Merge pull request #32 from cassiofb-dev/release/v0.1.1

Release/v0.1.1

backend/src/dags/feature_engineering/word_embeddings.py

@@ -8,7 +8,7 @@
 from dags import config
 import logging
 
-import logging
+import tensorflow_text
 
 # create logger
 logger = logging.getLogger(__name__)
@@ -25,7 +25,7 @@ def generate_embed_features(self,df,coluna = "new_text"):
 class USE_embedder(text_embedder):
     def __init__(self):
         super().__init__()
-        self.embedder = hub.load("https://tfhub.dev/google/universal-sentence-encoder/4")
+        self.embedder = hub.load("https://tfhub.dev/google/universal-sentence-encoder-multilingual-large/3")
 
     def embed_words(self,word_vector):
         embeddings = self.embedder(word_vector)

docker-compose.yml

@@ -7,7 +7,7 @@ services:
     working_dir: "/backend"
     command: sh -c "
       python -m venv venv && . venv/bin/activate && pip install -U pip &&
-      pip install -U "jax[cpu]" tensorflow-cpu tensorflow_hub scikit-learn nltk gensim pyarrow &&
+      pip install -U "jax[cpu]" tensorflow-cpu tensorflow_text tensorflow_hub scikit-learn nltk gensim pyarrow &&
       pip install -U pandas fastapi "uvicorn[standard]" orjson openpyxl colorlog pyspellchecker coloredlogs &&
       cd src && python requirements.py && python run.py task general_tasks task_correct_essays &&
       python run.py task general_tasks task_pipeline_essays && uvicorn api:app --host 0.0.0.0 --reload"