Comments (4)
#7359 broke models with more than 256 layers.
from llama.cpp.
Ooo I see... On purpose or as a consequence of supporting that model? Could it be patched or is it a hard limit?
from llama.cpp.
On purpose or as a consequence of supporting that model? Could it be patched or is it a hard limit?
It's a consequence of keeping llama_hparams
trivially copyable with a compile-time known size, while having layer-wise hyper-parameters. Increasing the limit to 512 would make llama_hparams
take 6.16 KiB
instead of 3.16 KiB
, but that's pretty much the only thing it changes. The size.
Making the layer-wise hparams take less space when not needed is something which I'll likely fix eventually, so that the limit only applies to models which need layer-wise hparams.
from llama.cpp.
Ooo I see... On purpose or as a consequence of supporting that model? Could it be patched or is it a hard limit?
It appears to be an arbitrary limit even though an int64 can handle an absurd 9x10^18 before overflowing. I don't know why but this seems to be something fairly unique to the machine learning space even though it makes the code needlessly brittle and user hostile.
from llama.cpp.
Related Issues (20)
- Feature Request: Pull from Ollama repo HOT 1
- Bug: Vulkan build no longer working with MSVC cmake on windows HOT 4
- Feature Request: Add support for new model conversion HOT 1
- Bug: failing ggml.c:12621: ne2 == ne02 during finetuning HOT 1
- llama : reimplement logging HOT 1
- Error converting SmolLM-1.7B-Instruct HOT 1
- Bug: [SYCL] Backend seems not stable HOT 5
- Bug: "main : failed to eval" with Self-extend and small context HOT 7
- Feature Request: Support llava with different vision/LM backbones
- Support Mistral-Nemo-Instruct-2407 128K HOT 29
- Bug: Multi-NPU execution error HOT 2
- Bug: export-lora does not accept GGUF files HOT 4
- Bug: null pointer defer in gguf_init_from_file HOT 5
- Bug: Tokenization adds a space to the first non-special token HOT 8
- Bug: The output content is different HOT 6
- Random seed possible problems. HOT 17
- The server was not compiled for multimodal or the model projector can't be loaded HOT 4
- [question] how do I save a loaded model? HOT 4
- Feature Request: Support and conversion of croos encoders HOT 2
- failed to quantize: requantizing from type q6_K is disabled
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 llama.cpp.