metadata

base_model: Rombo-Org/Rombo-LLM-V3.1-QWQ-32b
license: apache-2.0
tags:
  - llama-cpp
  - gguf-my-repo

Triangle104/Rombo-LLM-V3.1-QWQ-32b-Q4_K_M-GGUF

This model was converted to GGUF format from Rombo-Org/Rombo-LLM-V3.1-QWQ-32b using llama.cpp via the ggml.ai's GGUF-my-repo space. Refer to the original model card for more details on the model.

Rombo-LLM-V3.1-QWQ-32b is a Continued Finetune model (Merge only) of (Qwen/QwQ-32B) and its base model (Qwen/Qwen2.5-32B). This merge is done to decrease catastrophic forgetting during finetuning, and increase overall performance of the model. The tokenizers are taken from the QwQ-32B for thinking capabilities.

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 Triangle104/Rombo-LLM-V3.1-QWQ-32b-Q4_K_M-GGUF --hf-file rombo-llm-v3.1-qwq-32b-q4_k_m.gguf -p "The meaning to life and the universe is"

Server:

llama-server --hf-repo Triangle104/Rombo-LLM-V3.1-QWQ-32b-Q4_K_M-GGUF --hf-file rombo-llm-v3.1-qwq-32b-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

Step 2: Move into the llama.cpp folder and build it with LLAMA_CURL=1 flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).

cd llama.cpp && LLAMA_CURL=1 make

Step 3: Run inference through the main binary.

./llama-cli --hf-repo Triangle104/Rombo-LLM-V3.1-QWQ-32b-Q4_K_M-GGUF --hf-file rombo-llm-v3.1-qwq-32b-q4_k_m.gguf -p "The meaning to life and the universe is"

./llama-server --hf-repo Triangle104/Rombo-LLM-V3.1-QWQ-32b-Q4_K_M-GGUF --hf-file rombo-llm-v3.1-qwq-32b-q4_k_m.gguf -c 2048