Comments (2)
Ahh, I wish there was a option to do it entirely in memory, I also run this on an nvme, so it should be rather fast
from mergekit.
Yes, mergekit is currently single-threaded. PyTorch can do some internal threading on CPU - you may want to play with the OMP_NUM_THREADS environment variable.
The pauses you are seeing are probably shards of the output model being written to disk. I/O is a lot slower than the actual merge process.
from mergekit.
Related Issues (20)
- how to merge for different rope scaling? HOT 3
- Require later version of transformers HOT 1
- How to merge only at q_proj layers with SLERP? HOT 2
- `extract_lora.py` can't handle mismatched `lm_head` tensor due to added tokens HOT 10
- Questions about density gradient and weight gradient in Ties example
- Support for microsoft/Phi-3-vision-128k-instruct
- parameters: int8_mask: true ?? HOT 1
- Use logscale for operations dealing with norm layers?
- Evolutionary Merging out of memory HOT 3
- Need some help in merging same architectures, but with different tokens in their tokenizers HOT 4
- ValueError: operands could not be broadcast together with shapes (8192,28672) (8192,24576) HOT 1
- How to merge specific parameters
- New merge method / algorithm proposal: Geometric Median and TGMD merge
- "Missing required parameter weight" for LoRA merge. HOT 2
- pass eval arguments directly in mergekit-evo
- qwen2-0.5B cannot be merged into MoE HOT 1
- New method: MAP:Low-compute Model Merging with Amortized Pareto Fronts via Quadratic Approximation
- MOE models that based on Gemma cannot work HOT 3
- Sincerely.. I have no words...without insulting someone. HOT 2
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 mergekit.