vikarti.anatra/blt
1
0
Fork 0
mirror of https://github.com/facebookresearch/blt.git synced 2025-04-25 19:19:09 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
blt/bytelatent
History
Pedro Rodriguez 2cae41fe1f Get evals working again. 
- PPL/validation: Works now and uses multi-gpu. For some reason 1 GPU differs from multi-GPU, can debug in a followup PR
- Generation evals likely work, but are very slow, so disabled for now


Test Plan:
```
torchrun --nproc-per-node 8 -m bytelatent.eval config=../internal-blt/configs/eval.yaml
```
2025-02-28 00:41:01 +00:00
..
configs Remove byte tokenizer and add config args to switch between byte/patch packing () 2025-02-25 11:10:59 -08:00
data Pass mask in packing_iterator, correctly handle last batch, fix masking () 2025-02-27 11:41:47 -08:00
model Add vocab and seq len abstract fields () 2025-02-24 14:41:58 -08:00
plotting Add plotting code from paper () 2025-01-09 12:11:50 -08:00
preprocess Allow ArrowIterator to read from json () 2025-02-06 09:57:22 -08:00
tokenizers Remove byte tokenizer and add config args to switch between byte/patch packing () 2025-02-25 11:10:59 -08:00
.DS_Store Initial commit 2024-12-12 15:32:30 -08:00
__init__.py Initial commit 2024-12-12 15:32:30 -08:00
args.py Get evals working again. 2025-02-28 00:41:01 +00:00
base_transformer.py Initialize rope embeddings properly for the entropy model () 2025-02-25 15:35:25 -08:00
checkpoint.py Update checkpointing to use fsspec () 2025-02-06 09:41:58 -08:00
config_parser.py Make it possible to specify multiple config files () 2025-02-18 10:42:44 -08:00
constants.py Initial commit 2024-12-12 15:32:30 -08:00
distributed.py Get evals working again. 2025-02-28 00:41:01 +00:00
entropy_model.py Changes for training entropy model and correcting attention in local models () 2025-01-17 14:23:01 -08:00
eval.py Get evals working again. 2025-02-28 00:41:01 +00:00
float8.py Initial commit 2024-12-12 15:32:30 -08:00
generate.py Get evals working again. 2025-02-28 00:41:01 +00:00
logger.py Update checkpointing to use fsspec () 2025-02-06 09:41:58 -08:00
metrics.py Get evals working again. 2025-02-28 00:41:01 +00:00
norms.py Fix distributed all reduce grad norm () 2025-02-04 16:53:50 -08:00
optim.py Initial commit 2024-12-12 15:32:30 -08:00
print_config.py Make it possible to specify multiple config files () 2025-02-18 10:42:44 -08:00
probe.py Initial commit 2024-12-12 15:32:30 -08:00
profiling.py Initial commit 2024-12-12 15:32:30 -08:00
stool.py Allow ArrowIterator to read from json () 2025-02-06 09:57:22 -08:00
test_blt.py Initial codes and scripts for training entropy model () 2025-01-27 09:46:44 -08:00
test_config_parser.py Make it possible to specify multiple config files () 2025-02-18 10:42:44 -08:00
test_entropy_model.py Test first batch matches () 2025-02-13 10:05:08 -08:00
train.py Get evals working again. 2025-02-28 00:41:01 +00:00
transformer.py Initialize rope embeddings properly for the entropy model () 2025-02-25 15:35:25 -08:00