note: most qwen2 weights aren't divisible by 256, so this is really a q8/q5 quant.

main-horse/UI-TARS-72B-SFT-Q4_K_M-GGUF

This model was converted to GGUF format from bytedance-research/UI-TARS-72B-SFT using llama.cpp. Refer to the original model card for more details on the model.

Use with llama.cpp

Install llama.cpp through brew (works on Mac and Linux)

brew install llama.cpp

Invoke the llama.cpp server or the CLI.

CLI:

llama-cli --hf-repo main-horse/UI-TARS-72B-SFT-Q4_K_M-GGUF --hf-file UI-TARS-72B-SFT.Q4_K_M.gguf -p "The meaning to life and the universe is"

Server:

llama-server --hf-repo main-horse/UI-TARS-72B-SFT-Q4_K_M-GGUF --hf-file UI-TARS-72B-SFT.Q4_K_M.gguf -c 2048

Note: You can also use this checkpoint directly through the usage steps listed in the Llama.cpp repo as well. Step 1: Clone llama.cpp from GitHub.

git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp

Step 2: Build using CMake.

cmake -B build -DGGML_CUDA=ON -DGGML_CUDA_F16=1 -DGGML_CUDA_FA_ALL_QUANTS=1 -DCMAKE_CUDA_ARCHITECTURES=...
cmake --build build --config Release -j

Step 3: Run inference through the main binary.

./llama-server --hf-repo main-horse/UI-TARS-72B-SFT-Q4_K_M-GGUF --hf-file UI-TARS-72B-SFT.Q4_K_M.gguf -c 2048