Chatbot v0.2

[SeverusYixin]

No description provided.

[SeverusYixin]

Still have the Token authorization problem

[haesleinhuepf]

If you need assistance, feel free to open an issue and write down what the "Token authorization problem" is. ;-)

[SeverusYixin]

From Docker's log it looks like running on the CPU and needs some testing.

[SeverusYixin]

It will always auto change to the CPU and show me it don't have enough memory on docker as "Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s]
chatbot_backend exited with code 137"

[SeverusYixin]

The chat works(possible) after the update but is still on the CPU, making it slow. I'll try utilizing the laptop's NVIDIA RTX 500 Ada Generation GPU to improve performance.

[SeverusYixin]

The CPU is still computing...
Using the GPU should be a good option

[SeverusYixin]

Now, it got a good answer as:

"curl -X POST http://localhost:5002/api/chat -H "Content-Type: application/json" -d "{"query": "Tell me about bioimage analysis"}"
{
"response": "Based on the following documents, answer the user's question concisely and include relevant links.\n\n## Documents\n- Slides about FLUTE: a Python GUI for interactive phasor analysis of FLIM data: This presentation introduces the open source software to analyze FLIM data:\nFLUTE – (F)luorescence (L)ifetime (U)ltima(T)e (E)xplorer:\na Python GUI for interactive phasor analysis of FLIM data\n \nThe software is available on GitHub: https://github.com/LaboratoryOpticsBiosciences/FLUTE\nand it is published on Biological imaging Journal: Gottlieb, D., Asadipour, B., Kostina, P., Ung, T., & Stringari, C. (2023). FLUTE: A Python GUI for interactive phasor analysis of FLIM data. Biological Imaging, 1-22. doi:10.1017/S2633903X23000211\nThe lecture was part of the short talks on community developed FLIM-software at the German BioImaging workshop on FLIM in Munich. (URL: ['https://zenodo.org/records/10839310', 'https://doi.org/10.5281/zenodo.10839310'])\n- BioImage Analysis Notebooks: No description available (URL: https://haesleinhuepf.github.io/BioImageAnalysisNotebooks/intro.html)\n- BioImage Data Analysis: No description available (URL: https://analyticalscience.wiley.com/do/10.1002/was.00050003/full/bioimagedataanalysis.pdf)\n\n## Question\nTell me about bioimage analysis\n\n## Answer\nFluorescence lifetime imaging microscopy (FLIM) is a technique that combines fluorescence microscopy and time-resolved fluorescence spectroscopy to quantitatively measure the fluorescence lifetime of a fluorophore in a cellular sample. The fluorescence lifetime is a measure of the energy transfer rate between the excited and the ground state of the fluorophore. FLIM can be used to measure the dynamics of fluorescent proteins in live cells, which can provide insight into the function and regulation of biological processes.\n\nFLIM data is typically analyzed using specialized software, such as F",
"sources": [
{
"description": "This presentation introduces the open source software to analyze FLIM data:\nFLUTE – (F)luorescence (L)ifetime (U)ltima(T)e (E)xplorer:\na Python GUI for interactive phasor analysis of FLIM data\n \nThe software is available on GitHub: https://github.com/LaboratoryOpticsBiosciences/FLUTE\nand it is published on Biological imaging Journal: Gottlieb, D., Asadipour, B., Kostina, P., Ung, T., & Stringari, C. (2023). FLUTE: A Python GUI for interactive phasor analysis of FLIM data. Biological Imaging, 1-22. doi:10.1017/S2633903X23000211\nThe lecture was part of the short talks on community developed FLIM-software at the German BioImaging workshop on FLIM in Munich.",
"name": "Slides about FLUTE: a Python GUI for interactive phasor analysis of FLIM data",
"url": [
"https://zenodo.org/records/10839310",
"https://doi.org/10.5281/zenodo.10839310"
]
},
{
"description": "No description available",
"name": "BioImage Analysis Notebooks",
"url": "https://haesleinhuepf.github.io/BioImageAnalysisNotebooks/intro.html"
},
{
"description": "No description available",
"name": "BioImage Data Analysis",
"url": "https://analyticalscience.wiley.com/do/10.1002/was.00050003/full/bioimagedataanalysis.pdf"
}
]
}

C:\Users\huang>"

The current problem is that it still works on my CPU, so it is not efficient.