.. |
activation_checkpointing
|
42c1e916f6
feat(activation_checkpointing): add `non_reentrant_checkpoint` to support inputs require no grad (#4118)
|
1 year ago |
checkpoint_engine
|
c5edc91ecb
change partititon_name to partition_name (#3700)
|
1 year ago |
comm
|
f0463b4d1f
Pass correct node size for ZeRO++ (#4085)
|
1 year ago |
compression
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
data_pipeline
|
736bf1853b
bug fix (#3609)
|
1 year ago |
fp16
|
b354c28b76
polishing timers and log_dist (#3996)
|
1 year ago |
pipe
|
e20e4a9d02
clear redundant timers (#4308)
|
1 year ago |
swap_tensor
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
zero
|
60bf78454c
Fix incorrect assignment of self.quantized_nontrainable_weights (#4399)
|
1 year ago |
__init__.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
bf16_optimizer.py
|
5a5340d03b
remove UtilsBuilder load, use torch (un)flatten ops (#3728)
|
1 year ago |
config.py
|
aa4a7401f8
ZeRO-Inference refresh (#4197)
|
1 year ago |
config_utils.py
|
4d27225f3e
zero.Init() should pin params in GPU memory as requested (#2953)
|
1 year ago |
constants.py
|
0411a9f871
Expose Consecutive Hysteresis to Users (#3553)
|
1 year ago |
dataloader.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
eigenvalue.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
engine.py
|
9adc73ff65
Handle empty parameter groups (#4277)
|
1 year ago |
hybrid_engine.py
|
43188ff077
Pass missing positional arguments in `DeepSpeedHybridEngine.generate()` (#4026)
|
1 year ago |
lr_schedules.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
progressive_layer_drop.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
quantize.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
sparse_tensor.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
state_dict_factory.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |
utils.py
|
7f3e82fe09
do allgather only in shared optimizer states groups (#4167)
|
1 year ago |
weight_quantizer.py
|
b361c72761
Update DeepSpeed copyright license to Apache 2.0 (#3111)
|
1 year ago |