intel/ipex-llm - Accelerate local LLM inference and finetuning on Intel XPUs
https://github.com/intel/ipex-llm/
7d8612d806
* optimize logsoftmax * use ev instead of MKL * add clearState * revise precision of InceptionSpec and EvaluatorSpec using MKL * revise GoogleNet_v1 test precision * change buffer name |
||
|---|---|---|