first commit

.dockerignore

@@ -0,0 +1,89 @@
+# Git
+.git
+.gitignore
+.gitattributes
+
+
+# CI
+.codeclimate.yml
+.travis.yml
+.taskcluster.yml
+
+# Docker
+docker-compose.yml
+Dockerfile
+.docker
+.dockerignore
+
+# Byte-compiled / optimized / DLL files
+**/__pycache__/
+**/*.py[cod]
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.cache
+nosetests.xml
+coverage.xml
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Virtual environment
+.env
+.venv/
+venv/
+
+# PyCharm
+.idea
+
+# Python mode for VIM
+.ropeproject
+**/.ropeproject
+
+# Vim swap files
+**/*.swp
+
+# VS Code
+.vscode/

.gitignore

@@ -0,0 +1,30 @@
+.idea
+.ipynb_checkpoints
+.mypy_cache
+.vscode
+__pycache__
+.pytest_cache
+htmlcov
+dist
+site
+.coverage
+coverage.xml
+.netlify
+test.db
+log.txt
+Pipfile.lock
+env3.*
+env
+docs_build
+site_build
+venv
+docs.zip
+archive.zip
+
+# vim temporary files
+*~
+.*.sw?
+.cache
+
+# macOS
+.DS_Store

Dockerfile

@@ -0,0 +1,22 @@
+FROM python:3.11
+ENV HF_PORT=7860
+
+WORKDIR /code
+
+# Expose port 7860 (default huggingface)
+EXPOSE ${HF_PORT}
+
+# Install First Dependencies
+# "./requirements.txt" path-nya itu berdasarkan Dockerfile yang kita eksekusi (berati sudah di dalam src)
+COPY ./requirements.txt /code/requirements.txt
+
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+# Copy all files required
+# Instruction COPY, will create destination folder (this case "src") if didnt exist.
+COPY . /code/src
+
+# cause working dir it set from /code
+# and we have code in /code/src
+# for running command we user fastapi in "src.main:app"
+CMD ["uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "7860"]

data/44158_3_efficientnet_b2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d8b0bfcb84ceb12c3a7081538ad22eb52252ec9375c2b98e737e9ae08558f34
+size 31772794

data/class_names.txt

@@ -0,0 +1,90 @@
+a
+ba
+be
+bi
+bo
+bu
+ca
+ce
+ci
+co
+cu
+da
+de
+di
+do
+du
+e
+fa
+fe
+fi
+fo
+fu
+ga
+ge
+gi
+go
+gu
+ha
+he
+hi
+ho
+hu
+i
+ja
+je
+ji
+jo
+ju
+ka
+ke
+ki
+ko
+ku
+la
+le
+li
+lo
+lu
+ma
+me
+mi
+mo
+mu
+na
+ne
+ni
+no
+nu
+o
+pa
+pe
+pi
+po
+pu
+ra
+re
+ri
+ro
+ru
+sa
+se
+si
+so
+su
+ta
+te
+ti
+to
+tu
+u
+wa
+we
+wi
+wo
+wu
+ya
+ye
+yi
+yo
+yu

main.py

@@ -0,0 +1,15 @@
+from fastapi import FastAPI, Request
+from fastapi.responses import HTMLResponse
+
+from src.routers import api
+
+
+app = FastAPI()
+
+# Include all wraper router api.
+app.include_router(api.main_api)
+
+
+@app.get("/", response_class=HTMLResponse)
+def home(request: Request):
+    return f"Congratulations! Your API is working as expected. Now head over to <a href='{request.url}docs' target='_blank'>{request.url}docs</a>"

requirements.txt

@@ -0,0 +1,5 @@
+fastapi
+uvicorn[standard]
+torch
+torchvision
+python-multipart

routers/api.py

@@ -0,0 +1,6 @@
+from fastapi import APIRouter
+from src.routers.v1 import wrapper_v1
+
+main_api = APIRouter()
+
+main_api.include_router(wrapper_v1.router)

routers/v1/predict.py

@@ -0,0 +1,59 @@
+import torch
+
+from fastapi import APIRouter, UploadFile
+from PIL import Image
+from timeit import default_timer as timer 
+
+from src.utils.model import create_effnetb2_model
+from src.utils.helper import load_class_names
+from src.schemas.output_44158 import OutputBase
+
+
+
+router = APIRouter()
+
+@router.post("/predict", response_model=OutputBase)
+def inference(greetings: str, im: UploadFile):
+
+    # Load class names
+    class_names = load_class_names()
+    
+    # get image to PIL format
+    image = Image.open(im.file).convert('RGB')
+       
+    # Get model, transformation and calculate time to load it.
+    start_time = timer()
+    model, transforms = create_effnetb2_model(len(class_names))
+    end_time = timer()
+    print(f"[INFO] Total Loading model time: {end_time - start_time:.3f} seconds")
+
+    # Inference Time Start
+    inference_start_time = timer()
+
+    # turn image to same transformation
+    img = torch.unsqueeze(transforms(image), dim=0)
+
+    # Put model into evaluation mode and turn on inference mode
+    model.eval()
+    with torch.inference_mode():
+        # no need to send data to GPU, cause default is cpu.
+        pred_logits = model(img)
+        pred_probs = torch.softmax(pred_logits, dim=1)
+
+
+    # Create prediction labels 
+    predicted_label_index = torch.argmax(pred_probs).item() # get single index if not add dim.
+    predicted_label = class_names[predicted_label_index]
+
+    # Inference Time End
+    inference_start_end = timer()
+
+    # Get probability predictions
+    probability_pred = pred_probs[0][predicted_label_index].item()
+
+    return OutputBase(
+        message=greetings,
+        class_predicted=predicted_label,
+        prob=probability_pred,
+        inference_time=f"{inference_start_end - inference_start_time:.3f} Seconds"
+    )

routers/v1/wrapper_v1.py

@@ -0,0 +1,7 @@
+from fastapi import APIRouter
+
+from src.routers.v1 import predict
+
+router = APIRouter(prefix="/v1")
+
+router.include_router(predict.router)

schemas/output_44158.py

@@ -0,0 +1,7 @@
+from pydantic import BaseModel
+
+class OutputBase(BaseModel):
+    message: str
+    class_predicted: str
+    prob: float
+    inference_time: str

utils/helper.py

@@ -0,0 +1,11 @@
+from pathlib import Path
+
+def load_class_names():
+
+    file = "class_names.txt"
+    path_file = Path("./src/data") / file
+
+    with open(path_file,"r") as f:
+        class_names = [item.strip() for item in f.readlines()]
+
+    return class_names

utils/model.py

@@ -0,0 +1,49 @@
+import torchvision
+import torch
+import functools
+
+from torch import nn
+from pathlib import Path
+
+@functools.cache
+def create_effnetb2_model(num_class: int):
+    """Create a pytorch model for EfficientNetB2.
+
+    Making a EfficientNetB2 as Feature Extractor and also
+    can custom output class as need it.
+
+    Args:
+        num_class: A number of class, that will be output (head) of model.
+
+    Returns:
+        a tuple of (model, transforms) of EfficientNetB2.
+
+    """
+
+    # Get weights of ResNet50
+    weights_effnetb2 = torchvision.models.EfficientNet_B2_Weights.IMAGENET1K_V1
+
+    # Get transforms used in resnet
+    transforms = weights_effnetb2.transforms()
+
+    # making model
+    model = torchvision.models.efficientnet_b2(weights=weights_effnetb2)
+
+    # Freeze All layer
+    for param in model.parameters():
+        param.requires_grad = False
+
+    # Custom Output class
+    model.classifier = nn.Sequential(
+        nn.Dropout(p=0.3, inplace=True),
+        nn.Linear(in_features=1408,
+                  out_features=num_class)
+    )
+
+    # Load trained weights
+    path_model_weights = Path("./src/data") / "44158_3_efficientnet_b2.pth"
+    model.load_state_dict(
+        torch.load(path_model_weights, map_location=torch.device("cpu"))
+    )
+
+    return model, transforms