Index of /FlagScale_to_hygon/third_party/Megatron-LM/tests/functional_tests/test_cases/gpt/
../
gpt3_15b_8t_release/ 10-Sep-2025 08:16 -
gpt3_15b_8t_release_sm/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp1_d..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp1_f..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp1_r..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp2/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp2_fp16 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp2_r..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp4/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp1_pp4_r..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp4_pp1/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp4_pp1_r..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_mcore_tp4_pp1_r..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp1_dist_op..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp1_overlap..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp2/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp2_resume_..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp4/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp4_overlap..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp4_resume_..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp1_pp4_vp1_ove..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp2_pp2_overlap..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp2_pp2_resume_..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp4_pp1/ 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp4_pp1_overlap..> 10-Sep-2025 08:16 -
gpt3_345m_nightly_dgx_a100_1N8G_tp4_pp1_resume_..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_nondet_tp1..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp1_pp1_bf..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp1_pp1_fp..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp1_pp2_fp..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp2_pp2_fp..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp2_pp2_fp..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp2_pp2_na..> 10-Sep-2025 08:16 -
gpt3_345m_weekly_dgx_h100_1N8G_mcore_tp4_pp2_fp..> 10-Sep-2025 08:16 -
gpt3_7b_mr_dgx_a100_1N8G_tp1_pp4_memory_speed/ 10-Sep-2025 08:16 -
gpt3_7b_mr_dgx_a100_1N8G_tp4_pp1_memory_speed/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_dist_optimizer_no_mmap..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_frozen_resume_torch_di..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_resume_torch_dist_dist..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_resume_torch_dist_dist..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_resume_torch_dist_unif..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp1_uniform_full_recompute..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_cp4_a2a_p2p_nondetermi..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_resume_torch_dist_cp4_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_resume_torch_dist_rope..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_resume_torch_dist_rope..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_rope_embeddings_dgx_a1..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp2_rope_embeddings_interl..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_disable_bias_linear_dg..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_frozen_resume_torch_di..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_persistent_ckpt_disabl..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_resume_torch_dist_disa..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_resume_torch_dist_pers..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_resume_torch_dist_sequ..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_resume_torch_dist_swig..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_resume_torch_dist_unti..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_sequence_parallel_dgx_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_swiglu_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_untie_embeddings_and_o..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_calculate_per_toke..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_decoupled_lr_dgx_a..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_dist_optimizer_ove..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_dist_optimizer_ove..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_dist_optimizer_ove..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_dist_optimizer_ove..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_decou..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_resume_torch_dist_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_tunable_overlap_dg..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp1_uneven_pipeline_dg..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp1_pp4_vp2_account_for_embedd..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_cp2_nondeterministic_d..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_frozen_resume_torch_di..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_fsdp2_resume_torch_dis..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_multi_dist_optimizer_i..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_resume_torch_dist_cp2_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp1_resume_torch_dist_mult..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_calculate_per_toke..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_calculate_per_toke..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_etp4_calculate_per..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_etp4_calculate_per..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_etp4_dp_last_dgx_a..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_etp4_nondeterminis..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cp2_nondeterministic_d..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_cross_entropy_loss_fus..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_ddp_average_in_collect..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_defer_embedding_wgrad_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_mla_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_no_create_attention_ma..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_no_mmap_bin_files_dgx_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_cp2_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_cros..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_ddp_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_defe..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_dgx_..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_no_c..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_no_m..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_pp2_resume_torch_dist_resh..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp2_zp_z3_resume_torch_dist_dg..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_dist_optimizer_overlap..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_dist_optimizer_overlap..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_qk_layernorm_test_mode..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_resume_torch_dist_dist..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_resume_torch_dist_dist..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp1_resume_torch_dist_qk_l..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp2_frozen_resume_torch_di..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_te_tp4_pp2_resume_torch_dist_resh..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_tp2_pp2_resume_torch_dist_uninsta..> 10-Sep-2025 08:16 -
gpt3_mr_mcore_tp2_pp2_uninstall_te_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_te_tp2_pp2_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_te_tp2_pp2_resume_torch_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_tp1_pp4_vp1_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_tp1_pp4_vp1_resume_torch_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_tp2_pp2_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_mr_tp2_pp2_resume_torch_dgx_a100_1N8G/ 10-Sep-2025 08:16 -
gpt3_nightly_mcore_te_tp2_pp1_modelopt_distill_..> 10-Sep-2025 08:16 -