Create backPrompt.py

backPrompt.py

@@ -0,0 +1,343 @@
+import numpy as np
+import torch
+import torchvision.transforms as T
+from decord import VideoReader, cpu
+from PIL import Image
+from torchvision.transforms.functional import InterpolationMode
+from transformers import AutoModel, AutoTokenizer
+
+IMAGENET_MEAN = (0.485, 0.456, 0.406)
+IMAGENET_STD = (0.229, 0.224, 0.225)
+
+def build_transform(input_size):
+    MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
+    transform = T.Compose([
+        T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
+        T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
+        T.ToTensor(),
+        T.Normalize(mean=MEAN, std=STD)
+    ])
+    return transform
+
+
+def load_image(image_file, input_size=800, max_num=12):
+    image = Image.open(image_file).convert('RGB')
+    transform = build_transform(input_size=input_size)
+    pixel_values = [transform(image) for image in images]
+    pixel_values = torch.stack(pixel_values)
+    return pixel_values
+
+
+def main(image_path, model, tokenizer):
+    path = "OpenGVLab/InternVL2_5-4B"
+    model = AutoModel.from_pretrained(
+        path,
+        torch_dtype=torch.bfloat16,
+        load_in_8bit=True,
+        low_cpu_mem_usage=True,
+        use_flash_attn=True,
+        trust_remote_code=True).eval()
+    tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)
+    pixel_values = load_image(image_path, max_num=12).to(torch.bfloat16).cuda()
+    generation_config = dict(max_new_tokens=1024, do_sample=True)
+    
+    question = """<image>\n**Instruction:**
+    Analyze the image to extract values for the specified keys. Use the detailed descriptions below to determine the correct value for each key. Handle missing or ambiguous data as instructed.
+    
+    ---
+    ### Keys and Descriptions
+    
+    1. **`Surat Ketetapan Kewajiban Pembayaran PKB/BBNKB,SWDKLLJ DAN PNBP`**
+       - **Extract**: The value of the field is"Surat Ketetapan Kewajiban Pembayaran PKB/BBNKB, SWDKLLJ DAN PNBP" and this is title.
+       - **If the Field is Present** : "present"
+       - **If the Field is Absent**: `"null"`
+    
+    2. **`nomor_registrasi`**
+       - **Extract**: The "NOMOR REGISTRASI" field.
+       - **If the Field is Absent**: `"null"`
+       - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    3.  **`nama_pemilik`**
+       - **Extract**: The "NAMA PEMILIK" field.
+       - **If the Field is Absent**: `"null"`
+       - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    4. **`alamat`**
+       - **Extract**: The "ALAMAT" field and this is address may contain 1 or more lines format : "BUNTET PESANTREN RT/RW/015/005 DESA MERTAPADA
+    KULON KEC. ASTANAJAPURA KAB.CIREBON ".
+       - **If the Field is Absent**: `"null"`
+       - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    5. **`merk/type`**
+       - **Extract**: The "MERK/TYPE" field.
+       - **If the Field is Absent**: `"null"`
+       - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    6. **`jenis/model`**
+       - **Extract**: The "JENIS/MODEL" field.
+       - **If the Field is Absent**: `"null"`
+       - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    7.  **`tahun_pembuatan`**
+        - **Extract**: The "TAHUN PEMBUATAN/PERAKITAN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    8.  **`warna_kb`**
+        - **Extract**: The "WARNA KB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    9. **`isi_silinder`**
+        - **Extract**: The "ISI SILINDER/HP" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    10. **`nomor_rangka`**
+        - **Extract**: The "NOMOR RANGKA/NIK" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    11.**`nomor_mesin`**
+        - **Extract**: The "NOMOR MESIN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    12. **`no_bpkp`**
+        - **Extract**: The "NO BPKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    13. **`berlaku s/d`**
+        - **Extract**: The "BERLAKU S/D" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    14. **`bahan_bakar`**
+        - **Extract**: The "BAHAN BAKAR" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    15. **`warna_tnkb`**
+        - **Extract**: The "WARNA TNKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    
+    16.  **`kepemilikan_ke`**
+        - **Extract**: The "KEPEMILIKAN KE" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    17. **`no_registrasi_lama`**
+        - **Extract**: The "NO REGISTRASI LAMA" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    18. **`kode_njkb`**
+        - **Extract**: The "KODE NJKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    19. **`no`**
+        - **Extract**: The "NO." field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    20.**`asal_daerah`**
+        - **Extract**: The "ASAL DAERAH" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    21. **`no_urut`**
+        - **Extract**: The "NO. URUT" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    22. **`no_kohir`**
+        - **Extract**: The "NO. KOHIR" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    23. **`nik_no.hp`**
+        - **Extract**: The "NIK/NO.HP" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    24. **`pkok_bbnkb`**
+        - **Extract**: The "PKOK BBNKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    25. **`pkok_pkb`**
+        - **Extract**: The "POKOK PKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    26. **`pkok_swdkljj`**
+        - **Extract**: The "POKOK SWDKLJJ" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    27. **`pkok_penerbitan_stnk`**
+        - **Extract**: The "POKOK PENERBITAN STNK" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    28. **`pkok_penerbitan_tnkb`**
+        - **Extract**: The "POKOK PENERBITAN TNKB/NRKB PILIHAN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    29. **`pkok_jumlah`**
+        - **Extract**: The "POKOK JUMLAH" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    30. **`sanksi_adm_bbnkb`**
+        - **Extract**: The "SANKSI ADM BBNKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    31. **`sanksi_adm_pkb`**
+        - **Extract**: The "SANKSI ADM PKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    32. **`sanksi_adm_swdkljj`**
+        - **Extract**: The "SANKSI ADM SWDKLJJ" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    33. **`sanksi_adm_penerbitan_stnk`**
+        - **Extract**: The "SANKSI ADM PENERBITAN STNK" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    34. **`sanksi_adm_penerbitan_tnkb`**
+        - **Extract**: The "SANKSI ADM PENERBITAN TNKB/NRKB PILIHAN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    35. **`sanksi_adm_jumlah`**
+        - **Extract**: The "SANKSI ADM JUMLAH " field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    36. **`jumlah_bbnkb`**
+        - **Extract**: The "JUMLAH BBNKB" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    37. **`jumlah_pkb`**
+        - **Extract**: The "JUMLAH PKB " field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    38. **`jumlah_swdkljj`**
+        - **Extract**: The "JUMLAH SWDKLJJ" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    39. **`jumlah_penerbitan_stnk`**
+        - **Extract**: The "JUMLAH PENERBITAN STNK" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    40. **`jumlah_penerbitan_tnkb`**
+        - **Extract**: The "JUMLAH PENERBITAN TNKB/NRKB PILIHAN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    41. **`jumlah_jumlah`**
+        - **Extract**: The "JUMLAH JUMLAH" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    42. **`ditetapkan_tanggal`**
+        - **Extract**: The "DITETAPKAN TANGGAL" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    43. **`penetapan`**
+        - **Extract**: The "PENETAPAN" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    44. **`kasir`**
+        - **Extract**: The "KASIR" field.
+        - **If the Field is Absent**: `"null"`
+        - **If the Field is Present but No Value is Provided**: `"empty"`
+    
+    ---
+    ### Reasoning Process
+    For each key, explain your reasoning:
+    Indicate whether the field was present.
+    Justify the extracted value or the use of "null" or "empty" as per the conditions.
+    
+    ### Output Format
+    
+    ```json
+    {
+      "Surat_Ketetapan_Kewajiban_Pembayaran_PKB_BBNKB_SWDKLLJ_DAN_PNBP": "present OR null",
+      "nomor_registrasi": "<value> OR empty OR null",
+      "nama_pemilik": "<value> OR empty OR null",
+      "alamat": "<value> OR empty OR null",
+      "merk_type": "<value> OR empty OR null",
+      "jenis_model": "<value> OR empty OR null",
+      "tahun_pembuatan": "<value> OR empty OR null",
+      "warna_kb": "<value> OR empty OR null",
+      "isi_silinder": "<value> OR empty OR null",
+      "nomor_rangka": "<value> OR empty OR null",
+      "nomor_mesin": "<value> OR empty OR null",
+      "no_bpkp": "<value> OR empty OR null",
+      "berlaku_s_d": "<value> OR empty OR null",
+      "bahan_bakar": "<value> OR empty OR null",
+      "warna_tnkb": "<value> OR empty OR null",
+      "kepemilikan_ke": "<value> OR empty OR null",
+      "no_registrasi_lama": "<value> OR empty OR null",
+      "kode_njkb": "<value> OR empty OR null",
+      "no": "<value> OR empty OR null",
+      "asal_daerah": "<value> OR empty OR null",
+      "no_urut": "<value> OR empty OR null",
+      "no_kohir": "<value> OR empty OR null",
+      "nik_no_hp": "<value> OR empty OR null",
+      "pkok_bbnkb": "<value> OR empty OR null",
+      "pkok_pkb": "<value> OR empty OR null",
+      "pkok_swdkljj": "<value> OR empty OR null",
+      "pkok_penerbitan_stnk": "<value> OR empty OR null",
+      "pkok_penerbitan_tnkb": "<value> OR empty OR null",
+      "pkok_jumlah": "<value> OR empty OR null",
+      "sanksi_adm_bbnkb": "<value> OR empty OR null",
+      "sanksi_adm_pkb": "<value> OR empty OR null",
+      "sanksi_adm_swdkljj": "<value> OR empty OR null",
+      "sanksi_adm_penerbitan_stnk": "<value> OR empty OR null",
+      "sanksi_adm_penerbitan_tnkb": "<value> OR empty OR null",
+      "sanksi_adm_jumlah": "<value> OR empty OR null",
+      "jumlah_bbnkb": "<value> OR empty OR null",
+      "jumlah_pkb": "<value> OR empty OR null",
+      "jumlah_swdkljj": "<value> OR empty OR null",
+      "jumlah_penerbitan_stnk": "<value> OR empty OR null",
+      "jumlah_penerbitan_tnkb": "<value> OR empty OR null",
+      "jumlah_jumlah": "<value> OR empty OR null",
+      "ditetapkan_tanggal": "<value> OR empty OR null",
+      "penetapan": "<value> OR empty OR null",
+      "kasir": "<value> OR empty OR null"
+    }
+    }
+    Return Output:
+    Generate a JSON object:
+    {
+        "reasoning": "reasoning for each key",
+        "output JSON": "key-value pairs"
+    }
+    —
+    
+    """
+    response = model.chat(tokenizer, pixel_values, question, generation_config)
+    return (f'User: {question}\nAssistant: {response}')