Comments (4)
Hi, we only care about the relative values of scaling across channels. Multiplying the entire scaling factor vector by a fixed number does not affect the accuracy. However, it helps with numerical stability (a more proper range). Let me know if you have more questions.
from llm-awq.
Hi @MarsJacobs , the most important factor being the accuracy is higher. Our intuition is different compared to SmoothQuant: SmoothQuant wants to preserve the activation outliers for W8A8 quantization; we only want to introduce activation-awareness to the weights. Therefore, we want to use the average to reflect the overall effect from different tokens.
from llm-awq.
Thank you for the further clarification! I have one question. In the AWQ implementation, is there a particular reason for using the abs().mean() of weight and activation to explore the scale value? (For comparison, I understand that SmoothQuant utilizes the max value of weight and activation.)
Additional explanations would be greatly helpful in gaining a deeper understanding of AWQ. Thanks in advance!
from llm-awq.
Thank you for your response. Considering the motivational difference between SmoothQuant's shifting activation outliers and AWQ's activation-aware weight scaling, the reasoning behind the abs().mean() implementation becomes much clearer. Thank you for sharing your great work and for providing further answers!
from llm-awq.
Related Issues (20)
- reproduce Llama2 7b failure : RuntimeError: The expanded size of the tensor (4608) must match the existing size (4096) at non-singleton dimension 3. Target sizes: [65, 32, 512, 4608]. Tensor sizes: [65, 1, 512, 4096] HOT 3
- RuntimeError: Unknown Layout in CUDA Kernel Execution
- Use awq to quantize Deepseek-coder-33B-instruct model
- run_awq.<locals>.Catcher.forward() error
- KeyError: 'llava_llama' HOT 1
- Error while generating real quantized weights for VILA
- Weight int4 quantization, but actually it is int16 HOT 4
- Possible Bug in "_search_module_scale" Function
- AWQ for non-transformer layers?
- Out of memory in Jetson Orin NX 8GB
- Inquiry about Minimum GPU Requirements HOT 1
- when q-group-size = -1,the code will not run
- Weight Packing Format
- illegal memory access when input tokens < 8
- Grok-1 AWQ
- can awq support 3-bit,2-bit, 8-bit quantization? HOT 1
- awq_inference_engine is missing from source, so quantizing custom models fails HOT 2
- Support for Qwen models HOT 2
- AWQ for non-Transfomer Implementation HOT 3
- Error while Quantizing OWLv2 model
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 llm-awq.