fix cache dir

microsoft · Mar 26, 2024 · 98dc323 · 98dc323
1 parent 039f40d
commit 98dc323
Show file tree

Hide file tree

Showing 7 changed files with 17 additions and 174 deletions.
diff --git a/Dockerfile b/Dockerfile
@@ -10,7 +10,10 @@ ENV IS_DOCKER True
 
 # Install dependencies
 WORKDIR ./
-COPY . .
+COPY ./app ./app
+COPY ./.streamlit ./.streamlit
+COPY Dockerfile .
+COPY requirements.txt .
 RUN pip install -r requirements.txt
 
 # Run application

diff --git a/app/util/notebook_embedder.ipynb b/app/util/notebook_embedder.ipynb
diff --git a/app/workflows/question_answering/classes.py b/app/workflows/question_answering/classes.py
@@ -2,11 +2,12 @@
 import tiktoken
 import scipy.spatial.distance
 import workflows.question_answering.config as config
+import os
 
 
 import util.Embedder
 
-embedder = util.Embedder.create_embedder(cache=f'{config.cache_dir}\\qa_mine') #SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
+embedder = util.Embedder.create_embedder(cache=os.path.join(config.cache_dir,"qa_mine")) #SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
 encoder = tiktoken.get_encoding('cl100k_base')
 
 class Question:

diff --git a/app/workflows/question_answering/config.py b/app/workflows/question_answering/config.py
@@ -1,4 +1,6 @@
-cache_dir = '.\\cache\\question_answering'
+import os
+
+cache_dir = os.path.join(os.environ.get("CACHE_DIR", "cache"), "question_answering")
 
 chunk_size = 5000
 chunk_overlap = 0

diff --git a/app/workflows/question_answering/functions.py b/app/workflows/question_answering/functions.py
@@ -9,7 +9,7 @@
 import workflows.question_answering.classes as classes
 import workflows.question_answering.config as config
 
-embedder = util.Embedder.create_embedder(cache=f'{config.cache_dir}\qa_mine')
+embedder = util.Embedder.create_embedder(cache=f'{config.cache_dir}/qa_mine')
 encoder = tiktoken.get_encoding('cl100k_base')
 
 

diff --git a/app/workflows/record_matching/config.py b/app/workflows/record_matching/config.py
@@ -1,9 +1,11 @@
+import os
+
 att_val_sep = '=='
 list_sep = '; '
 max_rows_to_show = 1000
 entity_label = 'Entity'
-cache_dir = '.\\cache\\record_matching'
-outputs_dir = f'{cache_dir}\\outputs'
+cache_dir = os.path.join(os.environ.get("CACHE_DIR", "cache"), "record_matching")
+outputs_dir = os.path.join(cache_dir,"outputs")
 
 intro = """ \
 # Record Matching

diff --git a/app/workflows/risk_networks/config.py b/app/workflows/risk_networks/config.py
@@ -1,9 +1,10 @@
+import os
 att_val_sep = '=='
 list_sep = '; '
 max_rows_to_show = 1000
 entity_label = 'ENTITY'
-cache_dir = '.\\cache\\risk_networks'
-outputs_dir = f'{cache_dir}\\outputs'
+cache_dir = os.path.join(os.environ.get("CACHE_DIR", "cache"), "risk_networks")
+outputs_dir = os.path.join(cache_dir, "outputs")
 
 intro = """ \
 # Risk Networks