Commit graph

Select branches

Hide pull requests

EntilZha-patch-1

add_rope_fp32

apex_rmsnorm

artidoro-patch-1

assert_seq

consolidate_script

disable_raf

dist_init

ent_seq_plus_1

entropy_fix

entropy_init

eval_mask

fix_reload_state

fix_repro

fix_wandb

main

merge_sources

pr17

pr18

pr23

pr25

pr32

pr33

pr34

pr35

pr38

pr39

pr40

pr41

pr45

pr46

pr47

pr50

pr53

pr54

pr60

pr63

pr65

pr66

pr68

pr73

pr74

pr75

pr76

pr77

pr81

pr82

pr86

pr91

remove_selective_ac

sapling-pr-archive-EntilZha

sriniiyer-patch-1

stool_fix

str_id

#1

#13

#17

#18

#19

#23

#25

#26

#3

#32

#33

#34

#35

#38

#39

#40

#41

#42

#43

#44

#45

#46

#47

#48

#49

#50

#53

#54

#55

#56

#57

#58

#60

#62

#63

#65

#66

#68

#69

#7

#72

#73

#74

#75

#76

#77

#79

#80

#81

#82

#83

#85

#86

#90

#91

#92

#93

#94

#95

#95

#96

Mono Color

• 2749d0e435
  Merge 06df8e8651 into e299427ae4 Srinivasan Iyer 2025-04-09 02:30:23 +00:00
• e299427ae4
  Cast int sample id to str (#96) main Srinivasan Iyer 2025-04-08 17:45:45 -07:00
• 6a6a871c68 Cast int sample id to str str_id Srini Iyer 2025-04-09 00:21:48 +00:00
• 06df8e8651 +1 to seq len for entropy model ent_seq_plus_1 Srini Iyer 2025-04-09 00:20:02 +00:00
• 138c2f3494
  Init distributed when loading model (#94) Srinivasan Iyer 2025-04-08 13:57:28 -07:00
• a90d950d70 Init distributed when loading model dist_init Srini Iyer 2025-04-08 18:18:40 +00:00
• 19a3f7588d
  Fix eval mask (#93) Srinivasan Iyer 2025-04-08 10:31:40 -07:00
• a6acccdc71 Fix eval mask eval_mask Srini Iyer 2025-04-08 02:22:29 +00:00
• 8c1b1a78bb
  remove selective activation checkpointing (#92) Srinivasan Iyer 2025-04-07 19:20:41 -07:00
• 76fbba72b2 remove selective activation checkpointing remove_selective_ac Srini Iyer 2025-04-08 00:32:48 +00:00
• 1e78a49bf0
  update (#91) Pedro Rodriguez 2025-04-02 09:40:08 -07:00
• 60346fdde8
  Merge 8fe7bbfc4f into sapling-pr-archive-EntilZha sapling-pr-archive-EntilZha Pedro Rodriguez 2025-04-02 09:31:16 -07:00
• 8fe7bbfc4f update pr91 Pedro Rodriguez 2025-04-02 16:31:08 +00:00
• 264043b91e
  fix EntilZha-patch-1 Pedro Rodriguez 2025-04-02 09:16:42 -07:00
• b79eb3ef11
  Get generation working for BLT (#86) Pedro Rodriguez 2025-04-01 16:07:55 -07:00
• 661ccd3204
  Merge 0c09a840b5 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-20 19:13:56 -07:00
• 0c09a840b5 Get generation working for BLT pr86 Pedro Rodriguez 2025-03-21 02:13:35 +00:00
• 71b91ef68f
  Merge 06c52da75f into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-20 19:13:15 -07:00
• 06c52da75f Get generation working for BLT Pedro Rodriguez 2025-03-21 02:13:06 +00:00
• 2dcf48bdd9
  Fix in-place addition of patch_embds (#85) Hanna 2025-03-21 00:46:32 +01:00
• b460ef3cd3 Fix in-place addition of patch_embds Hanna Herasimchyk 2025-03-18 16:34:05 +01:00
• fc946a1918
  Some fixes for entropy model predictions (#83) Srinivasan Iyer 2025-03-13 10:28:42 -07:00
• 083656ce55
  Update ppl evals to work with blt model, in addition to entropy model (#82) Pedro Rodriguez 2025-03-13 10:23:31 -07:00
• 7a7b40b209
  Merge 719900d4bd into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-13 10:15:20 -07:00
• 719900d4bd Update ppl evals to work with blt model, in addition to entropy model pr82 Pedro Rodriguez 2025-03-13 17:14:53 +00:00
• f84ee635bd
  Update iterate_data (#81) Pedro Rodriguez 2025-03-13 10:14:41 -07:00
• ac372fa8d6 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-13 17:13:46 +00:00
• ae3b2cd8eb Update iterate_data pr81 Pedro Rodriguez 2025-03-13 00:23:54 +00:00
• f50157f3e2 Some fixes for entropy model predictions entropy_fix Srini Iyer 2025-03-13 05:12:56 +00:00
• 09276e85f6
  Merge a45f5f0880 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-12 17:25:38 -07:00
• a45f5f0880 Update ppl evals to work with blt model, in addition to entropy model Pedro Rodriguez 2025-03-13 00:25:09 +00:00
• d87ba751d3 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-13 00:24:10 +00:00
• 790e224b11 Update ppl evals to work with blt model, in addition to entropy model Pedro Rodriguez 2025-03-13 00:23:54 +00:00
• fe70785822 Update iterate_data Pedro Rodriguez 2025-03-13 00:23:54 +00:00
• c110f6be2a
  Add way to call consolidate (#80) Srinivasan Iyer 2025-03-11 16:53:33 -07:00
• a3973339f2 isort consolidate_script Srini Iyer 2025-03-11 23:53:01 +00:00
• 6160e83d41 black Srini Iyer 2025-03-11 23:51:23 +00:00
• 699d0be470 Add way to call consolidate Srini Iyer 2025-03-11 22:26:11 +00:00
• a5ceaaa226
  When merging configs, do not merge data sources (#79) Srinivasan Iyer 2025-03-11 11:03:24 -07:00
• e08957ffeb Add todo merge_sources Srini Iyer 2025-03-11 17:56:23 +00:00
• 662262a528 When merging configs, do not merge data sources Srini Iyer 2025-03-11 17:44:37 +00:00
• 7517ac2a9f
  Get evals working again. (#46) Pedro Rodriguez 2025-03-11 09:57:19 -07:00
• 63913e4dba
  Reduce per file resources arrow uses (#77) Pedro Rodriguez 2025-03-05 15:03:42 -08:00
• aec12c79e6
  Merge 880493e742 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 15:03:22 -08:00
• 880493e742 Reduce per file resources arrow uses pr77 Pedro Rodriguez 2025-03-05 23:03:14 +00:00
• 8f2cf8899d
  Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases (#75) Pedro Rodriguez 2025-03-05 15:02:57 -08:00
• bde475f8a0
  Merge a828594625 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 15:02:32 -08:00
• a828594625 Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases pr75 Pedro Rodriguez 2025-03-05 23:02:22 +00:00
• ea1fc75862
  Add approximate state persistence (#73) Pedro Rodriguez 2025-03-05 15:01:45 -08:00
• e78a24dc80
  Merge 34664fa7f1 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 14:49:24 -08:00
• 34664fa7f1 Reduce per file resources arrow uses Pedro Rodriguez 2025-03-05 22:49:04 +00:00
• 3c08d7e5d7
  Merge 3d44bd1b7a into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 13:23:32 -08:00
• 3d44bd1b7a Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases Pedro Rodriguez 2025-03-05 21:23:21 +00:00
• c3ad8b60f4 Add approximate state persistence pr73 Pedro Rodriguez 2025-03-05 21:23:21 +00:00
• 3114f52e82 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-05 21:16:34 +00:00
• 8c6103e6e7 Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases Pedro Rodriguez 2025-03-05 21:16:25 +00:00
• daea91e4a9 Add approximate state persistence Pedro Rodriguez 2025-03-05 21:16:24 +00:00
• 9bd51df961
  Fix rsync to not preserve original permissions, instead use destination (#76) Pedro Rodriguez 2025-03-05 11:49:41 -08:00
• 6bcefb0412 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-05 19:48:54 +00:00
• f05acb95fb Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases Pedro Rodriguez 2025-03-05 19:48:43 +00:00
• e60344da9e Add approximate state persistence Pedro Rodriguez 2025-03-05 19:48:42 +00:00
• 7e510088bc
  Merge 44668ef966 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 09:06:39 -08:00
• 44668ef966 Fix rsync to not preserve original permissions, instead use destination pr76 Pedro Rodriguez 2025-03-05 17:06:32 +00:00
• eea7f02949 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-05 17:01:54 +00:00
• 428bfe2f76
  Merge 1288483307 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-05 09:01:35 -08:00
• 1288483307 Fix rsync to not preserve original permissions, instead use destination Pedro Rodriguez 2025-03-05 17:01:23 +00:00
• f0636bf31c Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases Pedro Rodriguez 2025-03-04 01:42:53 +00:00
• 4756a88cdd Add approximate state persistence Pedro Rodriguez 2025-03-05 16:58:24 +00:00
• 4c82ed8732
  Merge abb4f7e6a4 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-03 17:43:54 -08:00
• abb4f7e6a4 Let process start before yielding preloaded prefetch buffer, avoid needlessly losing buffer in edge cases Pedro Rodriguez 2025-03-04 01:42:53 +00:00
• e0ddc2dc82 merge commit for archive created by Sapling Pedro Rodriguez 2025-03-04 01:19:29 +00:00
• 3e1df4ea4d Add approximate state persistence Pedro Rodriguez 2025-03-04 01:02:34 +00:00
• c727844e9d
  Correctly reset batch iterator at each arrow create_iter call. (#74) Pedro Rodriguez 2025-03-03 16:59:02 -08:00
• dd8557400c
  Merge f74aa7bd1a into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-03-03 15:32:39 -08:00
• f74aa7bd1a Correctly reset batch iterator at each arrow create_iter call. pr74 Pedro Rodriguez 2025-03-03 23:32:29 +00:00
• 9331363fc2
  Merge 967b23fd05 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-28 16:15:42 -08:00
• 967b23fd05 Add approximate state persistence Pedro Rodriguez 2025-03-01 00:15:32 +00:00
• a81de49649 merge commit for archive created by Sapling Pedro Rodriguez 2025-02-28 00:41:06 +00:00
• 2cae41fe1f Get evals working again. pr46 Pedro Rodriguez 2025-02-28 00:40:04 +00:00
• 0b12e91b3b merge commit for archive created by Sapling Pedro Rodriguez 2025-02-27 19:51:16 +00:00
• 57d04fa37d Minimal working eval Pedro Rodriguez 2025-02-27 19:42:39 +00:00
• 08b8c7cd05
  Pass mask in packing_iterator, correctly handle last batch, fix masking (#65) Pedro Rodriguez 2025-02-27 11:41:47 -08:00
• 0da051f4f9
  Initialize rope embeddings properly for the entropy model (#72) Srinivasan Iyer 2025-02-25 15:35:25 -08:00
• e668ac0280 Initialize rope embeddings properly for the entropy model entropy_init Srini Iyer 2025-02-25 20:36:20 +00:00
• 2d4f277596
  Merge 9446c1ee5c into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-25 11:15:02 -08:00
• a77878ae65
  Merge c7b40706f0 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-25 11:11:30 -08:00
• 9446c1ee5c Minimal working eval Pedro Rodriguez 2025-02-25 19:11:23 +00:00
• c7b40706f0 Pass mask in packing_iterator, correctly handle last batch, fix masking pr65 Pedro Rodriguez 2025-02-25 19:11:23 +00:00
• aeb95f12a1
  Remove byte tokenizer and add config args to switch between byte/patch packing (#68) Pedro Rodriguez 2025-02-25 11:10:59 -08:00
• 62cb8936ee merge commit for archive created by Sapling Pedro Rodriguez 2025-02-25 01:38:44 +00:00
• 52d5603b4f Pass mask in packing_iterator, correctly handle last batch, fix masking Pedro Rodriguez 2025-02-25 01:38:37 +00:00
• f48ad82d96
  Merge 6147207155 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-24 17:35:42 -08:00
• 6147207155 Pass mask in packing_iterator, correctly handle last batch, fix masking Pedro Rodriguez 2025-02-25 01:34:12 +00:00
• 2a04df1130
  Merge 3aaeb8ac14 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-24 17:34:18 -08:00
• 3aaeb8ac14 Pass mask in packing_iterator, correctly handle last batch, fix masking Pedro Rodriguez 2025-02-25 01:34:12 +00:00
• f3781cc0ca merge commit for archive created by Sapling Pedro Rodriguez 2025-02-25 00:04:43 +00:00
• edccc0873d Remove byte tokenizer and add config args to switch between byte/patch packing pr68 Pedro Rodriguez 2025-02-24 23:56:42 +00:00
• ff36aa8642
  Add vocab and seq len abstract fields (#66) Pedro Rodriguez 2025-02-24 14:41:58 -08:00
• bbd1edd90d
  Merge 4c6ee1aef0 into sapling-pr-archive-EntilZha Pedro Rodriguez 2025-02-24 14:41:33 -08:00
• 4c6ee1aef0 Add vocab and seq len abstract fields pr66 Pedro Rodriguez 2025-02-22 01:27:13 +00:00