Add new versions of pretraining and fine-tuning

tutorials-and-examples/nvidia-bionemo/README.md

@@ -2,8 +2,8 @@
 
 #### Pretraining ESM-2
 
-- [Pretraining ESM-2 LLM on GKE using BioNeMo Framework 2.0](./esm2/README.md#pretraining)
+- [Pretraining ESM-2 LLM on GKE using BioNeMo Framework 2.0](./pretraining/README.md)
 
-#### Fine-turning ESM-2
+#### Fine-tuning ESM-2
 
-- [Fine-tuning ESM-2 LLM on GKE using BioNeMo Framework 2.0](./esm2/README.md#fine-tuning)
+- [Fine-tuning ESM-2 LLM on GKE using BioNeMo Framework 2.0](./fine-tuning/README.md)

tutorials-and-examples/nvidia-bionemo/base/kustomization.yaml

@@ -0,0 +1,9 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- namespace.yaml
+- storage/storage-class.yaml
+- storage/pvcs.yaml
+- monitoring/tensorboard-deployment.yaml
+- monitoring/tensorboard-service.yaml
+- monitoring/rbac.yaml

tutorials-and-examples/nvidia-bionemo/base/monitoring/rbac.yaml

@@ -0,0 +1,29 @@
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: tensorboard-sa
+  namespace: bionemo-training
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: job-reader
+  namespace: bionemo-training
+rules:
+- apiGroups: ["batch"]
+  resources: ["jobs"]
+  verbs: ["get", "list", "watch"]
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: tensorboard-job-reader
+  namespace: bionemo-training
+subjects:
+- kind: ServiceAccount
+  name: tensorboard-sa
+  namespace: bionemo-training
+roleRef:
+  kind: Role
+  name: job-reader
+  apiGroup: rbac.authorization.k8s.io

tutorials-and-examples/nvidia-bionemo/base/monitoring/tensorboard-deployment.yaml

@@ -0,0 +1,33 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: tensorboard
+  namespace: bionemo-training
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: tensorboard
+  template:
+    metadata:
+      labels:
+        app: tensorboard
+    spec:
+      containers:
+      - name: tensorboard
+        image: tensorflow/tensorflow:latest
+        command:
+        - tensorboard
+        args:
+        - --logdir=/workspace/bionemo2/results/lightning_logs
+        - --port=6006
+        volumeMounts:
+        - name: bionemo-storage
+          mountPath: /workspace/bionemo2/results
+          subPath: tensorboard-logs
+          readOnly: true
+      volumes:
+      - name: bionemo-storage
+        persistentVolumeClaim:
+          claimName: bionemo-filestore
+      serviceAccountName: tensorboard-sa

tutorials-and-examples/nvidia-bionemo/base/monitoring/tensorboard-service.yaml

@@ -0,0 +1,12 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: tensorboard-service
+  namespace: bionemo-training
+spec:
+  selector:
+    app: tensorboard
+  ports:
+  - port: 6006
+    targetPort: 6006
+  type: ClusterIP

tutorials-and-examples/nvidia-bionemo/base/namespace.yaml

@@ -0,0 +1,4 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: bionemo-training

tutorials-and-examples/nvidia-bionemo/base/storage/kustomization.yaml

@@ -0,0 +1,5 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- storage-class.yaml
+- pvcs.yaml

tutorials-and-examples/nvidia-bionemo/base/storage/pvcs.yaml

@@ -0,0 +1,12 @@
+kind: PersistentVolumeClaim
+apiVersion: v1
+metadata:
+  name: bionemo-filestore
+  namespace: bionemo-training
+spec:
+  accessModes:
+  - ReadWriteMany
+  storageClassName: filestore-storage
+  resources:
+    requests:
+      storage: 1Ti

tutorials-and-examples/nvidia-bionemo/base/storage/storage-class.yaml

@@ -0,0 +1,10 @@
+apiVersion: storage.k8s.io/v1
+kind: StorageClass
+metadata:
+  name: filestore-storage
+provisioner: filestore.csi.storage.gke.io
+volumeBindingMode: Immediate
+allowVolumeExpansion: true
+parameters:
+  tier: BASIC_HDD
+  network: default