github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #15115] Deepseek V2 cloud cannot call functional tools for websearch in the graphical interface of ollama #56195

New Issue
Closed
opened 2026-04-29 10:25:02 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-29 10:25:02 -05:00
Owner
Copy Link

Originally created by @createne1 on GitHub (Mar 28, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/15115

What is the issue?

I found the recent ollama 0.18.3 cannot allow Deepseek V1-617b cloud and Deepseek V2 cloud to call functional tools for websearch. Some sample results:

Deepseek v3.2-cloud:
--(prompt)
Search difference in mac sequoia and mac tahoe in handling virtual memories

Thought for 6.0 seconds
--(answer)
<function_calls> macOS Tahoe virtual memory handling </function_calls>
(unable to continue)

Deepseek V3.1-cloud
--(prompt)
Search difference in mac sequoia and mac tahoe in handling virtual memories

I'll search for information about the differences between macOS Sequoia and macOS Tahoe in handling virtual memory.

Searching for macOS Sequoia vs macOS Tahoe virtual memory management differences…

Search results for macOS Sequoia vs macOS Tahoe virtual memory management differences
--(answer)
Based on the search results, I can see there are significant differences between macOS Sequoia and macOS Tahoe in how they handle virtual memory, particularly on Macs with limited RAM. Let me fetch more detailed information from one of the key articles.
(unable to continue)

Relevant log output

Deepseek v3.2-cloud:
--(prompt)
Search difference in mac sequoia and mac tahoe in handling virtual memories
--
Thought for 6.0 seconds
--(answer)
<function_calls> macOS Tahoe virtual memory handling </function_calls>
(unable to continue)

Deepseek V3.1-cloud:
--(prompt)
Search difference in mac sequoia and mac tahoe in handling virtual memories
--
I'll search for information about the differences between macOS Sequoia and macOS Tahoe in handling virtual memory.

Searching for macOS Sequoia vs macOS Tahoe virtual memory management differences…

Search results for macOS Sequoia vs macOS Tahoe virtual memory management differences
--(answer)
Based on the search results, I can see there are significant differences between macOS Sequoia and macOS Tahoe in how they handle virtual memory, particularly on Macs with limited RAM. Let me fetch more detailed information from one of the key articles.
(unable to continue)

OS

macOS

GPU

Apple

CPU

Apple

Ollama version

0.18.3

Originally created by @createne1 on GitHub (Mar 28, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/15115 ### What is the issue? I found the recent ollama 0.18.3 cannot allow Deepseek V1-617b cloud and Deepseek V2 cloud to call functional tools for websearch. Some sample results: Deepseek v3.2-cloud: --(prompt) Search difference in mac sequoia and mac tahoe in handling virtual memories -- Thought for 6.0 seconds --(answer) <function_calls> macOS Tahoe virtual memory handling </function_calls> (unable to continue) Deepseek V3.1-cloud --(prompt) Search difference in mac sequoia and mac tahoe in handling virtual memories -- I'll search for information about the differences between macOS Sequoia and macOS Tahoe in handling virtual memory. Searching for macOS Sequoia vs macOS Tahoe virtual memory management differences… Search results for macOS Sequoia vs macOS Tahoe virtual memory management differences --(answer) Based on the search results, I can see there are significant differences between macOS Sequoia and macOS Tahoe in how they handle virtual memory, particularly on Macs with limited RAM. Let me fetch more detailed information from one of the key articles. (unable to continue) ### Relevant log output ```shell Deepseek v3.2-cloud: --(prompt) Search difference in mac sequoia and mac tahoe in handling virtual memories -- Thought for 6.0 seconds --(answer) <function_calls> macOS Tahoe virtual memory handling </function_calls> (unable to continue) Deepseek V3.1-cloud: --(prompt) Search difference in mac sequoia and mac tahoe in handling virtual memories -- I'll search for information about the differences between macOS Sequoia and macOS Tahoe in handling virtual memory. Searching for macOS Sequoia vs macOS Tahoe virtual memory management differences… Search results for macOS Sequoia vs macOS Tahoe virtual memory management differences --(answer) Based on the search results, I can see there are significant differences between macOS Sequoia and macOS Tahoe in how they handle virtual memory, particularly on Macs with limited RAM. Let me fetch more detailed information from one of the key articles. (unable to continue) ``` ### OS macOS ### GPU Apple ### CPU Apple ### Ollama version 0.18.3
GiteaMirror added the bug label 2026-04-29 10:25:02 -05:00
GiteaMirror commented 2026-04-29 10:25:04 -05:00
Author
Owner
Copy Link

@createne1 commented on GitHub (Apr 4, 2026):

Now in ollama 0.20, the functional call of deepseek V3.2 failed again.

The answer is below (Same issue as above, no results for function_calls):

Thought for a moment

<function_calls> install ollama </function_calls>

<!-- gh-comment-id:4187082990 --> @createne1 commented on GitHub (Apr 4, 2026): Now in ollama 0.20, the functional call of deepseek V3.2 failed again. The answer is below (Same issue as above, no results for function_calls): Thought for a moment <function_calls> install ollama </function_calls>
GiteaMirror commented 2026-04-29 10:25:05 -05:00
Author
Owner
Copy Link

@createne1 commented on GitHub (Apr 6, 2026):

server log:
time=2026-04-06T12:57:31.568+02:00 level=INFO source=routes.go:1744 msg="server config" env="map[HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:65536 OLLAMA_DEBUG:INFO OLLAMA_DEBUG_LOG_REQUESTS:false OLLAMA_EDITOR: OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/Users/chengmac/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NO_CLOUD:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_REMOTES:[ollama.com] OLLAMA_SCHED_SPREAD:false http_proxy: https_proxy: no_proxy:]"
time=2026-04-06T12:57:31.569+02:00 level=INFO source=routes.go:1746 msg="Ollama cloud disabled: false"
time=2026-04-06T12:57:31.576+02:00 level=INFO source=images.go:499 msg="total blobs: 49"
time=2026-04-06T12:57:31.576+02:00 level=INFO source=images.go:506 msg="total unused blobs removed: 0"
time=2026-04-06T12:57:31.577+02:00 level=INFO source=routes.go:1802 msg="Listening on 127.0.0.1:11434 (version 0.20.0)"
time=2026-04-06T12:57:31.579+02:00 level=INFO source=runner.go:67 msg="discovering available GPUs..."
time=2026-04-06T12:57:31.580+02:00 level=INFO source=server.go:432 msg="starting runner" cmd="/Applications/Ollama.app/Contents/Resources/ollama runner --ollama-engine --port 49365"
time=2026-04-06T12:57:31.714+02:00 level=INFO source=types.go:42 msg="inference compute" id=0 filter_id=0 library=Metal compute=0.0 name=Metal description="Apple M3" libdirs="" driver=0.0 pci_id="" type=discrete total="16.0 GiB" available="16.0 GiB"
time=2026-04-06T12:57:31.714+02:00 level=INFO source=routes.go:1852 msg="vram-based default context" total_vram="16.0 GiB" default_num_ctx=4096
[GIN] 2026/04/06 - 12:57:31 | 200 | 97.417µs | 127.0.0.1 | GET "/api/version"
[GIN] 2026/04/06 - 12:57:31 | 200 | 104.958µs | 127.0.0.1 | GET "/api/version"
[GIN] 2026/04/06 - 12:57:31 | 200 | 31.834µs | 127.0.0.1 | GET "/api/version"
[GIN] 2026/04/06 - 12:57:31 | 200 | 6.31ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 12:57:31 | 200 | 56.109292ms | 127.0.0.1 | POST "/api/show"
[GIN] 2026/04/06 - 12:57:31 | 200 | 208.060666ms | 127.0.0.1 | POST "/api/me"
[GIN] 2026/04/06 - 12:57:31 | 200 | 208.713084ms | 127.0.0.1 | POST "/api/me"
[GIN] 2026/04/06 - 12:57:35 | 200 | 2.748ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 12:57:48 | 200 | 131.892125ms | 127.0.0.1 | POST "/api/show"
[GIN] 2026/04/06 - 12:57:53 | 200 | 183.593292ms | 127.0.0.1 | POST "/api/show"
[GIN] 2026/04/06 - 12:57:53 | 200 | 130.865917ms | 127.0.0.1 | POST "/api/show"
[GIN] 2026/04/06 - 12:58:03 | 200 | 9.30564125s | 127.0.0.1 | POST "/api/chat"
[GIN] 2026/04/06 - 12:58:05 | 200 | 9.525042ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 12:58:35 | 200 | 7.52975ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 12:59:05 | 200 | 5.365959ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 12:59:35 | 200 | 8.147291ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:00:05 | 200 | 5.815042ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:00:35 | 200 | 8.242625ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:01:05 | 200 | 9.310375ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:01:35 | 200 | 7.247ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:02:05 | 200 | 6.40575ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:02:36 | 200 | 8.030458ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:03:06 | 200 | 8.379083ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:03:36 | 200 | 5.085292ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:04:06 | 200 | 7.161041ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:04:36 | 200 | 6.21675ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:05:06 | 200 | 8.625417ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:05:36 | 200 | 4.78775ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:06:06 | 200 | 7.184125ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:06:36 | 200 | 5.488542ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:06:44 | 200 | 36.792µs | 127.0.0.1 | GET "/api/version"
[GIN] 2026/04/06 - 13:06:44 | 200 | 6.243042ms | 127.0.0.1 | GET "/api/tags"
[GIN] 2026/04/06 - 13:06:44 | 200 | 205.80675ms | 127.0.0.1 | POST "/api/me"

<!-- gh-comment-id:4194887432 --> @createne1 commented on GitHub (Apr 6, 2026): server log: time=2026-04-06T12:57:31.568+02:00 level=INFO source=routes.go:1744 msg="server config" env="map[HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:65536 OLLAMA_DEBUG:INFO OLLAMA_DEBUG_LOG_REQUESTS:false OLLAMA_EDITOR: OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/Users/chengmac/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NO_CLOUD:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_REMOTES:[ollama.com] OLLAMA_SCHED_SPREAD:false http_proxy: https_proxy: no_proxy:]" time=2026-04-06T12:57:31.569+02:00 level=INFO source=routes.go:1746 msg="Ollama cloud disabled: false" time=2026-04-06T12:57:31.576+02:00 level=INFO source=images.go:499 msg="total blobs: 49" time=2026-04-06T12:57:31.576+02:00 level=INFO source=images.go:506 msg="total unused blobs removed: 0" time=2026-04-06T12:57:31.577+02:00 level=INFO source=routes.go:1802 msg="Listening on 127.0.0.1:11434 (version 0.20.0)" time=2026-04-06T12:57:31.579+02:00 level=INFO source=runner.go:67 msg="discovering available GPUs..." time=2026-04-06T12:57:31.580+02:00 level=INFO source=server.go:432 msg="starting runner" cmd="/Applications/Ollama.app/Contents/Resources/ollama runner --ollama-engine --port 49365" time=2026-04-06T12:57:31.714+02:00 level=INFO source=types.go:42 msg="inference compute" id=0 filter_id=0 library=Metal compute=0.0 name=Metal description="Apple M3" libdirs="" driver=0.0 pci_id="" type=discrete total="16.0 GiB" available="16.0 GiB" time=2026-04-06T12:57:31.714+02:00 level=INFO source=routes.go:1852 msg="vram-based default context" total_vram="16.0 GiB" default_num_ctx=4096 [GIN] 2026/04/06 - 12:57:31 | 200 | 97.417µs | 127.0.0.1 | GET "/api/version" [GIN] 2026/04/06 - 12:57:31 | 200 | 104.958µs | 127.0.0.1 | GET "/api/version" [GIN] 2026/04/06 - 12:57:31 | 200 | 31.834µs | 127.0.0.1 | GET "/api/version" [GIN] 2026/04/06 - 12:57:31 | 200 | 6.31ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 12:57:31 | 200 | 56.109292ms | 127.0.0.1 | POST "/api/show" [GIN] 2026/04/06 - 12:57:31 | 200 | 208.060666ms | 127.0.0.1 | POST "/api/me" [GIN] 2026/04/06 - 12:57:31 | 200 | 208.713084ms | 127.0.0.1 | POST "/api/me" [GIN] 2026/04/06 - 12:57:35 | 200 | 2.748ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 12:57:48 | 200 | 131.892125ms | 127.0.0.1 | POST "/api/show" [GIN] 2026/04/06 - 12:57:53 | 200 | 183.593292ms | 127.0.0.1 | POST "/api/show" [GIN] 2026/04/06 - 12:57:53 | 200 | 130.865917ms | 127.0.0.1 | POST "/api/show" [GIN] 2026/04/06 - 12:58:03 | 200 | 9.30564125s | 127.0.0.1 | POST "/api/chat" [GIN] 2026/04/06 - 12:58:05 | 200 | 9.525042ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 12:58:35 | 200 | 7.52975ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 12:59:05 | 200 | 5.365959ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 12:59:35 | 200 | 8.147291ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:00:05 | 200 | 5.815042ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:00:35 | 200 | 8.242625ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:01:05 | 200 | 9.310375ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:01:35 | 200 | 7.247ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:02:05 | 200 | 6.40575ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:02:36 | 200 | 8.030458ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:03:06 | 200 | 8.379083ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:03:36 | 200 | 5.085292ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:04:06 | 200 | 7.161041ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:04:36 | 200 | 6.21675ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:05:06 | 200 | 8.625417ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:05:36 | 200 | 4.78775ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:06:06 | 200 | 7.184125ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:06:36 | 200 | 5.488542ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:06:44 | 200 | 36.792µs | 127.0.0.1 | GET "/api/version" [GIN] 2026/04/06 - 13:06:44 | 200 | 6.243042ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/04/06 - 13:06:44 | 200 | 205.80675ms | 127.0.0.1 | POST "/api/me"
GiteaMirror commented 2026-04-29 10:25:06 -05:00
Author
Owner
Copy Link

@createne1 commented on GitHub (Apr 7, 2026):

In the version 0.20.3. The bug is fixed.

<!-- gh-comment-id:4201271979 --> @createne1 commented on GitHub (Apr 7, 2026): In the version 0.20.3. The bug is fixed.
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 bug
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#56195
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?