feat(fabulinus) add fabulinus (#16909)

**Description**
Fabulinus is the server component of the TitanML Takeoff server.
⚒️ Fixes  # <!--(issue)-->

**⚙️ Type of change**

- [X] ⚙️ Feature/App addition
- [ ] 🪛 Bugfix
- [ ] ⚠️ Breaking change (fix or feature that would cause existing
functionality to not work as expected)
- [ ] 🔃 Refactor of current code

**🧪 How Has This Been Tested?**
<!--
Please describe the tests that you ran to verify your changes. Provide
instructions so we can reproduce. Please also list any relevant details
for your test configuration
-->

**📃 Notes:**
<!-- Please enter any other relevant information here -->

**✔️ Checklist:**

- [X] ⚖️ My code follows the style guidelines of this project
- [X] 👀 I have performed a self-review of my own code
- [ ] #️⃣ I have commented my code, particularly in hard-to-understand
areas
- [ ] 📄 I have made corresponding changes to the documentation
- [ ] ⚠️ My changes generate no new warnings
- [ ] 🧪 I have added tests to this description that prove my fix is
effective or that my feature works
- [ ] ⬆️ I increased versions for any altered app according to semantic
versioning

** App addition**

If this PR is an app addition please make sure you have done the
following.

- [ ] 🪞 I have opened a PR on
[truecharts/containers](https://github.com/truecharts/containers) adding
the container to TrueCharts mirror repo.
- [X] 🖼️ I have added an icon in the Chart's root directory called
`icon.png`

---

_Please don't blindly check all the boxes. Read them and only check
those that apply.
Those checkboxes are there for the reviewer to see what is this all
about and
the status of this PR with a quick glance._
This commit is contained in:
Xstar97TheNoob 2024-01-09 13:13:43 -05:00 committed by GitHub
parent 2803de1406
commit 478fb59e26
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
12 changed files with 2974 additions and 0 deletions

View File

@ -0,0 +1,30 @@
# Patterns to ignore when building packages.
# This supports shell glob matching, relative path matching, and
# negation (prefixed with !). Only one pattern per line.
.DS_Store
# Common VCS dirs
.git/
.gitignore
.bzr/
.bzrignore
.hg/
.hgignore
.svn/
# Common backup files
*.swp
*.bak
*.tmp
*~
# Various IDEs
.project
.idea/
*.tmproj
.vscode/
# OWNERS file for Kubernetes
OWNERS
# helm-docs templates
*.gotmpl
# docs folder
/docs
# icon
icon.png

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,37 @@
kubeVersion: ">=1.24.0-0"
apiVersion: v2
name: fabulinus
version: 0.0.1
appVersion: latest
description: Fabulinus is the server component of the TitanML Takeoff server.
home: https://truecharts.org/charts/dev/fabulinus
icon: https://truecharts.org/img/hotlink-ok/chart-icons/fabulinus.png
deprecated: false
sources:
- https://github.com/truecharts/charts/tree/master/charts/dev/fabulinus
- https://hub.docker.com/r/tytn/fabulinus
- https://github.com/titanml/takeoff-communitys
maintainers:
- name: TrueCharts
email: info@truecharts.org
url: https://truecharts.org
keywords:
- fabulinus
- machine-learning
dependencies:
- name: common
version: 17.2.21
repository: oci://tccr.io/truecharts
condition: ""
alias: ""
tags: []
import-values: []
annotations:
max_scale_version: 23.10.2
min_scale_version: 23.10.0
truecharts.org/SCALE-support: "true"
truecharts.org/category: ai
truecharts.org/max_helm_version: "3.13"
truecharts.org/min_helm_version: "3.12"
truecharts.org/train: dev
type: application

View File

@ -0,0 +1,28 @@
---
title: README
---
## General Info
TrueCharts can be installed as both _normal_ Helm Charts or as Apps on TrueNAS SCALE.
However only installations using the TrueNAS SCALE Apps system are supported.
For more information about this App, please check the docs on the TrueCharts [website](https://truecharts.org/charts/stable/jackett)
**This chart is not maintained by the upstream project and any issues with the chart should be raised [here](https://github.com/truecharts/charts/issues/new/choose)**
## Support
- Please check our [quick-start guides for TrueNAS SCALE](https://truecharts.org/manual/SCALE/guides/scale-intro).
- See the [Website](https://truecharts.org)
- Check our [Discord](https://discord.gg/tVsPTHWTtr)
- Open a [issue](https://github.com/truecharts/charts/issues/new/choose)
---
## Sponsor TrueCharts
TrueCharts can only exist due to the incredible effort of our staff.
Please consider making a [donation](https://truecharts.org/sponsor) or contributing back to the project any way you can!
_All Rights Reserved - The TrueCharts Project_

View File

@ -0,0 +1 @@
imageSelector: image

View File

@ -0,0 +1 @@
imageSelector: gpuImage

View File

@ -0,0 +1,17 @@
---
title: Fabulinus Installation Notes
---
Set the `Device` to **CPU** or **GPU** depending on the image selected.
Set the `Model` to a supported model from the following [list](https://docs.titanml.co/docs/titan-takeoff/experimentation/supported-models).
Set the `Quant Type` to the correct option when using a particular model.
Set the `Max Batch Size` to the maximum batch size a model can use.
::: note
If you are not using the batching endpoint, /generate, it is best practice to set `Disable Batching` to true. The batching service can interfere with the streaming service, and make it slower.
:::

Binary file not shown.

After

Width:  |  Height:  |  Size: 474 KiB

View File

@ -0,0 +1,160 @@
# Include{groups}
portals:
open:
# Include{portalLink}
path: "/docs"
chat:
# Include{portalLink}
path: "/demos/chat"
playground:
# Include{portalLink}
path: "/demos/playground"
questions:
# Include{global}
# Include{workload}
# Include{workloadDeployment}
# Include{replicas1}
# Include{podSpec}
# Include{containerMain}
- variable: imageSelector
label: Select Image
schema:
type: string
default: image
enum:
- value: image
description: CPU
- value: gpuImage
description: GPU
# Include{containerBasic}
# Include{containerAdvanced}
- variable: fabulinus
group: App Configuration
label: Fabulinus Configuration
schema:
additional_attrs: true
type: dict
attrs:
- variable: device
label: Device
description: The device to use.
schema:
type: string
default: cpu
enum:
- value: cpu
description: CPU
- value: gpu
description: GPU
- variable: model
label: Model
description: The model name to use.
schema:
type: string
default: "google/flan-t5-small"
required: true
- variable: quant_type
label: Quant Type
description: The quantization type to use.
schema:
type: string
default: int8
enum:
- value: int8
description: Int8
- value: float16
description: Float16
- value: bfloat16
description: Bfloat16
- value: int8_float16
description: Int8_float16
- value: int8_bfloat16
description: Int8_bfloat16
- variable: max_batch_size
label: Max Batch Size
description: The maximum batch size the model can use.
schema:
type: int
default: 32
min: 1
- variable: disable_batching
label: Disable Batching
description: Whether to allow batching or not. Set to False if not using the /generate end point.
schema:
type: boolean
default: true
# Include{containerConfig}
# Include{podOptions}
# Include{serviceRoot}
# Include{serviceMain}
# Include{serviceSelectorLoadBalancer}
# Include{serviceSelectorExtras}
- variable: main
label: "Main Service Port Configuration"
schema:
additional_attrs: true
type: dict
attrs:
- variable: port
label: "Port"
description: "This port exposes the container port on the service"
schema:
type: int
default: 10687
required: true
# Include{serviceExpertRoot}
# Include{serviceExpert}
# Include{serviceList}
# Include{persistenceRoot}
- variable: models
label: "App Models Storage"
description: "Stores the Application Models."
schema:
additional_attrs: true
type: dict
attrs:
# Include{persistenceBasic}
# Include{persistenceList}
# Include{ingressRoot}
- variable: main
label: "Main Ingress"
schema:
additional_attrs: true
type: dict
attrs:
# Include{ingressDefault}
# Include{ingressAdvanced}
# Include{ingressList}
# Include{securityContextRoot}
- variable: runAsUser
label: "runAsUser"
description: "The UserID of the user running the application"
schema:
type: int
default: 0
- variable: runAsGroup
label: "runAsGroup"
description: "The groupID of the user running the application"
schema:
type: int
default: 0
# Include{securityContextContainer}
# Include{securityContextAdvanced}
# Include{securityContextPod}
- variable: fsGroup
label: "fsGroup"
description: "The group that should own ALL storage."
schema:
type: int
default: 568
# Include{resources}
# Include{advanced}
# Include{addons}
# Include{codeserver}
# Include{netshoot}
# Include{vpn}
# Include{documentation}

View File

@ -0,0 +1 @@
{{- include "tc.v1.common.lib.chart.notes" $ -}}

View File

@ -0,0 +1 @@
{{ include "tc.v1.common.loader.all" . }}

View File

@ -0,0 +1,65 @@
image:
repository: tytn/fabulinus
pullPolicy: IfNotPresent
tag: latest-cpu@sha256:f8b30eaa5b61d6085fc715f40619733846927c2266a3e20523e93fb58afeef38
gpuImage:
repository: tytn/fabulinus
pullPolicy: IfNotPresent
tag: latest-gpu@sha256:9a826a578ca4157fbe3c091eb78aa35dfca6675f9fae24dcb90495ba4d9715d1
securityContext:
container:
readOnlyRootFilesystem: false
runAsUser: 0
runAsGroup: 0
service:
main:
ports:
main:
protocol: http
targetPort: 80
port: 10687
fabulinus:
# cpu | gpu
device: cpu
model: "google/flan-t5-small"
# int8 | float16 | bfloat16 | int8_float16 | int8_bfloat16
quant_type: int8
max_batch_size: 32
disable_batching: true
workload:
main:
podSpec:
containers:
main:
imageSelector: image
probes:
liveness:
enabled: true
type: http
path: /docs
readiness:
enabled: true
type: http
path: /docs
startup:
enabled: true
type: tcp
env:
TAKEOFF_DEVICE: "{{ .Values.fabulinus.device }}"
TAKEOFF_MODEL_NAME: "{{ .Values.fabulinus.model }}"
TAKEOFF_QUANT_TYPE: "{{ .Values.fabulinus.quant_type }}"
TAKEOFF_MAX_BATCH_SIZE: "{{ .Values.fabulinus.max_batch_size }}"
TAKEOFF_DISABLE_BATCHING: "{{ .Values.fabulinus.disable_batching }}"
persistence:
models:
enabled: true
mountPath: "/code/models"
portal:
open:
enabled: true