# llama.cpp

[llama.cpp](https://github.com/ggml-org/llama.cpp) is a C/C++ inference engine for deploying large language models locally. It's lightweight and doesn't require Python, CUDA, or other heavy server infrastructure. llama.cpp uses the [GGUF](https://huggingface.co/blog/ngxson/common-ai-model-formats#gguf) file format. GGUF supports quantized model weights and memory-mapping to reduce memory bandwidth on your device.

> [!TIP]
> Browse the [Hub](https://huggingface.co/models?apps=llama.cpp&sort=trending) for models already available in GGUF format.

llama.cpp can convert and run Transformers models as standalone C++ executables with the [convert_hf_to_gguf.py](https://github.com/ggml-org/llama.cpp/blob/master/convert_hf_to_gguf.py) script.

```bash
python3 convert_hf_to_gguf.py ./models/openai/gpt-oss-20b
```

The conversion process works as follows.

1. The script loads the model configuration with [AutoConfig.from_pretrained()](/docs/transformers/v5.1.0/en/model_doc/auto#transformers.AutoConfig.from_pretrained) and extracts the 
vocabulary with [AutoTokenizer.from_pretrained()](/docs/transformers/v5.1.0/en/model_doc/auto#transformers.AutoTokenizer.from_pretrained).
2. Based on the config's architecture field, the script selects a converter class from its internal registry. The registry maps Transformers architecture names (like [LlamaForCausalLM](/docs/transformers/v5.1.0/en/model_doc/llama2#transformers.LlamaForCausalLM)) to corresponding converter classes.
3. The converter class extracts config parameters, maps Transformers tensor names to GGUF tensor names, transforms tensors, and packages the vocabulary.
4. The output is a single GGUF file containing the model weights, tokenizer, and metadata.

Deploy the model locally from the command line with [llama-cli](https://github.com/ggml-org/llama.cpp/tree/master#llama-cli) or start a web UI with [llama-server](https://github.com/ggml-org/llama.cpp/tree/master#llama-server). Add the `-hf` flag to indicate the model is from the Hub.

```bash
llama-cli -hf openai/gpt-oss-20b
```

```bash
llama-server -hf ggml-org/gpt-oss-20b-GGUF
```

## Resources

- [llama.cpp](https://github.com/ggml-org/llama.cpp) documentation
- [Introduction to ggml](https://huggingface.co/blog/introduction-to-ggml) blog post