v.0.3.8 (build 4) doesn't work for me- solved

#2
by bmanske1 - opened

DeepSeek-R1-Distill-Qwen-32B-GGUF Q4_K_M with lmstudio 0.3.8 (build 4) will not run on my i7 with 32GB of memory and a RTX 4060 with 8GB. When selecting the model to load the tag QWEN2 is highlighted in red. I get a llama.cpp error loading vocabulary unknown pretokenizer.

Since this is a lmstudio-community model I really expected it to run.

I think I found the solution - sorry. (hint: update the runtimes at the gear icon)

bmanske1 changed discussion title from v.0.3.8 (build 4) doesn't work for me to v.0.3.8 (build 4) doesn't work for me- solved

Sign up or log in to comment