first

.idea/Gemma_2B_DataScience_RAFT_QA.iml

@@ -1,7 +1,9 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <module type="PYTHON_MODULE" version="4">
   <component name="NewModuleRootManager">
-    <content url="file://$MODULE_DIR$" />
+    <content url="file://$MODULE_DIR$">
+      <excludeFolder url="file://$MODULE_DIR$/.venv_custom_v" />
+    </content>
     <orderEntry type="inheritedJdk" />
     <orderEntry type="sourceFolder" forTests="false" />
   </component>

.idea/misc.xml

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="Black">
+    <option name="sdkName" value="Python 3.12 (Gemma_2B_DataScience_RAFT_QA)" />
+  </component>
+  <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.12 (Gemma_2B_DataScience_RAFT_QA)" project-jdk-type="Python SDK" />
+</project>

app.py

@@ -15,17 +15,11 @@ pipeline = transformers.pipeline(
     device="cuda",
 )
 
+def format_test_question(q):
+    return f"<bos><start_of_turn>user {q} <end_of_turn>model "
 @spaces.GPU
-def chat_function(message, history, system_prompt,max_new_tokens,temperature):
-    messages = [
-        {"role": "system", "content": system_prompt},
-        {"role": "user", "content": message},
-    ]
-    prompt = pipeline.tokenizer.apply_chat_template(
-        messages,
-        tokenize=False,
-        add_generation_prompt=True
-    )
+def chat_function(message, history,max_new_tokens,temperature):
+    prompt = format_test_question(message)
     print(prompt)
     temp = temperature + 0.1
     outputs = pipeline(
@@ -41,14 +35,14 @@ gr.ChatInterface(
     chat_function,
     chatbot=gr.Chatbot(height=400),
     textbox=gr.Textbox(placeholder="Enter message here", container=False, scale=7),
-    title="LLAMA 3 8B Chat",
+    title="Gemma 2B Data Science QA RAFT Demo",
     description="""
     This space is dedicated for chatting with Gemma 2B Finetuned for Data Science QA using RAFT. Find this model here: https://huggingface.co/ModularityAI/gemma-2b-datascience-it-raft
     Feel free to play with customization in the "Additional Inputs".
+    Fine tune Notebook: https://www.kaggle.com/code/hanzlajavaid/gemma-finetuning-raft-technique
     """,
-    theme="soft",
+    theme="Monochrome",
     additional_inputs=[
-        gr.Textbox("You are helpful AI.", label="System Prompt"),
         gr.Slider(512, 4096, label="Max New Tokens"),
         gr.Slider(0, 1, label="Temperature")
     ]