Skip to content

Issues: databricks/megablocks

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

27 Open 21 Closed
27 Open 21 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Cloning input x in megablocks.layers.glu.SparseGLU leads to different SDD outputs
#115 opened May 28, 2024 by cmsflash
1
Can we change self.blocking in dmoe.py from 128 to 64?
#114 opened May 24, 2024 by seanM29
2
_LOAD_BALANCING_LOSS returns empty list sometimes
#113 opened May 22, 2024 by exnx
1
Bad throughput with GLU
#110 opened May 17, 2024 by Muennighoff
1
1-expert worse than dense model
#107 opened May 8, 2024 by Muennighoff
OSError: Stale file handle with dMoE
#106 opened Apr 24, 2024 by Muennighoff
3
Add a fine-tune script for JetMoE
#105 opened Apr 17, 2024 by shamanez
2
ScatterMoE feature
#104 opened Apr 5, 2024 by ehartford
5
Implement Mixture of Depth and Experts (MoDE)
#103 opened Apr 5, 2024 by casper-hansen
2
Import dmoe model into other training script?
#101 opened Apr 2, 2024 by andrewnc
3
SFT Script and Hyperparameters used for DBRX-Instruct
#99 opened Mar 28, 2024 by alpayariyak
5
support amd/rocm enhancement New feature or request help wanted Extra attention is needed
#97 opened Mar 21, 2024 by ehartford
3
AMP + BF16 failing
#95 opened Jan 28, 2024 by jramapuram
2
Unsharding scripts for megablocks models
#94 opened Jan 21, 2024 by mayank31398
the wrong loss func was chosen at evaluation
#93 opened Jan 18, 2024 by peterjc123
1
2
Seeking a good multi-node training config
#92 opened Jan 15, 2024 by rpand002
3
selective router precision question Further information is requested
#91 opened Jan 14, 2024 by 152334H
1
Does this framework support SFT? question Further information is requested
#90 opened Jan 12, 2024 by banksy23
2
RuntimeError: Triton Error [CUDA]: invalid argument question Further information is requested
#88 opened Jan 10, 2024 by noob-ctrl
15
different load_balancing_loss with different pipeline_parallel_size question Further information is requested
#85 opened Jan 5, 2024 by bozheng-hit
8
How to integrate to transformers-based mixtral question Further information is requested
#84 opened Jan 3, 2024 by nxphi47
1
ParallelDroplessMLP initialises self.mlp twice enhancement New feature or request help wanted Extra attention is needed
#83 opened Jan 1, 2024 by 152334H
6
Why the second matrix of the mlp layer has the same shape of the first one? question Further information is requested
#81 opened Dec 29, 2023 by gouchangjiang
1
[BUG] Optimizer Weights Not Reloaded When Training with bf16 Pretrained Weights bug Something isn't working
#80 opened Dec 26, 2023 by RookieHong
1
Script for Full Fine-Tuning of Mixtral question Further information is requested
#68 opened Dec 20, 2023 by alpayariyak
1
ProTip! Adding no:label will show everything without a label.