Nadesh Seen
2 years ago
parent
6632d78875
commit
02accd6fd6
Binary file not shown.
@ -1,3 +0,0 @@
|
||||
AWS_ACCESS_KEY_ID="AKIAYFB773UVZSOAVZN4"
|
||||
AWS_SECRET_ACCESS_KEY="OZPLMjN/2ao6OlSd5PpIkT5d7cWD9WAP/DXSZbEs"
|
||||
AWS_REGION="ap-south-1"
|
||||
@ -0,0 +1 @@
|
||||
from . import *
|
||||
Binary file not shown.
Binary file not shown.
@ -1,12 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
# ./register.sh /decode-function /decode decode-action [SAMPLE USE]
|
||||
|
||||
function_dir_name=$1
|
||||
docker_image_name=$2
|
||||
|
||||
cd $function_dir_name
|
||||
|
||||
chmod -R 777 ./
|
||||
|
||||
./buildAndPush.sh $docker_image_name
|
||||
@ -1,550 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
|
||||
import sys
|
||||
import requests
|
||||
import uuid
|
||||
import re
|
||||
import subprocess
|
||||
import threading
|
||||
import queue
|
||||
import redis
|
||||
import pickle
|
||||
import json
|
||||
import os
|
||||
import time
|
||||
from requests.packages.urllib3.exceptions import InsecureRequestWarning
|
||||
from flask import Flask, request,jsonify,send_file
|
||||
requests.packages.urllib3.disable_warnings(InsecureRequestWarning)
|
||||
import pymongo
|
||||
|
||||
import trigger_gateway
|
||||
|
||||
|
||||
app = Flask(__name__)
|
||||
|
||||
action_url_mappings = {} #Store action->url mappings
|
||||
action_properties_mapping = {} #Stores the action name and its corresponding properties
|
||||
responses = []
|
||||
queue = []
|
||||
list_of_func_ids = []
|
||||
|
||||
def hello():
|
||||
print("Hello")
|
||||
|
||||
|
||||
def preprocess(filename):
|
||||
with open(filename) as f:
|
||||
lines = f.readlines()
|
||||
action_url_list = []
|
||||
for line in lines:
|
||||
line = line.replace("\n", "")
|
||||
line = line.replace("/guest/","")
|
||||
action_url_list.append(line)
|
||||
for item in action_url_list:
|
||||
action_name = item.split(' ')[0]
|
||||
url = item.split(' ')[1]
|
||||
action_url_mappings[action_name] = url
|
||||
|
||||
|
||||
def execute_thread(action,redis,url,json):
|
||||
reply = requests.post(url = url,json=json,verify=False)
|
||||
list_of_func_ids.append(reply.json()["activation_id"])
|
||||
redis.set(action+"-output",pickle.dumps(reply.json()))
|
||||
responses.append(reply.json())
|
||||
|
||||
|
||||
def handle_parallel(queue,redis,action_properties_mapping,parallel_action_list):
|
||||
thread_list = []
|
||||
output_list = [] # List to store the output of actions whose outputs are required by downstream operations
|
||||
|
||||
for action in parallel_action_list:
|
||||
action_names = action_properties_mapping[action]["outputs_from"]
|
||||
next_action = action_properties_mapping[action]["next"]
|
||||
if(next_action!=""):
|
||||
if next_action not in queue:
|
||||
queue.append(next_action)
|
||||
if(len(action_names)==1): # if only output of one action is required
|
||||
key = action_names[0]+"-output"
|
||||
output = pickle.loads(redis.get(key))
|
||||
action_properties_mapping[action]["arguments"] = output
|
||||
else:
|
||||
for item in action_names:
|
||||
key = item+"-output"
|
||||
output = pickle.loads(redis.get(key))
|
||||
output_list.append(output)
|
||||
|
||||
action_properties_mapping[action]["arguments"] = output_list
|
||||
|
||||
url = action_url_mappings[action]
|
||||
thread_list.append(threading.Thread(target=execute_thread, args=[action,redis,url,action_properties_mapping[action]["arguments"]]))
|
||||
for thread in thread_list:
|
||||
thread.start()
|
||||
for thread in thread_list:
|
||||
thread.join()
|
||||
action_properties_mapping[next_action]["arguments"] = responses
|
||||
return responses
|
||||
|
||||
def create_redis_instance():
|
||||
r = redis.Redis(host="10.129.28.219", port=6379, db=2)
|
||||
return r
|
||||
|
||||
def get_redis_contents(r):
|
||||
keys = r.keys()
|
||||
for key in keys:
|
||||
value = pickle.loads(r.get(key))
|
||||
if value is not None:
|
||||
print(f"{key.decode('utf-8')}: {json.dumps(value, indent=4)}")
|
||||
|
||||
def connect_mongo():
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dags"]
|
||||
return mycol
|
||||
|
||||
def get_dag_json(dag_name):
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dags"]
|
||||
query = {"name":dag_name}
|
||||
projection = {"_id": 0, "name": 1,"dag":1}
|
||||
document = mycol.find(query, projection)
|
||||
data = list(document)
|
||||
return data
|
||||
|
||||
def submit_dag_metadata(dag_metadata):
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dag_metadata"]
|
||||
try:
|
||||
cursor = mycol.insert_one(dag_metadata)
|
||||
# print("OBJECT ID GENERATED",cursor.inserted_id)
|
||||
data = {"message":"success"}
|
||||
return json.dumps(data)
|
||||
except Exception as err:
|
||||
data = {"message":"failed","reason":err}
|
||||
return json.dumps(data)
|
||||
|
||||
@app.route("/")
|
||||
def home():
|
||||
data = {"message": "Hello,welcome to create and manage serverless workflows.","author":"Anubhav Jana"}
|
||||
return jsonify(data)
|
||||
|
||||
@app.route('/view/functions', methods=['GET'])
|
||||
def list_actions():
|
||||
list_of_actions = []
|
||||
stream = os.popen(' wsk -i action list')
|
||||
actions = stream.read().strip().split(' ')
|
||||
for action in actions:
|
||||
if action=='' or action=='private' or action=='blackbox':
|
||||
continue
|
||||
else:
|
||||
list_of_actions.append(action.split('/')[2])
|
||||
data = {"list of available actions":list_of_actions}
|
||||
return jsonify(data)
|
||||
|
||||
|
||||
|
||||
@app.route('/register/trigger/',methods=['POST'])
|
||||
def register_trigger():
|
||||
trigger_json = request.json
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["trigger_store"]
|
||||
mycol = mydb["triggers"]
|
||||
try:
|
||||
cursor = mycol.insert_one(trigger_json)
|
||||
print("OBJECT ID GENERATED",cursor.inserted_id)
|
||||
if(trigger_json["type"]=="dag"):
|
||||
targets = trigger_json["dags"]
|
||||
elif(trigger_json["type"]=="function"):
|
||||
targets = trigger_json["functions"]
|
||||
data = {"message":"success","trigger_name":trigger_json["trigger_name"],"trigger":trigger_json["trigger"],"trigger_type":trigger_json["type"],"trigger_target":targets}
|
||||
return json.dumps(data)
|
||||
except Exception as e:
|
||||
print("Error--->",e)
|
||||
data = {"message":"fail","reason":e}
|
||||
return json.dumps(data)
|
||||
|
||||
|
||||
@app.route('/register/function/<function_name>',methods=['POST'])
|
||||
def register_function(function_name):
|
||||
list_of_file_keys = []
|
||||
document = {}
|
||||
function_dir = '/home/faasapp/Desktop/anubhav/function_modules' # Library of functions
|
||||
new_dir = function_name
|
||||
destination = os.path.join(function_dir, new_dir)
|
||||
# Create the directory
|
||||
os.makedirs(destination, exist_ok=True)
|
||||
files = request.files
|
||||
for filekey in files:
|
||||
if filekey!='description':
|
||||
list_of_file_keys.append(filekey)
|
||||
for key in list_of_file_keys:
|
||||
file = request.files[key]
|
||||
filename = file.filename
|
||||
# Save, copy, remove
|
||||
file.save(file.filename)
|
||||
shutil.copy(filename, destination)
|
||||
os.remove(filename)
|
||||
image_build_script = 'buildAndPush.sh'
|
||||
shutil.copy(image_build_script, destination)
|
||||
|
||||
# Prepare data
|
||||
document["function_name"] = function_name
|
||||
document["image_build_script"] = 'buildAndPush.sh'
|
||||
document["python_script"] = (request.files[list_of_file_keys[0]]).filename
|
||||
document["dockerfile"] = (request.files[list_of_file_keys[1]]).filename
|
||||
document["requirements.txt"] =(request.files[list_of_file_keys[2]]).filename
|
||||
|
||||
docker_image_name = "10.129.28.219:5000/"+function_name+"-image"
|
||||
api_name = "/"+function_name+"-api"
|
||||
path_name = "/"+function_name+"-path"
|
||||
password = '1234'
|
||||
# build docker image
|
||||
cmd = ["sudo", "-S", "/home/faasapp/Desktop/anubhav/controlplane/build_image.sh",destination,docker_image_name]
|
||||
# open subprocess with Popen
|
||||
process = subprocess.Popen(cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE, universal_newlines=True)
|
||||
|
||||
# pass password to standard input
|
||||
process.stdin.write(password + "\n")
|
||||
process.stdin.flush()
|
||||
|
||||
# wait for process to complete and get output
|
||||
output, errors = process.communicate()
|
||||
print("OUTPUT---------",output)
|
||||
print("ERRORS---------",errors)
|
||||
# if(errors):
|
||||
# print("There is error building docker file")
|
||||
# data = {"message":"fail","reason":"docker build failed"}
|
||||
# return json.dumps(data)
|
||||
# else:
|
||||
|
||||
# create action, register action with api, populate its mapping
|
||||
subprocess.call(['./create_action.sh',destination,docker_image_name,function_name])
|
||||
subprocess.call(['./register.sh',api_name,path_name,function_name])
|
||||
subprocess.call(['bash', './actions.sh'])
|
||||
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["function_store"]
|
||||
mycol = mydb["functions"]
|
||||
try:
|
||||
cursor = mycol.insert_one(document)
|
||||
print("OBJECT ID GENERATED",cursor.inserted_id)
|
||||
data = {"message":"success"}
|
||||
return json.dumps(data)
|
||||
except Exception as e:
|
||||
print("Error--->",e)
|
||||
data = {"message":"fail","reason":e}
|
||||
return json.dumps(data)
|
||||
|
||||
# data = {"message":"success"}
|
||||
# return json.dumps(data)
|
||||
|
||||
|
||||
@app.route('/register/dag/',methods=['POST'])
|
||||
def register_dag():
|
||||
dag_json = request.json
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dags"]
|
||||
try:
|
||||
cursor = mycol.insert_one(dag_json)
|
||||
print("OBJECT ID GENERATED",cursor.inserted_id)
|
||||
data = {"message":"success"}
|
||||
return json.dumps(data)
|
||||
except Exception as e:
|
||||
print("Error--->",e)
|
||||
data = {"message":"fail","reason":e}
|
||||
return json.dumps(data)
|
||||
|
||||
@app.route('/view/dag/<dag_name>',methods=['GET'])
|
||||
def view_dag(dag_name):
|
||||
dag_info_map = {}
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dags"]
|
||||
document = mycol.find({"name":dag_name})
|
||||
data = list(document)
|
||||
dag_info_list = []
|
||||
for items in data:
|
||||
dag_info_list = items["dag"]
|
||||
dag_info_map["DAG_Name--->>"] = items["name"]
|
||||
|
||||
dag_info_map["Number_of_nodes-->"] = len(dag_info_list)
|
||||
dag_info_map["Starting_Node-->"] = dag_info_list[0]["node_id"]
|
||||
|
||||
for dag_items in dag_info_list:
|
||||
node_info_map = {}
|
||||
if(len(dag_items["properties"]["outputs_from"])==0):
|
||||
node_info_map["get_outputs_from-->"] = "Starting action->No outputs consumed"
|
||||
else:
|
||||
node_info_map["get_outputs_from-->"] = dag_items["properties"]["outputs_from"]
|
||||
node_info_map["primitive_type"] = dag_items["properties"]["primitive"]
|
||||
if(dag_items["properties"]["primitive"]=="condition"):
|
||||
node_info_map["next_node_id_if_condition_true"] = dag_items["properties"]["branch_1"]
|
||||
node_info_map["next_node_id_if_condition_false"] = dag_items["properties"]["branch_2"]
|
||||
else:
|
||||
if(dag_items["properties"]["next"]!=""):
|
||||
node_info_map["next_node_id-->"] = dag_items["properties"]["next"]
|
||||
else:
|
||||
node_info_map["next_node_id-->"] = "Ending node_id of a path"
|
||||
dag_info_map[dag_items["node_id"]] = node_info_map
|
||||
response = {"dag_data":dag_info_map}
|
||||
formatted_json = json.dumps(response, indent=20)
|
||||
return formatted_json
|
||||
|
||||
@app.route('/view/dags',methods=['GET'])
|
||||
def view_dags():
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dags"]
|
||||
document = mycol.find()
|
||||
data = list(document)
|
||||
# Serialize the data to JSON
|
||||
json_data = json.dumps(data, default=str)
|
||||
json_string ='{"dag":'+str(json_data)+'}'
|
||||
data = json.loads(json_string)
|
||||
# Format the JSON string with indentation
|
||||
formatted_json = json.dumps(data, indent=4)
|
||||
return formatted_json
|
||||
|
||||
@app.route('/view/triggers',methods=['GET'])
|
||||
def view_triggers():
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["trigger_store"]
|
||||
mycol = mydb["triggers"]
|
||||
document = mycol.find()
|
||||
data = list(document)
|
||||
# Serialize the data to JSON
|
||||
json_data = json.dumps(data, default=str)
|
||||
json_string ='{"trigger":'+str(json_data)+'}'
|
||||
data = json.loads(json_string)
|
||||
# Format the JSON string with indentation
|
||||
formatted_json = json.dumps(data, indent=4)
|
||||
return formatted_json
|
||||
|
||||
@app.route('/view/trigger/<trigger_name>',methods=['GET'])
|
||||
def view_trigger(trigger_name):
|
||||
print(request.url)
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["trigger_store"]
|
||||
mycol = mydb["triggers"]
|
||||
query = {"trigger_name":trigger_name}
|
||||
projection = {"_id": 0,"trigger_name":1,"type":1,"trigger":1,"dags":1,"functions":1}
|
||||
document = mycol.find(query,projection)
|
||||
data = list(document)
|
||||
# print(data)
|
||||
json_data = json.dumps(data, default=str)
|
||||
json_string ='{"trigger":'+str(json_data)+'}'
|
||||
data = json.loads(json_string)
|
||||
formatted_json = json.dumps(data, indent=4)
|
||||
return formatted_json
|
||||
|
||||
# EXAMPLE URL: http://10.129.28.219:5001/view/activation/8d7df93e8f2940b8bdf93e8f2910b80f
|
||||
@app.route('/view/activation/<activation_id>', methods=['GET', 'POST'])
|
||||
def list_activations(activation_id):
|
||||
# activation_id = '74a7b6c707d14973a7b6c707d1a97392'
|
||||
cmd = ['wsk', '-i', 'activation', 'get', activation_id]
|
||||
result = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
|
||||
json_res = result.stdout.decode().split('\n')[1:] # Ignore first line of output
|
||||
res = json.loads('\n'.join(json_res))
|
||||
d={}
|
||||
d["action_name"] = res["name"]
|
||||
d["duration"] = res["duration"]
|
||||
d["status"] = res["response"]["status"]
|
||||
d["result"] = res["response"]["result"]
|
||||
return({"action_name":res["name"],
|
||||
"duration": res["duration"],
|
||||
"status": res["response"]["status"],
|
||||
"result":res["response"]["result"]
|
||||
})
|
||||
|
||||
# EXAMPLE URL: http://10.129.28.219:5001/view/76cc8a53-0a63-47bb-a5b5-9e6744f67c61
|
||||
@app.route('/view/<dag_id>',methods=['GET'])
|
||||
def view_dag_metadata(dag_id):
|
||||
myclient = pymongo.MongoClient("mongodb://127.0.0.1/27017")
|
||||
mydb = myclient["dag_store"]
|
||||
mycol = mydb["dag_metadata"]
|
||||
query = {"dag_id":dag_id}
|
||||
projection = {"_id": 0,"dag_id":1,"dag_name":1,"function_activation_ids":1}
|
||||
document = mycol.find(query, projection)
|
||||
data = list(document)
|
||||
response = {"dag_metadata":data}
|
||||
return json.dumps(response)
|
||||
|
||||
# EXAMPLE URL: http://10.129.28.219:5001/run/action/odd-even-action
|
||||
# http://10.129.28.219:5001/run/action/decode-function
|
||||
|
||||
@app.route('/run/action/<action_name>/', methods=['POST'])
|
||||
def execute_action(action_name):
|
||||
script_file = './actions.sh'
|
||||
subprocess.call(['bash', script_file])
|
||||
preprocess("action_url.txt")
|
||||
url = action_url_mappings[action_name]
|
||||
# json_data = json.loads(request.json)
|
||||
reply = requests.post(url = url,json = request.json,verify=False)
|
||||
return reply.json()
|
||||
|
||||
|
||||
|
||||
# EXAMPLE URL: http://10.129.28.219:5001/run/dag/odd-even-test/{"number":16}
|
||||
@app.route('/run/dag/<dag_name>/', methods=['GET', 'POST'])
|
||||
def execute_dag(dag_name):
|
||||
print("------------------------------------DAG START-----------------------------------------------")
|
||||
unique_id = uuid.uuid4()
|
||||
print("DAG UNIQUE ID----------",unique_id)
|
||||
dag_metadata={}
|
||||
dag_metadata["dag_id"] = str(unique_id)
|
||||
dag_metadata["dag_name"] = dag_name
|
||||
# list_of_func_ids = []
|
||||
######### Updates the list of action->url mapping ###################
|
||||
script_file = './actions.sh'
|
||||
subprocess.call(['bash', script_file])
|
||||
#####################################################################
|
||||
preprocess("action_url.txt")
|
||||
|
||||
### Create in-memory redis storage ###
|
||||
redis_instace = create_redis_instance()
|
||||
#######################################
|
||||
|
||||
action_properties_mapping = {} #Stores the action name and its corresponding properties
|
||||
|
||||
|
||||
dag_res = json.loads(json.dumps(get_dag_json(dag_name)))
|
||||
dag_data = dag_res[0]["dag"]
|
||||
for dag_item in dag_data:
|
||||
action_properties_mapping[dag_item["node_id"]] = dag_item["properties"]
|
||||
|
||||
flag = 0
|
||||
for dag_item in dag_data:
|
||||
if(flag==0): # To indicate the first action in the DAG
|
||||
queue.append(dag_item["node_id"])
|
||||
action_properties_mapping[dag_item["node_id"]]["arguments"] = request.json
|
||||
while(len(queue)!=0):
|
||||
flag=flag+1
|
||||
action = queue.pop(0)
|
||||
print("ACTION DEQUEUED FROM QUEUE : --->",action)
|
||||
##########################################################
|
||||
# HANDLE THE ACTION #
|
||||
##########################################################
|
||||
if isinstance(action, str):
|
||||
# if(isinstance(action_properties_mapping[action]['arguments'],list)):
|
||||
# pass
|
||||
json_data = action_properties_mapping[action]["arguments"]
|
||||
url = action_url_mappings[action]
|
||||
reply = requests.post(url = url,json=json_data,verify=False)
|
||||
list_of_func_ids.append(reply.json()["activation_id"])
|
||||
# print("Line 292------------",reply.json()["activation_id"])
|
||||
redis_instace.set(action+"-output",pickle.dumps(reply.json()))
|
||||
action_type = action_properties_mapping[action]["primitive"]
|
||||
|
||||
if(action_type=="condition"):
|
||||
branching_action = action_properties_mapping[action]["branch_1"]
|
||||
alternate_action = action_properties_mapping[action]["branch_2"]
|
||||
result=reply.json()["result"]
|
||||
condition_op = action_properties_mapping[action]["condition"]["operator"]
|
||||
if(condition_op=="equals"):
|
||||
if(isinstance(action_properties_mapping[action]["condition"]["target"], str)):
|
||||
target = action_properties_mapping[action]["condition"]["target"]
|
||||
else:
|
||||
target=int(action_properties_mapping[action]["condition"]["target"])
|
||||
|
||||
if(result==target):
|
||||
output_list = [] # List to store the output of actions whose outputs are required by downstream operations
|
||||
queue.append(branching_action)
|
||||
action_names = action_properties_mapping[branching_action]["outputs_from"] # Get the list of actions whose output will be used
|
||||
if(len(action_names)==1): # if only output of one action is required
|
||||
key = action_names[0]+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
action_properties_mapping[branching_action]["arguments"] = output
|
||||
else:
|
||||
for item in action_names:
|
||||
key = item+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
output_list.append(output)
|
||||
action_properties_mapping[branching_action]["arguments"] = output_list
|
||||
|
||||
else:
|
||||
output_list = [] # List to store the output of actions whose outputs are required by downstream operations
|
||||
queue.append(alternate_action)
|
||||
action_names = action_properties_mapping[alternate_action]["outputs_from"] # Get the list of actions whose output will be used
|
||||
if(len(action_names)==1): # if only output of one action is required
|
||||
key = action_names[0]+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
action_properties_mapping[alternate_action]["arguments"] = output
|
||||
else:
|
||||
for item in action_names:
|
||||
key = item+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
output_list.append(output)
|
||||
action_properties_mapping[alternate_action]["arguments"] = output_list
|
||||
|
||||
|
||||
if(condition_op=="greater_than"):
|
||||
pass
|
||||
if(condition_op=="greater_than_equals"):
|
||||
pass
|
||||
if(condition_op=="less_than"):
|
||||
pass
|
||||
if(condition_op=="less_than_equals"):
|
||||
pass
|
||||
elif(action_type=="serial"):
|
||||
next_action = action_properties_mapping[action]["next"]
|
||||
if(next_action!=""):
|
||||
output_list = [] # List to store the output of actions whose outputs are required by downstream operations
|
||||
queue.append(next_action)
|
||||
action_names = action_properties_mapping[next_action]["outputs_from"] # Get the list of actions whose output will be used
|
||||
if(len(action_names)==1): # if only output of one action is required
|
||||
key = action_names[0]+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
action_properties_mapping[next_action]["arguments"] = output
|
||||
else:
|
||||
for item in action_names:
|
||||
key = item+"-output"
|
||||
output = pickle.loads(redis_instace.get(key))
|
||||
output_list.append(output)
|
||||
action_properties_mapping[next_action]["arguments"] = output_list
|
||||
|
||||
elif(action_type=="parallel"):
|
||||
parallel_action_list = action_properties_mapping[action]["next"]
|
||||
queue.append(parallel_action_list)
|
||||
|
||||
|
||||
else:
|
||||
reply = handle_parallel(queue,redis_instace,action_properties_mapping,action)
|
||||
|
||||
|
||||
|
||||
|
||||
dag_metadata["function_activation_ids"] = list_of_func_ids
|
||||
# print("DAG SPEC AFTER WORKFLOW EXECUTION--------\n")
|
||||
# print(action_properties_mapping)
|
||||
# print('\n')
|
||||
submit_dag_metadata(dag_metadata)
|
||||
# print("DAG ID---->FUNC IDS",dag_metadata)
|
||||
print('\n')
|
||||
# print('INTERMEDIATE OUTPUTS FROM ALL ACTIONS-----\n')
|
||||
# get_redis_contents(redis_instace)
|
||||
# print('\n')
|
||||
redis_instace.flushdb()
|
||||
print("Cleaned up in-memory intermediate outputs successfully\n")
|
||||
|
||||
if(isinstance(reply,list)):
|
||||
return({"dag_id": dag_metadata["dag_id"],
|
||||
"result": reply
|
||||
})
|
||||
|
||||
else:
|
||||
return({
|
||||
"dag_id": dag_metadata["dag_id"],
|
||||
"result": reply.json()
|
||||
})
|
||||
# return({
|
||||
# "result": "success"
|
||||
# })
|
||||
|
||||
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
app.run(host='0.0.0.0', port=5001)
|
||||
|
||||
|
After Width: | Height: | Size: 19 KiB |
@ -1,51 +0,0 @@
|
||||
import os
|
||||
import boto3
|
||||
from botocore.exceptions import ClientError
|
||||
|
||||
aws_access_key_id = os.environ.get('AWS_ACCESS_KEY_ID')
|
||||
aws_secret_access_key = os.getenv('AWS_SECRET_ACCESS_KEY')
|
||||
aws_region = os.getenv('AWS_REGION')
|
||||
print(aws_access_key_id,aws_secret_access_key)
|
||||
|
||||
# s3 = boto3.client('s3', aws_access_key_id=aws_access_key_id,aws_secret_access_key=aws_secret_access_key,region_name=aws_region)
|
||||
|
||||
# upload_file_path = "dag_register.py"
|
||||
# bucket_name = 'dagit-store'
|
||||
# key_name = upload_file_path
|
||||
# folder_path = 'images'
|
||||
# folder_name = "images"
|
||||
# try:
|
||||
# s3.upload_file(upload_file_path,bucket_name,key_name)
|
||||
# s3.put_object_acl(Bucket=bucket_name, Key=key_name, ACL='public-read')
|
||||
# object_url = "https://dagit-store.s3.ap-south-1.amazonaws.com/"+key_name
|
||||
# print("Uploaded....\n")
|
||||
# print(object_url)
|
||||
# except ClientError as e:
|
||||
# print(e)
|
||||
# loop through files in folder
|
||||
# for subdir, dirs, files in os.walk(folder_path):
|
||||
# for file in files:
|
||||
# # get full path of file
|
||||
# file_path = os.path.join(subdir, file)
|
||||
# # get S3 object key
|
||||
# object_key = os.path.relpath(file_path, folder_path)
|
||||
# # upload file to S3
|
||||
# # s3.Object(bucket_name, object_key).upload_file(file_path)
|
||||
# # s3.upload_file(file_path,bucket_name,object_key)
|
||||
# s3.upload_file(file_path, bucket_name, f'{folder_name}/{file_path.split("/")[-1]}')
|
||||
# s3.put_object_acl(Bucket=bucket_name, Key=f'{folder_name}/{file_path.split("/")[-1]}', ACL='public-read')
|
||||
# print("Uploaded....\n")
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
# try:
|
||||
|
||||
# response = s3.generate_presigned_url('get_object',
|
||||
# Params={'Bucket': bucket_name,
|
||||
# 'Key': key_name},
|
||||
# ExpiresIn=3600)
|
||||
# print(response)
|
||||
# except ClientError as e:
|
||||
# print(e)
|
||||
@ -1,3 +1,5 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/image-blur-image
|
||||
wsk -i action create image-blur --docker 10.129.28.219:5000/image-blur-image --web=true --timeout=300000
|
||||
wsk -i action update image-blur --docker 10.129.28.219:5000/image-blur-image blur.py --web=true --timeout=300000
|
||||
|
||||
./register.sh /image-blur-api /image-blur-path image-blur --response-type=json
|
||||
@ -1,3 +1,3 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/image-denoise-image
|
||||
./register.sh /image-denoise-api /image-denoise-path image-denoise --response-type=j
|
||||
./register.sh /image-denoise-api /image-denoise-path image-denoise --response-type=json
|
||||
wsk -i action create image-denoise --docker 10.129.28.219:5000/image-denoise-image --web=true --timeout=300000
|
||||
|
||||
@ -1,3 +1,5 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/image-resize-image
|
||||
wsk -i action create image-resize --docker 10.129.28.219:5000/image-resize-image --web=true --timeout=300000
|
||||
./register.sh /image-resize-api /image-resize-path image-blur --response-type=json
|
||||
./register.sh /image-resize-api /image-resize-path image-resize --response-type=json
|
||||
|
||||
wsk -i action update image-resize --docker 10.129.28.219:5000/image-resize-image resize.py --web=true --timeout=300000
|
||||
|
||||
@ -0,0 +1,111 @@
|
||||
# # Dockerfile for Python whisk docker action
|
||||
# FROM openwhisk/dockerskeleton
|
||||
|
||||
# ENV FLASK_PROXY_PORT 8080
|
||||
|
||||
# ## Install our action's Python dependencies
|
||||
# ADD requirements.txt /action/requirements.txt
|
||||
# RUN apk --update add python py-pip openssl ca-certificates py-openssl wget git
|
||||
# RUN apk --update add --virtual build-dependencies libffi-dev openssl-dev python-dev py-pip build-base && apk add jpeg-dev zlib-dev libjpeg && pip install --upgrade pip
|
||||
# RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y
|
||||
# ENV PATH="/root/.cargo/bin:${PATH}"
|
||||
# # RUN pip install tokenizers
|
||||
# # RUN apk --update add world build-essential
|
||||
# RUN cd /action; pip install -r requirements.txt
|
||||
# # RUN pip install torch===1.4.0 torchvision===0.5.0 -f https://download.pytorch.org/whl/torch_stable.html
|
||||
# # RUN pip install torch===1.4.0 -f https://download.pytorch.org/whl/torch_stable.html
|
||||
# ENV USE_OPENMP=0
|
||||
# ENV OMP_NUM_THREADS=1
|
||||
# RUN apk --update add gcompat libc6-compat musl musl-dev
|
||||
# RUN apk add --no-cache libgcc ncurses-libs libstdc++
|
||||
# # ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib/python3.6/site-packages/torch/lib/:/usr/local/lib/python3.6/site-packages/torch/lib//
|
||||
|
||||
# RUN pip install torch==1.10.2+cpu -f https://download.pytorch.org/whl/torch_stable.html
|
||||
# # ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/ld-linux-x86-64.so.2:/usr/local/lib/python3.6/site-packages/torch/lib/
|
||||
# # ENV LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libgomp.so.1://usr/local/lib/python3.6/site-packages/torch/lib//libgomp-a34b3233.so.1
|
||||
# # ENV LD_LIBRARY_PATH=/usr/local/lib/python3.6/site-packages/torch/lib/:/usr/local/lib/python3.6/site-packages/torch/lib//libgomp-a34b3233.so.1
|
||||
# # ENV LD_LIBRARY_PATH=/usr/local/lib/python3.6/site-packages/torch/lib:$LD_LIBRARY_PATH
|
||||
|
||||
# # RUN pip install torch===2.0.0 torchvision===0.15.1
|
||||
|
||||
|
||||
# # Ensure source assets are not drawn from the cache after this date
|
||||
# ENV REFRESHED_AT 2016-09-05T13:59:39Z
|
||||
# # Add all source assets
|
||||
# ADD . /action
|
||||
# # Rename our executable Python action
|
||||
# ADD img_text.py /action/exec
|
||||
|
||||
# # Leave CMD as is for Openwhisk
|
||||
# CMD ["/bin/bash", "-c", "cd actionProxy && python3 -u actionproxy.py"]
|
||||
|
||||
|
||||
FROM ibmfunctions/action-python-v3.9
|
||||
|
||||
#download pip packages
|
||||
|
||||
RUN pip install --upgrade pip
|
||||
|
||||
RUN pip install transformers requests redis pilgram Pillow==6.2.2 zlib-state torch
|
||||
|
||||
ADD img_text.py /action/exec
|
||||
|
||||
# RUN python3 -c "from huggingface_hub import snapshot_download,hf_hub_download;REPO_ID = 'Salesforce/blip-image-captioning-base';snapshot_download(repo_id='Salesforce/blip-image-captioning-base',local_dir='/action/models/transformers')"
|
||||
|
||||
ADD models /action/models
|
||||
|
||||
CMD ["/bin/bash", "-c", "cd actionProxy && python3 -u actionproxy.py"]
|
||||
|
||||
#download pre-trained model from hugging face to docker image
|
||||
# RUN pip install sentence_transformers
|
||||
# RUN python3 -c "from transformers import BlipProcessor, BlipForConditionalGeneration;model = BlipForConditionalGeneration.from_pretrained('Salesforce/blip-image-captioning-base');model.save('models')"
|
||||
# COPY models .
|
||||
|
||||
|
||||
# ADD img_text.py .
|
||||
|
||||
# CMD ["python3","img_text.py"]
|
||||
|
||||
|
||||
# RUN ls /action/1/bin/models/transformers
|
||||
# CMD ['echo','/action/1/bin/models/transformers']
|
||||
# CMD ['cat','/action/1/bin/models/transformers/exec']
|
||||
# CMD ['ls','/action/1/bin/models/transformers']
|
||||
|
||||
# FROM python:3.9-slim-buster
|
||||
|
||||
# RUN apt-get update
|
||||
|
||||
# #download pip packages
|
||||
|
||||
# RUN pip install --upgrade pip
|
||||
|
||||
# RUN pip install transformers
|
||||
|
||||
# RUN pip install requests
|
||||
|
||||
# RUN pip install redis
|
||||
|
||||
# RUN pip install pilgram
|
||||
|
||||
# # RUN apt-get install jpeg-dev zlib-dev
|
||||
|
||||
# RUN pip install Pillow
|
||||
|
||||
# RUN pip install zlib-state
|
||||
|
||||
# RUN pip install torch
|
||||
|
||||
# RUN python3 -c "from huggingface_hub import snapshot_download,hf_hub_download;REPO_ID = 'Salesforce/blip-image-captioning-base';snapshot_download(repo_id='Salesforce/blip-image-captioning-base',local_dir='/models/transformers')"
|
||||
|
||||
# RUN ls /
|
||||
|
||||
|
||||
# FROM openwhisk/python3action
|
||||
# RUN apk update
|
||||
# # RUN apk add build-dependencies libffi-dev openssl-dev python-dev py-pip build-base
|
||||
# RUN pip3 install --upgrade pip
|
||||
# RUN pip3 install transformers redis requests boto3 Pillow torch pilgram
|
||||
# RUN python3 -c "from huggingface_hub import snapshot_download,hf_hub_download;REPO_ID = 'Salesforce/blip-image-captioning-base';snapshot_download(repo_id='Salesforce/blip-image-captioning-base',local_dir='/models/transformers')"
|
||||
# RUN ls /
|
||||
|
||||
@ -0,0 +1,9 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/img-text
|
||||
|
||||
wsk -i action create img-to-text --docker 10.129.28.219:5000/img-text img_text.py --web=true --timeout=300000
|
||||
|
||||
./register.sh /image-text-api /image-text-path img-to-text --response-type=json
|
||||
|
||||
wsk -i action update img-to-text --docker 10.129.28.219:5000/img-text img_text.py --timeout 300000
|
||||
|
||||
|
||||
@ -0,0 +1,91 @@
|
||||
import requests
|
||||
import threading
|
||||
import os
|
||||
import json
|
||||
import sys
|
||||
import requests
|
||||
import torch
|
||||
from PIL import Image
|
||||
from transformers import BlipProcessor, BlipForConditionalGeneration
|
||||
thread_list = []
|
||||
results = []
|
||||
|
||||
def process(image_url):
|
||||
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
# img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
raw_image = Image.open(requests.get(image_url, stream=True).raw).convert('RGB')
|
||||
inputs = processor(raw_image, return_tensors="pt")
|
||||
out = model.generate(**inputs)
|
||||
result = processor.decode(out[0], skip_special_tokens=True)
|
||||
results.append(result)
|
||||
def main(params):
|
||||
activation_id = os.environ.get('__OW_ACTIVATION_ID')
|
||||
# file = open('action/1/bin/exec__.py', 'r')
|
||||
# content = file.read()
|
||||
# print(content)
|
||||
print("Current directory----",os.getcwd())
|
||||
print("root----",os.listdir('/'))
|
||||
print("inside action----",os.listdir('/action/models/transformers'))
|
||||
print("current directory contents------",os.listdir(os.getcwd()))
|
||||
# print("/action/model contents------",os.listdir('../models/transformers'))
|
||||
|
||||
# print("action/1/bin------",os.listdir('/action/1/bin'))
|
||||
|
||||
|
||||
# params = json.loads(sys.argv[1])
|
||||
img_urls = params["image_url_links"]
|
||||
|
||||
modelpath = '/action/models/transformers'
|
||||
|
||||
# print("Line 32------",os.listdir(model_path))
|
||||
|
||||
# processor = BlipProcessor.from_pretrained(modelpath)
|
||||
# model = BlipForConditionalGeneration.from_pretrained(modelpath)
|
||||
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB')
|
||||
inputs = processor(raw_image, return_tensors="pt")
|
||||
out = model.generate(**inputs)
|
||||
result = processor.decode(out[0], skip_special_tokens=True)
|
||||
print(res)
|
||||
# # processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
# # model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
# img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
# raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB')
|
||||
# # raw_image = Image.open('puppy.jpg').convert('RGB')
|
||||
# inputs = processor(raw_image, return_tensors="pt")
|
||||
# out = model.generate(**inputs)
|
||||
# res = processor.decode(out[0], skip_special_tokens=True)
|
||||
# print(res)
|
||||
# # result = []
|
||||
# for image_url in img_urls:
|
||||
# thread_list.append(threading.Thread(target=process, args=[image_url]))
|
||||
# for thread in thread_list:
|
||||
# thread.start()
|
||||
# for thread in thread_list:
|
||||
# thread.join()
|
||||
|
||||
|
||||
# text = process(image_url)
|
||||
# result.append(text)
|
||||
|
||||
|
||||
|
||||
|
||||
#https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg
|
||||
|
||||
print(json.dumps({"activation_id": str(activation_id),
|
||||
"image_url_links": img_urls,
|
||||
"result":res
|
||||
}))
|
||||
|
||||
return({"activation_id": str(activation_id),
|
||||
"image_url_links": img_urls,
|
||||
"result":res
|
||||
})
|
||||
|
||||
if __name__ == "__main__":
|
||||
# print("Line 52.....",sys.argv)
|
||||
main(params)
|
||||
@ -0,0 +1,34 @@
|
||||
*.7z filter=lfs diff=lfs merge=lfs -text
|
||||
*.arrow filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin filter=lfs diff=lfs merge=lfs -text
|
||||
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
||||
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
||||
*.ftz filter=lfs diff=lfs merge=lfs -text
|
||||
*.gz filter=lfs diff=lfs merge=lfs -text
|
||||
*.h5 filter=lfs diff=lfs merge=lfs -text
|
||||
*.joblib filter=lfs diff=lfs merge=lfs -text
|
||||
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
||||
*.model filter=lfs diff=lfs merge=lfs -text
|
||||
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
||||
*.npy filter=lfs diff=lfs merge=lfs -text
|
||||
*.npz filter=lfs diff=lfs merge=lfs -text
|
||||
*.onnx filter=lfs diff=lfs merge=lfs -text
|
||||
*.ot filter=lfs diff=lfs merge=lfs -text
|
||||
*.parquet filter=lfs diff=lfs merge=lfs -text
|
||||
*.pb filter=lfs diff=lfs merge=lfs -text
|
||||
*.pickle filter=lfs diff=lfs merge=lfs -text
|
||||
*.pkl filter=lfs diff=lfs merge=lfs -text
|
||||
*.pt filter=lfs diff=lfs merge=lfs -text
|
||||
*.pth filter=lfs diff=lfs merge=lfs -text
|
||||
*.rar filter=lfs diff=lfs merge=lfs -text
|
||||
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
||||
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tflite filter=lfs diff=lfs merge=lfs -text
|
||||
*.tgz filter=lfs diff=lfs merge=lfs -text
|
||||
*.wasm filter=lfs diff=lfs merge=lfs -text
|
||||
*.xz filter=lfs diff=lfs merge=lfs -text
|
||||
*.zip filter=lfs diff=lfs merge=lfs -text
|
||||
*.zst filter=lfs diff=lfs merge=lfs -text
|
||||
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
||||
@ -0,0 +1,152 @@
|
||||
---
|
||||
pipeline_tag: image-to-text
|
||||
tags:
|
||||
- image-captioning
|
||||
languages:
|
||||
- en
|
||||
license: bsd-3-clause
|
||||
---
|
||||
|
||||
# BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
|
||||
|
||||
Model card for image captioning pretrained on COCO dataset - base architecture (with ViT base backbone).
|
||||
|
||||
|  |
|
||||
|:--:|
|
||||
| <b> Pull figure from BLIP official repo | Image source: https://github.com/salesforce/BLIP </b>|
|
||||
|
||||
## TL;DR
|
||||
|
||||
Authors from the [paper](https://arxiv.org/abs/2201.12086) write in the abstract:
|
||||
|
||||
*Vision-Language Pre-training (VLP) has advanced the performance for many vision-language tasks. However, most existing pre-trained models only excel in either understanding-based tasks or generation-based tasks. Furthermore, performance improvement has been largely achieved by scaling up the dataset with noisy image-text pairs collected from the web, which is a suboptimal source of supervision. In this paper, we propose BLIP, a new VLP framework which transfers flexibly to both vision-language understanding and generation tasks. BLIP effectively utilizes the noisy web data by bootstrapping the captions, where a captioner generates synthetic captions and a filter removes the noisy ones. We achieve state-of-the-art results on a wide range of vision-language tasks, such as image-text retrieval (+2.7% in average recall@1), image captioning (+2.8% in CIDEr), and VQA (+1.6% in VQA score). BLIP also demonstrates strong generalization ability when directly transferred to videolanguage tasks in a zero-shot manner. Code, models, and datasets are released.*
|
||||
|
||||
## Usage
|
||||
|
||||
You can use this model for conditional and un-conditional image captioning
|
||||
|
||||
### Using the Pytorch model
|
||||
|
||||
#### Running the model on CPU
|
||||
|
||||
<details>
|
||||
<summary> Click to expand </summary>
|
||||
|
||||
```python
|
||||
import requests
|
||||
from PIL import Image
|
||||
from transformers import BlipProcessor, BlipForConditionalGeneration
|
||||
|
||||
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
|
||||
img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB')
|
||||
|
||||
# conditional image captioning
|
||||
text = "a photography of"
|
||||
inputs = processor(raw_image, text, return_tensors="pt")
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
# >>> a photography of a woman and her dog
|
||||
|
||||
# unconditional image captioning
|
||||
inputs = processor(raw_image, return_tensors="pt")
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
>>> a woman sitting on the beach with her dog
|
||||
```
|
||||
</details>
|
||||
|
||||
#### Running the model on GPU
|
||||
|
||||
##### In full precision
|
||||
|
||||
<details>
|
||||
<summary> Click to expand </summary>
|
||||
|
||||
```python
|
||||
import requests
|
||||
from PIL import Image
|
||||
from transformers import BlipProcessor, BlipForConditionalGeneration
|
||||
|
||||
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base").to("cuda")
|
||||
|
||||
img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB')
|
||||
|
||||
# conditional image captioning
|
||||
text = "a photography of"
|
||||
inputs = processor(raw_image, text, return_tensors="pt").to("cuda")
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
# >>> a photography of a woman and her dog
|
||||
|
||||
# unconditional image captioning
|
||||
inputs = processor(raw_image, return_tensors="pt").to("cuda")
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
>>> a woman sitting on the beach with her dog
|
||||
```
|
||||
</details>
|
||||
|
||||
##### In half precision (`float16`)
|
||||
|
||||
<details>
|
||||
<summary> Click to expand </summary>
|
||||
|
||||
```python
|
||||
import torch
|
||||
import requests
|
||||
from PIL import Image
|
||||
from transformers import BlipProcessor, BlipForConditionalGeneration
|
||||
|
||||
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
|
||||
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base", torch_dtype=torch.float16).to("cuda")
|
||||
|
||||
img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg'
|
||||
raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB')
|
||||
|
||||
# conditional image captioning
|
||||
text = "a photography of"
|
||||
inputs = processor(raw_image, text, return_tensors="pt").to("cuda", torch.float16)
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
# >>> a photography of a woman and her dog
|
||||
|
||||
# unconditional image captioning
|
||||
inputs = processor(raw_image, return_tensors="pt").to("cuda", torch.float16)
|
||||
|
||||
out = model.generate(**inputs)
|
||||
print(processor.decode(out[0], skip_special_tokens=True))
|
||||
>>> a woman sitting on the beach with her dog
|
||||
```
|
||||
</details>
|
||||
|
||||
## BibTex and citation info
|
||||
|
||||
```
|
||||
@misc{https://doi.org/10.48550/arxiv.2201.12086,
|
||||
doi = {10.48550/ARXIV.2201.12086},
|
||||
|
||||
url = {https://arxiv.org/abs/2201.12086},
|
||||
|
||||
author = {Li, Junnan and Li, Dongxu and Xiong, Caiming and Hoi, Steven},
|
||||
|
||||
keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences},
|
||||
|
||||
title = {BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation},
|
||||
|
||||
publisher = {arXiv},
|
||||
|
||||
year = {2022},
|
||||
|
||||
copyright = {Creative Commons Attribution 4.0 International}
|
||||
}
|
||||
```
|
||||
@ -0,0 +1,169 @@
|
||||
{
|
||||
"_commit_hash": null,
|
||||
"architectures": [
|
||||
"BlipForConditionalGeneration"
|
||||
],
|
||||
"image_text_hidden_size": 256,
|
||||
"initializer_factor": 1.0,
|
||||
"logit_scale_init_value": 2.6592,
|
||||
"model_type": "blip",
|
||||
"projection_dim": 512,
|
||||
"text_config": {
|
||||
"_name_or_path": "",
|
||||
"add_cross_attention": false,
|
||||
"architectures": null,
|
||||
"attention_probs_dropout_prob": 0.0,
|
||||
"bad_words_ids": null,
|
||||
"begin_suppress_tokens": null,
|
||||
"bos_token_id": 30522,
|
||||
"chunk_size_feed_forward": 0,
|
||||
"cross_attention_hidden_size": null,
|
||||
"decoder_start_token_id": null,
|
||||
"diversity_penalty": 0.0,
|
||||
"do_sample": false,
|
||||
"early_stopping": false,
|
||||
"encoder_no_repeat_ngram_size": 0,
|
||||
"eos_token_id": 2,
|
||||
"exponential_decay_length_penalty": null,
|
||||
"finetuning_task": null,
|
||||
"forced_bos_token_id": null,
|
||||
"forced_eos_token_id": null,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_dropout_prob": 0.0,
|
||||
"hidden_size": 768,
|
||||
"id2label": {
|
||||
"0": "LABEL_0",
|
||||
"1": "LABEL_1"
|
||||
},
|
||||
"initializer_factor": 1.0,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 3072,
|
||||
"is_decoder": true,
|
||||
"is_encoder_decoder": false,
|
||||
"label2id": {
|
||||
"LABEL_0": 0,
|
||||
"LABEL_1": 1
|
||||
},
|
||||
"layer_norm_eps": 1e-12,
|
||||
"length_penalty": 1.0,
|
||||
"max_length": 20,
|
||||
"max_position_embeddings": 512,
|
||||
"min_length": 0,
|
||||
"model_type": "blip_text_model",
|
||||
"no_repeat_ngram_size": 0,
|
||||
"num_attention_heads": 12,
|
||||
"num_beam_groups": 1,
|
||||
"num_beams": 1,
|
||||
"num_hidden_layers": 12,
|
||||
"num_return_sequences": 1,
|
||||
"output_attentions": false,
|
||||
"output_hidden_states": false,
|
||||
"output_scores": false,
|
||||
"pad_token_id": 0,
|
||||
"prefix": null,
|
||||
"problem_type": null,
|
||||
"projection_dim": 768,
|
||||
"pruned_heads": {},
|
||||
"remove_invalid_values": false,
|
||||
"repetition_penalty": 1.0,
|
||||
"return_dict": true,
|
||||
"return_dict_in_generate": false,
|
||||
"sep_token_id": 102,
|
||||
"suppress_tokens": null,
|
||||
"task_specific_params": null,
|
||||
"temperature": 1.0,
|
||||
"tf_legacy_loss": false,
|
||||
"tie_encoder_decoder": false,
|
||||
"tie_word_embeddings": true,
|
||||
"tokenizer_class": null,
|
||||
"top_k": 50,
|
||||
"top_p": 1.0,
|
||||
"torch_dtype": null,
|
||||
"torchscript": false,
|
||||
"transformers_version": "4.26.0.dev0",
|
||||
"typical_p": 1.0,
|
||||
"use_bfloat16": false,
|
||||
"use_cache": true,
|
||||
"vocab_size": 30524
|
||||
},
|
||||
"torch_dtype": "float32",
|
||||
"transformers_version": null,
|
||||
"vision_config": {
|
||||
"_name_or_path": "",
|
||||
"add_cross_attention": false,
|
||||
"architectures": null,
|
||||
"attention_dropout": 0.0,
|
||||
"bad_words_ids": null,
|
||||
"begin_suppress_tokens": null,
|
||||
"bos_token_id": null,
|
||||
"chunk_size_feed_forward": 0,
|
||||
"cross_attention_hidden_size": null,
|
||||
"decoder_start_token_id": null,
|
||||
"diversity_penalty": 0.0,
|
||||
"do_sample": false,
|
||||
"dropout": 0.0,
|
||||
"early_stopping": false,
|
||||
"encoder_no_repeat_ngram_size": 0,
|
||||
"eos_token_id": null,
|
||||
"exponential_decay_length_penalty": null,
|
||||
"finetuning_task": null,
|
||||
"forced_bos_token_id": null,
|
||||
"forced_eos_token_id": null,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_size": 768,
|
||||
"id2label": {
|
||||
"0": "LABEL_0",
|
||||
"1": "LABEL_1"
|
||||
},
|
||||
"image_size": 384,
|
||||
"initializer_factor": 1.0,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 3072,
|
||||
"is_decoder": false,
|
||||
"is_encoder_decoder": false,
|
||||
"label2id": {
|
||||
"LABEL_0": 0,
|
||||
"LABEL_1": 1
|
||||
},
|
||||
"layer_norm_eps": 1e-05,
|
||||
"length_penalty": 1.0,
|
||||
"max_length": 20,
|
||||
"min_length": 0,
|
||||
"model_type": "blip_vision_model",
|
||||
"no_repeat_ngram_size": 0,
|
||||
"num_attention_heads": 12,
|
||||
"num_beam_groups": 1,
|
||||
"num_beams": 1,
|
||||
"num_channels": 3,
|
||||
"num_hidden_layers": 12,
|
||||
"num_return_sequences": 1,
|
||||
"output_attentions": false,
|
||||
"output_hidden_states": false,
|
||||
"output_scores": false,
|
||||
"pad_token_id": null,
|
||||
"patch_size": 16,
|
||||
"prefix": null,
|
||||
"problem_type": null,
|
||||
"projection_dim": 512,
|
||||
"pruned_heads": {},
|
||||
"remove_invalid_values": false,
|
||||
"repetition_penalty": 1.0,
|
||||
"return_dict": true,
|
||||
"return_dict_in_generate": false,
|
||||
"sep_token_id": null,
|
||||
"suppress_tokens": null,
|
||||
"task_specific_params": null,
|
||||
"temperature": 1.0,
|
||||
"tf_legacy_loss": false,
|
||||
"tie_encoder_decoder": false,
|
||||
"tie_word_embeddings": true,
|
||||
"tokenizer_class": null,
|
||||
"top_k": 50,
|
||||
"top_p": 1.0,
|
||||
"torch_dtype": null,
|
||||
"torchscript": false,
|
||||
"transformers_version": "4.26.0.dev0",
|
||||
"typical_p": 1.0,
|
||||
"use_bfloat16": false
|
||||
}
|
||||
}
|
||||
@ -0,0 +1,17 @@
|
||||
{
|
||||
"do_normalize": true,
|
||||
"do_resize": true,
|
||||
"image_mean": [
|
||||
0.48145466,
|
||||
0.4578275,
|
||||
0.40821073
|
||||
],
|
||||
"image_processor_type": "BlipImageProcessor",
|
||||
"image_std": [
|
||||
0.26862954,
|
||||
0.26130258,
|
||||
0.27577711
|
||||
],
|
||||
"processor_class": "BlipProcessor",
|
||||
"size": 384
|
||||
}
|
||||
@ -0,0 +1 @@
|
||||
../../../../../../.cache/huggingface/hub/models--Salesforce--blip-image-captioning-base/blobs/d6638651a5526cc2ede56f2b5104d6851b0755816d220e5e046870430180c767
|
||||
@ -0,0 +1,7 @@
|
||||
{
|
||||
"cls_token": "[CLS]",
|
||||
"mask_token": "[MASK]",
|
||||
"pad_token": "[PAD]",
|
||||
"sep_token": "[SEP]",
|
||||
"unk_token": "[UNK]"
|
||||
}
|
||||
@ -0,0 +1 @@
|
||||
../../../../../../.cache/huggingface/hub/models--Salesforce--blip-image-captioning-base/blobs/d0aaa4c0e003f599d8baa53a9dee85af14eef20554cf2f8113a2673e25a59f8c
|
||||
File diff suppressed because it is too large
Load Diff
@ -0,0 +1,17 @@
|
||||
{
|
||||
"cls_token": "[CLS]",
|
||||
"do_basic_tokenize": true,
|
||||
"do_lower_case": true,
|
||||
"mask_token": "[MASK]",
|
||||
"model_max_length": 512,
|
||||
"name_or_path": "bert-base-uncased",
|
||||
"never_split": null,
|
||||
"pad_token": "[PAD]",
|
||||
"processor_class": "BlipProcessor",
|
||||
"sep_token": "[SEP]",
|
||||
"special_tokens_map_file": null,
|
||||
"strip_accents": null,
|
||||
"tokenize_chinese_chars": true,
|
||||
"tokenizer_class": "BertTokenizer",
|
||||
"unk_token": "[UNK]"
|
||||
}
|
||||
File diff suppressed because it is too large
Load Diff
@ -0,0 +1,7 @@
|
||||
{
|
||||
"image_url_links": [
|
||||
"https://dagit-store.s3.ap-south-1.amazonaws.com/image_text_classification/flight.jpg",
|
||||
"https://dagit-store.s3.ap-south-1.amazonaws.com/image_text_classification/puppy.jpg",
|
||||
"https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg"
|
||||
]
|
||||
}
|
||||
@ -0,0 +1,7 @@
|
||||
requests
|
||||
redis
|
||||
torch
|
||||
pilgram
|
||||
Pillow==6.2.2
|
||||
zlib-state
|
||||
transformers
|
||||
@ -0,0 +1,25 @@
|
||||
# Dockerfile for Python whisk docker action
|
||||
FROM openwhisk/dockerskeleton
|
||||
|
||||
ENV FLASK_PROXY_PORT 8080
|
||||
|
||||
## Install our action's Python dependencies
|
||||
ADD requirements.txt /action/requirements.txt
|
||||
ENV AWS_ACCESS_KEY_ID="AKIAYFB773UVZSOAVZN4"
|
||||
ENV AWS_SECRET_ACCESS_KEY="OZPLMjN/2ao6OlSd5PpIkT5d7cWD9WAP/DXSZbEs"
|
||||
ENV AWS_REGION="ap-south-1"
|
||||
RUN apk --update add python py-pip openssl ca-certificates py-openssl wget
|
||||
RUN apk --update add --virtual build-dependencies libffi-dev openssl-dev python-dev py-pip build-base \
|
||||
&& apk add jpeg-dev zlib-dev libjpeg \
|
||||
&& pip install --upgrade pip
|
||||
RUN cd /action; pip install --no-cache-dir -r requirements.txt
|
||||
RUN pip install opencv-python
|
||||
# Ensure source assets are not drawn from the cache after this date
|
||||
ENV REFRESHED_AT 2016-09-05T13:59:39Z
|
||||
# Add all source assets
|
||||
ADD . /action
|
||||
# Rename our executable Python action
|
||||
ADD rotate.py /action/exec
|
||||
|
||||
# Leave CMD as is for Openwhisk
|
||||
CMD ["/bin/bash", "-c", "cd actionProxy && python3 -u actionproxy.py"]
|
||||
@ -0,0 +1,3 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/image-rotate-image
|
||||
wsk -i action create image-rotate --docker 10.129.28.219:5000/image-rotate-image --web=true --timeout=300000
|
||||
./register.sh /image-rotate-api /image-rotate-path image-rotate --response-type=json
|
||||
@ -0,0 +1,24 @@
|
||||
#!/bin/bash
|
||||
#
|
||||
# This script will build the docker image and push it to dockerhub.
|
||||
#
|
||||
# Usage: buildAndPush.sh imageName
|
||||
#
|
||||
# Dockerhub image names look like "username/appname" and must be all lower case.
|
||||
# For example, "janesmith/calculator"
|
||||
|
||||
IMAGE_NAME=$1
|
||||
echo "Using $IMAGE_NAME as the image name"
|
||||
|
||||
# Make the docker image
|
||||
docker build -t $IMAGE_NAME .
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Docker build failed"
|
||||
exit
|
||||
fi
|
||||
docker push $IMAGE_NAME
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Docker push failed"
|
||||
exit
|
||||
fi
|
||||
|
||||
@ -0,0 +1,5 @@
|
||||
requests
|
||||
boto3
|
||||
redis
|
||||
opencv-python
|
||||
|
||||
@ -0,0 +1,97 @@
|
||||
#!/usr/bin/env python3
|
||||
import requests
|
||||
import os
|
||||
import boto3
|
||||
import redis
|
||||
import pickle
|
||||
import json
|
||||
import cv2
|
||||
import sys
|
||||
|
||||
# Rotate an image by 90 degree clockwise
|
||||
def main():
|
||||
images_dir = "rotated-images"
|
||||
is_images_dir = os.path.isdir(images_dir)
|
||||
if(is_images_dir == False):
|
||||
os.mkdir(images_dir)
|
||||
r = redis.Redis(host="10.129.28.219", port=6379, db=2)
|
||||
activation_id = os.environ.get('__OW_ACTIVATION_ID')
|
||||
params = json.loads(sys.argv[1])
|
||||
rotated_result=[]
|
||||
try:
|
||||
decode_activation_id = params["activation_id"]
|
||||
parts = params["parts"]
|
||||
for i in range(0,parts):
|
||||
if os.path.exists(images_dir+'/rotated_image_'+str(i)+'.jpg'):
|
||||
os.remove(images_dir+'/rotated_image_'+str(i)+'.jpg')
|
||||
for i in range(0,parts):
|
||||
decode_output = "decode-output-image"+decode_activation_id+"-"+str(i)
|
||||
load_image = pickle.loads(r.get(decode_output))
|
||||
image_name = 'Image'+str(i)+'.jpg'
|
||||
with open(image_name, 'wb') as f:
|
||||
f.write(load_image)
|
||||
|
||||
img = cv2.imread(image_name)
|
||||
# Rotate the image by 90 degrees
|
||||
rotated = cv2.rotate(img, cv2.ROTATE_90_CLOCKWISE)
|
||||
output_image = images_dir+'/rotated_image_'+str(i)+'.jpg'
|
||||
cv2.imwrite(output_image, rotated)
|
||||
rotated_result.append('rotated_image_'+str(i)+'.jpg')
|
||||
except Exception as e: #If not running as a part of DAG workflow and implemented as a single standalone function
|
||||
image_url_list = params["image_url_links"]
|
||||
parts = len(image_url_list)
|
||||
for i in range(0,parts):
|
||||
if os.path.exists(images_dir+'/rotated_image_'+str(i)+'.jpg'):
|
||||
os.remove(images_dir+'/rotated_image_'+str(i)+'.jpg')
|
||||
for i in range(0,parts):
|
||||
response = requests.get(image_url_list[i])
|
||||
image_name = 'Image'+str(i)+'.jpg'
|
||||
with open(image_name, "wb") as f:
|
||||
f.write(response.content)
|
||||
img = cv2.imread(image_name)
|
||||
# Rotate the image by 90 degrees
|
||||
rotated = cv2.rotate(img, cv2.ROTATE_90_CLOCKWISE)
|
||||
output_image = images_dir+'/rotated_image_'+str(i)+'.jpg'
|
||||
cv2.imwrite(output_image, rotated)
|
||||
rotated_result.append('rotated_image_'+str(i)+'.jpg')
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
aws_access_key_id = os.getenv('AWS_ACCESS_KEY_ID')
|
||||
aws_secret_access_key = os.getenv('AWS_SECRET_ACCESS_KEY')
|
||||
aws_region = os.getenv('AWS_REGION')
|
||||
|
||||
|
||||
s3 = boto3.client('s3', aws_access_key_id=aws_access_key_id,aws_secret_access_key=aws_secret_access_key,region_name=aws_region)
|
||||
|
||||
|
||||
bucket_name = 'dagit-store'
|
||||
folder_path = images_dir
|
||||
folder_name = images_dir
|
||||
for subdir, dirs, files in os.walk(folder_path):
|
||||
for file in files:
|
||||
file_path = os.path.join(subdir, file)
|
||||
s3.upload_file(file_path, bucket_name, f'{folder_name}/{file_path.split("/")[-1]}')
|
||||
s3.put_object_acl(Bucket=bucket_name, Key=f'{folder_name}/{file_path.split("/")[-1]}', ACL='public-read')
|
||||
url_list=[]
|
||||
for image in rotated_result:
|
||||
url = "https://dagit-store.s3.ap-south-1.amazonaws.com/"+images_dir+"/"+image
|
||||
url_list.append(url)
|
||||
|
||||
print(json.dumps({"rotated_image_url_links":url_list,
|
||||
"activation_id": str(activation_id),
|
||||
"parts": parts
|
||||
|
||||
}))
|
||||
|
||||
return({"rotated_image_url_links":url_list,
|
||||
"activation_id": str(activation_id),
|
||||
"parts": parts
|
||||
|
||||
})
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@ -1,3 +1,7 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/encode-image-1
|
||||
wsk -i action create encode-action --docker 10.129.28.219:5000/encode-image-1
|
||||
wsk -i action update encode-action --docker 10.129.28.219:5000/encode-image-1 encode.py --timeout 400000
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/image-processing
|
||||
wsk -i action create encode-function --docker 10.129.28.219:5000/image-processing encode.py --web=true --timeout=420000 -m 4096
|
||||
wsk -i action update encode-function --docker 10.129.28.219:5000/image-processing encode.py --web=true --timeout=420000 -m 4096
|
||||
|
||||
|
||||
|
||||
./register.sh /encode-function /encode encode-function
|
||||
|
||||
@ -1,11 +1,3 @@
|
||||
opencv-python
|
||||
requests
|
||||
redis
|
||||
ffmpeg-python
|
||||
|
||||
zlib-state
|
||||
pilgram
|
||||
Pillow==6.2.2
|
||||
paramiko==2.11.0
|
||||
pycparser==2.21
|
||||
PyNaCl==1.5.0
|
||||
pysftp==0.2.9
|
||||
boto3
|
||||
@ -0,0 +1,11 @@
|
||||
sudo ./buildAndPush.sh 10.129.28.219:5000/test-s3
|
||||
wsk -i action create test-s3-trigger --docker 10.129.28.219:5000/test-s3 --web=true --timeout=300000
|
||||
wsk -i action update test-s3-trigger --docker 10.129.28.219:5000/test-s3 test.py --timeout 300000
|
||||
|
||||
./register.sh /image-rotate-api /image-rotate-path image-rotate --response-type=json
|
||||
|
||||
wsk -i trigger create myTrigger
|
||||
|
||||
wsk -i rule create myRule1 s3-trigger test-s3-trigger --param bucket dagit-store --param suffix .txt
|
||||
|
||||
wsk -i trigger create s3-trigger1 --feed /whisk.system/s3-trigger-feed/changes --param bucket dagit-store --param suffix .txt
|
||||
@ -1,4 +1,2 @@
|
||||
paramiko==2.11.0
|
||||
pycparser==2.21
|
||||
PyNaCl==1.5.0
|
||||
pysftp==0.2.9
|
||||
requests
|
||||
boto3
|
||||
Loading…
Reference in new issue