IntelliNode

This PR is example how to add new provider to intellinode: https://github.com/intelligentnode/IntelliNode/pull/86

@basilmusa kindly check the chatbot response structure below: you can access the none-stream model text response as follow: `results.text` ```json { "text": "string", "generation_id": "3fa85f64-5717-4562-b3fc-2c963f66afa6", "citations": [ { "start": 0,...

this is implemented

I left a few comments about the code lines. The requirements file looks good, as it does not use high-level modules. Also, I have the following questions: - What is...

I tested the PR and it is close to get done. Kindly work on the following: 1. Resolve the conflict with main. 2. Add a file called `test_deepseek_extend_wrapper.py` that uses...

I loaded the model using terminal and it seems the previous errors fixed, but **the model does not generated the tokens as expected**. When I run `test_code_generation`, following output printed:...

The token parsing does not work as expected and the model generate unknown characters! Example: Ġwild//================================================================================Erotìłķë³´.controllersPopupMenuáºĬĠEntryĠCoachingĠ׳×ŀצ×IJasureæ¯Ģä¸ĢçĤ¹ĠönerĠscop_COLLãģ¿ãģªintíķĺìĭľlope广大ĠBiosĠforsk봽裹BootstrapĠ×Ķ×Ĺ×ķ׾×Ļ×Ŀ);//Missionä¹Łä¸įè¦ģĠgetIntìīIJĠ|ĊåĨįçĶŁèĥ½æºIJnamaĠKernelReporter.pojoĠcollaborateumeà

Time passed.

Thanks for the attempt, it is good you used only low level dependencies. Kindly answer the following questions to help with the review: - There is no tokenizer implementation. How...

Deepseek use Byte-level BPE (BBPE), not whitespace word tokenizer. Please adjust.