richardliaw

import ray.data 

class DataGenerator:
    def __init__(self, permute_config):
        device = torch.device("cuda")
        self.model = Model().to(device)
        self.config = permute_config

    def __call__(self, input):
        for test_input in self.permute(permute_config, input):

richardliaw

import ray.data 

class DataGenerator:
    def __init__(self, permute_config):
        device = torch.device("cuda")
        self.model = Model().to(device)
        self.config = permute_config

    def __call__(self, input):
        for test_input in self.permute(permute_config, input):

richardliaw

import requests
from datetime import datetime
def get_issues_with_ray_in_title(repo_name):
    issues = []
    page = 1
    # headers = {'Authorization': 'token YOUR_GITHUB_TOKEN'}
    while True:
        issues_url = f"https://api.github.com/repos/{repo_name}/issues?page={page}&per_page=100&state=open"
        response = requests.get(issues_url)# headers=headers)
        if response.status_code == 200:

richardliaw

import requests
from datetime import datetime
def get_issues_with_ray_in_title(repo_name):
    issues = []
    page = 1
    # headers = {'Authorization': 'token YOUR_GITHUB_TOKEN'}
    while True:
        issues_url = f"https://api.github.com/repos/{repo_name}/issues?page={page}&per_page=100&state=open"
        response = requests.get(issues_url)# headers=headers)
        if response.status_code == 200:

richardliaw

import requests
from tabulate import tabulate
from datetime import datetime, timezone

# Retrieve open pull requests
open_url = 'https://api.github.com/repos/vllm-project/vllm/pulls?state=open&per_page=100'
# merged_url = 'https://api.github.com/repos/vllm-project/vllm/pulls?state=closed&per_page=100'
open_response = requests.get(open_url)
# merged_response = requests.get(merged_url)

richardliaw

diff --git a/examples/run.py b/examples/run.py
index 9a05434..d7946db 100644
--- a/examples/run.py
+++ b/examples/run.py
@@ -135,6 +135,12 @@ def parse_arguments(args=None):
                         choices=["hf", "nemo"],
                         help="The source of lora checkpoint.")
 
+    parser.add_argument(
+        '--run_profiling',

richardliaw

# An unique identifier for the head node and workers of this cluster.
cluster_name: gpu-docker

# The maximum number of workers nodes to launch in addition to the head.
max_workers: 2

docker:
    image: "rayproject/ray-ml:latest-gpu"
    container_name: "ray_nvidia_docker" # e.g. ray_docker

richardliaw

# An unique identifier for the head node and workers of this cluster.
cluster_name: gpu-docker

# The maximum number of workers nodes to launch in addition to the head
# node.
max_workers: 2

docker:
    image: "rayproject/ray-ml:latest-gpu"
    container_name: "ray_nvidia_docker" # e.g. ray_docker

richardliaw

# An unique identifier for the head node and workers of this cluster.
cluster_name: gpu-docker

# The maximum number of workers nodes to launch in addition to the head
# node.
max_workers: 2

# This executes all commands on all nodes in the docker container,
# and opens all the necessary ports to support the Ray cluster.
# Empty string means disabled.

richardliaw

# An unique identifier for the head node and workers of this cluster.
cluster_name: gpu-docker

# The maximum number of workers nodes to launch in addition to the head
# node.
max_workers: 2

# This executes all commands on all nodes in the docker container,
# and opens all the necessary ports to support the Ray cluster.
# Empty string means disabled.