Merge branch 'megatron-lm_dtk24.04' into 'main'

Megatron lm dtk24.04 See merge request OpenDAS/megatron-lm!1

Merge branch 'megatron-lm_dtk24.04' into 'main'
Megatron lm dtk24.04 See merge request OpenDAS/megatron-lm!1
3aca1415 · liangjing · 0024a5c6 · 1005e9d3 · 3aca1415 · 3aca1415
Commit 3aca1415 authored Apr 29, 2024 by liangjing
20 changed files
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_sequence_parallel.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_sequence_parallel.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_swiglu.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_swiglu.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_untie_embeddings_and_outputs.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp1_pp4_1nodes_50steps_core_enabled_untie_embeddings_and_outputs.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps_core_enabled.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps_core_enabled.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps_te_enabled.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp2_pp2_1nodes_50steps_te_enabled.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp4_pp1_1nodes_50steps.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp4_pp1_1nodes_50steps.json
--- a/tests/functional_tests/test_results/gpt3/gpt3_tp4_pp1_1nodes_50steps_core_enabled.json
+++ b/tests/functional_tests/test_results/gpt3/gpt3_tp4_pp1_1nodes_50steps_core_enabled.json
--- a/tests/functional_tests/test_scripts/bert/pretrain_bert_distributed_resume_checkpoint_test.sh
+++ b/tests/functional_tests/test_scripts/bert/pretrain_bert_distributed_resume_checkpoint_test.sh
--- a/tests/functional_tests/test_scripts/bert/pretrain_bert_distributed_test.sh
+++ b/tests/functional_tests/test_scripts/bert/pretrain_bert_distributed_test.sh
--- a/tests/functional_tests/test_scripts/bert/sbatch_bert_distributed_resume_checkpoint_test.sh
+++ b/tests/functional_tests/test_scripts/bert/sbatch_bert_distributed_resume_checkpoint_test.sh
--- a/tests/functional_tests/test_scripts/bert/sbatch_bert_distributed_test.sh
+++ b/tests/functional_tests/test_scripts/bert/sbatch_bert_distributed_test.sh
--- a/tests/functional_tests/test_scripts/gpt3/pretrain_gpt3_distributed_resume_checkpoint_test.sh
+++ b/tests/functional_tests/test_scripts/gpt3/pretrain_gpt3_distributed_resume_checkpoint_test.sh
--- a/tests/functional_tests/test_scripts/gpt3/pretrain_gpt3_distributed_test.sh
+++ b/tests/functional_tests/test_scripts/gpt3/pretrain_gpt3_distributed_test.sh
--- a/tests/functional_tests/test_scripts/gpt3/sbatch_gpt3_distributed_resume_checkpoint_test.sh
+++ b/tests/functional_tests/test_scripts/gpt3/sbatch_gpt3_distributed_resume_checkpoint_test.sh
--- a/tests/functional_tests/test_scripts/gpt3/sbatch_gpt3_distributed_test.sh
+++ b/tests/functional_tests/test_scripts/gpt3/sbatch_gpt3_distributed_test.sh
--- a/tests/unit_tests/models/__init__.py
+++ b/tests/unit_tests/models/__init__.py
--- a/tests/unit_tests/models/test_gpt_embedding.py
+++ b/tests/unit_tests/models/test_gpt_embedding.py
--- a/tests/unit_tests/models/test_gpt_model.py
+++ b/tests/unit_tests/models/test_gpt_model.py
--- a/tests/unit_tests/pipeline_parallel/__init__.py
+++ b/tests/unit_tests/pipeline_parallel/__init__.py