You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
llama.cpp/README.md

196 lines
7.7 KiB

2 years ago
# llama.cpp
Inference of [Facebook's LLaMA](https://github.com/facebookresearch/llama) model in pure C/C++
2 years ago
**Hot topics**
2 years ago
2 years ago
- Running on Windows: https://github.com/ggerganov/llama.cpp/issues/22
- Fix Tokenizer / Unicode support: https://github.com/ggerganov/llama.cpp/issues/11
2 years ago
2 years ago
## Description
The main goal is to run the model using 4-bit quantization on a MacBook
2 years ago
- Plain C/C++ implementation without dependencies
- Apple silicon first-class citizen - optimized via Arm Neon and Accelerate framework
- AVX2 support for x86 architectures
2 years ago
- 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.
2 years ago
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 and is not going to be maintained properly.
New features will probably be added mostly through community contributions, if any.
Supported platformst:
- [X] Mac OS
- [X] Linux
- [ ] Windows (soon)
2 years ago
---
2 years ago
Here is a typical run using LLaMA-7B:
2 years ago
```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:
2 years ago
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)
2 years ago
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 ...
2 years ago
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
2 years ago
llama_model_load: n_rot = 128
2 years ago
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
2 years ago
main: prompt: 'Building a website can be done in 10 simple steps:'
main: number of tokens in prompt = 15
2 years ago
1 -> ''
2 years ago
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 -> ':'
2 years ago
sampling parameters: temp = 0.800000, top_k = 40, top_p = 0.950000
2 years ago
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 users 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 users browser.
A website is known as a website when it is hosted
2 years ago
main: mem per token = 14434244 bytes
2 years ago
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
2 years ago
```
2 years ago
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
2 years ago
## Usage
Here are the step for the LLaMA-7B model:
2 years ago
```bash
# build this repo
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp
make
# 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
2 years ago
# convert the 7B model to ggml FP16 format
python3 convert-pth-to-ggml.py models/7B/ 1
# quantize the model to 4-bits
./quantize ./models/7B/ggml-model-f16.bin ./models/7B/ggml-model-q4_0.bin 2
# run the inference
./main -m ./models/7B/ggml-model-q4_0.bin -t 8 -n 128
```
For the bigger models, there are a few extra quantization steps. For example, for LLaMA-13B, converting to FP16 format
will create 2 ggml files, instead of one:
```bash
ggml-model-f16.bin
ggml-model-f16.bin.1
```
You need to quantize each of them separately like this:
```bash
./quantize ./models/13B/ggml-model-f16.bin ./models/13B/ggml-model-q4_0.bin 2
./quantize ./models/13B/ggml-model-f16.bin.1 ./models/13B/ggml-model-q4_0.bin.1 2
```
Everything else is the same. Simply run:
```bash
./main -m ./models/13B/ggml-model-q4_0.bin -t 8 -n 128
```
The number of files generated for each model is as follows:
```
7B -> 1 file
13B -> 2 files
30B -> 4 files
65B -> 8 files
```
When running the larger models, make sure you have enough disk space to store all the intermediate files.
2 years ago
## Limitations
- Not sure if my tokenizer is correct. There are a few places where we might have a mistake:
- https://github.com/ggerganov/llama.cpp/blob/26c084662903ddaca19bef982831bfb0856e8257/convert-pth-to-ggml.py#L79-L87
- https://github.com/ggerganov/llama.cpp/blob/26c084662903ddaca19bef982831bfb0856e8257/utils.h#L65-L69
In general, it seems to work, but I think it fails for unicode character support. Hopefully, someone can help with that
- I don't know yet how much the quantization affects the quality of the generated text
- Probably the token sampling can be improved
2 years ago
- 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 simlpy 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
2 years ago