metadata
license: apache-2.0
Unofficial GGUF Quantizations of Grok-1. Should just run with llama.cpp once the PR- Add grok-1 support #6204 is merged.
You need to merge the GGUF file using gguf-split before using them (there's already a PR to change this).
The splits now use PR: llama_model_loader: support multiple split/shard GGUFs.
Therefore no merging using gguf-split is needed anymore.
For now only Q2_K Quant, others (Q3_K, Q4_K, Q5_K & Q6_K) are prepared waiting to upload.