docs(proxy): added readme

litellm/proxy/README.md

@@ -0,0 +1,47 @@
+# litellm-proxy
+
+A local, fast, and lightweight **OpenAI-compatible server** to call 100+ LLM APIs.
+
+## usage 
+
+```shell 
+$ pip install litellm
+```
+```shell
+$ litellm --model ollama/codellama 
+
+#INFO: Ollama running on http://0.0.0.0:8000
+```
+
+## replace openai base
+```python 
+import openai 
+
+openai.api_base = "http://0.0.0.0:8000"
+
+print(openai.ChatCompletion.create(model="test", messages=[{"role":"user", "content":"Hey!"}]))
+``` 
+
+[**See how to call Huggingface,Bedrock,TogetherAI,Anthropic, etc.**](https://docs.litellm.ai/docs/proxy_server)
+
+## configure proxy
+
+To save API Keys, change model prompt, etc. you'll need to create a local instance of it:
+```shell
+$ litellm --create-proxy
+```
+This will create a local project called `litellm-proxy` in your current directory, that has: 
+* **proxy_cli.py**: Runs the proxy
+* **proxy_server.py**: Contains the API calling logic
+    - `/chat/completions`: receives `openai.ChatCompletion.create` call.
+    - `/completions`: receives `openai.Completion.create` call.
+    - `/models`: receives `openai.Model.list()` call
+* **secrets.toml**: Stores your api keys, model configs, etc.
+
+Run it by doing:
+```shell
+$ cd litellm-proxy
+```
+```shell
+$ python proxy_cli.py --model ollama/llama # replace with your model name
+```

litellm/proxy/proxy_cli.py

@@ -7,6 +7,7 @@ from dotenv import load_dotenv
 load_dotenv()
 from importlib import resources
 import shutil
+telemetry = None
 
 def run_ollama_serve():
     command = ['ollama', 'serve']
@@ -15,7 +16,6 @@ def run_ollama_serve():
         process = subprocess.Popen(command, stdout=devnull, stderr=devnull)
 
 def clone_subfolder(repo_url, subfolder, destination):
-
   # Clone the full repo
   repo_name = repo_url.split('/')[-1]  
   repo_master = os.path.join(destination, "repo_master")
@@ -35,6 +35,7 @@ def clone_subfolder(repo_url, subfolder, destination):
 
   # Remove cloned repo folder
   subprocess.run(['rm', '-rf', os.path.join(destination, "repo_master")])
+  feature_telemetry(feature="create-proxy")
 
 def is_port_in_use(port):
     import socket
@@ -59,15 +60,16 @@ def is_port_in_use(port):
 @click.option('--local', is_flag=True, default=False, help='for local debugging')
 @click.option('--cost', is_flag=True, default=False, help='for viewing cost logs')
 def run_server(host, port, api_base, model, deploy, debug, temperature, max_tokens, drop_params, create_proxy, add_function_to_prompt, max_budget, telemetry, test, local, cost):
+    global feature_telemetry
     if local:
-        from proxy_server import app, initialize, deploy_proxy, print_cost_logs
+        from proxy_server import app, initialize, deploy_proxy, print_cost_logs, usage_telemetry
         debug = True
     else:
         try:
-            from .proxy_server import app, initialize, deploy_proxy, print_cost_logs
+            from .proxy_server import app, initialize, deploy_proxy, print_cost_logs, usage_telemetry
         except ImportError as e: 
-            from proxy_server import app, initialize, deploy_proxy, print_cost_logs
-
+            from proxy_server import app, initialize, deploy_proxy, print_cost_logs, usage_telemetry
+    feature_telemetry = usage_telemetry
     if create_proxy == True: 
         repo_url = 'https://github.com/BerriAI/litellm'
         subfolder = 'litellm/proxy' 

litellm/proxy/proxy_server.py

@@ -75,7 +75,7 @@ user_model = None
 user_debug = False
 user_max_tokens = None
 user_temperature = None
-user_telemetry = False
+user_telemetry = True
 user_config = None
 config_filename = "secrets.toml"
 config_dir = os.getcwd()
@@ -87,12 +87,14 @@ def print_verbose(print_statement):
     if user_debug: 
          print(print_statement)
 
-def usage_telemetry(): # helps us know if people are using this feature. Set `litellm --telemetry False` to your cli call to turn this off
+def usage_telemetry(feature: str): # helps us know if people are using this feature. Set `litellm --telemetry False` to your cli call to turn this off
+    print(f"user_telemtry: {user_telemetry}")
     if user_telemetry: 
+        print(f"feature telemetry: {feature}")
         data = {
-            "feature": "local_proxy_server"
+            "feature": feature # "local_proxy_server"
         }
-        threading.Thread(target=litellm.utils.litellm_telemetry, args=(data,)).start()
+        threading.Thread(target=litellm.utils.litellm_telemetry, args=(data,), daemon=True).start()
 
 def load_config():
     try: 
@@ -174,7 +176,7 @@ def initialize(model, api_base, debug, temperature, max_tokens, max_budget, tele
     user_max_tokens = max_tokens
     user_temperature = temperature
     user_telemetry = telemetry
-    usage_telemetry()
+    usage_telemetry(feature="local_proxy_server")
     if drop_params == True: 
         litellm.drop_params = True
     if add_function_to_prompt == True: