model look sane.
+print("========== SAMPLE GENERATION ==============")
+dispatch_for_generation(model)
+messages = [
+    {
+        "role": "user",
+        "content": [
+            {
+                "type": "image",
+                "image": "http://images.cocodataset.org/train2017/000000231895.jpg",
+            },
+            {"type": "text", "text": "Please describe the animal in this image\n"},
+        ],
+    }
+]
+prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
+image_inputs, video_inputs = process_vision_info(messages)
+inputs = processor(
+    text=[prompt],
+    images=image_inputs,
+    videos=video_inputs,
+    padding=False,
+    max_length=MAX_SEQUENCE_LENGTH,
+    truncation=True,
+    return_tensors="pt",
+).to(model.device)
+output = model.generate(**inputs, max_new_tokens=100)
+print(processor.decode(output[0], skip_special_tokens=True))
+print("==========================================")
+
+
+# Save to disk compressed.
+SAVE_DIR = model_id.rstrip("/").split("/")[-1] + "-W4A16"
+model.save_pretrained(SAVE_DIR, save_compressed=True)
+processor.save_pretrained(SAVE_DIR)

From 6fcf755d11c96fc1dc7dec333fbf5474ec4538cf Mon Sep 17 00:00:00 2001
From: Kyle Sayers <kylesayrs@gmail.com>
Date: Wed, 28 Jan 2026 17:13:32 +0000
Subject: [PATCH 2/2] use more samples

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
---
 examples/multimodal_vision/qwen3_vl_example.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/examples/multimodal_vision/qwen3_vl_example.py b/examples/multimodal_vision/qwen3_vl_example.py
index 12bec4ba3a..1272eb88c8 100644
--- a/examples/multimodal_vision/qwen3_vl_example.py
+++ b/examples/multimodal_vision/qwen3_vl_example.py
@@ -17,8 +17,8 @@
 
 # Oneshot arguments
 DATASET_ID = "lmms-lab/flickr30k"
-DATASET_SPLIT = "test[:12]"
-NUM_CALIBRATION_SAMPLES = 12
+DATASET_SPLIT = "test[:512]"
+NUM_CALIBRATION_SAMPLES = 512
 MAX_SEQUENCE_LENGTH = 2048
 
 # Load dataset and preprocess.