[GH-ISSUE #7606] vram usage does not go back down after model unloads - stuck in Stopping... #30613

New Issue

GiteaMirror · 2026-04-22T10:25:55-05:00

GiteaMirror commented

2026-04-22 10:25:55 -05:00

Originally created by @CraftMaster163 on GitHub (Nov 11, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7606

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

when i set keep alive to 0 the memory usage does not go all the way back down. also it uses system ram when vram still avalible

gpu 7800xt
platform windows
rocm version 6.1

OS

Windows

GPU

AMD

CPU

AMD

Ollama version

0.4.1

Originally created by @CraftMaster163 on GitHub (Nov 11, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7606 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? when i set keep alive to 0 the memory usage does not go all the way back down. also it uses system ram when vram still avalible gpu 7800xt platform windows rocm version 6.1 ### OS Windows ### GPU AMD ### CPU AMD ### Ollama version 0.4.1

GiteaMirror added the amd bug labels 2026-04-22 10:25:55 -05:00

GiteaMirror closed this issue

2026-04-22 10:25:57 -05:00

GiteaMirror commented

2026-04-22 10:26:00 -05:00

@CraftMaster163 commented on GitHub (Nov 11, 2024):

server-3.log
app-4.log

@CraftMaster163 commented on GitHub (Nov 11, 2024): [server-3.log](https://github.com/user-attachments/files/17694531/server-3.log) [app-4.log](https://github.com/user-attachments/files/17694533/app-4.log)

GiteaMirror commented

2026-04-22 10:26:00 -05:00

@rick-github commented on GitHub (Nov 11, 2024):

What's the output of ollama ps?

@rick-github commented on GitHub (Nov 11, 2024): What's the output of `ollama ps`?

GiteaMirror commented

2026-04-22 10:26:01 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

screenshot of before and after closing ollama
and output of ollama ps

@CraftMaster163 commented on GitHub (Nov 13, 2024): ![image](https://github.com/user-attachments/assets/f6fd1af9-9f79-4b55-a61a-6dad3fcf531e) screenshot of before and after closing ollama and output of ollama ps ![image](https://github.com/user-attachments/assets/669e9805-883a-44c3-9840-fc3f631d1aef)

GiteaMirror commented

2026-04-22 10:26:01 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

here is when a model runs and stops

@CraftMaster163 commented on GitHub (Nov 13, 2024): ![image](https://github.com/user-attachments/assets/0f15959f-bee8-4188-af1b-34314dca4299) here is when a model runs and stops

GiteaMirror commented

2026-04-22 10:26:02 -05:00

@rick-github commented on GitHub (Nov 13, 2024):

Looks like VRAM usage returns the level it was before the model was run. Other processes are using VRAM - browser, media player, GUI, etc.

@rick-github commented on GitHub (Nov 13, 2024): Looks like VRAM usage returns the level it was before the model was run. Other processes are using VRAM - browser, media player, GUI, etc.

GiteaMirror commented

2026-04-22 10:26:03 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

no if i quit ollama it goes back down, that is ollama using it

@CraftMaster163 commented on GitHub (Nov 13, 2024): no if i quit ollama it goes back down, that is ollama using it

GiteaMirror commented

2026-04-22 10:26:04 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

i found a work around, for n8n(which is what im mainly using ollama for atm) just tell it to kill ollama and it forces it to dump the memory, but that should not be required

@CraftMaster163 commented on GitHub (Nov 13, 2024): i found a work around, for n8n(which is what im mainly using ollama for atm) just tell it to kill ollama and it forces it to dump the memory, but that should not be required

GiteaMirror commented

2026-04-22 10:26:04 -05:00

@dhiltgen commented on GitHub (Nov 13, 2024):

@CraftMaster163 can you check detailed processes to see if ollama_llama_server is still running when you see the memory leakage, or if only the main ollama process is running? When the model goes idle, we should shut down the ollama_llama_server which should release GPU memory, however we do GPU discovery in the main ollama process. That code should unload the GPU libraries when not actively querying to avoid additional VRAM usage.

Looking at your driver version, this might be a dup of #7107

Does the behavior disappear if you downgrade the driver to 24.8.1?

@dhiltgen commented on GitHub (Nov 13, 2024): @CraftMaster163 can you check detailed processes to see if `ollama_llama_server` is still running when you see the memory leakage, or if only the main `ollama` process is running? When the model goes idle, we should shut down the `ollama_llama_server` which should release GPU memory, however we do GPU discovery in the main `ollama` process. That code should unload the GPU libraries when not actively querying to avoid additional VRAM usage. Looking at your driver version, this might be a dup of #7107 Does the behavior disappear if you downgrade the driver to 24.8.1?

GiteaMirror commented

2026-04-22 10:26:05 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

yes i just checked and ollama_llama_server is still running, if i close it the vram goes down again

@CraftMaster163 commented on GitHub (Nov 13, 2024): yes i just checked and ollama_llama_server is still running, if i close it the vram goes down again

GiteaMirror commented

2026-04-22 10:26:06 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

also ollama.exe does not get rid of its vram usage, if i stop it, it restarts fine and goes back down to my normal vram usage

@CraftMaster163 commented on GitHub (Nov 13, 2024): also ollama.exe does not get rid of its vram usage, if i stop it, it restarts fine and goes back down to my normal vram usage

GiteaMirror commented

2026-04-22 10:26:06 -05:00

@dhiltgen commented on GitHub (Nov 13, 2024):

@CraftMaster163 can you clarify if ollama ps shows nothing running while ollama_llama_server is still running?

@dhiltgen commented on GitHub (Nov 13, 2024): @CraftMaster163 can you clarify if `ollama ps` shows nothing running while `ollama_llama_server` is still running?

GiteaMirror commented

2026-04-22 10:26:07 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

i assume /bye unloads model

@CraftMaster163 commented on GitHub (Nov 13, 2024): ![image](https://github.com/user-attachments/assets/f17f221c-21dc-4b79-a60e-17944cce479c) i assume /bye unloads model

GiteaMirror commented

2026-04-22 10:26:08 -05:00

@dhiltgen commented on GitHub (Nov 13, 2024):

i assume /bye unloads model

That just closes the client, but the model will stay loaded until the timeout expires. By default this is 5m but you can override. Can you verify if the VRAM usage drops back down once ollama ps no longer shows a model loaded?

https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-keep-a-model-loaded-in-memory-or-make-it-unload-immediately

@dhiltgen commented on GitHub (Nov 13, 2024): > i assume /bye unloads model That just closes the client, but the model will stay loaded until the timeout expires. By default this is 5m but you can override. Can you verify if the VRAM usage drops back down once `ollama ps` no longer shows a model loaded? https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-keep-a-model-loaded-in-memory-or-make-it-unload-immediately

GiteaMirror commented

2026-04-22 10:26:08 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

hm now it closes, when i removed the commands from n8n, it does stop but ollama.exe does not clear its cache, here is a photo showing the graph going down after closing ollama

@CraftMaster163 commented on GitHub (Nov 13, 2024): ![image](https://github.com/user-attachments/assets/59ad05e5-27bd-46f3-82fa-a9d5777624ab) hm now it closes, when i removed the commands from n8n, it does stop but ollama.exe does not clear its cache, here is a photo showing the graph going down after closing ollama ![image](https://github.com/user-attachments/assets/13ab5c1b-3cc8-46e6-93fc-59601b9c9af2)

GiteaMirror commented

2026-04-22 10:26:09 -05:00

@dhiltgen commented on GitHub (Nov 13, 2024):

hm now it closes, when i removed the commands from n8n, it does stop but ollama.exe does not clear its cache, here is a photo showing the graph going down after closing ollama

Can you clarify? When ollama ps shows no model loaded, does ollama_llama_server stop and does the VRAM usage drop back down to the baseline you area expecting? If so, this doesn't sound like a bug.

I'm not familiar with n8n, but if it calls the Ollama API and keeps the model loaded, that might explain what you are seeing.

@dhiltgen commented on GitHub (Nov 13, 2024): > hm now it closes, when i removed the commands from n8n, it does stop but ollama.exe does not clear its cache, here is a photo showing the graph going down after closing ollama Can you clarify? When `ollama ps` shows no model loaded, does `ollama_llama_server` stop and does the VRAM usage drop back down to the baseline you area expecting? If so, this doesn't sound like a bug. I'm not familiar with n8n, but if it calls the Ollama API and keeps the model loaded, that might explain what you are seeing.

GiteaMirror commented

2026-04-22 10:26:09 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

no i told it to unload model, there is a spike, it unloads the context but does not fully unload model, here is a screenshot of the three spikes. normal usage

after ollama runs with keep alive 0

and after i quit ollama

@CraftMaster163 commented on GitHub (Nov 13, 2024): no i told it to unload model, there is a spike, it unloads the context but does not fully unload model, here is a screenshot of the three spikes. normal usage ![image](https://github.com/user-attachments/assets/f7278702-cad9-4efa-81c4-71553d6947d9) after ollama runs with keep alive 0 ![image](https://github.com/user-attachments/assets/aa0fec5d-069a-4d13-a702-984224bfa254) and after i quit ollama ![image](https://github.com/user-attachments/assets/b53c0e81-88f3-4e83-97fb-f9bd860b4386)

GiteaMirror commented

2026-04-22 10:26:10 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

if i let my discord bot using ollama run for a while it goes to full usage, its just llama3.2:3b which does not need that much memory

@CraftMaster163 commented on GitHub (Nov 13, 2024): if i let my discord bot using ollama run for a while it goes to full usage, its just llama3.2:3b which does not need that much memory

GiteaMirror commented

2026-04-22 10:26:10 -05:00

@CraftMaster163 commented on GitHub (Nov 13, 2024):

https://drive.google.com/file/d/1x_CYMP45GIFcMIAYwTgio3_oYyrUsPNP/view?usp=sharing
here is a video of the issue

@CraftMaster163 commented on GitHub (Nov 13, 2024): https://drive.google.com/file/d/1x_CYMP45GIFcMIAYwTgio3_oYyrUsPNP/view?usp=sharing here is a video of the issue

GiteaMirror commented

2026-04-22 10:26:11 -05:00

@dhiltgen commented on GitHub (Apr 9, 2025):

Related issues #8178 #9617 #8969 #10119

@dhiltgen commented on GitHub (Apr 9, 2025): Related issues #8178 #9617 #8969 #10119

GiteaMirror commented

2026-04-22 10:26:12 -05:00

@dhiltgen commented on GitHub (Apr 9, 2025):

There seems to be a race somewhere in the scheduler under heavy load, possibly related to clients closing connections prematurely. If people are still seeing models get stuck in a "Stopping..." state in the ollama ps output and the model never actually unloads, please try running the server with OLLAMA_DEBUG=1 and share the logs including the model load, and eventual stuck state.

@dhiltgen commented on GitHub (Apr 9, 2025): There seems to be a race somewhere in the scheduler under heavy load, possibly related to clients closing connections prematurely. If people are still seeing models get stuck in a "Stopping..." state in the `ollama ps` output and the model never actually unloads, please try running the server with OLLAMA_DEBUG=1 and share the logs including the model load, and eventual stuck state.

GiteaMirror commented

2026-04-22 10:26:13 -05:00

@rick-github commented on GitHub (Apr 9, 2025):

Not necessarily high load. If the model loses coherence, then it will remain in the "Stopping..." state until the token limit count is exceeded or the client disconnects. If the context buffer is large and the client doesn't have a timeout, this can take a while.

$ ollama -v
ollama version is 0.6.5
$ ollama ps
NAME    ID    SIZE    PROCESSOR    UNTIL 
$ curl -s localhost:11434/api/generate -d '{"model":"gemma3:4b","prompt":'"$((echo write a story based on the following text: ; head -8000 /etc/dictionaries-common/words) | jq -sR)"',"options":{"num_ctx":32768}}' -o /dev/null &
[1] 3044530
$ ollama stop gemma3:4b ; date
Thu Apr 10 12:17:06 AM CEST 2025
$ while : ; do ollama ps ; date ; sleep 1m ; done
...
NAME         ID              SIZE      PROCESSOR    UNTIL       
gemma3:4b    c0494fe00251    6.4 GB    100% GPU     Stopping...    
Thu Apr 10 12:33:40 AM CEST 2025
...

@rick-github commented on GitHub (Apr 9, 2025): Not necessarily high load. If the model loses coherence, then it will remain in the "Stopping..." state until the token limit count is exceeded or the client disconnects. If the context buffer is large and the client doesn't have a timeout, this can take a while. ```console $ ollama -v ollama version is 0.6.5 $ ollama ps NAME ID SIZE PROCESSOR UNTIL $ curl -s localhost:11434/api/generate -d '{"model":"gemma3:4b","prompt":'"$((echo write a story based on the following text: ; head -8000 /etc/dictionaries-common/words) | jq -sR)"',"options":{"num_ctx":32768}}' -o /dev/null & [1] 3044530 $ ollama stop gemma3:4b ; date Thu Apr 10 12:17:06 AM CEST 2025 $ while : ; do ollama ps ; date ; sleep 1m ; done ... NAME ID SIZE PROCESSOR UNTIL gemma3:4b c0494fe00251 6.4 GB 100% GPU Stopping... Thu Apr 10 12:33:40 AM CEST 2025 ... ```

GiteaMirror commented

2026-04-22 10:26:13 -05:00

@Anaphylaxis commented on GitHub (Apr 18, 2025):

I'm experiencing an issue where loading multiple models (one after the other) seems to be persisting in the VRAM even after the model is completely stopped. Model after model will lower the GPU% and raise CPU%, and even with no models in ollama ps the shared VRAM is maxed out on my GPU after multiple sessions. Quitting ollama app.exe completely released all my VRAM, and on Ollama restart models are using 100% GPU again, so I'm 100% sure it's Ollama failing to reclaim the space, perhaps when a model is booted off for an embedding model. In the logs it does show that my available memory is less and less and less every time.

@Anaphylaxis commented on GitHub (Apr 18, 2025): I'm experiencing an issue where loading multiple models (one after the other) seems to be persisting in the VRAM even after the model is completely stopped. Model after model will lower the GPU% and raise CPU%, and even with no models in `ollama ps` the shared VRAM is maxed out on my GPU after multiple sessions. Quitting `ollama app.exe` completely released all my VRAM, and on Ollama restart models are using 100% GPU again, so I'm 100% sure it's Ollama failing to reclaim the space, perhaps when a model is booted off for an embedding model. In the logs it does show that my available memory is less and less and less every time.

GiteaMirror commented

2026-04-22 10:26:14 -05:00

@rick-github commented on GitHub (Apr 18, 2025):

Server logs and the output of nivida-smi will aid in debugging.

@rick-github commented on GitHub (Apr 18, 2025): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) and the output of `nivida-smi` will aid in debugging.

GiteaMirror commented

2026-04-22 10:26:15 -05:00

@Anaphylaxis commented on GitHub (Apr 18, 2025):

Server logs and the output of nivida-smi will aid in debugging.

I am using ROCm and Windows so there is no nvidia-smi for me, that's why I referenced the shared VRAM usage in Task Manager.
I didn't want to post hundreds of lines of logs if unnecessary, here they are.

ollamalog.txt

You can see we start out as memory.available="[23.7 GiB]" and towards the end we're memory.available="[12.7 GiB]" and 63 gpu layers that quickly starts dropping.
time=2025-04-18T10:47:45.183-04:00 level=INFO source=sched.go:187 msg="one or more GPUs detected that are unable to accurately report free memory - disabling default concurrency" could be related but I assume this is talking about the virtual display driver I'm using to remote into this machine.

@Anaphylaxis commented on GitHub (Apr 18, 2025): > [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) and the output of `nivida-smi` will aid in debugging. I am using ROCm and Windows so there is no `nvidia-smi` for me, that's why I referenced the shared VRAM usage in Task Manager. I didn't want to post hundreds of lines of logs if unnecessary, here they are. [ollamalog.txt](https://github.com/user-attachments/files/19813583/ollamalog.txt) You can see we start out as `memory.available="[23.7 GiB]"` and towards the end we're `memory.available="[12.7 GiB]" ` and 63 gpu layers that quickly starts dropping. `time=2025-04-18T10:47:45.183-04:00 level=INFO source=sched.go:187 msg="one or more GPUs detected that are unable to accurately report free memory - disabling default concurrency"` could be related but I assume this is talking about the virtual display driver I'm using to remote into this machine.

GiteaMirror commented

2026-04-22 10:26:16 -05:00

@rick-github commented on GitHub (Apr 18, 2025):

This might be a different issue in that you don't have models stuck in a "Stopping..." state. It also seems to affect both the old llama.cpp runner and the new go runner. I suggest opening a new ticket so this one doesn't get mixed in with a similar but different problem.

@rick-github commented on GitHub (Apr 18, 2025): This might be a different issue in that you don't have models stuck in a "Stopping..." state. It also seems to affect both the old llama.cpp runner and the new go runner. I suggest opening a new ticket so this one doesn't get mixed in with a similar but different problem.

GiteaMirror commented

2026-04-22 10:26:17 -05:00

@Anaphylaxis commented on GitHub (Apr 18, 2025):

Thanks, that's what I thought. I'm not sure whether to make a llama.cpp issue or an Ollama issue. I see some relevant issues but they're all closed.

@Anaphylaxis commented on GitHub (Apr 18, 2025): Thanks, that's what I thought. I'm not sure whether to make a llama.cpp issue or an Ollama issue. I see some relevant issues but they're all closed.

GiteaMirror commented

2026-04-22 10:26:17 -05:00

@rick-github commented on GitHub (Apr 18, 2025):

It seems to affect both runners so it might be a rocm issue.

@rick-github commented on GitHub (Apr 18, 2025): It seems to affect both runners so it might be a rocm issue.

Sign in to join this conversation.

Branches Tags

main

dhiltgen/ci

parth-launch-plan-gating

hoyyeva/anthropic-reference-images-path

parth-anthropic-reference-images-path

brucemacd/download-before-remove

hoyyeva/editor-config-repair

parth-mlx-decode-checkpoints

parth-launch-codex-app

hoyyeva/fix-codex-model-metadata-warning

hoyyeva/qwen

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

parth/update-claude-docs

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#30613