Releases: pymike00/tinychat
0.8.2
Hello world,
Happy to announce the release of tinychat 0.8.2 for Linux and Windows!
This version adds support for the newly released Claude 3.5 Sonnet by Anthropic and Mistral Codestral. Support for GPT-3.5 and Mistral Medium has been discontinued.
Claude 3.5 Sonnet Info: https://www.anthropic.com/news/claude-3-5-sonnet
Mistral Codestral Info: https://mistral.ai/news/codestral/
Important Note: Our executables are created with pyinstaller, which might cause them to be mistakenly flagged as malicious by some antiviruses (false positive). This is a well known issue and unfortunately no, manually compiling pyinstaller doesn't fix the issue.
If your antivirus flags tinychat, you may need to add an exception in order to use it. For your peace of mind, you can always compile the application yourself following our detailed instructions.
I don't own a Mac therefore I can't provide any builds for it, but it should run / be built with no issues on it too.
Feedback and questions are always welcome. Please feel free to report any issues you may encounter and offer any suggestion you may have regarding the improvement of the project. Contributions are also very much encouraged.
Most importantly,
Enjoy! ☕
0.8.1
Hello world,
Happy to announce the release of tinychat 0.8.1 for Linux and Windows!
This version adds support for the newly released GPT-4o model by OpenAI.
In case you missed it, among other things:
- GPT-4o is 2x faster at generating tokens than GPT-4 Turbo.
- GPT-4o is 50% cheaper than GPT-4 Turbo, across both input tokens ($5 per million) and output tokens ($15 per million).
Model Info: https://community.openai.com/t/announcing-gpt-4o-in-the-api/744700
Model Info: https://openai.com/index/hello-gpt-4o/
Important Note: Our executables are created with pyinstaller, which might cause them to be mistakenly flagged as malicious by some antiviruses (false positive). This is a well known issue and unfortunately no, manually compiling pyinstaller doesn't fix the issue.
If your antivirus flags tinychat, you may need to add an exception in order to use it. For your peace of mind, you can always compile the application yourself following our detailed instructions.
I don't own a Mac therefore I can't provide any builds for it, but it should run / be built with no issues on it too.
Feedback and questions are always welcome. Please feel free to report any issues you may encounter and offer any suggestion you may have regarding the improvement of the project. Contributions are also very much encouraged.
Most importantly,
Enjoy! ☕
0.8.0
Hello world,
Happy to announce the release of tinychat 0.8.0 for Linux and Windows!
This version adds support for the Llama3 8B and Llama3 70B models by Meta AI via Togheter API.
The value for the temperature parameter can now be changed via settings (please restart the application after changing it)
Important Note: Our executables are created with pyinstaller, which might cause them to be mistakenly flagged as malicious by some antiviruses (false positive). This is a well known issue and unfortunately no, manually compiling pyinstaller doesn't fix the issue.
If your antivirus flags tinychat, you may need to add an exception in order to use it. For your peace of mind, you can always compile the application yourself following our detailed instructions.
I don't own a Mac therefore I can't provide any builds for it, but it should run / be built with no issues on it too.
Feedback and questions are always welcome. Please feel free to report any issues you may encounter and offer any suggestion you may have regarding the improvement of the project. Contributions are also very much encouraged.
Most importantly,
Enjoy! ☕
0.7.5
Hello world,
Happy to announce the release of tinychat 0.7.5 for Linux and Windows!
This version adds support for Gemini 1.5 Pro and Claude 3 Sonnet.
Mistral Small, Mistral 7B and Mistral 8X7B are no longer among the available models, but you can always add them back by updating the Backend._models dictionary if needed. The decision to only keep the most powerful models was taken to reduce choice overload. Support for GPT-3.5 will also be discontinued in future releases.
Important Note: Our executables are created with pyinstaller, which might cause them to be mistakenly flagged as malicious by some antiviruses (false positive). This is a well known issue and unfortunately no, manually compiling pyinstaller doesn't fix the issue.
If your antivirus flags tinychat, you may need to add an exception in order to use it. For your peace of mind, you can always compile the application yourself following our detailed instructions.
I don't own a Mac therefore I can't provide any builds for it, but it should run / be built with no issues on it too.
Feedback and questions are always welcome. Please feel free to report any issues you may encounter and offer any suggestion you may have regarding the improvement of the project. Contributions are also very much encouraged.
Most importantly,
Enjoy! ☕
0.7.4
Hello world,
We're excited to announce the release of tinychat 0.7.4 for Linux and Windows!
This version offers support for the most recently released Claude 3 Opus model by Anthropic.
Important Note: Our executables are created with pyinstaller, which might cause them to be mistakenly flagged as malicious by some antiviruses (false positive). This is a well known issue and unfortunately no, manually compiling pyinstaller doesn't fix the issue.
If your antivirus flags tinychat, you may need to add an exception in order to use it. For your peace of mind, you can always compile the application yourself following our detailed instructions.
I don't own a Mac therefore I can't provide any builds for it, but it should run / be built with no issues on it too.
We welcome your feedback and questions! Please feel free to report any issues you may encounter and offer any suggestion you may have regarding the improvement of the project. Contributions are also very much welcome!
Most importantly,
Enjoy! ☕