use PaddleOCRVL dtype not match

[renoyuan]

🔎 Search before asking

• [x] I have searched the PaddleOCR Docs and found no similar bug report.
• [x] I have searched the PaddleOCR Issues and found no similar bug report.
• [x] I have searched the PaddleOCR Discussions and found no similar bug report.

🐛 Bug (问题描述)

AssertionError: Variable dtype not match, Variable [ linear_164.w_0 ] need tensor with dtype paddle.float32 but load tensor with dtype paddle.bfloat16 i try use gpu 2080 ti and cpu run ，but assert mistake.

🏃‍♂️ Environment (运行环境)

ubuntu 20 cuda 12.4 & cpu

🌰 Minimal Reproducible Example (最小可复现问题的Demo)

from paddleocr import PaddleOCRVL import paddle pipeline = PaddleOCRVL( device='cpu', vl_rec_model_dir='./model'
) output = pipeline.predict("./pp_ocr_vl_demo.png") for res in output: res.print() res.save_to_json(save_path="output") res.save_to_markdown(save_path="output")

[Bobholamovic]

PaddleOCR-VL has a minimum requirement for the compute capacity of GPUs. Please check the documentation for more details.

[flowcloudzhao]

直接转换成paddle.float32 value=value.astype(paddle.float32)

[Bobholamovic]

We are working on this. This will be fixed within this week.

[TingquanGao]

The issue has no response for a long time and will be closed. You can reopen or new another issue if are still confused.

---

From Bot