Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-01-12 21:37:19 +01:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Hongji Zhu 2024-05-23 22:09:41 +08:00 committed by GitHub
parent a491f45cbc
commit 7573b634a7
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 22 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
examples/minicpmv/README.md
View File

@ -1,13 +1,16 @@
## Instructions ## MiniCPM-Llama3-V 2.5
Download model files from huggingface to "MiniCPM-Llama3-V-2_5" folder.
Clone code ### Usage
Download [MiniCPM-Llama3-V-2_5](https://huggingface.co/openbmb/MiniCPM-Llama3-V-2_5) PyTorch model from huggingface to "MiniCPM-Llama3-V-2_5" folder.
Clone llama.cpp and checkout to branch `minicpm-v2.5`:
```bash ```bash
git clone -b minicpm-v2.5 https://github.com/OpenBMB/llama.cpp.git git clone -b minicpm-v2.5 https://github.com/OpenBMB/llama.cpp.git
cd llama.cpp cd llama.cpp
``` ```
Prepare the model Convert PyTorch model to gguf files (You can also download the converted [gguf](https://huggingface.co/openbmb/MiniCPM-Llama3-V-2_5-gguf) by us)
```bash ```bash
python ./examples/minicpmv/minicpmv-surgery.py -m ../MiniCPM-Llama3-V-2_5 python ./examples/minicpmv/minicpmv-surgery.py -m ../MiniCPM-Llama3-V-2_5
@ -17,15 +20,20 @@ python ./convert.py ../MiniCPM-Llama3-V-2_5/model --outtype f16 --vocab-type bp
# quantize int4 version # quantize int4 version
./quantize ../MiniCPM-Llama3-V-2_5/model/model-8B-F16.gguf ../MiniCPM-Llama3-V-2_5/model/ggml-model-Q4_K_M.gguf Q4_K_M ./quantize ../MiniCPM-Llama3-V-2_5/model/model-8B-F16.gguf ../MiniCPM-Llama3-V-2_5/model/ggml-model-Q4_K_M.gguf Q4_K_M
``` ```
Try to inference
Build for Linux or Mac
```bash ```bash
make make
make minicpmv-cli make minicpmv-cli
```
# run quantize f16 version Inference on Linux or Mac
```
# run f16 version
./minicpmv-cli -m ../MiniCPM-Llama3-V-2_5/model/model-8B-F16.gguf --mmproj ../MiniCPM-Llama3-V-2_5/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?" ./minicpmv-cli -m ../MiniCPM-Llama3-V-2_5/model/model-8B-F16.gguf --mmproj ../MiniCPM-Llama3-V-2_5/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?"
# run quantize int4 version # run quantized int4 version
./minicpmv-cli -m ../MiniCPM-Llama3-V-2_5/model/ggml-model-Q4_K_M.gguf --mmproj ../MiniCPM-Llama3-V-2_5/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?" ./minicpmv-cli -m ../MiniCPM-Llama3-V-2_5/model/ggml-model-Q4_K_M.gguf --mmproj ../MiniCPM-Llama3-V-2_5/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?"
# or run in interactive mode # or run in interactive mode
@ -34,8 +42,12 @@ make minicpmv-cli
### Android ### Android
#### Build for Android using Termux #### Build on Android device using Termux
[Termux](https://github.com/termux/termux-app#installation) is a method to execute `llama.cpp` on an Android device (no root required). We found that build on Android device would bring better runtime performance, so we recommend to build on device.
[Termux](https://github.com/termux/termux-app#installation) is a terminal app on Android device (no root required).
Install tools in Termux:
``` ```
apt update && apt upgrade -y apt update && apt upgrade -y
apt install git make cmake apt install git make cmake
@ -82,4 +94,4 @@ Now, you can start chatting:
``` ```
$cd /data/data/com.termux/files/home/bin $cd /data/data/com.termux/files/home/bin
$./minicpmv-cli -m ../model/ggml-model-Q4_K_M.gguf --mmproj ../model/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?" $./minicpmv-cli -m ../model/ggml-model-Q4_K_M.gguf --mmproj ../model/mmproj-model-f16.gguf -c 4096 --temp 0.7 --top-p 0.8 --top-k 100 --repeat-penalty 1.05 --image xx.jpg -p "What is in the image?"
``` ```