分支: brucemacd/install-path-clean

api

bmizerany/client-registry

bmizerany/client2resume

bmizerany/embedspeedup

bmizerany/fastverify

bmizerany/filepathnobuild

bmizerany/filepathwithcoloninhost

bmizerany/grammar

bmizerany/hrm

bmizerany/modenameenforcealphanum

bmizerany/nameswork

bmizerany/noseek

bmizerany/nosillyggufslurps

bmizerany/replacecolon

bmizerany/types/model/defaultfix

bmizerany/validatenames

bmizerany/x

bruce/iq-quants

brucemacd/allow-ollama

brucemacd/browser-key-register

brucemacd/check-key-register

brucemacd/check-key-register-structured-err

brucemacd/convert-cli

brucemacd/ctx-shift-err

brucemacd/doc-go-engine

brucemacd/done-reason

brucemacd/err-hint

brucemacd/err-no-vocab

brucemacd/forward-test

brucemacd/go_qwen2

brucemacd/install-path-clean

brucemacd/jomorganca/mistral

brucemacd/llama-mem-calc

brucemacd/logprobs

brucemacd/mistral

brucemacd/mistral-small-convert

brucemacd/new_runner_e2e

brucemacd/new_runner_graph_bench

brucemacd/new_runner_qwen2

brucemacd/next-bpe-bench

brucemacd/next-bpe-test

brucemacd/parallel-embed-models

brucemacd/push-name-validation

brucemacd/qwen2_5

brucemacd/rope-config

brucemacd/runner-completion

brucemacd/shim-grammar

brucemacd/tokenize

build_dist

cgo

cp-model

cuda-search

delete-fix

deletemodels

dhiltgen/remove_submodule

distribution

editor

fix-model-names

fix-unknown-model

format-config

go-opts

insecure-registry

jessegross/sample

jessegross/semaphore

jmorgan/sample-fix-sorting-extras

jmorganca/add-missing-symlink-eval

jmorganca/batch-embeddings

jmorganca/degin-1

jmorganca/done-reason

jmorganca/enable-fa

jmorganca/execstack

jmorganca/faster-releases

jmorganca/fix-gguf-error

jmorganca/fix-null-format

jmorganca/fix-proxy

jmorganca/ga

jmorganca/ggml-static

jmorganca/if-none-match

jmorganca/initcmake

jmorganca/limit

jmorganca/llama-bump

jmorganca/llama-cpp-7c26775

jmorganca/llama-cpp-8960fe8

jmorganca/llama-vit

jmorganca/mistral

jmorganca/mistral-wip

jmorganca/mistral3

jmorganca/mllama

jmorganca/mm

jmorganca/native

jmorganca/no-concat

jmorganca/no-error-template

jmorganca/openai-context

jmorganca/openai-fix-first-message

jmorganca/options

jmorganca/qwen2vl

jmorganca/replace-assets

jmorganca/temp-0-images

jmorganca/template-mistral

jmorganca/testing

jmorganca/vendor-081b29bd

jyan/auth

jyan/convert-prog

jyan/format

jyan/local

jyan/local2

jyan/ollama-v

jyan/p2

jyan/paligemma

jyan/palitest

jyan/parse-temp

jyan/progress

jyan/q4_4/8

jyan/quant3

jyan/quant4

jyan/quant5

jyan/reord-g

jyan/v0.146

language_support

license-layers

list-models

main

matt/examplemodelfiles

matt/streamingapi

mattw/airenamer

mattw/allmodelsonhuggingface

mattw/communitylinks

mattw/faq-context

mattw/howtoquant

mattw/noprune

mattw/python-functioncalling

mattw/quantcontext

mattw/selfqueryingretrieval

mattw/whatneedstorun

modelfile-readme

modelpath

modenameenforcealphanum

mxyng/api-models

mxyng/cmd-history

mxyng/create-context

mxyng/environ-2

mxyng/extra-args

mxyng/fix-memory

mxyng/fs-config

mxyng/func-checks

mxyng/gin-slog

mxyng/install

mxyng/layers-from-files

mxyng/mllama

mxyng/modelname-6

mxyng/modelname-7

mxyng/next-bert

mxyng/next-debug

mxyng/next-mlx

mxyng/no-deprecated-gpu-targets

mxyng/server-timestamp

mxyng/split-bin

mxyng/tune-concurrency

mxyng/update-registry-domain

native

nogogen

ollama.com

paligemma-support

parth/cmd-cleanup-SO

parth/constrained-sampling-json

parth/disallow-streaming-tools

parth/fix-default-to-warn-json

parth/fix-referencing-so

parth/log-probs

parth/openai-stream-usage

parth/sample-correctness-fix

parth/sample-fix-sorting

parth/sample-unmarshal-json-for-params

parth/sampling-structured-outputs

parth/set-context-size-openai

parth/templating

parth/tokenize-detokenize

pdevine/bfloat16

pdevine/convert-cohere2

pdevine/fix-template

pdevine/geems-2b

pdevine/gemma2

pdevine/ggla

pdevine/import-docs

pdevine/logging

pdevine/newlines

pdevine/ps-glitches

pdevine/showggmlinfo

progress-flicker

progressbar

pulse

readme-updates

remove-first

rename

revert-5963-revert-5924-mxyng/llama3.1-rope

rmdisplaylong

roy-embed-parallel

royh-embed-parallel

royh-imgembed

royh-ls

royh-name

royh-openai-delete

royh-openai-suffixdocs

royh-params

royh-precision

royh-show-rigid

royh-testdelete

royh/embed-viz

royh/ep-methods

royh/stream-tools

royh/whisper

scratch

shell

skip-list

stream-tools-stop

timeout

update-nous-hermes

upgrade-all

upload-progress

whitespace-detection

Mélony QIN 3f71ba406a Correct the kubernetes terminology (#3843)		1 年之前
..
README.md	3f71ba406a Correct the kubernetes terminology (#3843)	1 年之前
cpu.yaml	dccac8c8fa k8s example	1 年之前
gpu.yaml	0409c1fa59 docker: set PATH, LD_LIBRARY_PATH, and capabilities (#1336)	1 年之前

Deploy Ollama to Kubernetes

Prerequisites

Ollama: https://ollama.com/download
Kubernetes cluster. This example will use Google Kubernetes Engine.

Steps

Create the Ollama namespace, deployment, and service
```
kubectl apply -f cpu.yaml
```

(Optional) Hardware Acceleration

Hardware acceleration in Kubernetes requires NVIDIA's k8s-device-plugin which is deployed in Kubernetes in form of daemonset. Follow the link for more details.

Once configured, create a GPU enabled Ollama deployment.

kubectl apply -f gpu.yaml

Test

Port forward the Ollama service to connect and use it locally
```
kubectl -n ollama port-forward service/ollama 11434:80
```
Pull and run a model, for example orca-mini:3b
```
ollama run orca-mini:3b
```

README.md

Deploy Ollama to Kubernetes

Prerequisites

Steps

(Optional) Hardware Acceleration

Test