Skip to content
Closed
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 @@
17a67b9a97fb11a75933fd7f76ad76e1ac98a53d
10e7b74597b47a57449599bf2780bd9e82ef79ff
2 changes: 1 addition & 1 deletion 3rdparty/Megatron-LM
Submodule Megatron-LM updated 53 files
+400 −0 .claude/skills/build-and-test/SKILL.md
+1 −1 .github/copy-pr-bot.yaml
+19 −7 .github/oncall_schedule.json
+8 −0 .github/workflows/claude_review.yml
+2 −15 examples/academic_paper_scripts/sc21/run_figure_18.sh
+0 −1 examples/gpt3/gpt_config.yaml
+8 −1 megatron/core/dist_checkpointing/exchange_utils.py
+0 −6 megatron/core/dist_checkpointing/serialization.py
+39 −47 megatron/core/dist_checkpointing/strategies/base.py
+17 −35 megatron/core/dist_checkpointing/strategies/fully_parallel.py
+15 −32 megatron/core/dist_checkpointing/strategies/torch.py
+2 −0 megatron/core/full_cuda_graph.py
+7 −2 megatron/core/inference/config.py
+91 −6 megatron/core/inference/contexts/dynamic_context.py
+12 −0 megatron/core/inference/engines/dynamic_engine.py
+185 −13 megatron/core/optimizer/__init__.py
+60 −34 megatron/core/optimizer/clip_grads.py
+54 −41 megatron/core/optimizer/distrib_optimizer.py
+450 −0 megatron/core/optimizer/emerging_optimizers.py
+29 −6 megatron/core/optimizer/grad_scaler.py
+6 −8 megatron/core/optimizer/layer_wise_optimizer.py
+8 −386 megatron/core/optimizer/muon.py
+15 −10 megatron/core/optimizer/optimizer.py
+37 −37 megatron/core/optimizer/optimizer_config.py
+61 −0 megatron/core/optimizer/optimizer_cuda_graph.py
+4 −4 megatron/core/optimizer/qk_clip.py
+9 −3 megatron/core/optimizer_param_scheduler.py
+24 −0 megatron/core/tensor_parallel/layers.py
+231 −4 megatron/core/transformer/fsdp_dtensor_checkpoint.py
+1 −0 megatron/inference/utils.py
+19 −12 megatron/training/arguments.py
+15 −0 megatron/training/async_utils.py
+32 −10 megatron/training/checkpointing.py
+8 −0 megatron/training/config/training_config.py
+24 −38 megatron/training/training.py
+9 −0 megatron/training/utils.py
+0 −2 tests/functional_tests/test_cases/gpt/gpt3_mcore_te_tp2_pp1_gdn/model_config.yaml
+3 −1 tests/functional_tests/test_cases/gpt/gpt3_mcore_te_tp2_zp_z3_resume_fsdp_dtensor/model_config.yaml
+0 −2 tests/functional_tests/test_cases/hybrid/hybrid_mr_mcore_te_tp2_pp1_cp1_dgx_a100_1N8G/model_config.yaml
+0 −2 ...gpt3_mcore_te_tp1_pp2_resume_torch_dist_reshard_2x1x4_te_8experts2parallel_dist_optimizer/model_config.yaml
+166 −0 tests/functional_tests/test_cases/nemotron/nemotron3_super_release_g200/model_config.yaml
+6 −15 tests/unit_tests/dist_checkpointing/conftest.py
+0 −3 tests/unit_tests/dist_checkpointing/test_serialization.py
+39 −37 tests/unit_tests/dist_checkpointing/utils.py
+94 −0 tests/unit_tests/inference/contexts/test_dynamic_context.py
+27 −0 tests/unit_tests/inference/engines/test_dynamic_engine.py
+89 −0 tests/unit_tests/tensor_parallel/test_tp_attrs_without_init.py
+1,669 −0 tests/unit_tests/test_emerging_optimizers.py
+38 −23 tests/unit_tests/test_layer_wise_optimizer.py
+0 −787 tests/unit_tests/test_muon_optimizer.py
+3 −3 tests/unit_tests/test_optimizer.py
+644 −0 tests/unit_tests/transformer/test_fsdp_dtensor_checkpoint.py
+206 −20 tools/checkpoint/checkpoint_inspector.py
Loading
Loading