vikarti.anatra/blt
1
0
Fork 0
mirror of https://github.com/facebookresearch/blt.git synced 2025-02-23 21:42:14 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
49 commits 35 branches 0 tags 1.5 MiB
Commit graph

22 commits

Author SHA1 Message Date
Pedro Rodriguez 48df9ce785
Merge c6ef4285e2 into sapling-pr-archive-EntilZha 2025-02-04 10:03:26 -08:00
Pedro Rodriguez c6ef4285e2 Several changes to enable entropy model training/eval 
Summary:

- Make arrow iterator able to read from jsonl files, the entropies are omitted in this case
- Make the data/checkpoint code fsspec compatible
- Fix issues with all reduce with non-bf16 in dist_sum and norm computation.
- Minimal fixes to get eval to run, it is slow currently
- Add bpb numbers during training


Test Plan:
 2025-02-04 18:03:19 +00:00
Pedro Rodriguez 4ff8341738
Merge 11cad6c84d into sapling-pr-archive-EntilZha
Some checks are pending
Lint with Black / lint (push) Waiting to run
Details
Lint with isort / lint (push) Waiting to run
Details
2025-02-03 18:29:37 -08:00
Pedro Rodriguez 11cad6c84d WIP parallel copy script 
Summary:

Test Plan:
 2025-01-28 00:57:06 +00:00
Pedro Rodriguez 7044771a12
This includes fixes that make checkpointing and reloading work correctly. (#35)
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details 
It also batches in a first set of changes for fixing eval code

Summary:

Test Plan:
 2025-01-27 16:56:42 -08:00
Pedro Rodriguez 4db801a532
Merge caf82b924e into sapling-pr-archive-EntilZha
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details
2025-01-27 16:54:52 -08:00
Pedro Rodriguez caf82b924e This includes fixes that make checkpointing and reloading work correctly. 
It also batches in a first set of changes for fixing eval code

Summary:

Test Plan:
 2025-01-28 00:54:47 +00:00
Pedro Rodriguez c2f1e4845e
Merge e02ba763b0 into sapling-pr-archive-EntilZha 2025-01-27 16:38:54 -08:00
Pedro Rodriguez e02ba763b0 This includes fixes that make checkpointing and reloading work correctly. 
It also batches in a first set of changes for fixing eval code

Summary:

Test Plan:
 2025-01-28 00:38:46 +00:00
Pedro Rodriguez 7622d28b74
Initial codes and scripts for training entropy model (#34)
Some checks are pending
Lint with Black / lint (push) Waiting to run
Details
Lint with isort / lint (push) Waiting to run
Details 
Summary:

Test Plan:
 2025-01-27 09:46:44 -08:00
Pedro Rodriguez b1c12dd275
Merge 34ca1f7d4b into sapling-pr-archive-EntilZha
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details
2025-01-24 13:59:47 -08:00
Pedro Rodriguez 34ca1f7d4b Initial codes and scripts for training entropy model 
Summary:

Test Plan:
 2025-01-24 21:59:42 +00:00
Pedro Rodriguez fb09022e5e Initial codes and scripts for training entropy model 
Summary:

Test Plan:
 2025-01-24 21:55:41 +00:00
Pedro Rodriguez a809259e71
Use load_async flag to not start MP iterator (#33)
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details 
Summary:

Test Plan:
 2025-01-24 10:57:20 -08:00
Pedro Rodriguez bc42cebd7d
Update file check script to check sizes (#32)
Some checks failed
Lint with isort / lint (push) Has been cancelled
Details
Lint with Black / lint (push) Has been cancelled
Details 
Summary:

Test Plan:
 2025-01-22 13:06:46 -08:00
Ink 392117bff2
Fix realtime entropy patching (#26)
Some checks are pending
Lint with Black / lint (push) Waiting to run
Details
Lint with isort / lint (push) Waiting to run
Details 
* allow loading of the entropy model directly

* remove unused argument

* remove spammy warning

* allow patch_batch_size to be adjusted in the forward() method

* revert to original patcher style, fix warning

* allow grads when calculating entropies

* fix grad flow

* return preds from calculate_entropies()

* remove legacy arg

* fix an error with monotonicity and small sequence lengths

* ensure patcher is serializable

* revert patcher to original

* remove unused import
 2025-01-21 16:34:23 -08:00
Pedro Rodriguez 6ffeb66b53
Changes for training entropy model and correcting attention in local models (#25)
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details 
Summary:

- Refactor local model configs to be separate and clearer
- Add attention arguments and correct which attention is used in local models
- Preparation for being able to have an entropy train script
- Fix failing unit tests

Test Plan:
 2025-01-17 14:23:01 -08:00
Ink caec8d2621
allow flex-attention to be disabled (#19)
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details 
* allow flex-attention to silently fail

* allow flex-attn to be disabled via an env var
 2025-01-14 09:32:07 -08:00
Pedro Rodriguez 1da3dd9315
Update preprocess_entropies script to blt inference + add fsspec support (#23)
Some checks are pending
Lint with Black / lint (push) Waiting to run
Details
Lint with isort / lint (push) Waiting to run
Details 
Summary:

Test Plan:
 2025-01-13 15:28:14 -08:00
Pedro Rodriguez b0120da72f
Replace regular filesystem calls with fsspec + add s3 support (#18)
Some checks failed
Lint with Black / lint (push) Has been cancelled
Details
Lint with isort / lint (push) Has been cancelled
Details 
Summary:

For compatibility with either local/nfs or S3 datasets, swap to fsspec.

Add a tool to compare local and remote filesystems

Test Plan:

- Ran regular train script
- Ran with config with data in S3
 2025-01-10 11:04:41 -08:00
Pedro Rodriguez d4ddb95322
Add plotting code from paper (#17)
Some checks are pending
Lint with Black / lint (push) Waiting to run
Details
Lint with isort / lint (push) Waiting to run
Details 
Summary:

Test Plan:
 2025-01-09 12:11:50 -08:00
Pedro Rodriguez bcc039bb75 Initial commit 2024-12-12 15:32:30 -08:00