"enter/dglenter/model/vscode:/vscode.git/clone" did not exist on "539335cece9c9c7e9f21a0e433d8218990c73163"
Commit 9b174da8 authored by Deepak Narayanan's avatar Deepak Narayanan
Browse files

Merge branch 'pipeline_parallel_main' into 'main'

Pipeline parallelism and inter-layer model parallelism implementation

See merge request ADLR/megatron-lm!159
parents 3aacd955 6e83649f
This diff is collapsed.
...@@ -39,6 +39,8 @@ class RaceDataset(Dataset): ...@@ -39,6 +39,8 @@ class RaceDataset(Dataset):
print_rank_0(' >> total number of samples: {}'.format( print_rank_0(' >> total number of samples: {}'.format(
len(self.samples))) len(self.samples)))
self.sample_multiplier = NUM_CHOICES
def __len__(self): def __len__(self):
return len(self.samples) return len(self.samples)
......
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
...@@ -136,7 +136,7 @@ def get_args(): ...@@ -136,7 +136,7 @@ def get_args():
# some default/dummy values for the tokenizer # some default/dummy values for the tokenizer
args.rank = 0 args.rank = 0
args.make_vocab_size_divisible_by = 128 args.make_vocab_size_divisible_by = 128
args.model_parallel_size = 1 args.tensor_model_parallel_size = 1
return args return args
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment