issues
search
IBM
/
dolomite-engine
Dolomite Engine is a library for pretraining/finetuning LLMs
Apache License 2.0
23
stars
7
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
fix contiguous_count
#72
mayank31398
opened
17 hours ago
0
cute-kernels
#71
mayank31398
closed
1 day ago
0
efficient kernels
#70
mayank31398
closed
4 days ago
0
dim + 1 -> dim
#69
mayank31398
closed
6 days ago
0
efficient init for padding free transformer
#68
mayank31398
closed
6 days ago
0
efficient initialization for large models
#67
mayank31398
closed
6 days ago
0
efficient initialization for large models
#66
mayank31398
closed
6 days ago
0
drop torch version check
#65
mayank31398
closed
6 days ago
0
fix file lock hang
#64
mayank31398
closed
6 days ago
0
PP loss fix.
#63
shawntan
closed
1 week ago
0
fix broken backprop
#62
mayank31398
closed
1 week ago
0
Untranspose for FSDP-2
#61
mayank31398
closed
1 week ago
0
drop prompt tuning
#60
mayank31398
closed
1 week ago
0
step tracker
#59
mayank31398
opened
1 week ago
0
Issue with Fine-tuning Llama 3.1 8B model
#58
murthyrudra
opened
2 weeks ago
9
Begin MoE pipeline.
#57
shawntan
opened
3 weeks ago
0
Average gradients across gradient accumulation steps
#56
mayank31398
closed
1 week ago
0
frequency averaging
#55
mayank31398
closed
4 weeks ago
0
refactor dtensors
#54
mayank31398
closed
4 weeks ago
0
ValueError: offset must be non-negative and no greater than buffer length
#53
murthyrudra
closed
4 weeks ago
5
helper function for TP
#52
mayank31398
closed
1 month ago
0
drop fsdp-1 saving method
#51
mayank31398
closed
1 month ago
0
async TP
#50
mayank31398
closed
1 month ago
0
Stick-breaking attention model.
#49
shawntan
opened
1 month ago
0
fix slow start
#48
mayank31398
closed
1 month ago
0
fix flops with TP
#47
mayank31398
closed
1 month ago
0
C++ formatting
#46
mayank31398
closed
1 month ago
0
stable distillation
#45
mayank31398
closed
1 month ago
0
Refactored Shared Expert.
#44
shawntan
opened
1 month ago
0
Drop deepspeed
#43
mayank31398
closed
1 month ago
0
TP test
#42
shawntan
closed
1 month ago
0
MoE_TP training
#41
mayank31398
closed
1 month ago
0
Pipeline parallel
#40
mayank31398
closed
4 weeks ago
1
Expert parallel
#39
mayank31398
opened
1 month ago
0
Gate transpose without TP
#38
mayank31398
closed
1 month ago
0
Change function names
#37
mayank31398
closed
1 month ago
0
Gate transpose
#36
mayank31398
closed
1 month ago
0
Moe deltanet
#35
mayank31398
opened
1 month ago
0
transpose the gate
#34
mayank31398
closed
1 month ago
0
Dtensor cleanup
#33
mayank31398
closed
1 month ago
0
Scattermoe sp
#32
shawntan
closed
1 month ago
0
Enable ScatterMoE SP + some cleanup
#31
mayank31398
closed
1 month ago
0
Dimension swap.
#30
shawntan
closed
1 month ago
0
Scattermoe TP and SP.
#29
shawntan
closed
1 month ago
0
add power scheduler paper
#28
mayank31398
closed
2 months ago
0
fix tracking variance error
#27
mayank31398
closed
2 months ago
0
Deltanet test
#26
mayank31398
closed
2 months ago
0
merge deltanet changes
#25
mayank31398
closed
2 months ago
0
merge deltanet changes
#24
mayank31398
closed
2 months ago
0
use DTensor API for embedding matrix
#23
mayank31398
closed
2 months ago
0
Next