From 9fc674878dd17f57920688aa653f7bab9467fba1 Mon Sep 17 00:00:00 2001
From: Cyberes <cyberes@evulid.cc>
Date: Wed, 23 Aug 2023 16:11:32 -0600
Subject: [PATCH] allow disabling ssl verification

---
 config/config.yml                     | 1 +
 llm_server/llm/hf_textgen/generate.py | 2 +-
 llm_server/llm/info.py                | 4 ++--
 llm_server/llm/oobabooga/generate.py  | 2 +-
 llm_server/opts.py                    | 1 +
 server.py                             | 3 ++-
 6 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/config/config.yml b/config/config.yml
index 96ff5f9..e1a4a13 100644
--- a/config/config.yml
+++ b/config/config.yml
@@ -1,6 +1,7 @@
 # TODO: add this file to gitignore and add a .sample.yml
 
 log_prompts: true
+verify_ssl: false # Python request has issues with self-signed certs
 
 mode: oobabooga
 auth_required: false
diff --git a/llm_server/llm/hf_textgen/generate.py b/llm_server/llm/hf_textgen/generate.py
index a864580..385263e 100644
--- a/llm_server/llm/hf_textgen/generate.py
+++ b/llm_server/llm/hf_textgen/generate.py
@@ -35,7 +35,7 @@ def prepare_json(json_data: dict):
 def generate(json_data: dict):
     print(json.dumps(prepare_json(json_data)))
     # try:
-    r = requests.post(f'{opts.backend_url}/generate', json=prepare_json(json_data))
+    r = requests.post(f'{opts.backend_url}/generate', json=prepare_json(json_data), verify=opts.verify_ssl)
     print(r.text)
     # except Exception as e:
     #     return False, None, f'{e.__class__.__name__}: {e}'
diff --git a/llm_server/llm/info.py b/llm_server/llm/info.py
index 0f4ed9b..47104ec 100644
--- a/llm_server/llm/info.py
+++ b/llm_server/llm/info.py
@@ -6,7 +6,7 @@ from llm_server import opts
 def get_running_model():
     if opts.mode == 'oobabooga':
         try:
-            backend_response = requests.get(f'{opts.backend_url}/api/v1/model', timeout=3)
+            backend_response = requests.get(f'{opts.backend_url}/api/v1/model', timeout=3, verify=opts.verify_ssl)
         except Exception as e:
             return False, e
         try:
@@ -16,7 +16,7 @@ def get_running_model():
             return False, e
     elif opts.mode == 'hf-textgen':
         try:
-            backend_response = requests.get(f'{opts.backend_url}/info')
+            backend_response = requests.get(f'{opts.backend_url}/info', verify=opts.verify_ssl)
         except Exception as e:
             return False, e
         try:
diff --git a/llm_server/llm/oobabooga/generate.py b/llm_server/llm/oobabooga/generate.py
index ace030e..99260d6 100644
--- a/llm_server/llm/oobabooga/generate.py
+++ b/llm_server/llm/oobabooga/generate.py
@@ -5,7 +5,7 @@ from llm_server import opts
 
 def generate(json_data: dict):
     try:
-        r = requests.post(f'{opts.backend_url}/api/v1/generate', json=json_data)
+        r = requests.post(f'{opts.backend_url}/api/v1/generate', json=json_data, verify=opts.verify_ssl)
     except Exception as e:
         return False, None, f'{e.__class__.__name__}: {e}'
     if r.status_code != 200:
diff --git a/llm_server/opts.py b/llm_server/opts.py
index 9e961cb..b4893dc 100644
--- a/llm_server/opts.py
+++ b/llm_server/opts.py
@@ -10,3 +10,4 @@ auth_required = False
 log_prompts = False
 frontend_api_client = ''
 http_host = None
+verify_ssl = True
diff --git a/server.py b/server.py
index 28db475..87cc4e8 100644
--- a/server.py
+++ b/server.py
@@ -21,7 +21,7 @@ if config_path_environ:
 else:
     config_path = Path(script_path, 'config', 'config.yml')
 
-default_vars = {'mode': 'oobabooga', 'log_prompts': False, 'database_path': './proxy-server.db', 'auth_required': False, 'concurrent_gens': 3, 'frontend_api_client': ''}
+default_vars = {'mode': 'oobabooga', 'log_prompts': False, 'database_path': './proxy-server.db', 'auth_required': False, 'concurrent_gens': 3, 'frontend_api_client': '', 'verify_ssl': True}
 required_vars = ['token_limit']
 config_loader = ConfigLoader(config_path, default_vars, required_vars)
 success, config, msg = config_loader.load_config()
@@ -46,6 +46,7 @@ opts.log_prompts = config['log_prompts']
 opts.concurrent_gens = config['concurrent_gens']
 opts.frontend_api_client = config['frontend_api_client']
 opts.context_size = config['token_limit']
+opts.verify_ssl = config['verify_ssl']
 
 SemaphoreCheckerThread(concurrent_semaphore).start()