Comments (6)
python 3.8/3.9/3.10
pip install numpy tokenizers prompt_toolkit ninja
pip install torch --extra-index-url https://download.pytorch.org/whl/cu117 --upgrade (use 1.13.1)
pip install rwkv --upgrade
:)
from chatrwkv.
@bello7777 You probably need to adjust the strategy. If you're using the pull request:
model = RWKV(model=model_path, strategy='cuda fp16i8 *20 -> cuda fp16')
That's around line 28 in webui.py
from that pull. You didn't say what you're actually doing, so there's no way to know if you're saying it failed when trying to load the model for inference, when converting, whatever.
But the most likely solution is to find whatever is running and how it's setting the strategy and reduce the number of layers it will send to the GPU. For example, in the line above you could try using cuda fp16i8 *10 -> cuda fp16
instead which should roughly half the required GPU memory.
After you get it going, you can use other tools to see how much GPU memory you have available and adjust the setting according.
from chatrwkv.
There is no need to toss the environment, just use the container @bello7777
from chatrwkv.
Thanks mate, I will do it .
for the moment I just launch it on AWS ubuntu 18.04 ,
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 200.00 MiB (GPU 0; 14.62 GiB total capacity; 13.77 GiB already allocated; 163.94 MiB free; 13.97 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
can I try to reduce the batch sizes to smaller values if yes where are they?
from chatrwkv.
@soulteary
i tried to access your blog and guidelines but i could not,
could you give me the steps and version of Docker container so i can deploy it on ec2 server as a still have a problem with memory
from chatrwkv.
thanks solved and working
now moving to train the model
from chatrwkv.
Related Issues (20)
- IndexError: list index out of range on ubuntu 20.04 when run python v2/chat.py HOT 2
- "LayerNormKernelImpl" not implemented for 'Half'
- cpu运算乱码问题 HOT 1
- Question about the accuracy of smaller checkpoints HOT 2
- 很好的模型,但是 HOT 4
- 求助time-mixing部分,处理数值溢出代码的解释
- Feature Request: an option to use Positional Interpolation to extent CTX length at inference time HOT 2
- Add a support to "stop_words" in PIPELINE
- 开源中文NSFW微调模型
- demo ? HOT 2
- demo true error ? HOT 1
- 'No CUDA GPUs are available' in google colab with V100 GPU and high RAM HOT 2
- huggingface无法访问,模型无法下载 HOT 4
- Prompt for RAG with RWKV-4-World-7B-v1-20230626-ctx4096 HOT 1
- [Feature Request] text2music HOT 2
- RuntimeError: Error building extension 'wkv_cuda_v1' HOT 2
- How to write the RWKV in autogressive style like RNN HOT 2
- NameError: name 'PIPELINE' is not defined HOT 1
- 大哥,乱码了 HOT 1
- 回复总是截断了,如何让回复自然的结束 HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from chatrwkv.