Merge branch 'main' of https://huggingface.co/spaces/heikowagner/GPT-Docker

Dockerfile

@@ -38,7 +38,6 @@ RUN mkdir /nltk_data
 RUN mkdir /VectorStore
 RUN mkdir /app/.cache
 RUN mkdir /mymodels
-RUN mkdir /app/mymodels
 RUN ls -la
 RUN python run.py
 RUN chmod 777 /VectorStore
@@ -47,6 +46,7 @@ RUN chmod 777 /nltk_data
 RUN chmod 777 /.cache
 RUN chmod 777 /app/.cache
 RUN chmod 777 /app/mymodels
+RUN chmod 777 /app/VectorStore/
 CMD ["streamlit", "run", "app.py", "--server.port=7860"]
 #CMD ls -la
 EXPOSE 7860

app/load_model.py

@@ -89,7 +89,7 @@ def load_openai_model(temperature=0.9):
 def load_openai_embedding():
     return OpenAIEmbeddings()
 
-@st.cache_resource
+#@st.cache_resource
 def load_embedding(model_name):
     embeddings = HuggingFaceInstructEmbeddings(
             query_instruction="Represent the query for retrieval: ",

app/run.py

@@ -1,4 +1,4 @@
-# This script inits the models and adds an example collection to the Vectorstore
+# This script inits the models and adds an example collection to the Vectorstore 
 # %%
 import os
 import pathlib
@@ -13,7 +13,7 @@ with open(current_path+'/.openaiapikey', 'r') as reader:
 import load_model
 
 # %%
-#load_model.load_gpu_model("decapoda-research/llama-7b-hf") #Download local model
+load_model.load_gpu_model("decapoda-research/llama-7b-hf") #Download local model
 llm= load_model.load_openai_model()
 
 # %% 
@@ -26,7 +26,9 @@ metadata= {"loaded_docs":[], "Subject":"Heikos Papers", "model_name": ef.model_n
 selected_collection = client.create_collection(collection_name, embedding_function=ef, metadata=metadata) 
 
 docs_tarifs= [
-    "https://edoc.hu-berlin.de/bitstream/handle/18452/5294/33.pdf",    
+    "https://edoc.hu-berlin.de/bitstream/handle/18452/5294/33.pdf",
+    "https://arxiv.org/pdf/1702.03556v3.pdf",
+    "https://arxiv.org/pdf/1706.03762"
 ]
 
 # %%

app/utils.py

@@ -22,12 +22,12 @@ def format_result_set(result):
         for document in source_documents:
             st.write(format_document(document))
 
-@st.cache_resource
+#@st.cache_resource
 def get_chroma_client():
     return chromadb.Client(Settings(chroma_db_impl="duckdb+parquet",
                                     persist_directory=persist_directory
                                 ))
-@st.cache_data
+#@st.cache_data
 def retrieve_collections():
     client = get_chroma_client()
     all_collections = client.list_collections()

requirements.txt

@@ -1,7 +1,8 @@
 #git+https://github.com/hwchase17/langchain.git
 langchain==0.0.154
 git+https://github.com/huggingface/transformers.git
-git+https://github.com/chroma-core/chroma.git
+#git+https://github.com/chroma-core/chroma.git
+chromadb
 accelerate
 bitsandbytes
 InstructorEmbedding