diff --git a/megatron/neox_arguments/neox_args.py b/megatron/neox_arguments/neox_args.py index 6ae244431..a5ddf2766 100644 --- a/megatron/neox_arguments/neox_args.py +++ b/megatron/neox_arguments/neox_args.py @@ -688,11 +688,6 @@ class NeoXArgsOther(NeoXArgsTemplate): Include sequence id as attention bias. """ - concat_data: bool = True - """ - Concatenate sequences for training. - """ - @dataclass class NeoXArgsTokenizer(NeoXArgsTemplate): diff --git a/tools/convert_hf_to_sequential.py b/tools/convert_hf_to_sequential.py index 30419d2ea..40c93aa04 100644 --- a/tools/convert_hf_to_sequential.py +++ b/tools/convert_hf_to_sequential.py @@ -35,11 +35,11 @@ Example usage: (Converts the 70M Pythia model to NeoX format) ================================================================ OMPI_COMM_WORLD_RANK=0 CUDA_VISIBLE_DEVICES=0 python tools/convert_hf_to_sequential.py \ - --hf-model-name pythia-1b \ + --hf-model-name pythia-1.4b \ --revision 143000 \ - --output-dir /p/fastdata/mmlaion/hummingbird/checkpoints/1_neox \ - --cache-dir /p/fastdata/mmlaion/hummingbird/checkpoints/pythia1B \ - --config configs/1B.yml configs/hummingbird_streaming.yml \ + --output-dir /p/fastdata/mmlaion/hummingbird/checkpoints/1_4_final \ + --cache-dir /p/fastdata/mmlaion/hummingbird/checkpoints/pythia_final \ + --config configs/pythia/1-4B.yml configs/hummingbird_streaming.yml \ --test @@ -47,16 +47,16 @@ NOTE: This requires manually changing the arguments below. ================================================================ CUDA_VISIBLE_DEVICES=0,1,2,3 python ./deepy.py tools/convert_hf_to_sequential.py \ - -d configs 1B.yml hummingbird_streaming.yml + -d configs pythia/1-4B.yml hummingbird_streaming.yml """ MULTI_GPU_ARGS = " ".join( [ - "--hf-model-name pythia-1b", + "--hf-model-name pythia-1.4b", "--revision 143000", - "--output-dir /p/fastdata/mmlaion/hummingbird/checkpoints/1_neox", - "--cache-dir /p/fastdata/mmlaion/hummingbird/checkpoints/pythia1B", - "--config configs/1B.yml configs/hummingbird_streaming.yml", + "--output-dir /p/fastdata/mmlaion/hummingbird/checkpoints/1_4_final", + "--cache-dir /p/fastdata/mmlaion/hummingbird/checkpoints/pythia_final", + "--config configs/pythia/1-4B.yml configs/hummingbird_streaming.yml", "--test", ] )