ai: Do not abuse this services.

NOTICE

@@ -10,3 +10,56 @@ Based on the Qwen: QwQ-32B model developed by Alibaba Cloud.
 
 Original model: Qwen: QwQ-32B
 Available at: https://huggingface.co/Qwen/QwQ-32B
+
+
+---------------------------------------------------------------------------------------------------------
+
+
+NOTICE OF ACCEPTABLE USE AND SERVICE TERMS FOR JARVIS AI MULTI PLATFORM
+
+Welcome to JARVIS, a free multi platform AI service provider.
+
+This notice outlines the terms and conditions for using JARVIS’s services.
+By accessing or using JARVIS, you agree to abide by the following terms and
+are fully responsible for your use of the platform.
+
+1. Purpose of JARVIS
+JARVIS is designed to provide free access to various AI models, It is intended for
+educational, personal, research, and non-commercial use.
+
+2. Prohibited Uses
+You agree not to misuse the JARVIS platform in any way, including but not limited to:
+
+- Illegal Activity: Using JARVIS for any illegal, unlawful, or criminal purposes.
+- Malicious Intent: Generating, sharing, or distributing harmful content such as:
+  - Malware, ransomware, or phishing tools.
+  - Hateful, abusive, discriminatory, or violent content.
+  - False or misleading information.
+- Security Violations: Attempting to hack, exploit, or tamper with JARVIS’s systems or services.
+- Impersonation: Pretending to be another person, organization, or falsely claiming affiliation.
+- Automated Abuse: Using bots or scripts to excessively access or overload the system.
+- Bypassing Limitations: Attempting to override or manipulate usage restrictions or safeguards.
+
+3. User Responsibilities
+As a user of JARVIS, you are solely responsible for:
+- Ensuring that your use of the platform complies with local laws and regulations.
+- Avoiding the generation or distribution of harmful or illegal content.
+- Respecting the rights and safety of others, including intellectual property rights.
+
+4. Monitoring and Enforcement
+To maintain service integrity:
+- JARVIS may monitor usage patterns to detect abuse.
+- In cases of serious violation, legal action may be taken.
+
+5. Limitation of Liability
+JARVIS is provided as-is with no guarantees or warranties.
+
+6. Changes and Updates
+This notice may be updated periodically. Continued use of JARVIS implies agreement to the latest version
+of the terms.
+
+7. Agreement
+By using JARVIS, you acknowledge that you:
+- Have read and understood this notice.
+- Agree not to misuse the platform.
+- Accept full responsibility for your interactions with the service.

jarvis.py

@@ -203,7 +203,7 @@ def extract_file_content(fp):
             return f"{fp}: {e}"
 
 async def fetch_response_stream_async(host, key, model, msgs, cfg, sid, stop_event):
-    for t in [1, 2]:
+    for t in [0.5, 1]:
         try:
             async with httpx.AsyncClient(timeout=t) as client:
                 async with client.stream("POST", host, json={**{"model": model, "messages": msgs, "session_id": sid, "stream": True}, **cfg}, headers={"Authorization": f"Bearer {key}"}) as response:
@@ -296,6 +296,8 @@ async def respond_async(multi, history, model_display, sess, custom_prompt):
         display_text = ""
         content_started = False
         async for typ, chunk in chat_with_model_async(history, inp, model_display, sess, custom_prompt):
+            if sess.stop_event.is_set():
+                break
             if typ == "reasoning":
                 if content_started:
                     continue
@@ -362,4 +364,4 @@ with gr.Blocks(fill_height=True, fill_width=True, title=AI_TYPES["AI_TYPE_4"], h
     system_prompt.change(fn=lambda x: x, inputs=[system_prompt], outputs=[custom_prompt_state])
     msg.submit(fn=respond_async, inputs=[msg, user_history, selected_model, user_session, custom_prompt_state], outputs=[chatbot, msg, user_session], api_name=INTERNAL_AI_GET_SERVER)
     msg.stop(fn=stop_response, inputs=[user_history, user_session], outputs=[chatbot, msg, user_session])
-jarvis.queue(default_concurrency_limit=3).launch(max_file_size="1mb")
+jarvis.queue(default_concurrency_limit=2).launch(max_file_size="1mb")