Spaces:
Runtime error
Runtime error
Todd Deshane
commited on
Update app.py
Browse files
app.py
CHANGED
@@ -59,40 +59,57 @@ def process_file(file: cl.AskFileMessage):
|
|
59 |
return texts
|
60 |
|
61 |
|
|
|
62 |
@cl.on_chat_start
|
63 |
async def on_chat_start():
|
64 |
files = None
|
65 |
|
66 |
# Wait for the user to upload a file
|
67 |
-
while files
|
|
|
68 |
files = await cl.AskFileMessage(
|
69 |
-
content="Please upload a PDF file to begin!",
|
70 |
-
accept=["application/pdf"],
|
71 |
-
max_size_mb=20,
|
72 |
timeout=180,
|
73 |
).send()
|
74 |
|
75 |
file = files[0]
|
76 |
|
77 |
-
|
78 |
-
|
79 |
-
)
|
80 |
await msg.send()
|
81 |
|
82 |
-
#
|
83 |
-
texts =
|
84 |
-
|
85 |
-
print(texts[0])
|
86 |
|
87 |
-
#
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
texts, embeddings, metadatas=metadatas
|
94 |
-
)
|
95 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
message_history = ChatMessageHistory()
|
97 |
|
98 |
memory = ConversationBufferMemory(
|
@@ -102,7 +119,6 @@ async def on_chat_start():
|
|
102 |
return_messages=True,
|
103 |
)
|
104 |
|
105 |
-
# Create a chain that uses the Chroma vector store
|
106 |
chain = ConversationalRetrievalChain.from_llm(
|
107 |
ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0, streaming=True),
|
108 |
chain_type="stuff",
|
|
|
59 |
return texts
|
60 |
|
61 |
|
62 |
+
|
63 |
@cl.on_chat_start
|
64 |
async def on_chat_start():
|
65 |
files = None
|
66 |
|
67 |
# Wait for the user to upload a file
|
68 |
+
while files is None:
|
69 |
+
# Note: This now accepts both text/plain and application/pdf files
|
70 |
files = await cl.AskFileMessage(
|
71 |
+
content="Please upload a text or PDF file to begin!",
|
72 |
+
accept=["text/plain", "application/pdf"],
|
73 |
+
max_size_mb=20, # Assuming PDFs might be larger
|
74 |
timeout=180,
|
75 |
).send()
|
76 |
|
77 |
file = files[0]
|
78 |
|
79 |
+
# Notify the user that their file is being processed
|
80 |
+
msg = cl.Message(content=f"Processing `{file.name}`...")
|
|
|
81 |
await msg.send()
|
82 |
|
83 |
+
# Initialize an empty list for texts, this will be populated based on file type
|
84 |
+
texts = []
|
|
|
|
|
85 |
|
86 |
+
# Check the file type and process accordingly
|
87 |
+
if file.content_type == "text/plain":
|
88 |
+
# Handle text file
|
89 |
+
with open(file.path, "r", encoding="utf-8") as f:
|
90 |
+
text = f.read()
|
91 |
+
texts.append(text) # Add the text to the texts list
|
|
|
|
|
92 |
|
93 |
+
# Update the user about the text file
|
94 |
+
await cl.Message(
|
95 |
+
content=f"`{file.name}` uploaded, it contains {len(text)} characters!"
|
96 |
+
).send()
|
97 |
+
|
98 |
+
elif file.content_type == "application/pdf":
|
99 |
+
# Handle PDF file
|
100 |
+
texts = process_file(file) # Assuming process_file() is a function you've defined to extract text from PDF
|
101 |
+
|
102 |
+
# Create metadata for each chunk
|
103 |
+
metadatas = [{"source": f"{i}-pl"} for i in range(len(texts))]
|
104 |
+
|
105 |
+
# Create a Chroma vector store
|
106 |
+
embeddings = OpenAIEmbeddings()
|
107 |
+
docsearch = await cl.make_async(Chroma.from_texts)(
|
108 |
+
texts, embeddings, metadatas=metadatas
|
109 |
+
)
|
110 |
+
|
111 |
+
# The rest of your setup, like creating the chain, goes here
|
112 |
+
# This part is unchanged from your second snippet
|
113 |
message_history = ChatMessageHistory()
|
114 |
|
115 |
memory = ConversationBufferMemory(
|
|
|
119 |
return_messages=True,
|
120 |
)
|
121 |
|
|
|
122 |
chain = ConversationalRetrievalChain.from_llm(
|
123 |
ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0, streaming=True),
|
124 |
chain_type="stuff",
|