diff --git a/examples/benchmarks/bert/yamls/finetuning/glue/hf-bert-base-uncased.yaml b/examples/benchmarks/bert/yamls/finetuning/glue/hf-bert-base-uncased.yaml index 661cb5fb4..bd7f94dc9 100644 --- a/examples/benchmarks/bert/yamls/finetuning/glue/hf-bert-base-uncased.yaml +++ b/examples/benchmarks/bert/yamls/finetuning/glue/hf-bert-base-uncased.yaml @@ -7,7 +7,7 @@ parallel: true # Basic run configuration, additional details will be added to this name for each GLUE task, and each random seed base_run_name: hf-bert-base-uncased-glue-finetuning # Determines how runs are saved and logged in W&B default_seed: 19 -precision: bf16 +precision: amp_bf16 # Tokenizer for dataset creation tokenizer_name: bert-base-uncased @@ -19,7 +19,7 @@ model: tokenizer_name: ${tokenizer_name} # Loading -starting_checkpoint_load_path: # Fill this in with the composer checkpoint from the end of pre-training a HF BERT +starting_checkpoint_load_path: # Fill this in with the composer checkpoint from the end of pre-training a HF BERT local_pretrain_checkpoint_folder: ./local-bert-checkpoints/ # Saving diff --git a/examples/benchmarks/bert/yamls/finetuning/glue/mcloud_run.yaml b/examples/benchmarks/bert/yamls/finetuning/glue/mcloud_run.yaml index 28168d19f..29e775abb 100644 --- a/examples/benchmarks/bert/yamls/finetuning/glue/mcloud_run.yaml +++ b/examples/benchmarks/bert/yamls/finetuning/glue/mcloud_run.yaml @@ -12,19 +12,19 @@ name: mosaic-bert-base-uncased-glue-finetuning image: mosaicml/pytorch:1.13.1_cu117-python3.10-ubuntu20.04 compute: - gpus: 8 # Number of GPUs to use + gpus: 8 # Number of GPUs to use ## These configurations are optional # cluster: TODO # Name of the cluster to use for this run # gpu_type: a100_80gb # Type of GPU to use. We use a100_80gb in our experiments integrations: -- integration_type: git_repo - git_repo: mosaicml/examples - git_branch: v0.0.4 # use your branch - # git_commit: # OR use your commit hash - pip_install: -e .[bert] - ssh_clone: false # Should be true if using a private repo + - integration_type: git_repo + git_repo: mosaicml/examples + git_branch: v0.0.4 # use your branch + # git_commit: # OR use your commit hash + pip_install: -e .[bert] + ssh_clone: false # Should be true if using a private repo command: | cd examples/examples/bert python glue.py /mnt/config/parameters.yaml @@ -43,7 +43,7 @@ parameters: base_run_name: # If left blank, will be read from top YAML name default_seed: 19 - precision: bf16 + precision: amp_bf16 # Tokenizer for dataset creation tokenizer_name: bert-base-uncased diff --git a/examples/benchmarks/bert/yamls/finetuning/glue/mosaic-bert-base-uncased.yaml b/examples/benchmarks/bert/yamls/finetuning/glue/mosaic-bert-base-uncased.yaml index 80911e0d0..99dec55f8 100644 --- a/examples/benchmarks/bert/yamls/finetuning/glue/mosaic-bert-base-uncased.yaml +++ b/examples/benchmarks/bert/yamls/finetuning/glue/mosaic-bert-base-uncased.yaml @@ -7,7 +7,7 @@ parallel: true # Basic run configuration, additional details will be added to this name for each GLUE task, and each random seed base_run_name: mosaic-bert-base-uncased-glue-finetuning # Determines how runs are saved and logged in W&B default_seed: 19 -precision: bf16 +precision: amp_bf16 # Tokenizer for dataset creation tokenizer_name: bert-base-uncased