Indus 3.0 Demo

app.py

@@ -1,64 +1,43 @@
+import os, tarfile
+from huggingface_hub import hf_hub_download
+from transformers import AutoTokenizer, AutoModelForCausalLM
 import gradio as gr
-from huggingface_hub import InferenceClient
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+# At startup, download and extract model
+archive_path = hf_hub_download(
+    repo_id="SamOrion/Llama_3.2_3b_Hindi_Pruned",
+    filename="llama-3.2-3b-hindi-pruned.tar.gz",
+    repo_type="model"
 )
-
+extract_dir = "/data/model"
+os.makedirs(extract_dir, exist_ok=True)
+with tarfile.open(archive_path, "r:gz") as tar:
+    tar.extractall(path=extract_dir)
+
+# Load from the extracted folder
+tokenizer = AutoTokenizer.from_pretrained(extract_dir)
+model     = AutoModelForCausalLM.from_pretrained(extract_dir, torch_dtype="auto", device_map="auto")
+
+def chat_fn(prompt, history):
+    history = history or []
+    history.append({"role": "user", "content": prompt})
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=100)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    history.append({"role": "assistant", "content": response})
+    return history, ""
+
+with gr.Blocks() as demo:
+    gr.Markdown("## 🌐 Indus 3.0 Demo")
+    chat = gr.Chatbot(type="messages")
+    msg = gr.Textbox(placeholder="Type here...")
+    clear = gr.Button("Clear")
+    msg.submit(chat_fn, [msg, chat], [chat, msg])
+    clear.click(lambda: None, None, chat)
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=True
+    )

requirements.txt

@@ -1 +1,6 @@
-huggingface_hub==0.25.2
+huggingface_hub==0.25.2
+torch>=2.0.0
+transformers>=4.30.0
+gradio>=3.0.0
+accelerate>=0.20.0
+bitsandbytes>=0.39.0