update configs

allenai · Oct 8, 2023 · 7906bd2 · 7906bd2
1 parent c110059
commit 7906bd2
Show file tree

Hide file tree

Showing 4 changed files with 17 additions and 0 deletions.
diff --git a/configs/mcli/v1-mix-medium-mitch-ish.yaml b/configs/mcli/v1-mix-medium-mitch-ish.yaml
@@ -12,9 +12,11 @@ integrations:
 command: |-
   pip install urllib3==1.26.17
   pip freeze
+  mkdir -p /root/.cache/torch/
 
   export OMP_NUM_THREADS=8
   export LOG_FILTER_TYPE=local_rank0_only
+  export OLMO_NO_SSL=1  # we get SSLErrors all the time on this cluster
   #export PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:128
 
   cd LLM

diff --git a/configs/mcli/v1-mix-medium.yaml b/configs/mcli/v1-mix-medium.yaml
@@ -10,10 +10,17 @@ integrations:
     pip_install: -e .
     ssh_clone: true
 command: |-
+  pip install urllib3==1.26.17
+  pip freeze
+  mkdir -p /root/.cache/torch/
+
   export OMP_NUM_THREADS=8
   export LOG_FILTER_TYPE=local_rank0_only
+  export OLMO_NO_SSL=1  # we get SSLErrors all the time on this cluster
   #export PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:128
+
   cd LLM
+
   torchrun \
   --master_addr $MASTER_ADDR \
   --master_port $MASTER_PORT \

diff --git a/configs/mcli/v1_5-mix-medium-mitch-ish.yaml b/configs/mcli/v1_5-mix-medium-mitch-ish.yaml
@@ -12,6 +12,7 @@ integrations:
 command: |-
   pip install urllib3==1.26.17
   pip freeze
+  mkdir -p /root/.cache/torch/
 
   export OMP_NUM_THREADS=8
   export LOG_FILTER_TYPE=local_rank0_only

diff --git a/configs/mcli/v1_5-mix-medium.yaml b/configs/mcli/v1_5-mix-medium.yaml
@@ -10,10 +10,17 @@ integrations:
     pip_install: -e .
     ssh_clone: true
 command: |-
+  pip install urllib3==1.26.17
+  pip freeze
+  mkdir -p /root/.cache/torch/
+
   export OMP_NUM_THREADS=8
   export LOG_FILTER_TYPE=local_rank0_only
+  export OLMO_NO_SSL=1  # we get SSLErrors all the time on this cluster
   #export PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:128
+
   cd LLM
+
   torchrun \
   --master_addr $MASTER_ADDR \
   --master_port $MASTER_PORT \