[GH-ISSUE #14662] why the ollama get response very slow? (QWEN3.5 35B A3B) #35254

New Issue

GiteaMirror · 2026-04-22T19:38:28-05:00

GiteaMirror commented

2026-04-22 19:38:28 -05:00

Originally created by @acer1204 on GitHub (Mar 6, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/14662

What is the issue?

My hardware is an NVIDIA DGX Spark, running Ollama 0.17.7 and the latest version of OpenWebUI.

During testing, I noticed that the first prompt responds quickly. OpenWebUI immediately shows “think” and then generates the response.

However, when I ask a second question, it takes 30 to 60 seconds before the frontend webpage shows “think” and the response begins to appear.

This issue does not occur when I run GPT-OSS 120B.

I test this problem in my RTX 3090 have the same problem.

i see the nvidia-smi, it use 100% cuda but not thing to show.

if restart ollama, the first qustion response quick.

Relevant log output

OS

Windows, Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.17.7

Originally created by @acer1204 on GitHub (Mar 6, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/14662 ### What is the issue? My hardware is an NVIDIA DGX Spark, running Ollama 0.17.7 and the latest version of OpenWebUI. During testing, I noticed that the first prompt responds quickly. OpenWebUI immediately shows “think” and then generates the response. However, when I ask a second question, it takes 30 to 60 seconds before the frontend webpage shows “think” and the response begins to appear. This issue does not occur when I run GPT-OSS 120B. <img width="1062" height="112" alt="Image" src="https://github.com/user-attachments/assets/32e1f49c-98e7-437f-b9f8-f38d8a165e8c" /> I test this problem in my RTX 3090 have the same problem. i see the nvidia-smi, it use 100% cuda but not thing to show. if restart ollama, the first qustion response quick. ### Relevant log output ```shell ``` ### OS Windows, Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.17.7

GiteaMirror added the bug label 2026-04-22 19:38:28 -05:00

GiteaMirror commented

2026-04-22 19:38:31 -05:00

@BeatWolf commented on GitHub (Mar 6, 2026):

I'm currently trying to debug the same thing with qwen3.5:2b. The time to first token is huge, but i don't see any model loading in the logs etc.
What is also strange, this is with using openwebui, even the second message in the chat is slow, not only the first. So something seems wrong

@BeatWolf commented on GitHub (Mar 6, 2026): I'm currently trying to debug the same thing with qwen3.5:2b. The time to first token is huge, but i don't see any model loading in the logs etc. What is also strange, this is with using openwebui, even the second message in the chat is slow, not only the first. So something seems wrong

GiteaMirror commented

2026-04-22 19:38:34 -05:00

@VistritPandey commented on GitHub (Mar 6, 2026):

+1 the same issue on M3 Ultra, it is taking twice the time to generate, when compared to qwen3-vl:8b and takes almost generates token at a speed of 40~50 tokens/sec (for 35b:a3b, 9b, and 4b), which ideally shouldn't be the case considering the size difference

@VistritPandey commented on GitHub (Mar 6, 2026): +1 the same issue on `M3 Ultra`, it is taking twice the time to generate, when compared to `qwen3-vl:8b` and takes almost generates token at a speed of 40~50 tokens/sec (for 35b:a3b, 9b, and 4b), which ideally shouldn't be the case considering the size difference

GiteaMirror commented

2026-04-22 19:38:35 -05:00

@SquareSphere commented on GitHub (Mar 7, 2026):

I have the same issue with the combination of Open WebUI and Ollama using Qwen 3.5 9b BUT if I try to run locally through the ollama UI it doesn't have the same issue

@SquareSphere commented on GitHub (Mar 7, 2026): I have the same issue with the combination of Open WebUI and Ollama using Qwen 3.5 9b BUT if I try to run locally through the ollama UI it doesn't have the same issue

GiteaMirror commented

2026-04-22 19:38:36 -05:00

@leegimblett commented on GitHub (Mar 9, 2026):

I have similar experience with qwen3.5:27b . Works great with the ollama cli with the gpu use falling to 0 soon after the output is finished. When using the api (Open WebUI) the first prompt is answered quickly but the second one hangs forever and the GPU flat lines at around 50% after the prompt is finished and stays there for a long time after.

@leegimblett commented on GitHub (Mar 9, 2026): I have similar experience with qwen3.5:27b . Works great with the ollama cli with the gpu use falling to 0 soon after the output is finished. When using the api (Open WebUI) the first prompt is answered quickly but the second one hangs forever and the GPU flat lines at around 50% after the prompt is finished and stays there for a long time after.

GiteaMirror commented

2026-04-22 19:38:38 -05:00

@leegimblett commented on GitHub (Mar 9, 2026):

I think my experience is Open WebUI related. I don't see it with out API clients + when I turn off thinking in WebUI the issue disappears. Web UI is sending requests in the background after the LLM completes its main response (title generation etc). The Qwen3.5 27B model does a lot of thinking so takes a long time. You can change the model used by the background processes (admin/interface)which can be a lot smaller than the main model. I used a Qwen3.5 0.8B and it still took a long time until I turned off that models thinking, then the problem went away.

@leegimblett commented on GitHub (Mar 9, 2026): I think my experience is Open WebUI related. I don't see it with out API clients + when I turn off thinking in WebUI the issue disappears. Web UI is sending requests in the background after the LLM completes its main response (title generation etc). The Qwen3.5 27B model does a lot of thinking so takes a long time. You can change the model used by the background processes (admin/interface)which can be a lot smaller than the main model. I used a Qwen3.5 0.8B and it still took a long time until I turned off that models thinking, then the problem went away.

GiteaMirror commented

2026-04-22 19:38:38 -05:00

@luke2023 commented on GitHub (Mar 9, 2026):

I am experiencing the same issue, on 5090 + windows 11 pro + open web ui

@luke2023 commented on GitHub (Mar 9, 2026): I am experiencing the same issue, on 5090 + windows 11 pro + open web ui

GiteaMirror commented

2026-04-22 19:38:39 -05:00

@rawflecat commented on GitHub (Mar 9, 2026):

I think my experience is Open WebUI related. I don't see it with out API clients + when I turn off thinking in WebUI the issue disappears. Web UI is sending requests in the background after the LLM completes its main response (title generation etc). The Qwen3.5 27B model does a lot of thinking so takes a long time. You can change the model used by the background processes (admin/interface)which can be a lot smaller than the main model. I used a Qwen3.5 0.8B and it still took a long time until I turned off that models thinking, then the problem went away.

Just went through this over the weekend with Open WebUI. Ended up spinning up a VM in order to run a small version of gemma to use for the backend processes. Helped tremendously, but the overhead of having to ship that to another model is a pain. Doing it on the same box is out of the questions. I'm not going to unload and reload a 40 GB+ model just to generate a chat title in under 60 seconds.

@rawflecat commented on GitHub (Mar 9, 2026): > I think my experience is Open WebUI related. I don't see it with out API clients + when I turn off thinking in WebUI the issue disappears. Web UI is sending requests in the background after the LLM completes its main response (title generation etc). The Qwen3.5 27B model does a lot of thinking so takes a long time. You can change the model used by the background processes (admin/interface)which can be a lot smaller than the main model. I used a Qwen3.5 0.8B and it still took a long time until I turned off that models thinking, then the problem went away. Just went through this over the weekend with Open WebUI. Ended up spinning up a VM in order to run a small version of gemma to use for the backend processes. Helped tremendously, but the overhead of having to ship that to another model is a pain. Doing it on the same box is out of the questions. I'm not going to unload and reload a 40 GB+ model just to generate a chat title in under 60 seconds.

GiteaMirror commented

2026-04-22 19:38:40 -05:00

@BeatWolf commented on GitHub (Mar 9, 2026):

i think its multiple issues. webui doing unnecessary thinking calls on large models but also there is likely something wrong with qwrn3.5 and or ollama that there are super high amount of thinking tokens for the simplest requests.

@BeatWolf commented on GitHub (Mar 9, 2026): i think its multiple issues. webui doing unnecessary thinking calls on large models but also there is likely something wrong with qwrn3.5 and or ollama that there are super high amount of thinking tokens for the simplest requests.

GiteaMirror commented

2026-04-22 19:38:41 -05:00

@luke2023 commented on GitHub (Mar 9, 2026):

I am experience an abnormally long time to first token on all Qwen 3.5 models (second and third request would be even worse), other models are working fine.

@luke2023 commented on GitHub (Mar 9, 2026): I am experience an abnormally long time to first token on all Qwen 3.5 models (second and third request would be even worse), other models are working fine.

GiteaMirror commented

2026-04-22 19:38:42 -05:00

@luke2023 commented on GitHub (Mar 9, 2026):

Can confirm only happens on Qwen3.5 series. Qwen3 and GPT-OSS works fine

@luke2023 commented on GitHub (Mar 9, 2026): Can confirm only happens on Qwen3.5 series. Qwen3 and GPT-OSS works fine

GiteaMirror commented

2026-04-22 19:38:42 -05:00

@Megara18 commented on GitHub (Apr 3, 2026):

Ollama + OpenWeb UI, Can confirm same problem with the new Gemma4. Slow responses and long thinking times for even say "Hi"

@Megara18 commented on GitHub (Apr 3, 2026): Ollama + OpenWeb UI, Can confirm same problem with the new Gemma4. Slow responses and long thinking times for even say "Hi"

GiteaMirror commented

2026-04-22 19:38:42 -05:00

@beykansen commented on GitHub (Apr 5, 2026):

same issue with my asus gx10. qwen3.5 and gemma4 do that.

@beykansen commented on GitHub (Apr 5, 2026): same issue with my asus gx10. qwen3.5 and gemma4 do that.

GiteaMirror commented

2026-04-22 19:38:43 -05:00

@Joni1717 commented on GitHub (Apr 6, 2026):

Same issue with qwen3.5 35b and 27b on 2xRTX4090

@Joni1717 commented on GitHub (Apr 6, 2026): Same issue with qwen3.5 35b and 27b on 2xRTX4090

GiteaMirror commented

2026-04-22 19:38:43 -05:00

@Megara18 commented on GitHub (Apr 16, 2026):

It happens too with Qwen3.6... 40 Seconds of THINKING just for say "Hola" (Hello) is pretty insane... it prevents me to try frontier models with Ollama... near all new models haves this problem.

And is not about hardware, my server can move up +200B models at high speed.

Using Ollama lastest on Docker + Vulkan.

@Megara18 commented on GitHub (Apr 16, 2026): <img width="538" height="133" alt="Image" src="https://github.com/user-attachments/assets/d08967a7-4d69-477b-a937-a05707463302" /> It happens too with Qwen3.6... 40 Seconds of THINKING just for say "Hola" (Hello) is pretty insane... it prevents me to try frontier models with Ollama... near all new models haves this problem. And is not about hardware, my server can move up +200B models at high speed. Using Ollama lastest on Docker + Vulkan.

GiteaMirror commented

2026-04-22 19:38:44 -05:00

@Lemondsky commented on GitHub (Apr 21, 2026):

Same issue, ollama + openwebui + qwen (no matter the model) takes like 8 min before ecen start thinking. Prompt processing and tokengeneration fast as hell, but every message takes 8-9 minutes before it ecen get visible in ollama logs and starts. Without openwebui, direktly calling ollama api, response is fast af and nearly no processing time. Something seems extremely bugged in combination with qwen, makes openwebui useless for me at the time.

@Lemondsky commented on GitHub (Apr 21, 2026): Same issue, ollama + openwebui + qwen (no matter the model) takes like 8 min before ecen start thinking. Prompt processing and tokengeneration fast as hell, but every message takes 8-9 minutes before it ecen get visible in ollama logs and starts. Without openwebui, direktly calling ollama api, response is fast af and nearly no processing time. Something seems extremely bugged in combination with qwen, makes openwebui useless for me at the time.

GiteaMirror commented

2026-04-22 19:38:44 -05:00

@Joni1717 commented on GitHub (Apr 22, 2026):

For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request).

Here’s how I fixed it:
Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking).
Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled).

Result:
Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker.

@Joni1717 commented on GitHub (Apr 22, 2026): For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request). Here’s how I fixed it: Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking). Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled). Result: Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker.

GiteaMirror commented

2026-04-22 19:38:45 -05:00

@Megara18 commented on GitHub (Apr 22, 2026):

For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request).

Here’s how I fixed it: Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking). Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled).

Result: Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker.

No. Its not that. I already have specific models for the minor task already setup. And this is happening specific with frontier models like Gemma and Qwen.

Other tested 120B and 220B models do not have this problem.

@Megara18 commented on GitHub (Apr 22, 2026): > For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request). > > Here’s how I fixed it: Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking). Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled). > > Result: Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker. No. Its not that. I already have specific models for the minor task already setup. And this is happening specific with frontier models like Gemma and Qwen. Other tested 120B and 220B models do not have this problem.

GiteaMirror commented

2026-04-22 19:38:45 -05:00

@Lemondsky commented on GitHub (Apr 22, 2026):

For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request).

Here’s how I fixed it:
Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking).
Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled).

Result:
Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker.

To me its definetly related to openwebui. Tried some other tools, none has this issue, it happens only if i use openwebui with qwen and ollama.
There ist no token generation or anything happen for the ~8 minutes, cant see any kind of communication between qwen/ollama and openwebui in the logs, also the graphiccards dont do anything for the time, hanging arround 28 watts each. After the ~8 minutes they all go up to ~160 watts and start prompt processing and generating / thinking ..
Gemma and everything works fine.
I just use another interface for qwen now until its fixed.
Otherwise, my hardware is pretty exotic, maybe theres something colliding or bugging at my system with this specific combination, cant tell for sure

@Lemondsky commented on GitHub (Apr 22, 2026): > For me, the issue wasn’t related to Ollama. It’s caused by qwen3.5/3.6 sometimes taking a very long time to “think.” Open WebUI generates tags, titles, and follow-up questions after the initial response, and each of these is a separate request using the same model. Since the model spends a lot of time on each request, the delays quickly add up (sometimes around a minute per request). > > Here’s how I fixed it: > Go to Admin Settings → Models and select qwen3.5/3.6. Create a clone of the model. In the clone’s settings, disable “Thinking” (Ollama Thinking). > Then go to Admin Settings → Interface and select the cloned qwen3.6 model (with Thinking disabled). > > Result: > Your main responses are still generated with the original qwen3.6 model, while tags, titles, and follow-up questions are handled by the faster version without Thinking—making the whole process much quicker. To me its definetly related to openwebui. Tried some other tools, none has this issue, it happens only if i use openwebui with qwen and ollama. There ist no token generation or anything happen for the ~8 minutes, cant see any kind of communication between qwen/ollama and openwebui in the logs, also the graphiccards dont do anything for the time, hanging arround 28 watts each. After the ~8 minutes they all go up to ~160 watts and start prompt processing and generating / thinking .. Gemma and everything works fine. I just use another interface for qwen now until its fixed. Otherwise, my hardware is pretty exotic, maybe theres something colliding or bugging at my system with this specific combination, cant tell for sure

Sign in to join this conversation.

Branches Tags

main

parth-mlx-decode-checkpoints

dhiltgen/ci

hoyyeva/editor-config-repair

parth-launch-codex-app

hoyyeva/fix-codex-model-metadata-warning

hoyyeva/qwen

hoyyeva/launch-backup-ux

parth/hide-claude-desktop-till-release

hoyyeva/opencode-image-modality

parth-add-claude-code-autoinstall

release_v0.22.0

pdevine/manifest-list

codex/fix-codex-model-metadata-warning

pdevine/addressable-manifest

brucemacd/launch-fetch-reccomended

jmorganca/llama-compat

launch-copilot-cli

hoyyeva/opencode-thinking

release_v0.20.7

parth-auto-save-backup

parth-test

jmorganca/gemma4-audio-replacements

fix-manifest-digest-on-pull

hoyyeva/vscode-improve

brucemacd/install-server-wait

brucemacd/download-before-remove

parth/update-claude-docs

parth-anthropic-reference-images-path

brucemac/start-ap-install

pdevine/mlx-update

pdevine/qwen35_vision

drifkin/api-show-fallback

mintlify/image-generation-1773352582

hoyyeva/server-context-length-local-config

jmorganca/faster-reptition-penalties

jmorganca/convert-nemotron

parth-pi-thinking

pdevine/sampling-penalties

jmorganca/fix-create-quantization-memory

dongchen/resumable_transfer_fix

pdevine/sampling-cache-error

jessegross/mlx-usage

hoyyeva/openclaw-config

hoyyeva/app-html

pdevine/qwen3next

brucemacd/sign-sh-install

brucemacd/tui-update

brucemacd/usage-api

jmorganca/launch-empty

fix-app-dist-embed

mxyng/mlx-compile

mxyng/mlx-quant

mxyng/mlx-glm4.7

mxyng/mlx

brucemacd/simplify-model-picker

jmorganca/qwen3-concurrent

fix-glm-4.7-flash-mla-config

drifkin/qwen3-coder-opening-tag

brucemacd/usage-cli

fix-cuda12-fattn-shmem

ollama-imagegen-docs

parth/fix-multiline-inputs

brucemacd/config-docs

mxyng/model-files

mxyng/simple-execute

fix-imagegen-ollama-models

mxyng/async-upload

jmorganca/lazy-no-dtype-changes

imagegen-auto-detect-create

parth/decrease-concurrent-download-hf

fix-mlx-quantize-init

jmorganca/x-cleanup

usage

imagegen-readme

jmorganca/glm-image

mlx-gpu-cd

jmorganca/imagegen-modelfile

parth/agent-skills

parth/agent-allowlist

parth/signed-in-offline

parth/agents

parth/fix-context-chopping

improve-cloud-flow

parth/add-models-websearch

parth/prompt-renderer-mcp

jmorganca/native-settings

jmorganca/download-stream-hash

jmorganca/client2-rebased

brucemacd/oai-chat-req-multipart

jessegross/multi_chunk_reserve

grace/additional-omit-empty

grace/mistral-3-large

mxyng/tokenizer2

mxyng/tokenizer

jessegross/flash

hoyyeva/windows-nacked-app

mxyng/cleanup-attention

grace/deepseek-parser

hoyyeva/remember-unsent-prompt

parth/add-lfs-pointer-error-conversion

parth/olmo2-test2

hoyyeva/ollama-launchagent-plist

nicole/olmo-model

parth/olmo-test

mxyng/remove-embedded

parth/render-template

jmorganca/intellect-3

parth/remove-prealloc-linter

jmorganca/cmd-eval

nicole/nomic-embed-text-fix

mxyng/lint-2

hoyyeva/add-gemini-3-pro-preview

hoyyeva/load-model-list

mxyng/expand-path

mxyng/environ-2

hoyyeva/deeplink-json-encoding

parth/improve-tool-calling-tests

hoyyeva/conversation

hoyyeva/assistant-edit-response

hoyyeva/thinking

origin/brucemacd/invalid-char-i-err

parth/improve-tool-calling

jmorganca/required-omitempty

grace/qwen3-vl-tests

mxyng/iter-client

parth/docs-readme

nicole/embed-test

pdevine/integration-benchstat

parth/remove-generate-cmd

parth/add-toolcall-id

mxyng/server-tests

jmorganca/glm-4.6

jmorganca/gin-h-compat

drifkin/stable-tool-args

pdevine/qwen3-more-thinking

parth/add-websearch-client

nicole/websearch_local

jmorganca/qwen3-coder-updates

grace/deepseek-v3-migration-tests

mxyng/fix-create

jmorganca/cloud-errors

pdevine/parser-tidy

revert-12233-parth/simplify-entrypoints-runner

parth/enable-so-gpt-oss

brucemacd/qwen3vl

jmorganca/readme-simplify

parth/gpt-oss-structured-outputs

revert-12039-jmorganca/tools-braces

mxyng/embeddings

mxyng/gguf

mxyng/benchmark

mxyng/types-null

parth/move-parsing

mxyng/gemma2

jmorganca/docs

mxyng/16-bit

mxyng/create-stdin

pdevine/authorizedkeys

mxyng/quant

parth/opt-in-error-context-window

brucemacd/cache-models

brucemacd/runner-completion

jmorganca/llama-update-6

brucemacd/benchmark-list

brucemacd/partial-read-caps

parth/deepseek-r1-tools

mxyng/omit-array

parth/tool-prefix-temp

brucemacd/runner-test

jmorganca/qwen25vl

brucemacd/model-forward-test-ext

parth/python-function-parsing

jmorganca/cuda-compression-none

drifkin/num-parallel

drifkin/chat-truncation-fix

jmorganca/sync

parth/python-tools-calling

drifkin/array-head-count

brucemacd/create-no-loop

parth/server-enable-content-stream-with-tools

qwen25omni

mxyng/v3

brucemacd/ropeconfig

jmorganca/silence-tokenizer

parth/sample-so-test

parth/sampling-structured-outputs

brucemacd/doc-go-engine

parth/constrained-sampling-json

jmorganca/mistral-wip

brucemacd/mistral-small-convert

parth/sample-unmarshal-json-for-params

brucemacd/jomorganca/mistral

pdevine/bfloat16

jmorganca/mistral

brucemacd/mistral

pdevine/logging

parth/sample-correctness-fix

parth/sample-fix-sorting

jmorgan/sample-fix-sorting-extras

jmorganca/temp-0-images

brucemacd/parallel-embed-models

brucemacd/shim-grammar

jmorganca/fix-gguf-error

bmizerany/nameswork

jmorganca/faster-releases

bmizerany/validatenames

brucemacd/err-no-vocab

brucemacd/rope-config

brucemacd/err-hint

brucemacd/qwen2_5

brucemacd/logprobs

brucemacd/new_runner_graph_bench

progress-flicker

brucemacd/forward-test

brucemacd/go_qwen2

pdevine/gemma2

jmorganca/add-missing-symlink-eval

mxyng/next-debug

parth/set-context-size-openai

brucemacd/next-bpe-bench

brucemacd/next-bpe-test

brucemacd/new_runner_e2e

brucemacd/new_runner_qwen2

pdevine/convert-cohere2

brucemacd/convert-cli

parth/log-probs

mxyng/next-mlx

mxyng/cmd-history

parth/templating

parth/tokenize-detokenize

brucemacd/check-key-register

bmizerany/grammar

jmorganca/vendor-081b29bd

mxyng/func-checks

jmorganca/fix-null-format

parth/fix-default-to-warn-json

jmorganca/qwen2vl

jmorganca/no-concat

parth/cmd-cleanup-SO

brucemacd/check-key-register-structured-err

parth/openai-stream-usage

parth/fix-referencing-so

stream-tools-stop

jmorganca/degin-1

brucemacd/install-path-clean

brucemacd/push-name-validation

brucemacd/browser-key-register

jmorganca/openai-fix-first-message

jmorganca/fix-proxy

jessegross/sample

parth/disallow-streaming-tools

dhiltgen/remove_submodule

jmorganca/ga

jmorganca/mllama

pdevine/newlines

pdevine/geems-2b

jmorganca/llama-bump

mxyng/modelname-7

mxyng/gin-slog

mxyng/modelname-6

jyan/convert-prog

jyan/quant5

paligemma-support

pdevine/import-docs

jmorganca/openai-context

jyan/paligemma

jyan/p2

jyan/palitest

bmizerany/embedspeedup

jmorganca/llama-vit

brucemacd/allow-ollama

royh/ep-methods

royh/whisper

mxyng/api-models

mxyng/fix-memory

jyan/q4_4/8

jyan/ollama-v

royh/stream-tools

roy-embed-parallel

bmizerany/hrm

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

royh/embed-viz

jyan/local2

jyan/auth

jyan/local

jyan/parse-temp

jmorganca/template-mistral

jyan/reord-g

royh-openai-suffixdocs

royh-imgembed

royh-embed-parallel

jyan/quant4

royh-precision

jyan/progress

pdevine/fix-template

jyan/quant3

pdevine/ggla

mxyng/update-registry-domain

jmorganca/ggml-static

mxyng/create-context

jyan/v0.146

mxyng/layers-from-files

build_dist

bmizerany/noseek

royh-ls

royh-name

timeout

mxyng/server-timestamp

bmizerany/nosillyggufslurps

royh-params

jmorganca/llama-cpp-7c26775

royh-openai-delete

royh-show-rigid

jmorganca/enable-fa

jmorganca/no-error-template

jyan/format

royh-testdelete

bmizerany/fastverify

language_support

pdevine/ps-glitches

brucemacd/tokenize

bruce/iq-quants

bmizerany/filepathwithcoloninhost

mxyng/split-bin

bmizerany/client-registry

jmorganca/if-none-match

native

jmorganca/native

jmorganca/batch-embeddings

jmorganca/initcmake

jmorganca/mm

pdevine/showggmlinfo

modenameenforcealphanum

bmizerany/modenameenforcealphanum

jmorganca/done-reason

jmorganca/llama-cpp-8960fe8

ollama.com

bmizerany/filepathnobuild

bmizerany/types/model/defaultfix

rmdisplaylong

nogogen

bmizerany/x

modelfile-readme

bmizerany/replacecolon

jmorganca/limit

jmorganca/execstack

jmorganca/replace-assets

mxyng/tune-concurrency

jmorganca/testing

whitespace-detection

jmorganca/options

upgrade-all

scratch

cuda-search

mattw/airenamer

mattw/allmodelsonhuggingface

mattw/quantcontext

mattw/whatneedstorun

brucemacd/llama-mem-calc

mattw/faq-context

mattw/communitylinks

mattw/noprune

mattw/python-functioncalling

rename

mxyng/install

pulse

remove-first

editor

mattw/selfqueryingretrieval

cgo

mattw/howtoquant

api

matt/streamingapi

format-config

mxyng/extra-args

shell

update-nous-hermes

cp-model

upload-progress

fix-unknown-model

fix-model-names

delete-fix

insecure-registry

ls

deletemodels

progressbar

readme-updates

license-layers

skip-list

list-models

modelpath

matt/examplemodelfiles

distribution

go-opts

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/ollama#35254