Tempus-labs

Tempus V1.0.2 Release: Enhanced `QuantAIAgent.chat_stream` Method**

🛠 **Bug Fixes**
**1. Redundant Output:**
- Fixed an issue in the `chat_stream` method where repeated message chunks were generated during streaming.
- Implemented a unique message ID tracking system (`_printed` set) to ensure only new and unique messages are yielded.

**2. Inconsistent Output:**
- Enhanced message handling to maintain a consistent and smooth streaming experience.
- Improved the logic for processing streamed responses, preventing duplication and ensuring a coherent flow of AI-generated messages.

---

✨ **Improvements**
**🔍 Real-Time State Tracking:**
- Added logging of the `dialog_state` during streaming, providing better visibility into the current state of AI interactions.

**⚡ Optimized Streaming Performance:**
- The method now efficiently manages message IDs using a set, avoiding redundant outputs.
- Enhanced `conversation_history` management by accurately appending `HumanMessage` and `AIMessage` objects, ensuring complete and accurate conversation context.

---

📈 **Impact**
This update significantly enhances the performance and stability of the `QuantAIAgent` class, delivering a more reliable and consistent AI-driven market analysis experience within the Solana ecosystem.

💡 **Developers Note:**
For detailed implementation and usage examples, refer to the updated [Documentation](https://tempuslabs.gitbook.io/tempus) in the repository.

**Full Changelog**: https://github.com/Tempus-Labs/tempus/commits/v1.0.2

Basic usage:
python
from tempus.agents import QuantAIAgent

Initialize agent with Deepseek
agent = QuantAIAgent(llm_provider="deepseek", model_name="deepseek-chat")

Analyze pump.fun market
response = agent.chat("Analyze the top 10 trending ai meta on Pump.fun")
print(response)

Stream chat for analyze $ai16z coin
for chunk in agent.chat_stream("What about $ai16z coin?"):
chunk.pretty_print()

Documentation
For detailed documentation and examples, visit our [Documentation](https://tempuslabs.gitbook.io/tempus).

Feedback and Support
We welcome feedback and contributions! Please open issues on our GitHub repository for bug reports or feature requests.

Note: This is our first stable release, and we are actively working on expanding the documentation and adding more features. Stay tuned for updates!

**Full Changelog**: https://github.com/Tempus-Labs/tempus/commits/v1.0.0