Qwen3-VL-30B-A3B-Thinking-ablation-v1.1-test-500

Qwen3-VL-30B-A3B-Thinking-ablation-v1.1-test-500 is a hybrid abliterated (v1.1) variant of the Qwen3-VL-30B-A3B-Thinking model, developed for advanced scientific and technological multimodal reasoning. This model integrates refined ablation techniques within the Qwen3-VL-MoE (Mixture of Experts) framework, enhancing precision, factual grounding, and interpretive clarity across complex analytical contexts. It exhibits a strict moderation bias toward photo-sensitive or sensual imagery, aligning with its design goal of maintaining ethical visual reasoning and domain-focused adaptability across multimodal tasks.

Quick Start with Transformers

from transformers import Qwen3VLMoeForConditionalGeneration, AutoProcessor
from qwen_vl_utils import process_vision_info
import torch

model = Qwen3VLMoeForConditionalGeneration.from_pretrained(
    "prithivMLmods/Qwen3-VL-30B-A3B-Thinking-ablation-v1.1-test-500",
    torch_dtype="auto",
    device_map="auto"
)

processor = AutoProcessor.from_pretrained("prithivMLmods/Qwen3-VL-30B-A3B-Thinking-ablation-v1.1-test-500")

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Provide a detailed caption and reasoning for this image."},
        ],
    }
]

text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)

inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
).to("cuda")

generated_ids = model.generate(**inputs, max_new_tokens=128)

generated_ids_trimmed = [
    out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]

output_text = processor.batch_decode(
    generated_ids_trimmed,
    skip_special_tokens=True,
    clean_up_tokenization_spaces=False
)

print(output_text)