david-j-smith comments

Repositories
Issues
Comments

Results 2 comments of


                                            david-j-smith

LLamaSharp.Backend.OpenCL 0.11.2: GGML_ASSERT llama.cpp:14093: hparams.n_embd_head_v % ggml_blck_size(type_v) == 0

Yes, I tested with llama.cpp release b2303 (commit 3ab8b3a) and the latest (binary) release b2589 (commit 1ff4d9f). Both work as expected and for all GGUF models. However, I have noticed...

LLamaSharp.Backend.OpenCL 0.11.2: GGML_ASSERT llama.cpp:14093: hparams.n_embd_head_v % ggml_blck_size(type_v) == 0

I had hoped that the iGPU would be around the same level of CPU performance, but more efficient (which on my PC actually means: much quieter 🤣). But the CUDA...