ML app added, SDK changed to gradio

Dockerfile

@@ -1,13 +0,0 @@
-FROM python:3.10
-
-ADD requirements.txt .
-
-RUN pip install -r requirements.txt
-
-RUN rm requirements.txt
-
-ADD app.py .
-
-EXPOSE 8001
-
-CMD python app.py

README.md

@@ -3,11 +3,11 @@ title: SummaryDemo
 emoji: 🔥
 colorFrom: yellow
 colorTo: yellow
-sdk: docker
+sdk: gradio
 sdk_version: 4.24.0
 app_file: app.py
 pinned: false
-app_port: 8001
+#app_port: 8001
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -1,13 +1,50 @@
+import torch
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+from peft import PeftModel
 import gradio as gr
 
-def greet(name):
-    return "Hello " + name + "!!"
+username = "harikach24"
 
-iface = gr.Interface(fn=greet,
-                     inputs="text",
-                     outputs="text",
-                     title="Simple Application",
-                     description="via docker",
+base_checkpoint = username + '/dialogue_Summary'
+
+peft_model_id = username + '/dialogue_Summary_peft'
+
+# Load tokenizer
+tokenizer = AutoTokenizer.from_pretrained(base_checkpoint)
+
+# Load Base model
+peft_model_base = AutoModelForSeq2SeqLM.from_pretrained(base_checkpoint)
+
+# Load PEFT model
+loaded_peft_model = PeftModel.from_pretrained(
+            model=peft_model_base, model_id = peft_model_id, is_trainable=False,)
+
+def generate_summary(dialogue):
+    llm = loaded_peft_model
+    """Prepare prompt -> tokenize -> generate output using LLM -> detokenize
+output"""
+
+    input_prompt = f"""
+                    Summarize the following conversation.
+                    {dialogue}
+                    Summary:
+                    """
+
+    input_ids = tokenizer(input_prompt, return_tensors='pt')
+
+    tokenized_output = llm.generate(input_ids=input_ids['input_ids'],
+                                    min_length=30, max_length=200, )
+
+    summary = tokenizer.decode(tokenized_output[0], skip_special_tokens=True)
+
+    return summary
+
+# Create the Gradio demo
+iface = gr.Interface(fn=generate_summary,
+                     inputs="textbox",
+                     outputs="textbox",
+                     title="Dialogue Summarization",
+                     description="via gradio",
                      allow_flagging="never",
                      )
-iface.launch(server_name = "0.0.0.0", server_port = 8001)
+iface.launch()

requirements.txt

@@ -1,4 +1,6 @@
 # Dependencies for application to be added to this file
 
 numpy
-gradio
+gradio
+transformers[torch]
+peft