Giters
ridgerchu
/
matmulfreellm
Implementation for MatMul-free LM.
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
2942
Watchers:
44
Issues:
35
Forks:
187
ridgerchu/matmulfreellm Issues
How to use FusedBitLinear layers for my own models for fast inference?
Closed
2 months ago
Comments count
3
Indecies in the arxiv paper reversed?
Closed
2 months ago
Comments count
1
No reduction in VRAM usage
Updated
2 months ago
Comments count
7
How to use FusedBitLinear layers for my own models for fast inference?
Closed
2 months ago
How can we say matmul-free enables parallel computation?
Updated
3 months ago
Comments count
1
AttributeError: module 'triton.language.math' has no attribute 'round'
Updated
3 months ago
Comments count
2
How do I train my models?
Updated
4 months ago
Comments count
8
Does matmulfreellm support Windows 10?
Updated
4 months ago
Comments count
6
Larger models available or planned?
Updated
4 months ago
Comments count
5
Question about ”fused_recurrent_hgrn“ function
Closed
4 months ago
Comments count
2
About the FPGA Performance Metrics
Updated
4 months ago
Comments count
2
training model with huggingface
Updated
5 months ago
Comments count
2
triton error using generate.py
Closed
5 months ago
Comments count
2
Discrepency in code and paper related to HGRNBitAttention
Updated
5 months ago
Comments count
1
Can Meta's llama model be quantized to matmulfree?
Closed
5 months ago
Comments count
1
LLVM ERROR: Cannot select: intrinsic %llvm.nvvm.shfl.sync.bfly.i32
Updated
5 months ago
Comments count
5
FPGA implementation
Updated
6 months ago
Comments count
4
Any pretrained model bigger than 2.7B?
Closed
6 months ago
Comments count
3
Reproduce the results of the code in the paper
Updated
6 months ago
Comments count
4
Convert existing model to mmfree
Updated
6 months ago
Comments count
1
how to train or ft?
Updated
6 months ago
Comments count
4
Failure to train.
Closed
7 months ago
Comments count
4
Error on pip install: ModuleNotFoundError: No module named 'torch'
Closed
7 months ago
Comments count
2
scaling_law.png
Updated
6 months ago
Comments count
1
MLGRU
Updated
6 months ago
Comments count
1
tried to train
Closed
6 months ago
Comments count
9
Why Transformer++
Updated
6 months ago
Comments count
1
why FusedBitLinear.forward() use F.linear() with float16 inputs?
Updated
6 months ago
Comments count
3
Baseline weights?
Closed
6 months ago
Comments count
2
Ternary weight values
Closed
7 months ago
Comments count
5
Rocm(7900xtx) GPU fail
Updated
7 months ago
Comments count
12
Devcloud notebook
Updated
7 months ago
Comments count
1
Add topic tags
Closed
7 months ago
Comments count
1
Please add a LICENSE file.
Closed
7 months ago
Comments count
1