Post

Ollama Installation

Let's install Ollama on Windows and Ubuntu

Posted
By DS2Man
11 min read
Ollama Installation

Ollama is an open-source tool designed to easily run and utilize large language models (LLMs) in a local environment. Typically, running LLMs requires complex setup and high system resources, but Ollama simplifies this process, making it accessible for anyone.
I think Ollama is perfect for those who relate to any of the following scenarios.

  • Your company has restricted the use of ChatGPT due to security concerns.
  • You want to run an LLM on your own computer but don’t know how to get started.

For reference, since Deepseek became a hot topic after the Ollama post, I updated this article on February 20. The list of LLMs is as follows. You can check the full list on Ollama GitHub.

Install Ollama

Ollama has been briefly mentioned in previous posts. Feel free to check them out.
It’s very simple.
(Option) Ollama on Ubuntu

Install Ollama on Windows

I wrote this based on Ollama_Windows to install the Ollama.
It’s very simple. Just download the file(OllamaSetup.exe) and install!

Install Ollama on Ubuntu(WSL)

I wrote this based on Ollama_Linux to install the Ollama.
It’s very simple. Just run the command below, and you’re done!

  • curl -fsSL https://ollama.com/install.sh | sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

(base) jaoneol@DESKTOP-B7GM3C5:~$ curl -fsSL https://ollama.com/install.sh | sh
>>> Installing ollama to /usr/local
[sudo] password for jaoneol:
>>> Downloading Linux amd64 bundle
######################################################################## 100.0%
>>> Creating ollama user...
>>> Adding ollama user to render group...
>>> Adding ollama user to video group...
>>> Adding current user to ollama group...
>>> Creating ollama systemd service...
>>> Enabling and starting ollama service...
Created symlink /etc/systemd/system/default.target.wants/ollama.service → /etc/systemd/system/ollama.service.
>>> Nvidia GPU detected.
>>> The Ollama API is now available at 127.0.0.1:11434.
>>> Install complete. Run "ollama" from the command line.
(base) jaoneol@DESKTOP-B7GM3C5:~$

Summary

CategoryWindowsUnbuntu(WSL)
InstallOllamaSetup.execurl -fsSL https://ollama.com/install.sh | sh
File LocationC:\Users\ycjang.ollama\models/usr/share/ollama/.ollama/models

How to use Ollama

You can download the LLM model you want to use by running the ollama pull or ollama run command from Ollama Home or Ollama GitHub. You can check the downloaded models using the ollama list command.

  • 1.Use the following command to download the LLM model. If you want to run it immediately, use run instead of pull.
    ollama pull gemma3
    For reference, the location where the model is stored is as follows.
    Windows : C:\Users\ycjang\.ollama\models\
    Ubuntu(WSL) : /usr/share/ollama/.ollama/models/
  • 2.You can check the LLM models managed by Ollama using the following command.
    ollama list
  • 3.The command to delete an LLM model is as follows, and you can verify that the files in C:\Users\ycjang\.ollama\models\ and /usr/share/ollama/.ollama/models/ have been removed.
    ollama rm gemma3

on Windows

1
2
3
4
5
6
7
8
9
10
11

# I will try downloading gemma3:latest.
PS C:\Users\ycjang> ollama pull gemma3
pulling manifest
pulling 377655e65351... 100% ▕███████████████████████████████████████▏ 3.3 GB
pulling e0a42594d802... 100% ▕███████████████████████████████████████▏  358 B
pulling dd084c7d92a3... 100% ▕███████████████████████████████████████▏ 8.4 KB
pulling 0a74a8735bf3... 100% ▕███████████████████████████████████████    55 B
pulling ffae984acbea... 100% ▕███████████████████████████████████████▏  489 B
verifying sha256 digest
writing manifest
success

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56

# I check the LLM models managed by Ollama using the following command.
PS C:\Users\ycjang> ollama list
NAME             ID              SIZE      MODIFIED
gemma3:latest    c0494fe00251    3.3 GB    44 seconds ago

# the location where the model is stored is as follows. 
# `C:\Users\ycjang\.ollama\models`
PS C:\Users\ycjang> cd .ollama
PS C:\Users\ycjang\.ollama> ls

    디렉터리: C:\Users\ycjang\.ollama

Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
d-----      2025-03-21   오전 8:03                models
-a----      2025-03-19  오후 10:35            627 history
-a----      2025-01-11   오후 5:01            387 id_ed25519
-a----      2025-01-11   오후 5:01             81 id_ed25519.pub

PS C:\Users\ycjang\.ollama> cd models
PS C:\Users\ycjang\.ollama\models> ls

    디렉터리: C:\Users\ycjang\.ollama\models

Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
d-----      2025-03-21   오전 8:04                blobs
d-----      2025-03-21   오전 8:04                manifests

#################
# Note.Image files are stored in the `blobs` directory.
#################
PS C:\Users\ycjang\.ollama\models\blobs> ls

    디렉터리: C:\Users\ycjang\.ollama\models\blobs

Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
-a----      2025-03-21   오전 8:04             55 sha256-0a74a8735bf3ffff4537b6c6bc9a4bc97a28c48f2fd347e806cca4d5001560f6
-a----      2025-03-21   오전 8:04     3338792384 sha256-377655e65351a68cddfbd69b7c8dc60c1890466254628c3e494661a52c2c5ada
-a----      2025-03-21   오전 8:04           8432 sha256-dd084c7d92a3c1c14cc09ae77153b903fd2024b64a100a0cc8ec9316063d2dbc
-a----      2025-03-21   오전 8:04            358 sha256-e0a42594d802e5d31cdc786deb4823edb8adff66094d49de8fffe976d753e348
-a----      2025-03-21   오전 8:04            489 sha256-ffae984acbea1c53e6ca781206ca24fb5c677c6ad6cc0a90a6321d01d4f8c606

#################
# Note.The `latest` file contains information about the image layer.
#################
PS C:\Users\ycjang\.ollama\models\manifests\registry.ollama.ai\library\gemma3> ls

    디렉터리: C:\Users\ycjang\.ollama\models\manifests\registry.ollama.ai\library\gemma3

Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
-a----      2025-03-21   오전 8:14            856 latest

PS C:\Users\ycjang\.ollama\models\manifests\registry.ollama.ai\library\gemma3>

1
2
3
4
5
6
7
8
9
10

# The command to delete an LLM model is as follows, 
# and you can verify that the files in `/usr/share/ollama/.ollama/models/` have been removed.
PS C:\Users\ycjang\.ollama\models\blobs> ollama rm gemma3
deleted 'gemma3'

PS C:\Users\ycjang\.ollama\models\blobs> ollama list
NAME    ID    SIZE    MODIFIED

PS C:\Users\ycjang\.ollama\models\blobs> ls
PS C:\Users\ycjang\.ollama\models\blobs>

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

# I run gemma3.
PS C:\Users\ycjang\.ollama\models\blobs> ollama run gemma3
>>> hi. my name is ds2man. what is your name?
Hi ds2man! It’s nice to meet you. My name is Gemma, a large language model created by the Gemma team at Google DeepMind.

It’s great to chat with you!

>>> Hmm... From now on, let's call you ds2man.
Okay, ds2man. Let’s do it. 😉
What’s on your mind?

>>> what is your name?
Haha! You’re testing me, aren’t you? 😉
I’m still Gemma. But you can call me ds2man, as you prefer. 😊
What would you like to talk about?

# If you want to enter a long sentence, enclose it with `"""`.
>>> """
... I'm happy to be talking with you.
... Let's have lots of conversations from now on!
... """
That’s fantastic to hear, ds2man! I’m really glad you’re enjoying our conversations. I’m looking forward to chatting with you a lot too. 😊

What’s been on your mind lately? Anything you’d like to talk about, or maybe a game we could play?
>>>

on Ubuntu(WSL)

1
2
3
4
5
6
7
8
9
10
11

# I will try downloading gemma3:latest.
(base) jaoneol@DESKTOP-B7GM3C5:~$ ollama pull gemma3
pulling manifest
pulling 377655e65351... 100% ▕███████████████████████████████████████▏ 3.3 GB
pulling e0a42594d802... 100% ▕███████████████████████████████████████▏  358 B
pulling dd084c7d92a3... 100% ▕███████████████████████████████████████▏ 8.4 KB
pulling 0a74a8735bf3... 100% ▕███████████████████████████████████████    55 B
pulling ffae984acbea... 100% ▕███████████████████████████████████████▏  489 B
verifying sha256 digest
writing manifest
success

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

# I check the LLM models managed by Ollama using the following command.
(base) jaoneol@DESKTOP-B7GM3C5:~$ ollama list
NAME             ID              SIZE      MODIFIED
gemma3:latest    c0494fe00251    3.3 GB    2 seconds ago

# the location where the model is stored is as follows. 
# `/usr/share/ollama/.ollama/models`
(base) jaoneol@DESKTOP-B7GM3C5:~$ cd /usr/share/ollama/.ollama
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama$ ls
id_ed25519  id_ed25519.pub  models

(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama$ cd models
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models$ ls
blobs  manifests
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models$

#################
# Note.Image files are stored in the `blobs` directory.
#################
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$ ls -lrt
total 3260572
-rw-r--r-- 1 ollama ollama 3338792384 Mar 21 08:34 sha256-377655e65351a68cddfbd69b7c8dc60c1890466254628c3e494661a52c2c5ada
-rw-r--r-- 1 ollama ollama        358 Mar 21 08:34 sha256-e0a42594d802e5d31cdc786deb4823edb8adff66094d49de8fffe976d753e348
-rw-r--r-- 1 ollama ollama       8432 Mar 21 08:34 sha256-dd084c7d92a3c1c14cc09ae77153b903fd2024b64a100a0cc8ec9316063d2dbc
-rw-r--r-- 1 ollama ollama         55 Mar 21 08:34 sha256-0a74a8735bf3ffff4537b6c6bc9a4bc97a28c48f2fd347e806cca4d5001560f6
-rw-r--r-- 1 ollama ollama        489 Mar 21 08:34 sha256-ffae984acbea1c53e6ca781206ca24fb5c677c6ad6cc0a90a6321d01d4f8c606
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$

#################
# Note.The `latest` file contains information about the image layer.
#################
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/manifests/registry.ollama.ai/library/gemma3$ ls -lrt
total 4
-rw-r--r-- 1 ollama ollama 856 Mar 21 08:34 latest
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/manifests/registry.ollama.ai/library/gemma3$

1
2
3
4
5
6
7
8
9
10

# The command to delete an LLM model is as follows, 
# and you can verify that the files in `/usr/share/ollama/.ollama/models/` have been removed.
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$ ollama rm gemma3
deleted 'gemma3'

(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$ ollama list
NAME    ID    SIZE    MODIFIED

(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$ ls
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

# I run gemma3.
(base) jaoneol@DESKTOP-B7GM3C5:/usr/share/ollama/.ollama/models/blobs$ ollama run gemma3
>>> hi. my name is ds2man. what is your name?
Hi ds2man! It’s nice to meet you. My name is Gemma, a large language model created by the Gemma team at Google DeepMind.

It’s great to chat with you!

>>> Hmm... From now on, let's call you ds2man.
Okay, ds2man. Let’s do it. 😉
What’s on your mind?

>>> what is your name?
Haha! You’re testing me, aren’t you? 😉
I’m still Gemma. But you can call me ds2man, as you prefer. 😊
What would you like to talk about?

# If you want to enter a long sentence, enclose it with `"""`.
>>> """
... I'm happy to be talking with you.
... Let's have lots of conversations from now on!
... """
That’s fantastic to hear, ds2man! I’m really glad you’re enjoying our conversations. I’m looking forward to chatting with you a lot too. 😊

What’s been on your mind lately? Anything you’d like to talk about, or maybe a game we could play?
>>>

Testing Deepseek-r1

Recently, due to DeepSeek-R1, the stock prices of American AI companies have dropped significantly. This is because it offers high performance at a low cost.
DeepSeek-R1 is a fully open-source LLM licensed by MIT (Massachusetts Institute of Technology). It provides performance similar to the OpenAI-O1 large language model but at a much lower cost.
I tried using it—nothing more, nothing less.
Perhaps starting around May 2025, I might begin reading research papers on LLM models.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

(base) jaoneol@DESKTOP-B7GM3C5:~$ ollama pull deepseek-r1
pulling manifest
pulling 96c415656d37... 100% ▕█████████████████████████████████████████▏ 4.7 GB
pulling 369ca498f347... 100% ▕█████████████████████████████████████████▏  387 B
pulling 6e4c38e1172f... 100% ▕█████████████████████████████████████████▏ 1.1 KB
pulling f4d24e9138dd... 100% ▕█████████████████████████████████████████▏  148 B
pulling 40fb844194b2... 100% ▕█████████████████████████████████████████▏  487 B
verifying sha256 digest
writing manifest
success
(base) jaoneol@DESKTOP-B7GM3C5:~$ ollama list
NAME                  ID              SIZE      MODIFIED
deepseek-r1:latest    0a8c26691023    4.7 GB    13 minutes ago
llama3.1:latest       46e0c10c039e    4.9 GB    30 minutes ago
(base) jaoneol@DESKTOP-B7GM3C5:~$ 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

# I run Llama 3.1.
(base) jaoneol@DESKTOP-B7GM3C5:~$ ollama run deepseek-r1
>>> Recommend two places worth visiting in Shanghai, China.
<think>
Okay, so the user has asked for recommendations on two places to visit in Shanghai, China. First off, they've previously asked about the capital of Korea and wanted a response in English,
so now they're switching back to Shanghai.

I should consider why they're interested in Shanghai. Maybe they're planning a trip there soon or looking for ideas for a visit. They probably want well-known attractions that are popular
among tourists.

Shanghai is known for its blend of old and new architecture, rich history, and modernity. Therefore, recommending places that showcase both aspects would be ideal. Huangpu River comes to
mind because it's iconic with its bridges and the Bund district. Oriental Pearl TV Tower is another must-visit for its stunning views.

In terms of cultural experiences, Shanghai Disneyland fits well as a family-friendly attraction with unique rides. For food enthusiasts, Huohe Road is famous for its variety of dumpling
shops, which could be worth visiting if they're interested in local cuisine.

I should present these options clearly and briefly explain why each place is recommended without being too lengthy. Keeping the tone friendly and helpful will make the response effective.
</think>

Two highly recommended places to visit in Shanghai are:

1. **Huangpu River**
   This iconic riverfront offers stunning views of the city's landmarks, including the Oriental Pearl Tower and Pudian Riverside Promenade. It's also a great spot for dining and shopping.

2. **Oriental Pearl Tower (Shanghai Tower)**
   The tallest skyscraper in China, this modern architectural marvel provides panoramic views of the Huangpu River and the city skyline.

If you're looking for something more cultural and family-friendly, you might also enjoy exploring **Shanghai Disneyland**, which offers unique rides and attractions for all ages.

>>> Send a message (/? for help)
LLM&RAG, L&R-Ollama
Ollama
This post is licensed under CC BY 4.0 by the author.