bloomz.cpp下载bloomz.cpp源码下载

bloomz.cpp

C/C++

1.0.0

下载

布卢姆兹.cpp

在纯 C/C++ 中推理 HuggingFace 的类 BLOOM 模型。

该存储库是在@ggerganov 令人惊叹的 llama.cpp 存储库之上构建的，以支持 BLOOM 模型。它支持可以使用BloomForCausalLM.from_pretrained()加载的所有模型。

布卢姆兹-7b1

演示

布卢姆兹-7b1

用法

首先，您需要克隆存储库并构建它：

git clone https://github.com/NouamaneTazi/bloomz.cpp
cd bloomz.cpp
make

转换重量

然后，您必须将模型权重转换为 ggml 格式。任何 BLOOM 模型都可以转换。

Hub 上托管的一些权重已经转换。您可以在此处找到该列表。

否则，转换权重的最快方法是使用此转换器工具。它是托管在 Huggingface Hub 上的一个空间，可以为您转换和量化权重并将其上传到您选择的存储库。

如果您愿意，您可以手动转换机器上的重量：

 # install required libraries
python3 -m pip install torch numpy transformers accelerate

# download and convert the 7B1 model to ggml FP16 format
python3 convert-hf-to-ggml.py bigscience/bloomz-7b1 ./models 
# Note: you can add --use-f32 to convert to FP32 instead of FP16

您也可以选择将模型量化为 4 位。

./quantize ./models/ggml-model-bloomz-7b1-f16.bin ./models/ggml-model-bloomz-7b1-f16-q4_0.bin 2

运行推理

最后，您可以运行推理。

./main -m ./models/ggml-model-bloomz-7b1-f16-q4_0.bin -t 8 -n 128

你的输出应该是这样的：

make && ./main -m models/ggml-model-bloomz-7b1-f16-q4_0.bin  -p ' Translate "Hi, how are you?" in French: ' -t 8 -n 256

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 13.1.6 (clang-1316.0.21.2.5)
I CXX:      Apple clang version 13.1.6 (clang-1316.0.21.2.5)

make: Nothing to be done for ` default ' .
main: seed = 1678899845
llama_model_load: loading model from ' models/ggml-model-bloomz-7b1-f16-q4_0.bin ' - please wait ...
llama_model_load: n_vocab = 250880
llama_model_load: n_ctx   = 512
llama_model_load: n_embd  = 4096
llama_model_load: n_mult  = 1
llama_model_load: n_head  = 32
llama_model_load: n_layer = 30
llama_model_load: f16     = 2
llama_model_load: n_ff    = 16384
llama_model_load: n_parts = 1
llama_model_load: ggml ctx size = 5312.64 MB
llama_model_load: memory_size =   480.00 MB, n_mem = 15360
llama_model_load: loading model part 1/1 from ' models/ggml-model-bloomz-7b1-f16-q4_0.bin '
llama_model_load: ............................................. done
llama_model_load: model size =  4831.16 MB / num tensors = 366

main: prompt: ' Translate " Hi, how are you? " in French: '
main: number of tokens in prompt = 11
153772 -> ' Translate '
 17959 -> ' " H'
    76 -> 'i'
 98257 -> ', '
 20263 -> 'how'
  1306 -> ' are'
  1152 -> ' you'
  2040 -> '?'
     5 -> ' " '
   361 -> ' in '
196427 -> ' French: '

sampling parameters: temp = 0.800000, top_k = 40, top_p = 0.950000, repeat_last_n = 64, repeat_penalty = 1.300000


Translate "Hi, how are you?" in French: Bonjour, comment ça va?</s> [end of text]


main: mem per token = 24017564 bytes
main:     load time =  3092.29 ms
main:   sample time =     2.40 ms
main:  predict time =  1003.04 ms / 59.00 ms per token
main:    total time =  5307.23 ms

高级用法

以下是可用选项的列表：

usage: ./main [options]

options:
  -h, --help            show this help message and exit
  -s SEED, --seed SEED  RNG seed (default: -1)
  -t N, --threads N     number of threads to use during computation (default: 4)
  -p PROMPT, --prompt PROMPT
                        prompt to start generation with (default: random)
  -n N, --n_predict N   number of tokens to predict (default: 128)
  --top_k N             top-k sampling (default: 40)
  --top_p N             top-p sampling (default: 0.9)
  --repeat_last_n N     last n tokens to consider for penalize (default: 64)
  --repeat_penalty N    penalize repeat sequence of tokens (default: 1.3)
  --temp N              temperature (default: 0.8)
  -b N, --batch_size N  batch size for prompt processing (default: 8)
  -m FNAME, --model FNAME
                        model path (default: models/ggml-model-bloomz-7b1-f16-q4_0.bin)