.. |
comm
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
inference
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
monitor
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
ops
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
profiling
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
runtime
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
__init__.py
|
4912e0ad7e
Various ZeRO Stage3 Optimizations + Improvements (including bfloat16 support) (#1453)
|
2 年之前 |
common.py
|
1a71e77dc2
Fix for distributed tests on pytorch>=1.12 (#2141)
|
2 年之前 |
ds_batch_config.json
|
a10e4811fe
force set lf instead of crlf (https://github.com/pre-commit/pre-commit-hooks#mixed-line-ending) (#1598)
|
2 年之前 |
gpt2-merges.txt
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
gpt2-vocab.json
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
megatron_model.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
modeling.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
modelingpreln.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
multi_output_model.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
simple_model.py
|
36ad3119d5
DeepSpeed comm backend v1 (#1985)
|
2 年之前 |
test_activation_checkpointing.py
|
36ad3119d5
DeepSpeed comm backend v1 (#1985)
|
2 年之前 |
test_aio.py
|
36ad3119d5
DeepSpeed comm backend v1 (#1985)
|
2 年之前 |
test_autocast.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_autotuning.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_bf16.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_checkpointing.py
|
e669aaf55b
Trajepl/nebula ckpt engine (#2085)
|
2 年之前 |
test_coalesced_collectives.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_compression.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_config.py
|
5997589683
Refactor ZeRO configs to use Pydantic (#2004)
|
2 年之前 |
test_configurable_parallel.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_csr.py
|
fcb3ca5e66
Proposal of how we might use sparse tensors for gradients (#1484)
|
3 年之前 |
test_cuda_backward.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_cuda_forward.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_curriculum_learning.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_data.py
|
4912e0ad7e
Various ZeRO Stage3 Optimizations + Improvements (including bfloat16 support) (#1453)
|
2 年之前 |
test_ds_arguments.py
|
8326aff279
Improve doc string for add_XXX_arguments (#32)
|
4 年之前 |
test_ds_config.py
|
5997589683
Refactor ZeRO configs to use Pydantic (#2004)
|
2 年之前 |
test_ds_initialize.py
|
4912e0ad7e
Various ZeRO Stage3 Optimizations + Improvements (including bfloat16 support) (#1453)
|
2 年之前 |
test_dynamic_loss_scale.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_elastic.py
|
1ed5aa96a8
Elastic Training support in DeepSpeed (#2153) (#2156)
|
2 年之前 |
test_fp16.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_get_optim_files.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_groups.py
|
1e61c7a860
fix: Fix undefined variable in _create_expert_data_and_model_parallel and make it easier to understand (#1826)
|
2 年之前 |
test_ignore_unused_parameters.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_init_on_device.py
|
aa88137b8d
Add Inference support for running the BigScience-BLOOM Architecture (#2083)
|
2 年之前 |
test_lr_schedulers.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_moe.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_moe_tp.py
|
5fe9d61065
Tensor parallelism for Mixture of Experts (#2074)
|
2 年之前 |
test_multi_output_model.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_onebit.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_partition.py
|
36ad3119d5
DeepSpeed comm backend v1 (#1985)
|
2 年之前 |
test_pipe.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_pipe_module.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_pipe_schedule.py
|
65c2f974d8
Pipeline parallel training engine. (#392)
|
4 年之前 |
test_pld.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_reshape_checkpoint.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_run.py
|
4cf970e6bb
Add codespell to pre-commit checks (#1717)
|
2 年之前 |
test_runtime_utils.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_sparse_attention.py
|
b442264dc9
formatting fix for #1962
|
2 年之前 |
test_sparse_grads.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_zero.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
test_zero_config.py
|
5997589683
Refactor ZeRO configs to use Pydantic (#2004)
|
2 年之前 |
test_zero_context.py
|
46401b3884
[zero-3] shutdown zero.Init from within ds.init (#2150)
|
2 年之前 |
test_zero_tiled.py
|
316c4a43e0
Add flake8 to pre-commit checks (#2051)
|
2 年之前 |
util.py
|
80d0a32f0b
Checkpoint reshaping (#1953)
|
2 年之前 |