new flask-based server for running scripts on huggingface

.ipynb_checkpoints/app-checkpoint.py

@@ -0,0 +1,18 @@
+from flask import Flask,request,render_template
+import os
+
+app=Flask(__name__)
+
+@app.route('/', methods=['GET', 'POST'])
+def home():
+    return render_template
+def run_script():
+    script=request.args.get('script')
+    os.system(script+' > ./out.txt')
+    with open('./out.txt','r') as f: output=f.read()
+    return output
+
+if __name__=='__main__':
+    app.run()
+
+

README.md

@@ -12,9 +12,5 @@ license: unknown
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
-Simple example of a gradio UI that calls anvilserver.py running on any
-machine with internet connection. Can be a laptop also.
-Just push app.py and requirements.txt to gradio and 
-pip install anvil_uplink on your server (e.g. laptop) machine
-and run anvilserver.py there. The gradio UI calls the server_function
-published within anvilserver.py. This uses anvil as a router.
+Flask server to run on huggingface.
+Just push app.py and requirements.txt to repo and call via REST API

anvilserver.py

@@ -1,9 +0,0 @@
-import anvil.server
-
-anvil.server.connect('55MH4EBKM22EP4E6D5T6CVSL-VGO5X4SM6JEXGJVT')
-
-@anvil.server.callable
-def server_function(arg):
-    return arg+' called anvilserver.py'
-
-anvil.server.wait_forever()

app.py

@@ -1,50 +1,32 @@
-import gradio as gr
-from threading import Thread
-import time
-import anvil.server
-from registration import register,get_register,func_reg
-from library import get_file,get_files
+from flask import Flask,request,render_template
 import os
-anvil.server.connect('55MH4EBKM22EP4E6D5T6CVSL-VGO5X4SM6JEXGJVT')
-register(get_file)
-register(get_files)
 
-# with gr.Blocks() as block:
-#     textbox = gr.inputs.Textbox(label='Function Register')
-#     button = gr.Button(value="Show Function Calls")
-#     button.click(get_register,inputs=None,outputs=[textbox])
-
-# block.launch()
-import json
-import ast
-def my_inference_function(name):
-#   print(ast.literal_eval(name)['name'])
-    os.system(name+' > ./out.txt')
+app=Flask(__name__)
+MESSAGED={'title':'Script Server',
+          'messageL':['script in "script" parameter or field of json input to /run will be run']}
+
+@app.route('/run',methods=['GET','POST'])
+def run_script():
+    script=''
+    # print(request.method)
+    print(request)
+    if request.method=='GET': 
+        script=request.args.get('script')
+        print('I am in get')
+    elif request.method=='POST':
+        print('I am in post')
+        data=request.get_json()
+        if 'script' in data: script=data['script']
+    if script=='' or script is None: return 'INVALID'
+    os.system(script+' > ./out.txt')
     with open('./out.txt','r') as f: output=f.read()
     return output
-#  return "Input Data: " + name + ", stay tuned for ML models from this API"
-
-gradio_interface = gr.Interface(
-  fn=my_inference_function,
-  inputs="text",
-  outputs="text",
-  title="REST API with Gradio and Huggingface Spaces",
-  description='''Inputs should be json of test item e.g., as a dictionary; 
-  output right now is just returning the input; later label will be returned.
-
-  This is how to call the API from Python: 
-
-  import requests
-
-  response = requests.post("https://gmshroff-gmserver.hf.space/run/predict", json={
-  "data": [
-    "\<put some json string here\>",
-  ]}).json()
 
-  data = response["data"])
+@app.route('/',methods=['GET', 'POST'])
+def home():
+    return render_template('home.html',messageD=MESSAGED)
 
-  ''')
+if __name__=='__main__':
+    app.run()
 
-gradio_interface.launch()
-# anvil.server.wait_forever()
 

client.py

@@ -1,8 +0,0 @@
-from gradio_client import Client
-
-client = Client("https://gmshroff-gmserver.hf.space/")
-result = client.predict(
-				"{'name':'Gautam'}",	# str in 'name' Textbox component
-				api_name="/predict"
-)
-print(result)

library.py

@@ -1,26 +0,0 @@
-from registration import update_reg
-from anvil.tables import app_tables
-from PyPDF2 import PdfReader
-from io import BytesIO
-# from notebook_funcs import get_message
-
-def get_files():
-    update_reg('get_files')
-    retval=''
-    query=app_tables.files.search()
-    for r in query:
-        retval+=r['filename']+'\n'
-    return retval
-
-def get_file(name):
-    update_reg('get_file')
-    retval=''
-    r=app_tables.files.get(filename=name)
-    bts=r['file'].get_bytes()
-    sio=BytesIO(bts)
-    with sio as f:
-        reader=PdfReader(f)
-        fpage=reader.pages[0].extract_text()
-    retval+=r['filename']+'\n'+ fpage
-    return retval
-    

out.txt

@@ -1 +1 @@
-/Users/a112956/TryCode/huggingface/gmserver
+Now please copy the file to server via: scp ../../data/validation_subset_int8.parquet gms@gms1:/home/gms/numerai/data/.

registration.py

@@ -1,21 +0,0 @@
-import anvil.server
-
-func_reg={}
-
-def register(function):
-    global func_reg
-    func={}
-    func['func']=function
-    func['calls']=0
-    func_reg[function.__name__]=func
-    anvil.server.callable(function)
-
-def update_reg(name):
-    global func_reg
-    func_reg[name]['calls']+=1
-
-def get_register():
-    regstr=''
-    for f in func_reg:
-        regstr+=f + ' called ' + str(func_reg[f]['calls']) + ' times\n'
-    return regstr

requirements.txt

@@ -1,2 +1,2 @@
-anvil_uplink
-PyPDF2
+numerapi
+flask

static/dummy.txt

@@ -0,0 +1 @@
+dummy file

templates/home.html

@@ -0,0 +1,40 @@
+<html>
+ <head> 
+    <title>gmserver</title> 
+    <style>
+    * {
+      box-sizing: border-box;
+    }
+    /* Create two equal columns that floats next to each other */
+    .columnleft {
+      float: left;
+      padding: 10px;
+      width: 15%;
+    }
+    .columncenter {
+      float: left;
+      padding: 10px;
+      width: 80%;
+    }    
+    /* Clear floats after the columns */
+    .row:after {
+      content: "";
+      display: table;
+      clear: both;
+    }
+    </style>
+ </head>
+ <body>
+  <div class="row">
+  <div class="columnleft">
+  <img src="/static/logo.jpeg" width="50px">
+  </div>
+  <div class="columncenter" align="center">
+   <h1>{{messageD['title']}}</h1>
+    {% for message in messageD['messageL']:%}
+     <h4>{{message}}</h4>
+    {% endfor %}
+  </div>
+  </div>
+ </body>
+</html>

test.ipynb

@@ -0,0 +1,84 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import requests\n",
+    "import json\n",
+    "from urllib.request import urlretrieve\n",
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "headers = {'Content-Type': 'application/json'}\n",
+    "url='http://127.0.0.1:5000/run'\n",
+    "body={\"script\":\"python update_valdata.py\"}"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "response=requests.post(url=url,data=json.dumps(body),headers = {'Content-Type': 'application/json'})"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "response.content"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "urlretrieve(url='http://127.0.0.1:5000/static/validation_subset_int8.parquet',filename='./validation_subset_int8.parquet')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df=pd.read_parquet('/tmp/validation_subset_int8.parquet')"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "py39numerai",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

test_rest.py

@@ -1,9 +0,0 @@
-import requests
-
-response = requests.post("https://gmshroff-gmserver.hf.space/run/predict", json={
-  "data": [
-    "<some json string>",
-]}).json()
-
-data = response["data"]
-print(data)

update_valdata.py

@@ -0,0 +1,34 @@
+#!/usr/bin/env python
+# coding: utf-8
+
+# In[ ]:
+
+
+from numerapi import NumerAPI
+import os
+import pandas as pd
+import numpy as np
+
+
+# In[ ]:
+
+
+napi = NumerAPI()
+data_path='./static/'
+
+
+# In[ ]:
+
+
+napi.download_dataset("v4.2/validation_int8.parquet", data_path+"validation_int8.parquet")
+validation_data=pd.read_parquet(data_path+"validation_int8.parquet")
+recent_eras=list(validation_data.loc[validation_data['data_type']=='validation']['era'].unique()[-2:])
+validation_subset=validation_data[validation_data['era'].isin(recent_eras)]
+validation_subset.to_parquet(data_path+"validation_subset_int8.parquet",index=False)
+
+
+# In[ ]:
+
+
+print("Now please copy the file to server via: scp ../../data/validation_subset_int8.parquet gms@gms1:/home/gms/numerai/data/.")
+