forked from ggml-org/llama.cpp
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
9b4a15b
commit e750735
Showing
4 changed files
with
94 additions
and
241 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,223 +1,46 @@ | ||
| # llama.cpp | ||
|
|
||
| [](https://github.com/ggerganov/llama.cpp/actions) | ||
| [](https://opensource.org/licenses/MIT) | ||
|
|
||
| Inference of [LLaMA](https://arxiv.org/abs/2302.13971) model in pure C/C++ | ||
|
|
||
| **Hot topics:** | ||
|
|
||
| - Cache input prompts for faster initialization: https://github.com/ggerganov/llama.cpp/issues/64 | ||
| - Create a `llama.cpp` logo: https://github.com/ggerganov/llama.cpp/issues/105 | ||
|
|
||
| ## Description | ||
|
|
||
| The main goal is to run the model using 4-bit quantization on a MacBook | ||
|
|
||
| - Plain C/C++ implementation without dependencies | ||
| - Apple silicon first-class citizen - optimized via ARM NEON | ||
| - AVX2 support for x86 architectures | ||
| - Mixed F16 / F32 precision | ||
| - 4-bit quantization support | ||
| - Runs on the CPU | ||
|
|
||
| This was [hacked in an evening](https://github.com/ggerganov/llama.cpp/issues/33#issuecomment-1465108022) - I have no idea if it works correctly. | ||
| Please do not make conclusions about the models based on the results from this implementation. | ||
| For all I know, it can be completely wrong. This project is for educational purposes. | ||
| New features will probably be added mostly through community contributions. | ||
|
|
||
| Supported platforms: | ||
|
|
||
| - [X] Mac OS | ||
| - [X] Linux | ||
| - [X] Windows (via CMake) | ||
|
|
||
| --- | ||
|
|
||
| Here is a typical run using LLaMA-7B: | ||
|
|
||
| ```java | ||
| make -j && ./main -m ./models/7B/ggml-model-q4_0.bin -p "Building a website can be done in 10 simple steps:" -t 8 -n 512 | ||
| I llama.cpp build info: | ||
| I UNAME_S: Darwin | ||
| I UNAME_P: arm | ||
| I UNAME_M: arm64 | ||
| I CFLAGS: -I. -O3 -DNDEBUG -std=c11 -fPIC -pthread -DGGML_USE_ACCELERATE | ||
| I CXXFLAGS: -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -pthread | ||
| I LDFLAGS: -framework Accelerate | ||
| I CC: Apple clang version 14.0.0 (clang-1400.0.29.202) | ||
| I CXX: Apple clang version 14.0.0 (clang-1400.0.29.202) | ||
|
|
||
| make: Nothing to be done for `default'. | ||
| main: seed = 1678486056 | ||
| llama_model_load: loading model from './models/7B/ggml-model-q4_0.bin' - please wait ... | ||
| llama_model_load: n_vocab = 32000 | ||
| llama_model_load: n_ctx = 512 | ||
| llama_model_load: n_embd = 4096 | ||
| llama_model_load: n_mult = 256 | ||
| llama_model_load: n_head = 32 | ||
| llama_model_load: n_layer = 32 | ||
| llama_model_load: n_rot = 128 | ||
| llama_model_load: f16 = 2 | ||
| llama_model_load: n_ff = 11008 | ||
| llama_model_load: ggml ctx size = 4529.34 MB | ||
| llama_model_load: memory_size = 512.00 MB, n_mem = 16384 | ||
| llama_model_load: .................................... done | ||
| llama_model_load: model size = 4017.27 MB / num tensors = 291 | ||
| main: prompt: 'Building a website can be done in 10 simple steps:' | ||
| main: number of tokens in prompt = 15 | ||
| 1 -> '' | ||
| 8893 -> 'Build' | ||
| 292 -> 'ing' | ||
| 263 -> ' a' | ||
| 4700 -> ' website' | ||
| 508 -> ' can' | ||
| 367 -> ' be' | ||
| 2309 -> ' done' | ||
| 297 -> ' in' | ||
| 29871 -> ' ' | ||
| 29896 -> '1' | ||
| 29900 -> '0' | ||
| 2560 -> ' simple' | ||
| 6576 -> ' steps' | ||
| 29901 -> ':' | ||
| sampling parameters: temp = 0.800000, top_k = 40, top_p = 0.950000 | ||
| Building a website can be done in 10 simple steps: | ||
| 1) Select a domain name and web hosting plan | ||
| 2) Complete a sitemap | ||
| 3) List your products | ||
| 4) Write product descriptions | ||
| 5) Create a user account | ||
| 6) Build the template | ||
| 7) Start building the website | ||
| 8) Advertise the website | ||
| 9) Provide email support | ||
| 10) Submit the website to search engines | ||
| A website is a collection of web pages that are formatted with HTML. HTML is the code that defines what the website looks like and how it behaves. | ||
| The HTML code is formatted into a template or a format. Once this is done, it is displayed on the user's browser. | ||
| The web pages are stored in a web server. The web server is also called a host. When the website is accessed, it is retrieved from the server and displayed on the user's computer. | ||
| A website is known as a website when it is hosted. This means that it is displayed on a host. The host is usually a web server. | ||
| A website can be displayed on different browsers. The browsers are basically the software that renders the website on the user's screen. | ||
| A website can also be viewed on different devices such as desktops, tablets and smartphones. | ||
| Hence, to have a website displayed on a browser, the website must be hosted. | ||
| A domain name is an address of a website. It is the name of the website. | ||
| The website is known as a website when it is hosted. This means that it is displayed on a host. The host is usually a web server. | ||
| A website can be displayed on different browsers. The browsers are basically the software that renders the website on the user’s screen. | ||
| A website can also be viewed on different devices such as desktops, tablets and smartphones. Hence, to have a website displayed on a browser, the website must be hosted. | ||
| A domain name is an address of a website. It is the name of the website. | ||
| A website is an address of a website. It is a collection of web pages that are formatted with HTML. HTML is the code that defines what the website looks like and how it behaves. | ||
| The HTML code is formatted into a template or a format. Once this is done, it is displayed on the user’s browser. | ||
| A website is known as a website when it is hosted | ||
|
|
||
| main: mem per token = 14434244 bytes | ||
| main: load time = 1332.48 ms | ||
| main: sample time = 1081.40 ms | ||
| main: predict time = 31378.77 ms / 61.41 ms per token | ||
| main: total time = 34036.74 ms | ||
| ``` | ||
|
|
||
| And here is another demo of running both LLaMA-7B and [whisper.cpp](https://github.com/ggerganov/whisper.cpp) on a single M1 Pro MacBook: | ||
|
|
||
| https://user-images.githubusercontent.com/1991296/224442907-7693d4be-acaa-4e01-8b4f-add84093ffff.mp4 | ||
| # Alpaca.cpp | ||
|
|
||
| ## Usage | ||
| Run a fast ChatGPT-like model locally on your device. The screencast below is not sped up and running on an M2 Macbook Air with 4GB of weights. | ||
|
|
||
| Here are the step for the LLaMA-7B model: | ||
|
|
||
| ```bash | ||
| # build this repo | ||
| git clone https://github.com/ggerganov/llama.cpp | ||
| cd llama.cpp | ||
| make | ||
| [](https://asciinema.org/a/dfJ8QXZ4u978Ona59LPEldtKK) | ||
|
|
||
| # obtain the original LLaMA model weights and place them in ./models | ||
| ls ./models | ||
| 65B 30B 13B 7B tokenizer_checklist.chk tokenizer.model | ||
|
|
||
| # install Python dependencies | ||
| python3 -m pip install torch numpy sentencepiece | ||
| This combines the [LLaMA foundation model](https://github.com/facebookresearch/llama) with an [open reproduction](https://github.com/tloen/alpaca-lora) of [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) a fine-tuning of the base model to obey instructions (akin to the [RLHF](https://huggingface.co/blog/rlhf) used to train ChatGPT). | ||
|
|
||
| # convert the 7B model to ggml FP16 format | ||
| python3 convert-pth-to-ggml.py models/7B/ 1 | ||
| ## Get started | ||
|
|
||
| # quantize the model to 4-bits | ||
| ./quantize.sh 7B | ||
|
|
||
| # run the inference | ||
| ./main -m ./models/7B/ggml-model-q4_0.bin -t 8 -n 128 | ||
| ``` | ||
| git clone https://github.com/antimatter15/alpaca.cpp | ||
| cd alpaca.cpp | ||
| When running the larger models, make sure you have enough disk space to store all the intermediate files. | ||
|
|
||
| TODO: add model disk/mem requirements | ||
|
|
||
| ### Interactive mode | ||
|
|
||
| If you want a more ChatGPT-like experience, you can run in interactive mode by passing `-i` as a parameter. | ||
| In this mode, you can always interrupt generation by pressing Ctrl+C and enter one or more lines of text which will be converted into tokens and appended to the current context. You can also specify a *reverse prompt* with the parameter `-r "reverse prompt string"`. This will result in user input being prompted whenever the exact tokens of the reverse prompt string are encountered in the generation. A typical use is to use a prompt which makes LLaMa emulate a chat between multiple users, say Alice and Bob, and pass `-r "Alice:"`. | ||
|
|
||
| Here is an example few-shot interaction, invoked with the command | ||
| make chat | ||
| ./chat | ||
| ``` | ||
| ./main -m ./models/13B/ggml-model-q4_0.bin -t 8 -n 256 --repeat_penalty 1.0 --color -i -r "User:" \ | ||
| -p \ | ||
| "Transcript of a dialog, where the User interacts with an Assistant named Bob. Bob is helpful, kind, honest, good at writing, and never fails to answer the User's requests immediately and with precision. | ||
|
|
||
| User: Hello, Bob. | ||
| Bob: Hello. How may I help you today? | ||
| User: Please tell me the largest city in Europe. | ||
| Bob: Sure. The largest city in Europe is Moscow, the capital of Russia. | ||
| User:" | ||
| You can download the weights for `ggml-alpaca-7b-14.bin` with BitTorrent `magnet:?xt=urn:btih:5aaceaec63b03e51a98f04fd5c42320b2a033010&dn=ggml-alpaca-7b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce` | ||
|
|
||
| ``` | ||
| Note the use of `--color` to distinguish between user input and generated text. | ||
|
|
||
|  | ||
| Alternatively you can download them with IPFS. | ||
|
|
||
| ### Android | ||
|
|
||
| You can easily run `llama.cpp` on Android device with [termux](https://play.google.com/store/apps/details?id=com.termux). | ||
| First, obtain the [Android NDK](https://developer.android.com/ndk) and then build with CMake: | ||
| ``` | ||
| $ mkdir build-android | ||
| $ cd build-android | ||
| $ export NDK=<your_ndk_directory> | ||
| $ cmake -DCMAKE_TOOLCHAIN_FILE=$NDK/build/cmake/android.toolchain.cmake -DANDROID_ABI=arm64-v8a -DANDROID_PLATFORM=android-23 -DCMAKE_C_FLAGS=-march=armv8.4a+dotprod .. | ||
| $ make | ||
| # any of these will work | ||
| wget -O ggml-alpaca-7b-q4.bin -c https://gateway.estuary.tech/gw/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC | ||
| wget -O ggml-alpaca-7b-q4.bin -c https://ipfs.io/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC | ||
| wget -O ggml-alpaca-7b-q4.bin -c https://cloudflare-ipfs.com/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC | ||
| ``` | ||
| Install [termux](https://play.google.com/store/apps/details?id=com.termux) on your device and run `termux-setup-storage` to get access to your SD card. | ||
| Finally, copy the `llama` binary and the model files to your device storage. Here is a demo of an interactive session running on Pixel 5 phone: | ||
|
|
||
| https://user-images.githubusercontent.com/271616/225014776-1d567049-ad71-4ef2-b050-55b0b3b9274c.mp4 | ||
|
|
||
| Save the `ggml-alpaca-7b-14.bin` file in the same directory as your `./chat` executable. | ||
|
|
||
| ## Limitations | ||
| The weights are based on the published fine-tunes from `alpaca-lora`, converted back into a pytorch checkpoint with a [modified script](https://github.com/tloen/alpaca-lora/pull/19) and then quantized with llama.cpp the regular way. | ||
|
|
||
| - We don't know yet how much the quantization affects the quality of the generated text | ||
| - Probably the token sampling can be improved | ||
| - The Accelerate framework is actually currently unused since I found that for tensor shapes typical for the Decoder, | ||
| there is no benefit compared to the ARM_NEON intrinsics implementation. Of course, it's possible that I simply don't | ||
| know how to utilize it properly. But in any case, you can even disable it with `LLAMA_NO_ACCELERATE=1 make` and the | ||
| performance will be the same, since no BLAS calls are invoked by the current implementation | ||
| ## Credit | ||
|
|
||
| ### Contributing | ||
| This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tatsu-lab/stanford_alpaca) (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and a modified version of [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA. | ||
|
|
||
| - Contributors can open PRs | ||
| - Collaborators can push to branches in the `llama.cpp` repo | ||
| - Collaborators will be invited based on contributions | ||
|
|
||
| ### Coding guidelines | ||
| ## Disclaimer | ||
|
|
||
| - Avoid adding third-party dependencies, extra files, extra headers, etc. | ||
| - Always consider cross-compatibility with other operating systems and architectures | ||
| - Avoid fancy looking modern STL constructs, use basic `for` loops, avoid templates, keep it simple | ||
| - There are no strict rules for the code style, but try to follow the patterns in the code (indentation, spaces, etc.). Vertical alignment makes things more readable and easier to batch edit | ||
| - Clean-up any trailing whitespaces, use 4 spaces indentation, brackets on same line, `void * ptr`, `int & a` | ||
| - See [good first issues](https://github.com/ggerganov/llama.cpp/issues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22) for tasks suitable for first contributions | ||
| Note that the model weights are only to be used for research purposes, as they are derivative of LLaMA, and uses the published instruction data from the Stanford Alpaca project which is generated by OpenAI, which itself disallows the usage of its outputs to train competing models. | ||
|
|
||
| ### Misc | ||
|
|
||
| - Practice your C++ typing skills: https://typing-battles.ggerganov.com |
Oops, something went wrong.