Commits · 10c6ad06f6e7016219a5426e0570b1ca4ecfcfe2 · OpenDAS / Megatron-LM

08 Feb, 2022 1 commit
- fixed t5 'get_num_layers()' · 2fadaa50
  Lawrence McAfee authored Feb 08, 2022
  
  2fadaa50
04 Feb, 2022 1 commit
- renamed argument; 'embed' -> 'embedding' · c04c4977
  Lawrence McAfee authored Feb 04, 2022
  
  c04c4977
01 Feb, 2022 2 commits
- comments, cleanup. · b93bef00
  Lawrence McAfee authored Feb 01, 2022
  
  b93bef00
- found root source of t5 issue (fast layer norm) · bea16fa3
  Lawrence McAfee authored Feb 01, 2022
  
  bea16fa3
31 Jan, 2022 2 commits
- working for t5 [ encoder embedding only ] · 3af6725d
  Lawrence McAfee authored Jan 31, 2022
  
  3af6725d
- added 'no-op' layer, to replace transformer layer when num_layers == 0. · 1fa6990c
  Lawrence McAfee authored Jan 31, 2022
  
  1fa6990c
29 Jan, 2022 3 commits
- narrowed issue to pipeline rank 0, virtual pipeline rank >= 1 · 5bc9f889
  Lawrence McAfee authored Jan 28, 2022
  
  5bc9f889
- typo fix · 2b628f96
  Vijay Korthikanti authored Jan 28, 2022
  
  2b628f96
- Revert incorrect fix. · cd499559
  Jared Casper authored Jan 28, 2022
  
  cd499559
28 Jan, 2022 2 commits
- more naming cleanup · 641408f5
  Vijay Korthikanti authored Jan 28, 2022
  
  641408f5
- changing class name AnnealingLR to OptimizerParamScheduler · 04ecc834
  Vijay Korthikanti authored Jan 28, 2022
  
  04ecc834
27 Jan, 2022 1 commit
- address review comments · 53931b8b
  Vijay Korthikanti authored Jan 27, 2022
  
  53931b8b
26 Jan, 2022 2 commits
- further clarified viewless tensor comment in transformer.py · d16e2a24
  Lawrence McAfee authored Jan 26, 2022
  
  d16e2a24
- address review comments · 8acbbe25
  Vijay Korthikanti authored Jan 26, 2022
  
  8acbbe25
25 Jan, 2022 2 commits
- limit 'make_viewless_tensor()' to case of micro_batch_size == 1; added comment · 24369dd6
  Lawrence McAfee authored Jan 25, 2022
  
  24369dd6
- working with interleaving · 804ed2e6
  Lawrence McAfee authored Jan 24, 2022
  
  804ed2e6
24 Jan, 2022 4 commits
- added args.transformer_pipeline_model_parallel_size · a06af061
  Lawrence McAfee authored Jan 24, 2022
  
  a06af061
- fixed args.virtual_pipeline_model_parallel_size · c2b7d0b3
  Lawrence McAfee authored Jan 24, 2022
  
  c2b7d0b3
- minor fixes · 7e810e41
  Vijay Korthikanti authored Jan 24, 2022
  
  7e810e41
- working when no interleaving · 33dc8e9c
  Lawrence McAfee authored Jan 24, 2022
  
  33dc8e9c
22 Jan, 2022 1 commit
- second phase of vision code merge · d8c85650
  Vijay Korthikanti authored Jan 21, 2022
  
  d8c85650
20 Jan, 2022 1 commit
- address review comments · 798b6a64
  Vijay Korthikanti authored Jan 20, 2022
  
  798b6a64
19 Jan, 2022 2 commits
- using all_gather instead of gather (nccl does not support gather) · b037a69e
  Vijay Korthikanti authored Jan 19, 2022
  
  b037a69e
- flag for data parallel random initialization · a7ee77ea
  Vijay Korthikanti authored Jan 19, 2022
  
  a7ee77ea
15 Jan, 2022 2 commits
- minor fix · fd8dd9c0
  Vijay Korthikanti authored Jan 14, 2022
  
  fd8dd9c0
- checkpointing rng_state of all data parallel ranks · 09d38884
  Vijay Korthikanti authored Jan 14, 2022
  
  09d38884
14 Jan, 2022 1 commit
- address review comments · 343dc97a
  Vijay Korthikanti authored Jan 14, 2022
  
  343dc97a
13 Jan, 2022 7 commits
- handle output_tensor == None · 7f1c5917
  Lawrence McAfee authored Jan 13, 2022
  
  7f1c5917
- more cleanup · 8fc5e323
  Lawrence McAfee authored Jan 13, 2022
  
  8fc5e323
- renamed 'free' -> 'deallocate' · c1e4526b
  Lawrence McAfee authored Jan 13, 2022
  
  c1e4526b
- removed uses are args.deallocate_pipeline_output · d10f81c5
  Lawrence McAfee authored Jan 13, 2022
  
  d10f81c5
- removed argument flag; also, converted torch.FloatTenstor -> torch.empty in previous commit · 18846a0a
  Lawrence McAfee authored Jan 13, 2022
  
  18846a0a
- fixed bug for no-interleave pipeline schedule · 1f387c2c
  Lawrence McAfee authored Jan 13, 2022
  
  1f387c2c
- minor fixes · 3b29f0e7
  Vijay Korthikanti authored Jan 13, 2022
  
  3b29f0e7
12 Jan, 2022 1 commit
- Phase1 merge: vit optimizations + dataset enhancements + scaled_softmax kernel · 7a77abd9
  Vijay Korthikanti authored Jan 12, 2022
  
  7a77abd9
11 Jan, 2022 4 commits
- removed several 'assert_viewless_tensor' calls. · 0c8e8cce
  Lawrence McAfee authored Jan 11, 2022
  
  0c8e8cce
- added comments · a1fe4805
  Lawrence McAfee authored Jan 11, 2022
  
  a1fe4805
- partially cleaned · 806422e5
  Lawrence McAfee authored Jan 11, 2022
  
  806422e5
- jan 11 alpha · 05042081
  Lawrence McAfee authored Jan 11, 2022
  
  05042081
10 Jan, 2022 1 commit
- loss matches; memory savings for multi-node (tested n3, n16) · 270d6412
  Lawrence McAfee authored Jan 10, 2022
  
  270d6412