issues
search
HazyResearch
/
safari
Convolutions for Sequence Modeling
Apache License 2.0
848
stars
70
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
How to reproduce the Hyena-Imagenet experiment result simillar with the paper?
#46
Diagreen
opened
4 months ago
1
How to visualize Hyena Matrix H(u) from the Paper
#45
ZHymLumine
opened
6 months ago
1
About the squash operator of long convolution
#44
ahmed-fau
closed
6 months ago
2
Hyena
#43
JulesGM
closed
7 months ago
0
Non-causal implementation of language model for synthetic datasets
#42
Karami-m
opened
7 months ago
5
Accuracy on CIFAR is not similar to that in the paper
#41
Karami-m
closed
6 months ago
4
Hyena seems forward leakage?
#40
datatalkv
closed
7 months ago
9
dropout_add_layer_norm is not installed
#39
oferballin
opened
9 months ago
1
learn_ifft in long_conv.py
#38
robflynnyh
closed
9 months ago
1
RWKV
#37
Alchemy5
opened
9 months ago
0
What is the suggested config for running LRA exps with Hyena?
#36
pone7
opened
9 months ago
0
low train accuracy (10% / 40%) on synthetic language modeling tasks for H3
#35
YilunKuang
closed
11 months ago
2
Fix WT103 configs
#34
DanFu09
closed
11 months ago
0
Bidirectional Hyena
#33
starmpcc
closed
11 months ago
1
Does Hyena support BERT style LLM?
#32
sherdencooper
opened
11 months ago
3
Encoder decoder
#31
him4318
opened
11 months ago
0
Inconsistent between implementation and paper descriptions
#30
hoangmit
opened
1 year ago
1
Downstream evaluation on SuperGLUE
#29
moranant
opened
1 year ago
2
configs for Hyena Wikitext103 experiments
#28
xiaobo-guo
opened
1 year ago
10
fftconv compiling issue
#27
leeyeehoo
closed
1 year ago
3
RuntimeError: u must have shape (batch_size, H, L)
#26
kaansancak
opened
1 year ago
3
Relative positional encodings with Hyena
#25
fransilvionGenomica
opened
1 year ago
2
What is the suggested way to implement an autograd version of nn.functional.conv1d?
#24
penguinshin
opened
1 year ago
0
Add mentions to external resources
#23
Zymrael
closed
11 months ago
0
Fix Hyena YAML syntax
#22
janEbert
closed
1 year ago
1
Fix Hyena commands
#21
janEbert
closed
1 year ago
0
Hyena YAML parsing fails
#20
janEbert
closed
1 year ago
1
Make repo a python package
#19
tbenst
opened
1 year ago
1
Bug in dataset caching
#18
PgLoLo
closed
1 year ago
1
Potential reasons for stuck building wheel for droput-layer-norm
#17
radarFudan
closed
1 year ago
1
How to set config properly to launch an experiment on wk103 (of any model)?
#16
Chord-Chen-30
opened
1 year ago
2
GPU memory requirements of Hyena
#15
mgaido91
closed
11 months ago
6
Fix paths to completely install FlashAttention
#14
YuigaWada
closed
1 year ago
2
Pickling error with standalone_cifar
#13
carbonox-infernox
opened
1 year ago
1
Question about the model size
#12
fransilvionGenomica
closed
1 year ago
2
Question about synthetic dataset.
#11
SSamDav
opened
1 year ago
0
Question about long sequence lengths with Hyena
#10
VivekPanyam
opened
1 year ago
9
Order of hyena operation
#9
sylee0124
closed
1 year ago
6
Long latency when loading fft_conv kernel for the first time
#8
sylee0124
opened
1 year ago
1
Questions about bidirectional-version of H3
#7
alstonlo
closed
1 year ago
4
RuntimeError: Expected fft_size >= 16 && fft_size <= 16384
#6
JCBrouwer
opened
1 year ago
4
Question about discrepancy between implementations available in the repo and related papers
#5
sylee0124
opened
1 year ago
3
Question concerning FFT operation.
#4
veritas9872
closed
1 year ago
9
Arguments for simple H3 model than could learn Causal or Masked LM modeling?
#3
kingb12
closed
1 year ago
5
H3 / LongConvKernel - l_max=None isn't working
#2
grazder
closed
9 months ago
3
Update standalone_hyena.py
#1
andrewnc
closed
1 year ago
1