lxkarthi

3 ms·

1. 1.▲
   by lxkarthi 1mo ago

   I can run Deepseek flash 0731 version (ds4, esl3) on single DGX spark. getting around ~20 tok/s. It's great. Quantized version of this model would probably run on the DGX spark. I am excited to wait for quantized models that fits
2. 2.▲
   by lxkarthi 3y ago

   NVIDIA announced that RAPIDS cuDF can now bring GPU acceleration to 9.5M million pandas users without requiring them to change their code. python -m cudf.pandas script.py More info at https://rapids.ai/cudf-pandas/
3. 3.▲
   NVIDIA Accelerates Pandas Nearly 150x with Zero Code Changes with RAPIDS cuDF(developer.nvidia.com)

   6 points by lxkarthi 3y ago|1 comments