Redo packaging

.github/workflows/python-publish.yml

@@ -0,0 +1,37 @@
+name: Release
+
+on:
+  push:
+    branches:
+    - main
+jobs:
+  deploy:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v2
+    - uses: actions-ecosystem/action-regex-match@v2
+      id: regex-match
+      with:
+        text: ${{ github.event.head_commit.message }}
+        regex: '^Release ([^ ]+)'
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: '3.8'
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install setuptools wheel twine
+    - name: Release
+      if: ${{ steps.regex-match.outputs.match != '' }}
+      uses: softprops/action-gh-release@v1
+      with:
+        tag_name: ${{ steps.regex-match.outputs.group1 }}
+    - name: Build and publish
+      if: ${{ steps.regex-match.outputs.match != '' }}
+      env:
+        TWINE_USERNAME: __token__
+        TWINE_PASSWORD: ${{ secrets.PYPI_PASSWORD }}
+      run: |
+        python setup.py sdist bdist_wheel
+        twine upload dist/*

Makefile

@@ -0,0 +1,3 @@
+install: ## [Local development] Upgrade pip, install requirements, install package.
+	python -m pip install -U pip
+	python -m pip install -e .

Model Cards/M-BERT Base 69/README.md

@@ -13,7 +13,7 @@
 To use this model along with the original CLIP vision encoder follow the [main page usage instructions](https://github.com/FreddeFrallan/Multilingual-CLIP) to download the additional linear weights.
 Once this is done, you can load and use the model with the following code
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 model = multilingual_clip.load_model('M-BERT-Base-69')
 embeddings = model(['Älgen är skogens konung!', 'Wie leben Eisbären in der Antarktis?', 'Вы знали, что все белые медведи левши?'])

Model Cards/M-BERT Base ViT-B/README.md

@@ -13,7 +13,7 @@
 To use this model along with the original CLIP vision encoder follow the [main page usage instructions](https://github.com/FreddeFrallan/Multilingual-CLIP) to download the additional linear weights.
 Once this is done, you can load and use the model with the following code
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 model = multilingual_clip.load_model('M-BERT-Base-ViT-B')
 embeddings = model(['Älgen är skogens konung!', 'Wie leben Eisbären in der Antarktis?', 'Вы знали, что все белые медведи левши?'])

Model Cards/M-BERT Distil 40/README.md

@@ -13,7 +13,7 @@
 To use this model along with the original CLIP vision encoder follow the [main page usage instructions](https://github.com/FreddeFrallan/Multilingual-CLIP) to download the additional linear weights.
 Once this is done, you can load and use the model with the following code
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 model = multilingual_clip.load_model('M-BERT-Distil-40')
 embeddings = model(['Älgen är skogens konung!', 'Wie leben Eisbären in der Antarktis?', 'Вы знали, что все белые медведи левши?'])

Model Cards/Swe-CLIP 2M/README.md

@@ -13,7 +13,7 @@
 To use this model along with the original CLIP vision encoder follow the [main page usage instructions](https://github.com/FreddeFrallan/Multilingual-CLIP) to download the additional linear weights.
 Once this is done, you can load and use the model with the following code
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 model = multilingual_clip.load_model('Swe-CLIP-2M')
 embeddings = model(['Älgen är skogens konung!', 'Alla isbjörnar är vänsterhänta'])

Model Cards/Swe-CLIP 500k/README.md

@@ -13,7 +13,7 @@
 To use this model along with the original CLIP vision encoder follow the [main page usage instructions](https://github.com/FreddeFrallan/Multilingual-CLIP) to download the additional linear weights.
 Once this is done, you can load and use the model with the following code
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 model = multilingual_clip.load_model('Swe-CLIP-500k')
 embeddings = model(['Älgen är skogens konung!', 'Alla isbjörnar är vänsterhänta'])

README.md

@@ -35,12 +35,22 @@ While it is possible that other versions works equally fine, we have worked with
 * Python = 3.6.9
 * Transformers = 4.8.1
 
-## Inference Example
+## Install
+
+`pip install mclip torch`
+
+You can also choose to `pip install tensorflow` instead of torch.
+
+
+## Inference Usage
+
 Inference code for Tensorflow is also available in [inference_example.py](https://github.com/FreddeFrallan/Multilingual-CLIP/blob/main/inference_example.py)
+
 ```python
-from src import pt_multilingual_clip
+from mclip import pt_multilingual_clip
+import transformers
 
-exampleTexts = [
+texts = [
     'Three blind horses listening to Mozart.',
     'Älgen är skogens konung!',
     'Wie leben Eisbären in der Antarktis?',
@@ -55,6 +65,17 @@ tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)
 embeddings = model.forward(texts, tokenizer)
 print(embeddings.shape)
 ```
+
+## Install for development
+
+Setup a virtualenv:
+
+```
+python3 -m venv .env
+source .env/bin/activate
+pip install -e .
+```
+
 ## Pre-trained Models
 Every text encoder is a [Huggingface](https://huggingface.co/) available transformer, with an additional linear layer on top. For more information of a specific model, click the Model Name to see its model card.
 <br>
@@ -101,7 +122,7 @@ $ bash legacy_get-weights.sh
 
 ### Inference
 ```python
-from src import multilingual_clip
+from mclip import multilingual_clip
 
 print(multilingual_clip.AVAILABLE_MODELS.keys())
 
@@ -137,11 +158,11 @@ Every text encoder is a [Huggingface](https://huggingface.co/) available transfo
   </details>
 
 ## Training a new model
-[This folder](https://github.com/FreddeFrallan/Multilingual-CLIP/tree/main/src/TeacherLearning) contains the code used for training the above models. If you wsh to train your own model you must do the following things:
+[This folder](https://github.com/FreddeFrallan/Multilingual-CLIP/tree/main/mclip/TeacherLearning) contains the code used for training the above models. If you wsh to train your own model you must do the following things:
 
 * Prepare a set of translated sentence pairs from English -> Your Language(s)
 * Compute regular CLIP-Text embeddings for the English sentences.
-* Edit [Training.py](https://github.com/FreddeFrallan/Multilingual-CLIP/blob/main/src/TeacherLearning/Training.py) to load your data.
+* Edit [Training.py](https://github.com/FreddeFrallan/Multilingual-CLIP/blob/main/mclip/TeacherLearning/Training.py) to load your data.
 * Train a new CLIP-Text encoder via Teacher Learning 
 
 ### Pre-computed CLIP Embeddings & Translaton Data

inference_example.py

@@ -2,7 +2,7 @@
 
 
 def tf_example(texts, model_name='M-CLIP/XLM-Roberta-Large-Vit-L-14'):
-    import tf_multilingual_clip
+    from mclip import tf_multilingual_clip
 
     model = tf_multilingual_clip.MultiLingualCLIP.from_pretrained(model_name)
     tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)
@@ -13,7 +13,7 @@ def tf_example(texts, model_name='M-CLIP/XLM-Roberta-Large-Vit-L-14'):
 
 
 def pt_example(texts, model_name='M-CLIP/XLM-Roberta-Large-Vit-L-14'):
-    import pt_multilingual_clip
+    from mclip import pt_multilingual_clip
 
     model = pt_multilingual_clip.MultilingualCLIP.from_pretrained(model_name)
     tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)

legacy_inference.py

@@ -1,4 +1,4 @@
-from src.legacy_multilingual_clip import MultilingualClip
+from mclip.legacy_multilingual_clip import MultilingualClip
 
 model_path = 'M-CLIP/Swedish-500k'
 tok_path = 'M-CLIP/Swedish-500k'

src/pt_multilingual_clip.py → mclip/pt_multilingual_clip.py

@@ -1,4 +1,4 @@
-from src import Config_MCLIP
+from mclip import Config_MCLIP
 import transformers
 import torch
 

src/tf_multilingual_clip.py → mclip/tf_multilingual_clip.py

@@ -1,4 +1,4 @@
-from src import Config_MCLIP
+from mclip import Config_MCLIP
 import tensorflow as tf
 import transformers
 

requirements.txt

@@ -0,0 +1 @@
+transformers

setup.py

@@ -0,0 +1,38 @@
+from setuptools import setup, find_packages
+from pathlib import Path
+import os
+
+if __name__ == "__main__":
+    with Path(Path(__file__).parent, "README.md").open(encoding="utf-8") as file:
+        long_description = file.read()
+
+    def _read_reqs(relpath):
+        fullpath = os.path.join(os.path.dirname(__file__), relpath)
+        with open(fullpath) as f:
+            return [s.strip() for s in f.readlines() if (s.strip() and not s.startswith("#"))]
+
+    REQUIREMENTS = _read_reqs("requirements.txt")
+
+    setup(
+        name="mclip",
+        packages=find_packages(),
+        include_package_data=True,
+        version="1.0.0",
+        license="MIT",
+        description="OpenAI CLIP text encoders for multiple languages!",
+        long_description=long_description,
+        long_description_content_type="text/markdown",
+        author="Fredrik Carlsson",
+        author_email="FreddeFc@gmail.com",
+        url="https://github.com/FreddeFrallan/Multilingual-CLIP",
+        data_files=[(".", ["README.md"])],
+        keywords=["machine learning"],
+        install_requires=REQUIREMENTS,
+        classifiers=[
+            "Development Status :: 4 - Beta",
+            "Intended Audience :: Developers",
+            "Topic :: Scientific/Engineering :: Artificial Intelligence",
+            "License :: OSI Approved :: MIT License",
+            "Programming Language :: Python :: 3.6",
+        ],
+    )