diff --git a/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_711.sh b/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_711.sh new file mode 100644 index 0000000000000000000000000000000000000000..4451858b422bb8657c5c76bc2a2f19561cb50a4d --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_711.sh @@ -0,0 +1,25 @@ +#!/bin/bash +set -m +sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalbh_splitab_splitae +echo "above deepspeed$" +FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/ +#TOKENIZER=facebook/nllb-200-distilled-600M +#TOKENIZER_TYPE=HuggingFaceTokenizer +TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k +TOKENIZER=$TOKENIZER_PATH/all.model +VOCAB_FILE=$TOKENIZER_PATH/all.vocab +TOKENIZER_TYPE=SentencePieceTokenizer +mkdir -p $FINAL_DIR/tokenizer711/ +cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed +echo "inside deepspeed" +pwd +python3 tools/preprocess_data.py \ + --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalbh_splitab_splitae \ + --output-prefix $FINAL_DIR/tokenizer711/ \ + --tokenizer-model $TOKENIZER \ + --vocab-file $VOCAB_FILE \ + --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\ + --append-eod --workers 8 # --partitions 16 #--chunk-size 50 + + # --merge-file $MERGES_FILE \ + diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_100.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_100.yaml new file mode 100644 index 0000000000000000000000000000000000000000..72cbac8217e7a66f616bc03bd86b0047015a7168 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_100.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-100 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-100 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-100-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_100.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_101.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_101.yaml new file mode 100644 index 0000000000000000000000000000000000000000..b98e7ea31dc0be5297ee2d4db6798b5b8494c22e --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_101.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-101 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-101 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-101-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_101.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_102.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_102.yaml new file mode 100644 index 0000000000000000000000000000000000000000..ce966386eaf172266e5ddf537d8b3a771f608fe5 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_102.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-102 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-102 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-102-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_102.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_103.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_103.yaml new file mode 100644 index 0000000000000000000000000000000000000000..b02ba1aa8c7db31843dffa59cf32f0f889ffbbff --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_103.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-103 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-103 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-103-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_103.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_104.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_104.yaml new file mode 100644 index 0000000000000000000000000000000000000000..af733000b8c462362943f269cbb0843919d08727 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_104.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-104 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-104 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-104-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_104.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_106.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_106.yaml new file mode 100644 index 0000000000000000000000000000000000000000..420931f4197e1df38203517a7780571837908c14 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_106.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-106 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-106 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-106-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_106.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_107.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_107.yaml new file mode 100644 index 0000000000000000000000000000000000000000..5dc0bc8f2c453895eba9bb739204d4c6b5d431f6 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_107.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-107 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-107 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-107-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_107.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_108.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_108.yaml new file mode 100644 index 0000000000000000000000000000000000000000..717dc9f884ec9367a61b611bf294792949f00680 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_108.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-108 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-108 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-108-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_108.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_110.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_110.yaml new file mode 100644 index 0000000000000000000000000000000000000000..34862dc26f22d21d928b80988c0aeb454f4f5b8e --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_110.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-110 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-110 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-110-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_110.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_113.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_113.yaml new file mode 100644 index 0000000000000000000000000000000000000000..adaa683000959f81e7b86dbd3c104d6beab592aa --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_113.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-113 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-113 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-113-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_113.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_114.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_114.yaml new file mode 100644 index 0000000000000000000000000000000000000000..2ba662b8159fa06a090e5e1fa0a6bbac2ac2bfa3 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_114.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-114 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-114 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-114-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_114.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_115.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_115.yaml new file mode 100644 index 0000000000000000000000000000000000000000..dd6791fd96a4d27c760e5a9aff06fffaa804018a --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_115.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-115 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-115 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-115-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_115.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_116.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_116.yaml new file mode 100644 index 0000000000000000000000000000000000000000..50cf4261f6d42e35e97c6caa9e8271f60889c614 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_116.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-116 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-116 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-116-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_116.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_117.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_117.yaml new file mode 100644 index 0000000000000000000000000000000000000000..ac41297fac6db01908438fd615870845a983836f --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_117.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-117 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-117 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-117-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_117.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_118.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_118.yaml new file mode 100644 index 0000000000000000000000000000000000000000..5e4461efadc6189cf44987a7b016b352c25ff120 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_118.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-118 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-118 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-118-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_118.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_120.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_120.yaml new file mode 100644 index 0000000000000000000000000000000000000000..5fbb29e4220d884af98d697fe787b7ed9ab29839 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_120.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-120 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-120 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-120-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_120.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_122.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_122.yaml new file mode 100644 index 0000000000000000000000000000000000000000..85fcacc3ec177a05131c011b8e8ee7c3e1f93764 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_122.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-122 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-122 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-122-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_122.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_124.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_124.yaml new file mode 100644 index 0000000000000000000000000000000000000000..93eb1f62809fc56bec18876bc3be4ca2c6f0f422 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_124.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-124 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-124 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-124-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_124.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_125.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_125.yaml new file mode 100644 index 0000000000000000000000000000000000000000..2de688a2f003c9adc7f2017c39dd77ef23187aa4 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_125.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-125 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-125 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-125-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_125.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_128.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_128.yaml new file mode 100644 index 0000000000000000000000000000000000000000..18a00ac5de2951dd14a0d23720828dcce1e354d2 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_128.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-128 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-128 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-128-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_128.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_129.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_129.yaml new file mode 100644 index 0000000000000000000000000000000000000000..bbb5be658a49af8f6a83e128bd4b68e9c6962641 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_129.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-129 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-129 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-129-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_129.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_130.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_130.yaml new file mode 100644 index 0000000000000000000000000000000000000000..9b49f509da63aa85ee3d4f988997c14dd571145b --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_130.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-130 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-130 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-130-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_130.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_131.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_131.yaml new file mode 100644 index 0000000000000000000000000000000000000000..0e4355b2f9b4ccd8c416da66dcda01593f8a7781 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_131.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-131 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-131 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-131-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_131.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_132.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_132.yaml new file mode 100644 index 0000000000000000000000000000000000000000..674c2a8cf0e61a564ae25c58895d9e4e0bdf0e53 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_132.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-132 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-132 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-132-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_132.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_133.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_133.yaml new file mode 100644 index 0000000000000000000000000000000000000000..7db62ef7e848a4ad8f249c1fcd866a2247379839 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_133.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-133 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-133 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-133-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_133.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_134.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_134.yaml new file mode 100644 index 0000000000000000000000000000000000000000..ddb79847a862f0a428e9c02701ac44d867871fec --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_134.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-134 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-134 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-134-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_134.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_135.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_135.yaml new file mode 100644 index 0000000000000000000000000000000000000000..1ad11850dc6fde6ca1e8a90675d4a0701f1152d0 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_135.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-135 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-135 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-135-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_135.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_136.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_136.yaml new file mode 100644 index 0000000000000000000000000000000000000000..a16ef892f53ffcde6fb7b9f0942e3e30afc0e0ae --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_136.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-136 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-136 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-136-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_136.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_137.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_137.yaml new file mode 100644 index 0000000000000000000000000000000000000000..7ae54794b13aa9929998bac98417a7ff635273c2 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_137.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-137 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-137 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-137-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_137.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_138.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_138.yaml new file mode 100644 index 0000000000000000000000000000000000000000..462a9eea50d664611b6dd14c6063bef9a496d51b --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_138.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-138 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-138 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-138-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_138.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_139.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_139.yaml new file mode 100644 index 0000000000000000000000000000000000000000..b20ec18322261c04b324d5c24d9446b46f6508fd --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_139.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-139 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-139 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-139-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_139.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_141.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_141.yaml new file mode 100644 index 0000000000000000000000000000000000000000..bca693691816386d02e133db917706b9e2f70918 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_141.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-141 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-141 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-141-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_141.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_142.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_142.yaml new file mode 100644 index 0000000000000000000000000000000000000000..b6563b2491dd1a971c95eb2650b976078e676754 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_142.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-142 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-142 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-142-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_142.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_143.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_143.yaml new file mode 100644 index 0000000000000000000000000000000000000000..928c2eb8651130702d8eda7ea9cd4b445e169655 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_143.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-143 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-143 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-143-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_143.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_145.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_145.yaml new file mode 100644 index 0000000000000000000000000000000000000000..9d8a1d1337290ac86e61cc30cc1217d656dd4201 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_145.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-145 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-145 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-145-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_145.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_147.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_147.yaml new file mode 100644 index 0000000000000000000000000000000000000000..7f5e80c3cb007e5cf1b7ae0815465ca30c22044a --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_147.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-147 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-147 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-147-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_147.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_148.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_148.yaml new file mode 100644 index 0000000000000000000000000000000000000000..ded5faae3d0fb3aa3a9b63a5f48124b32454d0a8 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_148.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-148 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-148 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-148-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_148.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_149.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_149.yaml new file mode 100644 index 0000000000000000000000000000000000000000..a49c7dd0cb6e151037a34a17527a462d00611893 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_149.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-149 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-149 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-149-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_149.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_150.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_150.yaml new file mode 100644 index 0000000000000000000000000000000000000000..d7183ae0965caef0c4c36f7bedabcd7e24ca0a8a --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_150.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-150 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-150 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-150-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_150.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_151.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_151.yaml new file mode 100644 index 0000000000000000000000000000000000000000..43e643508e22b2209101732901b40648c5254a34 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_151.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-151 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-151 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-151-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_151.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_152.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_152.yaml new file mode 100644 index 0000000000000000000000000000000000000000..af11134f02daf85a7933a87177ab0398cc5defb7 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_152.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-152 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-152 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-152-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_152.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_153.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_153.yaml new file mode 100644 index 0000000000000000000000000000000000000000..1db6a3997474e6737054e82e60ed9e6f22655ecd --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_153.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-153 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-153 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-153-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_153.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_155.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_155.yaml new file mode 100644 index 0000000000000000000000000000000000000000..43c48800f219e53ade21e23704b2583b454c9069 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_155.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-155 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-155 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-155-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_155.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_156.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_156.yaml new file mode 100644 index 0000000000000000000000000000000000000000..ec3db2bef7b04dc4c288b6e613a12af602302220 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_156.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-156 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-156 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-156-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_156.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_157.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_157.yaml new file mode 100644 index 0000000000000000000000000000000000000000..c33165f029346aefadf54d35b1d1438508d30033 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_157.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-157 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-157 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-157-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_157.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_158.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_158.yaml new file mode 100644 index 0000000000000000000000000000000000000000..c726525b290affea3fc9a56540f54b4d2e09ca0b --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_158.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-158 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-158 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-158-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_158.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_159.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_159.yaml new file mode 100644 index 0000000000000000000000000000000000000000..81c84105247be4144e010c7453a449b6d9733579 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_159.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-159 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-159 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-159-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_159.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_160.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_160.yaml new file mode 100644 index 0000000000000000000000000000000000000000..f519cb9d6c7c8bdf71aef5598977e5f54feeeacd --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_160.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-160 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-160 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-160-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_160.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_162.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_162.yaml new file mode 100644 index 0000000000000000000000000000000000000000..95de1cdf68325b711e466d76a0b75e58b25379f6 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_162.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-162 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-162 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-162-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_162.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_163.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_163.yaml new file mode 100644 index 0000000000000000000000000000000000000000..f3dda66fadd8aeb930f408808c5873426a6cd2b4 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_163.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-163 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-163 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-163-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_163.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_165.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_165.yaml new file mode 100644 index 0000000000000000000000000000000000000000..301ecbafc13c6063135ed1094f64e01875e337b8 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_165.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-165 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-165 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-165-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_165.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_166.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_166.yaml new file mode 100644 index 0000000000000000000000000000000000000000..fea8de88ef2d3b993912ec346c06ea6346a2b7d5 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_166.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-166 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-166 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-166-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_166.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_168.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_168.yaml new file mode 100644 index 0000000000000000000000000000000000000000..e00a99fc7d3d037f6952c05842046ae1d11b7b0b --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_168.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-168 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-168 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-168-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_168.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_169.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_169.yaml new file mode 100644 index 0000000000000000000000000000000000000000..fc2af2ed17322fa84f6c5a4a106cf24fb4fc9bba --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_169.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-169 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-169 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-169-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_169.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_170.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_170.yaml new file mode 100644 index 0000000000000000000000000000000000000000..3d87c2d218e3f4248d45c297d62580cc51d6c2e8 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_170.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-170 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-170 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-170-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_170.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_171.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_171.yaml new file mode 100644 index 0000000000000000000000000000000000000000..9cf621bed2ec4c54be8f950955e1c864f164117e --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_171.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-171 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-171 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-171-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_171.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_173.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_173.yaml new file mode 100644 index 0000000000000000000000000000000000000000..5508200fa31eb38b7a80e7d9511cba28d1b7c215 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_173.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-173 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-173 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-173-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_173.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_174.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_174.yaml new file mode 100644 index 0000000000000000000000000000000000000000..efc25ec044a9a7aa9192780fcbf1011fb2b68331 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_174.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-174 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-174 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-174-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_174.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_175.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_175.yaml new file mode 100644 index 0000000000000000000000000000000000000000..f0c158f8b4d295737013bdb9b25ab928b37dbadc --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_175.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-175 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-175 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-175-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_175.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_179.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_179.yaml new file mode 100644 index 0000000000000000000000000000000000000000..b941c54ffbe60fdd4ba90a922df5469e29804b82 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_179.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-179 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-179 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-179-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_179.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_182.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_182.yaml new file mode 100644 index 0000000000000000000000000000000000000000..828fc1307249710677a0b435ea5085c976f68a9b --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_182.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-182 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-182 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-182-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_182.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_184.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_184.yaml new file mode 100644 index 0000000000000000000000000000000000000000..179a314da4bec940ab60ece0ddf4fb40639dc724 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_184.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-184 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-184 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-184-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_184.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_185.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_185.yaml new file mode 100644 index 0000000000000000000000000000000000000000..68401733570ba28f717591347495cd6941b97c46 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_185.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-185 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-185 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-185-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_185.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_186.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_186.yaml new file mode 100644 index 0000000000000000000000000000000000000000..70f26f9756b8fb60e42d065650f11b2b98fd09e3 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_186.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-186 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-186 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-186-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_186.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_187.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_187.yaml new file mode 100644 index 0000000000000000000000000000000000000000..081b6a6bb264a65165a5203fa500be15cbc79862 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_187.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-187 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-187 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-187-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_187.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_188.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_188.yaml new file mode 100644 index 0000000000000000000000000000000000000000..57728cb0d5942ec36c6c735bdbb5e6d1047479f8 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_188.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-188 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-188 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-188-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_188.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_190.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_190.yaml new file mode 100644 index 0000000000000000000000000000000000000000..470ab1b6ed18fadb0e4452cf7674f2f06dc8cbaf --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_190.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-190 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-190 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-190-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_190.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_191.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_191.yaml new file mode 100644 index 0000000000000000000000000000000000000000..3f12c2bfd2493f01049e601edcd8a42c7cd59166 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_191.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-191 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-191 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-191-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_191.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_194.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_194.yaml new file mode 100644 index 0000000000000000000000000000000000000000..533015fba09a612310238b8b82d1f5c236e3c23f --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_194.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-194 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-194 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-194-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_194.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_195.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_195.yaml new file mode 100644 index 0000000000000000000000000000000000000000..f1ff460404bd8ce8e104402b665c4d4162622907 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_195.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-195 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-195 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-195-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_195.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_196.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_196.yaml new file mode 100644 index 0000000000000000000000000000000000000000..9d6d0fe1a9979571cd0ed0b4754196e52ebd6e33 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_196.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-196 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-196 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-196-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_196.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_197.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_197.yaml new file mode 100644 index 0000000000000000000000000000000000000000..5be32c6f1fde4b3ca230c91d5bbf73995c5d7038 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_197.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-197 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-197 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-197-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_197.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_198.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_198.yaml new file mode 100644 index 0000000000000000000000000000000000000000..53ee645c8400489bf306978a5e7d56e4c3912488 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_198.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-198 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-198 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-198-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_198.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock diff --git a/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_199.yaml b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_199.yaml new file mode 100644 index 0000000000000000000000000000000000000000..38bf1a14c3c347027528104ea8374733af7088d3 --- /dev/null +++ b/hn_eng_bn/tok_en-bn-hn/tok_jobs_2/tokenizer_199.yaml @@ -0,0 +1,104 @@ +apiVersion: kubeflow.org/v2beta1 +kind: MPIJob +metadata: + name: peacock-tokenizer-en-bn-hn-199 + namespace: peacock + labels: + app: peacock-tokenizer-en-bn-hn-199 +spec: + slotsPerWorker: 8 + runPolicy: + backoffLimit: 1 + cleanPodPolicy: Running + mpiReplicaSpecs: + Launcher: + replicas: 1 + template: + spec: + hostIPC: true + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-tokenizer-en-bn-hn-199-launcher + imagePullPolicy: IfNotPresent + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock + command: ["/bin/bash", "-c"] + args: + - >- + + export SYNAPSE_VERSION="1.15.1"; + export WORKER_DIR="/mnt/weka/peacock/experiments/llama"; + + export MEGATRON_SETUP_CMD="$WORKER_DIR/launch/setup.sh" + export TOKENIZER_CMD=/mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_199.sh + HOSTSFILE=$OMPI_MCA_orte_default_hostfile; + echo "HOSTSFILE=$HOSTSFILE"; + MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; + NUM_NODES=$(wc -l < $HOSTSFILE); + CARDS_PER_NODE=8; + N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); + echo "MPI_ROOT=$MPI_ROOT"; + echo "N_CARDS=$N_CARDS"; + echo "MASTER_ADDR=$MASTER_ADDR"; + sleep 20; + + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $MEGATRON_SETUP_CMD; + + mpirun -np $N_CARDS -npernode 8 \ + --tag-output \ + --allow-run-as-root \ + --prefix $MPI_ROOT \ + -x WORKER_DIR=$WORKER_DIR \ + -x SYNAPSE_VERSION=$SYNAPSE_VERSION $TOKENIZER_CMD; + + + Worker: + replicas: 1 + template: + spec: + volumes: + - name: work-dir + persistentVolumeClaim: + claimName: peacock-fs-pvc + tolerations: + - key: "habana.ai/gaudi" + operator: "Exists" + effect: "NoSchedule" + - key: "k8s/namespace" + operator: "Equal" + value: "peacock" + effect: "NoSchedule" + hostIPC: true + containers: + - image: vault.habana.ai/gaudi-docker/1.15.1/ubuntu22.04/habanalabs/pytorch-installer-2.2.0:latest + name: peacock-llama-worker + imagePullPolicy: IfNotPresent + securityContext: + capabilities: + add: + - SYSLOG + resources: + limits: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + requests: + habana.ai/gaudi: 8 + hugepages-2Mi: 300Gi + memory: 700Gi + cpu: 150 + volumeMounts: + - name: work-dir + mountPath: /mnt/weka/peacock