github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #10905] Deepseek-R1 Qwen 3 8B Distill #32930

New Issue
Closed
opened 2026-04-22 14:53:35 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-04-22 14:53:35 -05:00
Owner
Copy Link

Originally created by @numinousmuses on GitHub (May 29, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10905

https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B

Originally created by @numinousmuses on GitHub (May 29, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10905 https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B
GiteaMirror added the model label 2026-04-22 14:53:35 -05:00
GiteaMirror commented 2026-04-22 14:53:58 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 29, 2025):

https://ollama.com/library/deepseek-r1:8b-0528-qwen3-q4_K_M

<!-- gh-comment-id:2920701957 --> @rick-github commented on GitHub (May 29, 2025): https://ollama.com/library/deepseek-r1:8b-0528-qwen3-q4_K_M
GiteaMirror commented 2026-04-22 14:54:00 -05:00
Author
Owner
Copy Link

@djouallah commented on GitHub (May 30, 2025):

do you know how to turn off thinking for this model

<!-- gh-comment-id:2920846517 --> @djouallah commented on GitHub (May 30, 2025): do you know how to turn off thinking for this model
GiteaMirror commented 2026-04-22 14:54:01 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 30, 2025):

Use ollama 0.9.0 and /set nothink or --think=false.

$ ollama run deepseek-r1:8b 
>>> /set nothink
Set 'nothink' mode.
>>> hello
Hello! 👋 How can I assist you today?
$ ollama run deepseek-r1:8b hello --think=false
Hello! 👋 How can I assist you today?

If you are using the API, set think to false:

$ curl -s localhost:11434/api/generate  -d '{"model":"deepseek-r1:8b","prompt":"hello","stream":false,"think":true}' | jq 'del(.context)'
{
  "model": "deepseek-r1:8b",
  "created_at": "2025-05-30T00:13:30.319562655Z",
  "response": "Hello! 👋 How can I help you today?",
  "thinking": "Okay, the user just said “hello” with a simple lowercase. Hmm, this is probably an opening message or maybe they're testing how I respond.\n\nSince it's such a basic greeting without any context, I should keep my response warm but neutral - not too formal since that might feel stiff for casual chat, and not overly familiar either because we don't know each other yet. The lowercase suggests they might be in a relaxed mood or just typing quickly at first.\n\nI notice they didn't share their name or ask anything specific. This could mean either:\n- They're new to chatting with AI assistants\n- They're being deliberately vague for some reason\n- Or it's an accidental message\n\nThe safest approach is to greet them and leave the conversation open-ended, giving them space to steer the interaction if they want. No need to dive deep unless they ask something substantive.\n\nAdding a cheerful emoji feels appropriate here - it sets a friendly tone without being pushy. The “feel free” part gives permission for either casual or serious conversations depending on what they choose next.\n",
  "done": true,
  "done_reason": "stop",
  "total_duration": 3353366981,
  "load_duration": 283025290,
  "prompt_eval_count": 3,
  "prompt_eval_duration": 3958820,
  "eval_count": 231,
  "eval_duration": 3065828714
}
$ curl -s localhost:11434/api/generate  -d '{"model":"deepseek-r1:8b","prompt":"hello","stream":false,"think":false}' | jq 'del(.context)'
{
  "model": "deepseek-r1:8b",
  "created_at": "2025-05-30T00:13:37.963268234Z",
  "response": "Hello! How can I assist you today?",
  "done": true,
  "done_reason": "stop",
  "total_duration": 436042414,
  "load_duration": 296689403,
  "prompt_eval_count": 7,
  "prompt_eval_duration": 6456547,
  "eval_count": 10,
  "eval_duration": 132208872
}
<!-- gh-comment-id:2920868887 --> @rick-github commented on GitHub (May 30, 2025): Use ollama 0.9.0 and `/set nothink` or `--think=false`. ```console $ ollama run deepseek-r1:8b >>> /set nothink Set 'nothink' mode. >>> hello Hello! 👋 How can I assist you today? $ ollama run deepseek-r1:8b hello --think=false Hello! 👋 How can I assist you today? ``` If you are using the API, set `think` to `false`: ```console $ curl -s localhost:11434/api/generate -d '{"model":"deepseek-r1:8b","prompt":"hello","stream":false,"think":true}' | jq 'del(.context)' { "model": "deepseek-r1:8b", "created_at": "2025-05-30T00:13:30.319562655Z", "response": "Hello! 👋 How can I help you today?", "thinking": "Okay, the user just said “hello” with a simple lowercase. Hmm, this is probably an opening message or maybe they're testing how I respond.\n\nSince it's such a basic greeting without any context, I should keep my response warm but neutral - not too formal since that might feel stiff for casual chat, and not overly familiar either because we don't know each other yet. The lowercase suggests they might be in a relaxed mood or just typing quickly at first.\n\nI notice they didn't share their name or ask anything specific. This could mean either:\n- They're new to chatting with AI assistants\n- They're being deliberately vague for some reason\n- Or it's an accidental message\n\nThe safest approach is to greet them and leave the conversation open-ended, giving them space to steer the interaction if they want. No need to dive deep unless they ask something substantive.\n\nAdding a cheerful emoji feels appropriate here - it sets a friendly tone without being pushy. The “feel free” part gives permission for either casual or serious conversations depending on what they choose next.\n", "done": true, "done_reason": "stop", "total_duration": 3353366981, "load_duration": 283025290, "prompt_eval_count": 3, "prompt_eval_duration": 3958820, "eval_count": 231, "eval_duration": 3065828714 } $ curl -s localhost:11434/api/generate -d '{"model":"deepseek-r1:8b","prompt":"hello","stream":false,"think":false}' | jq 'del(.context)' { "model": "deepseek-r1:8b", "created_at": "2025-05-30T00:13:37.963268234Z", "response": "Hello! How can I assist you today?", "done": true, "done_reason": "stop", "total_duration": 436042414, "load_duration": 296689403, "prompt_eval_count": 7, "prompt_eval_duration": 6456547, "eval_count": 10, "eval_duration": 132208872 } ```
GiteaMirror commented 2026-04-22 14:54:04 -05:00
Author
Owner
Copy Link

@djouallah commented on GitHub (May 30, 2025):

thanks, found the issue, for some reason in windows, the app did not auto update, and i was on version 8 :(

<!-- gh-comment-id:2920926403 --> @djouallah commented on GitHub (May 30, 2025): thanks, found the issue, for some reason in windows, the app did not auto update, and i was on version 8 :(
GiteaMirror commented 2026-04-22 14:54:04 -05:00
Author
Owner
Copy Link

@comeonzhj commented on GitHub (May 30, 2025):

I have updated to the latest version of Ollama, but it still prompts that the version does not support the new model.

ollama run deepseek-r1:8b
pulling manifest 
Error: pull model manifest: 412: 

The model you are attempting to pull requires a newer version of Ollama.

Please download the latest version at:

	https://ollama.com/download

ollama --version
ollama version is 0.6.5
Warning: client version is 0.9.0
<!-- gh-comment-id:2921050472 --> @comeonzhj commented on GitHub (May 30, 2025): I have updated to the latest version of Ollama, but it still prompts that the version does not support the new model. ```bash ollama run deepseek-r1:8b pulling manifest Error: pull model manifest: 412: The model you are attempting to pull requires a newer version of Ollama. Please download the latest version at: https://ollama.com/download ``` --- ```bash ollama --version ollama version is 0.6.5 Warning: client version is 0.9.0 ```
GiteaMirror commented 2026-04-22 14:54:06 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 30, 2025):

Restart the server.

<!-- gh-comment-id:2921437825 --> @rick-github commented on GitHub (May 30, 2025): Restart the server.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label model
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#32930
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?