---
inference: false
language:
- en
tags:
- llama
- text-generation-inference
---

# TogetherAI's Llama-2-7B-32K-Instruct GGML

These files are GGML format model files for [lmsys's Longchat-7B-v1.5-32k](https://huggingface.co/togethercomputer/Llama-2-7B-32K-Instruct).

GGML files are for CPU + GPU inference using [llama.cpp](https://github.com/ggerganov/llama.cpp) and libraries and UIs which support this format, such as:
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui)
* [KoboldCpp](https://github.com/LostRuins/koboldcpp)
* [ParisNeo/GPT4All-UI](https://github.com/ParisNeo/gpt4all-ui)
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python)
* [ctransformers](https://github.com/marella/ctransformers)

## How to run in `llama.cpp`

I use the following command line, adjust for your tastes and needs:

```
./main -t 2 -ngl 32 -m Llama-2-7B-32K-Instruct.ggmlv3.q4_0.bin --color --rope-freq-scale 0.25 --rope-freq-base 347620 -c 32768 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "[INST]\nWrite a poem about cats\n[\INST]\n\n"
```
Change `-t 2` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`.

Change `-ngl 26` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.

If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`, you can use `--interactive-first` to start in interactive mode.

## Compatibility

I have uploded bothe the original llama.cpp quant methods (`q4_0, q4_1, q5_0, q5_1, q8_0`) as well as the new k-quant methods (`q2_K, q3_K_S, q3_K_M, q3_K_L, q4_K_S, q4_K_M, q5_K_S, q6_K`).

Please refer to [llama.cpp](https://github.com/ggerganov/llama.cpp) and [TheBloke](https://huggingface.co/TheBloke)'s GGML models for further explanation.

## How to run in `text-generation-webui`

Further instructions here: [text-generation-webui/docs/llama.cpp-models.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/llama.cpp-models.md).

<!-- footer start -->
## Thanks

Thanks to [TheBloke](https://huggingface.co/TheBloke) for inspiration and providing almost all of the readme here!

Thanks to [TogetherAI](https://huggingface.co/togethercomputer) for providing checkpoints of the model.

Thanks to [Georgi Gerganov](https://github.com/ggerganov) and all of the awesome people in the AI community.