Comments (11)
can you try the alpaca-native-enhanced model
and can you tell me how many cores are there in your system
from alpaca-turbo.
I've had the same problem with the "ggml-model-q4_0.bin" but had no issue with "ggml-model-q4_0_unfiltered.bin".
So it's probably about the model.
from alpaca-turbo.
I'm having a similar problem. Win10, ggml Alpaca 7B downloaded from huggingface:
https://huggingface.co/Pi3141
from alpaca-turbo.
Same issue here with: gpt4all-lora-quantized.bin and ggml-alpaca-7b-q4.bin
from alpaca-turbo.
Using 2 files from Pi3141, gpt4-x-alpaca-native-13B-ggml worked, alpaca-native-7B-ggml didn't load
gpt4-x-alpaca-native-13B-ggml:
'models\ggml-model-q4_1.bin',
Working... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00
alpaca-native-7B-ggml:
'models\ggml-model-q4_0.bin',
ERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoR
from alpaca-turbo.
Similar, won't load any of my 3 quantized 7B and 13B alpaca variants that worked in dalai or alpaca.cpp.
...
'-m',
'models\ggml-model-q4_0.bin',
'--interactive-first'
]
ERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoRERRoR
from alpaca-turbo.
Same here ggml-alpaca-7b-q4.bin not loading
from alpaca-turbo.
Same here, llama7B, llama13B, alpaca,... - all working locally with llama.cpp on the commandline. All hanging on load.
Parameters for invoking llama.cpp commandline seem right and commandline status shows apparent completion:
Working... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00
But web-ui dialogue hangs with loading.
EDIT: running on macOS/Apple Silicon via current git-clone + copying templates folder from ZIP.
from alpaca-turbo.
Same here, endless -loading model "ggml-model-q4_0.bin"
from alpaca-turbo.
Updated / totally edited for better clarification.
- I'm on macOS/Apple silicon. Running the current/latest llama.cpp with several models from terminal. It all works fine in terminal, even when testing in alpaca-turbo's environment with its parameters from the terminal
- Because I want the latest llama.cpp+models, I can't just run the docker or other images. Needed to git-clone (+ copy templates folder from ZIP). Linked my working llama.cpp "main" to ./bin/mac, and its models' *.bin into ./models with nice names for the UI.
Cause of model-hang for me:
Alpaca-Turbo expects a certain string "RSTsr" to be displayed by it's invocation of llama.cpp once its finished loading (set in alpaca_turbo.py line 46). The current llama.cpp + llama 7B e.g. does not show this string in its invocation. So the load hangs. For me it seems, it's not how the model is, it's more how Alpaca-Turbo was constructed. The current llama.cpp+models produces very different responses - I'm trying to understand alpaca_turbo.py, so that maybe I can fix it (at least for me).
Hope this clarifies/helps.
from alpaca-turbo.
Update - I got it to work (most of the time) on my Mac by changing alpaca_turbo.py quite a bit. But I don't think it is mergeable into a pull-request, because my solution seems to be one-off just for my situation.
from alpaca-turbo.
Related Issues (20)
- Clarify Open Source License HOT 1
- Support reverse prompting (-r flag in llama.cpp) HOT 1
- Your instructions are completely wrong HOT 2
- How to load leaked LLaMA weights? HOT 1
- Can you add some more how to setup with like Portainer or in proxmox HOT 1
- alpaca_turbo.py typo on line 224 directoryh HOT 2
- Alpaca Install EXE detected as a trojan HOT 3
- Wrong port in docker-compose.yml file HOT 1
- alpaca_turbo.exe got virus when download and scan by antivirus HOT 1
- TemplateNotFound jinja2.exceptions.TemplateNotFound: index.html HOT 1
- Error Loading model HOT 3
- Copied the release Alpaca-Turbo-beta_v0.6 HOT 6
- Error using docker-compose HOT 4
- UI is weird in dark mode HOT 1
- Can't run in Hyper-V VM
- jinja2.exceptions.TemplateNotFound: index.html
- [Errno 13] Permission denied: 'bin/mac' HOT 2
- Loading models issue?
- Is this an abandoned project ?
- “mac” can’t be opened because Apple cannot check it for malicious software.
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 alpaca-turbo.