Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .main.commit
Original file line number Diff line number Diff line change
@@ -1 +1 @@
9ed8b0c4ac8e65d80fb35eecb3656c96b30df943
f456199700bc4ad56881cb76b9d7140b82841da4
2 changes: 1 addition & 1 deletion 3rdparty/Megatron-LM
Submodule Megatron-LM updated 28 files
+23 −6 .github/workflows/cicd-approve-test-queue.yml
+71 −18 .github/workflows/cicd-main.yml
+1 −5 docker/common/install_source_wheels.sh
+29 −0 megatron/core/distributed/distributed_data_parallel_config.py
+11 −6 megatron/core/distributed/fsdp/mcore_fsdp_adapter.py
+29 −0 megatron/core/distributed/fsdp/src/megatron_fsdp/distributed_data_parallel_config.py
+22 −0 megatron/core/extensions/transformer_engine.py
+0 −27 megatron/core/inference/engines/dynamic_engine.py
+33 −35 megatron/core/inference/text_generation_server/dynamic_text_gen_server/endpoints/chat_completions.py
+11 −23 megatron/core/inference/text_generation_server/dynamic_text_gen_server/endpoints/completions.py
+1 −1 megatron/core/inference/text_generation_server/dynamic_text_gen_server/text_generation_server.py
+38 −20 megatron/core/inference/utils.py
+4 −1 megatron/core/models/gpt/gpt_model.py
+130 −0 megatron/core/resharding/planner.py
+44 −10 megatron/core/resharding/utils.py
+24 −3 megatron/core/ssm/mamba_mixer.py
+5 −3 megatron/core/transformer/attention.py
+100 −100 megatron/core/transformer/cuda_graphs.py
+8 −7 megatron/core/transformer/moe/moe_layer.py
+6 −3 megatron/core/transformer/multi_token_prediction.py
+3 −0 megatron/core/transformer/transformer_config.py
+3 −12 megatron/training/training.py
+0 −3 pyproject.toml
+3 −3 tests/unit_tests/distributed/megatron_fsdp/test_mcore_fully_sharded_data_parallel.py
+2 −0 tests/unit_tests/pipeline_parallel/test_fine_grained_activation_offloading.py
+338 −55 tests/unit_tests/resharding/test_model_swap.py
+7 −2 tests/unit_tests/transformer/test_vision_cuda_graphs.py
+0 −16 uv.lock
Loading
Loading