tieubaoca

PaddleOCR-VL-1.6-Vietnamese-Merged

README

License: apache-2.0

Loading and inference

The verified environment used Python 3.12, PyTorch 2.11.0+cu128, Transformers 4.55.4, and a CUDA GPU.

python

import os os.environ.setdefault("CUBLAS_WORKSPACE_CONFIG", ":4096:8") import torchfrom PIL import Imagefrom transformers import AutoModelForCausalLM, AutoProcessor model_id = "tieubaoca/PaddleOCR-VL-1.6-Vietnamese-Merged"processor = AutoProcessor.from_pretrained(    model_id,    trust_remote_code=True,    use_fast=False,)model = AutoModelForCausalLM.from_pretrained(    model_id,    trust_remote_code=True,    torch_dtype=torch.bfloat16,    attn_implementation="eager",).to("cuda").eval() messages = [    {        "role": "user",        "content": [            {"type": "image", "image": "document.png"},            {"type": "text", "text": "OCR:"},        ],    }]prompt = processor.apply_chat_template(    messages,    tokenize=False,    add_generation_prompt=True,)with Image.open("document.png") as source:    image = source.convert("RGB").copy()inputs = processor(text=[prompt], images=[image], return_tensors="pt")inputs = {name: value.to("cuda") for name, value in inputs.items()}prompt_length = inputs["input_ids"].shape[1] with torch.inference_mode():    generated_ids = model.generate(        **inputs,        do_sample=False,        num_beams=1,        max_new_tokens=256,        use_cache=True,    )text = processor.batch_decode(    generated_ids[:, prompt_length:],    skip_special_tokens=True,    clean_up_tokenization_spaces=False,)[0].strip()print(text)

Evaluation

Deterministic greedy decoding with the native processor and chat template, 32 fixed validation samples from each of five sources, and 256 maximum new tokens.

Table with columns: Model, Samples, CER, Exact match, Normalized edit distance
Model	Samples	CER	Exact match	Normalized edit distance
PaddleOCR-VL-1.6 base	160	0.234065	0.137500	0.755668
Vietnamese merged model	160	0.066029	0.512500	0.941240

Merged per-source metrics:

Table with columns: Source, Samples, CER, Exact match, Normalized edit distance
Source	Samples	CER	Exact match	Normalized edit distance
source-000	32	0.091364	0.062500	0.910921
source-001	32	0.096000	0.781250	0.910813
source-002	32	0.070355	0.312500	0.932029

Merge verification

All 126 LoRA-targeted tensors were merged.
The exported weights matched the in-memory merged model after BF16 output casting within rtol=0.0005 and atol=0.0005.
Exported and in-memory merged logits were identical on the verification fixture (max_abs_error=0.0, equal argmax).

The original PaddleOCR-VL-1.6 license is included as LICENSE.

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Model Details

Model Provider

tieubaoca

Model Tree

Base

PaddlePaddle/PaddleOCR-VL-1.6

Fine-tuned

this model

Input Modalities

TextImage

Output Modalities

Text

Supported Functionality