[GH-ISSUE #1736] Download slows to a crawl at 99% #26751

New Issue

GiteaMirror · 2026-04-22T03:16:26-05:00

GiteaMirror commented

2026-04-22 03:16:26 -05:00

Originally created by @Pugio on GitHub (Dec 29, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/1736

For every model I've downloaded, the speed saturates my bandwidth (~13MB/sec) until it hits 98/99%. Then the download slows to a few tens of KB/s and takes hour(s) to finish.

I've tried multiple models and this behavior happens each time. Happy to debug, but I'm not sure what to try.

I'm in Australia, in case that matters.

Originally created by @Pugio on GitHub (Dec 29, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/1736 For every model I've downloaded, the speed saturates my bandwidth (~13MB/sec) until it hits 98/99%. Then the download slows to a few tens of KB/s and takes hour(s) to finish. <img width="884" alt="image" src="https://github.com/jmorganca/ollama/assets/286180/e47037e1-aea8-4a13-a6fc-7841baa0db6c"> I've tried multiple models and this behavior happens each time. Happy to debug, but I'm not sure what to try. I'm in Australia, in case that matters.

GiteaMirror added the networking registry bug labels 2026-04-22 03:16:26 -05:00

GiteaMirror commented

2026-04-22 03:16:29 -05:00

@Sully233 commented on GitHub (Dec 29, 2023):

I'm also experiencing this exact issue.

The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed.

@Sully233 commented on GitHub (Dec 29, 2023): I'm also experiencing this exact issue. The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed.

GiteaMirror commented

2026-04-22 03:16:32 -05:00

@pdevine commented on GitHub (Jan 2, 2024):

@Pugio can you run ollama --version? Also, can you take a look at the logs? I'm curious if there is something like:

[GIN] 2024/01/02 - 14:04:51 | 200 |      32.625µs |       127.0.0.1 | HEAD     "/"
2024/01/02 14:04:53 download.go:123: downloading e9e56e8bb5f0 in 64 413 MB part(s)
2024/01/02 14:05:53 download.go:162: e9e56e8bb5f0 part 22 attempt 0 failed: unexpected EOF, retrying in 1s
2024/01/02 14:05:53 download.go:162: e9e56e8bb5f0 part 46 attempt 0 failed: unexpected EOF, retrying in 1s
2024/01/02 14:10:26 download.go:123: downloading 43070e2d4e53 in 1 11 KB part(s)
2024/01/02 14:10:28 download.go:123: downloading ed11eda7790d in 1 30 B part(s)
2024/01/02 14:10:31 download.go:123: downloading 9dec05e9b2db in 1 484 B part(s)
[GIN] 2024/01/02 - 14:10:44 | 200 |         5m53s |       127.0.0.1 | POST     "/api/pull"

@pdevine commented on GitHub (Jan 2, 2024): @Pugio can you run `ollama --version`? Also, can you take a look at the logs? I'm curious if there is something like: ``` [GIN] 2024/01/02 - 14:04:51 | 200 | 32.625µs | 127.0.0.1 | HEAD "/" 2024/01/02 14:04:53 download.go:123: downloading e9e56e8bb5f0 in 64 413 MB part(s) 2024/01/02 14:05:53 download.go:162: e9e56e8bb5f0 part 22 attempt 0 failed: unexpected EOF, retrying in 1s 2024/01/02 14:05:53 download.go:162: e9e56e8bb5f0 part 46 attempt 0 failed: unexpected EOF, retrying in 1s 2024/01/02 14:10:26 download.go:123: downloading 43070e2d4e53 in 1 11 KB part(s) 2024/01/02 14:10:28 download.go:123: downloading ed11eda7790d in 1 30 B part(s) 2024/01/02 14:10:31 download.go:123: downloading 9dec05e9b2db in 1 484 B part(s) [GIN] 2024/01/02 - 14:10:44 | 200 | 5m53s | 127.0.0.1 | POST "/api/pull" ```

GiteaMirror commented

2026-04-22 03:16:33 -05:00

@Pugio commented on GitHub (Jan 2, 2024):

Version 0.1.17.
Tinyllama pulled fine, so tried it on Mistral and got the slowdown. I'm 3.8/4.1 GB and the download speed went from 13MB/s (saturating my connection) to a consistent 600-700KB/s.

The only log entries for this pull are:

2024/01/03 10:31:36 download.go:123: downloading e8a35b5937a5 in 42 100 MB part(s)
2024/01/03 10:34:22 download.go:162: e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s
[GIN] 2024/01/03 - 10:38:34 | 200 |      90.875µs |       127.0.0.1 | GET      "/api/version"

EDIT: PatchingInitiative's trick of cancelling and restarting the pull has worked for me fairly well in the past few days, though for some models I needed to do that a couple of times.

@Pugio commented on GitHub (Jan 2, 2024): Version `0.1.17`. Tinyllama pulled fine, so tried it on Mistral and got the slowdown. I'm 3.8/4.1 GB and the download speed went from 13MB/s (saturating my connection) to a consistent 600-700KB/s. The only log entries for this `pull` are: ``` 2024/01/03 10:31:36 download.go:123: downloading e8a35b5937a5 in 42 100 MB part(s) 2024/01/03 10:34:22 download.go:162: e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s [GIN] 2024/01/03 - 10:38:34 | 200 | 90.875µs | 127.0.0.1 | GET "/api/version" ``` EDIT: PatchingInitiative's trick of cancelling and restarting the `pull` has worked for me fairly well in the past few days, though for some models I needed to do that a couple of times.

GiteaMirror commented

2026-04-22 03:16:33 -05:00

@Pugio commented on GitHub (Jan 3, 2024):

Additional logs after download finished:

2024/01/03 10:41:14 download.go:162: e8a35b5937a5 part 23 attempt 0 failed: unexpected EOF, retrying in 1s
2024/01/03 10:47:21 download.go:162: e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s
2024/01/03 10:50:58 download.go:162: e8a35b5937a5 part 39 attempt 0 failed: unexpected EOF, retrying in 1s
2024/01/03 10:51:08 download.go:123: downloading 43070e2d4e53 in 1 11 KB part(s)
2024/01/03 10:51:12 download.go:123: downloading f9b1e3196ecf in 1 483 B part(s)
[GIN] 2024/01/03 - 10:51:16 | 200 |        19m42s |       127.0.0.1 | POST     "/api/pull"

@Pugio commented on GitHub (Jan 3, 2024): Additional logs after download finished: ``` 2024/01/03 10:41:14 download.go:162: e8a35b5937a5 part 23 attempt 0 failed: unexpected EOF, retrying in 1s 2024/01/03 10:47:21 download.go:162: e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s 2024/01/03 10:50:58 download.go:162: e8a35b5937a5 part 39 attempt 0 failed: unexpected EOF, retrying in 1s 2024/01/03 10:51:08 download.go:123: downloading 43070e2d4e53 in 1 11 KB part(s) 2024/01/03 10:51:12 download.go:123: downloading f9b1e3196ecf in 1 483 B part(s) [GIN] 2024/01/03 - 10:51:16 | 200 | 19m42s | 127.0.0.1 | POST "/api/pull" ```

GiteaMirror commented

2026-04-22 03:16:34 -05:00

@raidmax21 commented on GitHub (Jan 6, 2024):

Same issue here.

It seems to happen on all the models I download.

It starts out very quick using almost all of my 1gbit connection, but once it passes about 98% it drops right off to below 1mbps.

On a side note I'm using this in a docker on unraid 6.12.4.

ollama version is 0.1.18

root@43f90abcec0d:/# ollama run llama2:13b-text pulling manifest pulling a42778cb0676... 99% ▕█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 7.3 GB/7.4 GB 42 KB/s 18m17s

@raidmax21 commented on GitHub (Jan 6, 2024): Same issue here. It seems to happen on all the models I download. It starts out very quick using almost all of my 1gbit connection, but once it passes about 98% it drops right off to below 1mbps. On a side note I'm using this in a docker on unraid 6.12.4. `ollama version is 0.1.18` `` root@43f90abcec0d:/# ollama run llama2:13b-text pulling manifest pulling a42778cb0676... 99% ▕█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 7.3 GB/7.4 GB 42 KB/s 18m17s ``

GiteaMirror commented

2026-04-22 03:16:34 -05:00

@FustarReal commented on GitHub (Jan 9, 2024):

I tried to download a model, it went to 48% with 26 mb/s then it slowed to 1 mb/s then I restarted my wifi and it went back to 20% only downloading at 4mb/s, tried to reset my wifi and stop the download as someone suggested and it didnt work, it still downloads only at 4mb/s

@FustarReal commented on GitHub (Jan 9, 2024): I tried to download a model, it went to 48% with 26 mb/s then it slowed to 1 mb/s then I restarted my wifi and it went back to 20% only downloading at 4mb/s, tried to reset my wifi and stop the download as someone suggested and it didnt work, it still downloads only at 4mb/s

GiteaMirror commented

2026-04-22 03:16:35 -05:00

@pdevine commented on GitHub (Jan 9, 2024):

cc @mxyng

@pdevine commented on GitHub (Jan 9, 2024): cc @mxyng

GiteaMirror commented

2026-04-22 03:16:35 -05:00

@mxyng commented on GitHub (Jan 11, 2024):

Some context:

Ollama downloads large files in parts with multiple concurrent workers. This maximizes transfer speed allow users to get their files faster.

The problem seems to be certain parts stall completely and zero data is received from the backend. The connection itself is still healthy so it doesn't trigger a retry. When this happens, the part stands out as other parts finish making it very visible in the last few percentages.

The linked PR aims to mitigate this by detecting these stalls and preemptively reset the connection before it can become a visible problem. The real solution is to address the storage backend so it never gets into this stalled state.

Aside: it seems the EOFs are at least correlated with this stalling behaviour. While the EOFs themselves will not cause any problems (the request will retry and continue where it left off), it's an indicator something's not quite right with the storage backend.

@mxyng commented on GitHub (Jan 11, 2024): Some context: Ollama downloads large files in parts with multiple concurrent workers. This maximizes transfer speed allow users to get their files faster. The problem seems to be certain parts stall completely and zero data is received from the backend. The connection itself is still healthy so it doesn't trigger a retry. When this happens, the part stands out as other parts finish making it very visible in the last few percentages. The linked PR aims to mitigate this by detecting these stalls and preemptively reset the connection before it can become a visible problem. The real solution is to address the storage backend so it never gets into this stalled state. Aside: it seems the EOFs are at least correlated with this stalling behaviour. While the EOFs themselves will not cause any problems (the request will retry and continue where it left off), it's an indicator something's not quite right with the storage backend.

GiteaMirror commented

2026-04-22 03:16:36 -05:00

@ewebgh33 commented on GitHub (Jan 31, 2024):

**I have this same issue today and it's very VERY frustrating.
Running latest - just did an update.

I've pulled two 70b models and now a 34b model, and each time it rips until it gets to around 96%
Then it takes an hour to do the last few %.**
As seen in the screenshot it also sits there for 10mins or so (2min at the time I took the screenshot) when it's at 100% already. If it's at 100% why does it keep downloading for 10mins?

Why is this??
I wish we could download models directly into a folder, and not have to wait for Ollama to pull them itself.

Aside -
On that note, I also wish this because I have a bunch of models already, but it seems Ollama wants them in it's own special hashed (no file names) format, instead of just downloading a model from TheBloke directly to the folder, like can be done with textgen-webui (oobabooga).

So if you want to use a model with both apps or share models (heaven forbid, because we're all testing apps each day), we end up with each app wanting to download it's own version of the model and the terabytes just vanish from drives.

@ewebgh33 commented on GitHub (Jan 31, 2024): ![WindowsTerminal_b9NOWnxluf](https://github.com/ollama/ollama/assets/123797054/569b8cd1-5f7d-47ea-b425-bb166d156eee) **I have this same issue today and it's very VERY frustrating. Running latest - just did an update. I've pulled two 70b models and now a 34b model, and each time it rips until it gets to around 96% Then it takes an hour to do the last few %.** As seen in the screenshot it also sits there for 10mins or so (2min at the time I took the screenshot) when it's at 100% already. If it's at 100% why does it keep downloading for 10mins? Why is this?? I wish we could download models directly into a folder, and not have to wait for Ollama to pull them itself. Aside - On that note, I also wish this because I have a bunch of models already, but it seems Ollama wants them in it's own special hashed (no file names) format, instead of just downloading a model from TheBloke directly to the folder, like can be done with textgen-webui (oobabooga). So if you want to use a model with both apps or share models (heaven forbid, because we're all testing apps each day), we end up with each app wanting to download it's own version of the model and the terabytes just vanish from drives.

GiteaMirror commented

2026-04-22 03:16:37 -05:00

@pdevine commented on GitHub (Jan 31, 2024):

@EmmaWebGH sorry you're running into this. It's as @mxyng mentioned though; one of the streams of data is hung, but it's reporting healthy. You should be able to hit Ctrl + c and then re-pull the image again and it should resume where it left off. Can you double check that you're running 0.1.22?

@pdevine commented on GitHub (Jan 31, 2024): @EmmaWebGH sorry you're running into this. It's as @mxyng mentioned though; one of the streams of data is hung, but it's reporting healthy. You should be able to hit `Ctrl + c` and then re-pull the image again and it should resume where it left off. Can you double check that you're running `0.1.22`?

GiteaMirror commented

2026-04-22 03:16:37 -05:00

@ewebgh33 commented on GitHub (Jan 31, 2024):

@pdevine OK I'll try that next time.
I was afraid it wouldn't resume so yesterday I didn't try that solution! :)

@ewebgh33 commented on GitHub (Jan 31, 2024): @pdevine OK I'll try that next time. I was afraid it wouldn't resume so yesterday I didn't try that solution! :)

GiteaMirror commented

2026-04-22 03:16:37 -05:00

@raidmax21 commented on GitHub (Feb 1, 2024):

There should be a fix in the next release ~ https://github.com/ollama/ollama/pull/1916

@raidmax21 commented on GitHub (Feb 1, 2024): There should be a fix in the next release ~ https://github.com/ollama/ollama/pull/1916

GiteaMirror commented

2026-04-22 03:16:38 -05:00

@marklysze commented on GitHub (Feb 3, 2024):

I just updated to 0.1.23 and tried to pull a model and it started off at full speed but near the end it slowed down to a crawl (as it had previously). So for me the fix didn't work (I'm in Australia).

Ctrl+C and running the command again finished the download at good speeds.

@marklysze commented on GitHub (Feb 3, 2024): I just updated to 0.1.23 and tried to pull a model and it started off at full speed but near the end it slowed down to a crawl (as it had previously). So for me the fix didn't work (I'm in Australia). Ctrl+C and running the command again finished the download at good speeds.

GiteaMirror commented

2026-04-22 03:16:39 -05:00

@raidmax21 commented on GitHub (Feb 3, 2024):

Gave the new version a test and sadly, 0.1.23 did not solve this issue.

@raidmax21 commented on GitHub (Feb 3, 2024): Gave the new version a test and sadly, 0.1.23 did not solve this issue.

GiteaMirror commented

2026-04-22 03:16:39 -05:00

@cyphbt commented on GitHub (Feb 27, 2024):

Stop the downloading process, and retry that will continue with high speed network.

@cyphbt commented on GitHub (Feb 27, 2024): Stop the downloading process, and retry that will continue with high speed network.

GiteaMirror commented

2026-04-22 03:16:40 -05:00

@rhuanbarros commented on GitHub (Mar 11, 2024):

I'm using Ollama version 0.1.28 and encountering the same issue. I tried to download the model using the REST API, but the download stalled. So I tried again, but this time I connected to the terminal and attempted to download. This time it worked fine. So perhaps the problem lies solely with using the REST API

@rhuanbarros commented on GitHub (Mar 11, 2024): I'm using Ollama version 0.1.28 and encountering the same issue. I tried to download the model using the REST API, but the download stalled. So I tried again, but this time I connected to the terminal and attempted to download. This time it worked fine. So perhaps the problem lies solely with using the REST API

GiteaMirror commented

2026-04-22 03:16:40 -05:00

@ZaneA commented on GitHub (Mar 12, 2024):

I am getting this with 0.1.28, and I am using the terminal and not the REST API (yet) :)

Hitting ctrl-c and running again helps it along as suggested.

@ZaneA commented on GitHub (Mar 12, 2024): I am getting this with 0.1.28, and I am using the terminal and not the REST API (yet) :) Hitting ctrl-c and running again helps it along as suggested.

GiteaMirror commented

2026-04-22 03:16:41 -05:00

@Kuiriel commented on GitHub (Mar 12, 2024):

Same issue with the latest Ollama as well. Also Ctrl-C and running it again helped. Thank you.

@Kuiriel commented on GitHub (Mar 12, 2024): Same issue with the latest Ollama as well. Also Ctrl-C and running it again helped. Thank you.

GiteaMirror commented

2026-04-22 03:16:41 -05:00

@Qualzz commented on GitHub (Mar 14, 2024):

Confirming the same issue with high 8gbps speed. Downloading mixtral model in a few seconds at around 800mb/s then at 98/99% is goes down to around 100Kb/s and takes forever. Ctrl C and running again works fine.

@Qualzz commented on GitHub (Mar 14, 2024): Confirming the same issue with high 8gbps speed. Downloading mixtral model in a few seconds at around 800mb/s then at 98/99% is goes down to around 100Kb/s and takes forever. Ctrl C and running again works fine.

GiteaMirror commented

2026-04-22 03:16:42 -05:00

@marschr commented on GitHub (Mar 25, 2024):

0.1.29 here and still getting this issue.. Ctrl+C and resuming the download seems to work, but something is flaky..

@marschr commented on GitHub (Mar 25, 2024): 0.1.29 here and still getting this issue.. `Ctrl+C` and resuming the download seems to work, but something is flaky..

GiteaMirror commented

2026-04-22 03:16:42 -05:00

@frosty1492 commented on GitHub (Mar 29, 2024):

I'm having this same issue. I have a slow link, only 10 mbs but I have plenty of time and the connection is extremely stable. Old DSL server line. The download stops after it downloads about 38 MB. By using the terminal commands ctrl C and then restarting the with the command; ollama pull llama2:70b I can get about 80 mb before it craps out. I have concluded that this is an issue with either Ollama or the module server. It seems to be a handshake error. Yes, I am aware that this is a larger file but my system is very capable of managing it. I just live in an area where we do not have great internet service.

@frosty1492 commented on GitHub (Mar 29, 2024): I'm having this same issue. I have a slow link, only 10 mbs but I have plenty of time and the connection is extremely stable. Old DSL server line. The download stops after it downloads about 38 MB. By using the terminal commands ctrl C and then restarting the with the command; ollama pull llama2:70b I can get about 80 mb before it craps out. I have concluded that this is an issue with either Ollama or the module server. It seems to be a handshake error. Yes, I am aware that this is a larger file but my system is very capable of managing it. I just live in an area where we do not have great internet service.

GiteaMirror commented

2026-04-22 03:16:43 -05:00

@chuklee commented on GitHub (Apr 2, 2024):

I have the same issue, i have this log on my docker:
ollama | time=2024-04-02T14:58:04.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying"ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 3 stalled; retrying"
ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 14 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 41 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 33 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 4 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 16 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 12 stalled; retrying"
ollama | time=2024-04-02T14:59:10.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 10 stalled; retrying"
ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T14:59:50.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 10 stalled; retrying"
ollama | time=2024-04-02T14:59:50.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 16 stalled; retrying"
ollama | time=2024-04-02T14:59:51.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 28 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 9 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 7 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T14:59:51.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 41 stalled; retrying"
ollama | time=2024-04-02T15:00:05.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:00:05.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T15:00:07.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 8 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 24 stalled; retrying"
ollama | time=2024-04-02T15:00:09.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 32 stalled; retrying"
ollama | time=2024-04-02T15:00:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 12 stalled; retrying"
ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T15:00:14.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 21 stalled; retrying"
ollama | time=2024-04-02T15:00:37.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:00:38.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T15:00:52.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T15:00:54.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 32 stalled; retrying"
ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 27 stalled; retrying"
ollama | time=2024-04-02T15:00:58.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 4 stalled; retrying"
ollama | time=2024-04-02T15:01:02.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T15:01:04.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T15:01:06.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 0 stalled; retrying"
ollama | time=2024-04-02T15:01:14.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:01:16.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 22 stalled; retrying"
ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 19 stalled; retrying"
ollama | time=2024-04-02T15:01:22.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T15:01:23.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 33 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 14 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying"
ollama | time=2024-04-02T15:01:24.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:01:28.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 27 stalled; retrying"
ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 3 stalled; retrying"
ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 0 stalled; retrying"
ollama | time=2024-04-02T15:01:34.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T15:01:35.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying"
ollama | time=2024-04-02T15:04:36.697Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 11 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:04:37.802Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:04:41.698Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 28 stalled; retrying"
ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T15:05:16.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:05:25.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 7 stalled; retrying"
ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 9 stalled; retrying"
ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:05:59.854Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 36 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:07:06.904Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:07:38.614Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 21 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:08:52.542Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 15 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:09:14.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 8 stalled; retrying"
ollama | time=2024-04-02T15:09:16.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:09:21.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:09:22.699Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T15:09:26.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:09:48.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 22 stalled; retrying"
ollama | time=2024-04-02T15:09:48.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T15:09:48.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:09:49.615Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 21 stalled; retrying"
ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying"

@chuklee commented on GitHub (Apr 2, 2024): I have the same issue, i have this log on my docker: ollama | time=2024-04-02T14:58:04.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying"ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying" ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying" ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 3 stalled; retrying" ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying" ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying" ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying" ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 14 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 41 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 33 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying" ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying" ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying" ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying" ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 4 stalled; retrying" ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 16 stalled; retrying" ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 12 stalled; retrying" ollama | time=2024-04-02T14:59:10.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 10 stalled; retrying" ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying" ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying" ollama | time=2024-04-02T14:59:50.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 10 stalled; retrying" ollama | time=2024-04-02T14:59:50.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 16 stalled; retrying" ollama | time=2024-04-02T14:59:51.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 28 stalled; retrying" ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 9 stalled; retrying" ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 7 stalled; retrying" ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying" ollama | time=2024-04-02T14:59:51.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 41 stalled; retrying" ollama | time=2024-04-02T15:00:05.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying" ollama | time=2024-04-02T15:00:05.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying" ollama | time=2024-04-02T15:00:07.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying" ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying" ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 8 stalled; retrying" ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying" ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 24 stalled; retrying" ollama | time=2024-04-02T15:00:09.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 32 stalled; retrying" ollama | time=2024-04-02T15:00:09.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying" ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 12 stalled; retrying" ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying" ollama | time=2024-04-02T15:00:14.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 21 stalled; retrying" ollama | time=2024-04-02T15:00:37.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying" ollama | time=2024-04-02T15:00:38.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 5 stalled; retrying" ollama | time=2024-04-02T15:00:52.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying" ollama | time=2024-04-02T15:00:54.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 32 stalled; retrying" ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying" ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 27 stalled; retrying" ollama | time=2024-04-02T15:00:58.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 4 stalled; retrying" ollama | time=2024-04-02T15:01:02.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying" ollama | time=2024-04-02T15:01:04.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying" ollama | time=2024-04-02T15:01:06.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 0 stalled; retrying" ollama | time=2024-04-02T15:01:14.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying" ollama | time=2024-04-02T15:01:16.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying" ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 22 stalled; retrying" ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 19 stalled; retrying" ollama | time=2024-04-02T15:01:22.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying" ollama | time=2024-04-02T15:01:23.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying" ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 33 stalled; retrying" ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 14 stalled; retrying" ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying" ollama | time=2024-04-02T15:01:24.947Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 40 stalled; retrying" ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying" ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 25 stalled; retrying" ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying" ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying" ollama | time=2024-04-02T15:01:28.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 27 stalled; retrying" ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 3 stalled; retrying" ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 0 stalled; retrying" ollama | time=2024-04-02T15:01:34.946Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 18 stalled; retrying" ollama | time=2024-04-02T15:01:35.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying" ollama | time=2024-04-02T15:04:36.697Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 11 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:04:37.802Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:04:41.698Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying" ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 36 stalled; retrying" ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 34 stalled; retrying" ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 37 stalled; retrying" ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 28 stalled; retrying" ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying" ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying" ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 31 stalled; retrying" ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 38 stalled; retrying" ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying" ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying" ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 6 stalled; retrying" ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 39 stalled; retrying" ollama | time=2024-04-02T15:05:16.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 29 stalled; retrying" ollama | time=2024-04-02T15:05:25.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 7 stalled; retrying" ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 9 stalled; retrying" ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying" ollama | time=2024-04-02T15:05:59.854Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 36 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:07:06.904Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:07:38.614Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 21 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:08:52.542Z level=INFO source=download.go:178 msg="e8a35b5937a5 part 15 attempt 0 failed: unexpected EOF, retrying in 1s" ollama | time=2024-04-02T15:09:14.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 8 stalled; retrying" ollama | time=2024-04-02T15:09:16.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 17 stalled; retrying" ollama | time=2024-04-02T15:09:21.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 23 stalled; retrying" ollama | time=2024-04-02T15:09:22.699Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 11 stalled; retrying" ollama | time=2024-04-02T15:09:26.944Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying" ollama | time=2024-04-02T15:09:48.942Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 22 stalled; retrying" ollama | time=2024-04-02T15:09:48.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 20 stalled; retrying" ollama | time=2024-04-02T15:09:48.945Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 35 stalled; retrying" ollama | time=2024-04-02T15:09:49.615Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 21 stalled; retrying" ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 30 stalled; retrying" ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 msg="e8a35b5937a5 part 13 stalled; retrying"

GiteaMirror commented

2026-04-22 03:16:43 -05:00

@frosty1492 commented on GitHub (Apr 2, 2024):

I don't know why we can't just download a file from the server using
regular download protocols. It seems that for some reason Ollama requires
the Ollama app to pull the files. Doesn't make sense. I download work
press and Mediawiki files then install the add ons etc.

On Tue, Apr 2, 2024 at 10:10 AM chuklee @.***> wrote:

I have the same issue, i have this log on my docker:
ollama | time=2024-04-02T14:58:04.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 34 stalled; retrying"ollama |
time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 3 stalled; retrying"
ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 14 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 41 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 33 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 4 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 16 stalled; retrying"
ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 12 stalled; retrying"
ollama | time=2024-04-02T14:59:10.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 10 stalled; retrying"
ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T14:59:50.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 10 stalled; retrying"
ollama | time=2024-04-02T14:59:50.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 16 stalled; retrying"
ollama | time=2024-04-02T14:59:51.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 28 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 9 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 7 stalled; retrying"
ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T14:59:51.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 41 stalled; retrying"
ollama | time=2024-04-02T15:00:05.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:00:05.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T15:00:07.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 13 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 8 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 24 stalled; retrying"
ollama | time=2024-04-02T15:00:09.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 32 stalled; retrying"
ollama | time=2024-04-02T15:00:09.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 12 stalled; retrying"
ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T15:00:14.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 21 stalled; retrying"
ollama | time=2024-04-02T15:00:37.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:00:38.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 5 stalled; retrying"
ollama | time=2024-04-02T15:00:52.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T15:00:54.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 32 stalled; retrying"
ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 27 stalled; retrying"
ollama | time=2024-04-02T15:00:58.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 4 stalled; retrying"
ollama | time=2024-04-02T15:01:02.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T15:01:04.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T15:01:06.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 0 stalled; retrying"
ollama | time=2024-04-02T15:01:14.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:01:16.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 22 stalled; retrying"
ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 19 stalled; retrying"
ollama | time=2024-04-02T15:01:22.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T15:01:23.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 33 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 14 stalled; retrying"
ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 34 stalled; retrying"
ollama | time=2024-04-02T15:01:24.947Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 40 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 25 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:01:28.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 27 stalled; retrying"
ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 3 stalled; retrying"
ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 0 stalled; retrying"
ollama | time=2024-04-02T15:01:34.946Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 18 stalled; retrying"
ollama | time=2024-04-02T15:01:35.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 13 stalled; retrying"
ollama | time=2024-04-02T15:04:36.697Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 11 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:04:37.802Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:04:41.698Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 36 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 34 stalled; retrying"
ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 37 stalled; retrying"
ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 28 stalled; retrying"
ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 31 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 38 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 6 stalled; retrying"
ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 39 stalled; retrying"
ollama | time=2024-04-02T15:05:16.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 29 stalled; retrying"
ollama | time=2024-04-02T15:05:25.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 7 stalled; retrying"
ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 9 stalled; retrying"
ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:05:59.854Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 36 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:07:06.904Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:07:38.614Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 21 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:08:52.542Z level=INFO source=download.go:178
msg="e8a35b5937a5 part 15 attempt 0 failed: unexpected EOF, retrying in 1s"
ollama | time=2024-04-02T15:09:14.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 8 stalled; retrying"
ollama | time=2024-04-02T15:09:16.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 17 stalled; retrying"
ollama | time=2024-04-02T15:09:21.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 23 stalled; retrying"
ollama | time=2024-04-02T15:09:22.699Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 11 stalled; retrying"
ollama | time=2024-04-02T15:09:26.944Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:09:48.942Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 22 stalled; retrying"
ollama | time=2024-04-02T15:09:48.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 20 stalled; retrying"
ollama | time=2024-04-02T15:09:48.945Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 35 stalled; retrying"
ollama | time=2024-04-02T15:09:49.615Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 21 stalled; retrying"
ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 30 stalled; retrying"
ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250
msg="e8a35b5937a5 part 13 stalled; retrying"

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2032318618,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6AAAD42IQS53DGV63TY3LC7ZAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZSGMYTQNRRHA
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Apr 2, 2024): I don't know why we can't just download a file from the server using regular download protocols. It seems that for some reason Ollama requires the Ollama app to pull the files. Doesn't make sense. I download work press and Mediawiki files then install the add ons etc. On Tue, Apr 2, 2024 at 10:10 AM chuklee ***@***.***> wrote: > I have the same issue, i have this log on my docker: > ollama | time=2024-04-02T14:58:04.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 34 stalled; retrying"ollama | > time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 11 stalled; retrying" > ollama | time=2024-04-02T14:59:07.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 29 stalled; retrying" > ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 3 stalled; retrying" > ollama | time=2024-04-02T14:59:08.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 25 stalled; retrying" > ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 37 stalled; retrying" > ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 20 stalled; retrying" > ollama | time=2024-04-02T14:59:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 14 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 40 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 41 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 31 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 33 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 35 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 18 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 39 stalled; retrying" > ollama | time=2024-04-02T14:59:09.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 38 stalled; retrying" > ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 6 stalled; retrying" > ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 5 stalled; retrying" > ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 4 stalled; retrying" > ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 16 stalled; retrying" > ollama | time=2024-04-02T14:59:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 12 stalled; retrying" > ollama | time=2024-04-02T14:59:10.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 10 stalled; retrying" > ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 39 stalled; retrying" > ollama | time=2024-04-02T14:59:49.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 40 stalled; retrying" > ollama | time=2024-04-02T14:59:50.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 10 stalled; retrying" > ollama | time=2024-04-02T14:59:50.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 16 stalled; retrying" > ollama | time=2024-04-02T14:59:51.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 28 stalled; retrying" > ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 9 stalled; retrying" > ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 7 stalled; retrying" > ollama | time=2024-04-02T14:59:51.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 25 stalled; retrying" > ollama | time=2024-04-02T14:59:51.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 41 stalled; retrying" > ollama | time=2024-04-02T15:00:05.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 17 stalled; retrying" > ollama | time=2024-04-02T15:00:05.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 38 stalled; retrying" > ollama | time=2024-04-02T15:00:07.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 23 stalled; retrying" > ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 13 stalled; retrying" > ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 8 stalled; retrying" > ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 36 stalled; retrying" > ollama | time=2024-04-02T15:00:08.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 24 stalled; retrying" > ollama | time=2024-04-02T15:00:09.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 32 stalled; retrying" > ollama | time=2024-04-02T15:00:09.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 6 stalled; retrying" > ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 12 stalled; retrying" > ollama | time=2024-04-02T15:00:10.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 5 stalled; retrying" > ollama | time=2024-04-02T15:00:14.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 21 stalled; retrying" > ollama | time=2024-04-02T15:00:37.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 30 stalled; retrying" > ollama | time=2024-04-02T15:00:38.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 5 stalled; retrying" > ollama | time=2024-04-02T15:00:52.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 18 stalled; retrying" > ollama | time=2024-04-02T15:00:54.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 32 stalled; retrying" > ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 20 stalled; retrying" > ollama | time=2024-04-02T15:00:57.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 27 stalled; retrying" > ollama | time=2024-04-02T15:00:58.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 4 stalled; retrying" > ollama | time=2024-04-02T15:01:02.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 25 stalled; retrying" > ollama | time=2024-04-02T15:01:04.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 31 stalled; retrying" > ollama | time=2024-04-02T15:01:06.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 0 stalled; retrying" > ollama | time=2024-04-02T15:01:14.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 6 stalled; retrying" > ollama | time=2024-04-02T15:01:16.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 39 stalled; retrying" > ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 22 stalled; retrying" > ollama | time=2024-04-02T15:01:17.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 19 stalled; retrying" > ollama | time=2024-04-02T15:01:22.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 37 stalled; retrying" > ollama | time=2024-04-02T15:01:23.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 36 stalled; retrying" > ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 33 stalled; retrying" > ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 14 stalled; retrying" > ollama | time=2024-04-02T15:01:24.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 34 stalled; retrying" > ollama | time=2024-04-02T15:01:24.947Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 40 stalled; retrying" > ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 29 stalled; retrying" > ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 25 stalled; retrying" > ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 35 stalled; retrying" > ollama | time=2024-04-02T15:01:25.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 17 stalled; retrying" > ollama | time=2024-04-02T15:01:28.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 27 stalled; retrying" > ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 3 stalled; retrying" > ollama | time=2024-04-02T15:01:34.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 0 stalled; retrying" > ollama | time=2024-04-02T15:01:34.946Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 18 stalled; retrying" > ollama | time=2024-04-02T15:01:35.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 13 stalled; retrying" > ollama | time=2024-04-02T15:04:36.697Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 11 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:04:37.802Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 24 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:04:41.698Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 11 stalled; retrying" > ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 36 stalled; retrying" > ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 34 stalled; retrying" > ollama | time=2024-04-02T15:04:52.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 37 stalled; retrying" > ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 28 stalled; retrying" > ollama | time=2024-04-02T15:04:53.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 23 stalled; retrying" > ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 17 stalled; retrying" > ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 31 stalled; retrying" > ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 38 stalled; retrying" > ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 29 stalled; retrying" > ollama | time=2024-04-02T15:04:53.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 35 stalled; retrying" > ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 6 stalled; retrying" > ollama | time=2024-04-02T15:04:53.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 39 stalled; retrying" > ollama | time=2024-04-02T15:05:16.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 29 stalled; retrying" > ollama | time=2024-04-02T15:05:25.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 7 stalled; retrying" > ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 9 stalled; retrying" > ollama | time=2024-04-02T15:05:47.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 30 stalled; retrying" > ollama | time=2024-04-02T15:05:59.854Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 36 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:07:06.904Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 6 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:07:38.614Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 21 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:08:52.542Z level=INFO source=download.go:178 > msg="e8a35b5937a5 part 15 attempt 0 failed: unexpected EOF, retrying in 1s" > ollama | time=2024-04-02T15:09:14.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 8 stalled; retrying" > ollama | time=2024-04-02T15:09:16.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 17 stalled; retrying" > ollama | time=2024-04-02T15:09:21.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 23 stalled; retrying" > ollama | time=2024-04-02T15:09:22.699Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 11 stalled; retrying" > ollama | time=2024-04-02T15:09:26.944Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 35 stalled; retrying" > ollama | time=2024-04-02T15:09:48.942Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 22 stalled; retrying" > ollama | time=2024-04-02T15:09:48.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 20 stalled; retrying" > ollama | time=2024-04-02T15:09:48.945Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 35 stalled; retrying" > ollama | time=2024-04-02T15:09:49.615Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 21 stalled; retrying" > ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 30 stalled; retrying" > ollama | time=2024-04-02T15:10:05.943Z level=INFO source=download.go:250 > msg="e8a35b5937a5 part 13 stalled; retrying" > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2032318618>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6AAAD42IQS53DGV63TY3LC7ZAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZSGMYTQNRRHA> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:16:44 -05:00

@dinakar17 commented on GitHub (Apr 3, 2024):

The simple solution is to run your command prompt or Powershell as administrator. It will work like a charm.

@dinakar17 commented on GitHub (Apr 3, 2024): The simple solution is to run your command prompt or Powershell as administrator. It will work like a charm.

GiteaMirror commented

2026-04-22 03:16:45 -05:00

@frosty1492 commented on GitHub (Apr 3, 2024):

I use Linux and have tried as root. No success.

On Wed, Apr 3, 2024 at 2:20 AM Dinakar Chennupati @.***>
wrote:

The simple solution is to run your command prompt or Powershell as
administrator. It will work like a charm.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2033742075,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6H5JMKXHVQCGBDU2GDY3OUTBAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZTG42DEMBXGU
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Apr 3, 2024): I use Linux and have tried as root. No success. On Wed, Apr 3, 2024 at 2:20 AM Dinakar Chennupati ***@***.***> wrote: > The simple solution is to run your command prompt or Powershell as > administrator. It will work like a charm. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2033742075>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6H5JMKXHVQCGBDU2GDY3OUTBAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZTG42DEMBXGU> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:16:45 -05:00

@wlhee commented on GitHub (Apr 9, 2024):

+1, it's pretty confusing about why Ollama pull slows down at 99%.

@wlhee commented on GitHub (Apr 9, 2024): +1, it's pretty confusing about why Ollama pull slows down at 99%.

GiteaMirror commented

2026-04-22 03:16:46 -05:00

@pdevine commented on GitHub (Apr 10, 2024):

@wlhee see my comment above. I think it's a problem we're having w/ Cloudflare. The only way to really get around it is to hit Ctrl+C and retry; you won't lose any progress.

@pdevine commented on GitHub (Apr 10, 2024): @wlhee see my [comment above](https://github.com/ollama/ollama/issues/1736#issuecomment-1919592282). I think it's a problem we're having w/ Cloudflare. The only way to really get around it is to hit Ctrl+C and retry; you won't lose any progress.

GiteaMirror commented

2026-04-22 03:16:46 -05:00

@wlhee commented on GitHub (Apr 10, 2024):

@wlhee see my comment above. I think it's a problem we're having w/ Cloudflare. The only way to really get around it is to hit Ctrl+C and retry; you won't lose any progress.

Thanks for the tip. ctrl+c and retry does resume the progress, but I still see the extreme slowness issue.

More importantly, I'd like to have the root cause to this problem as it would affect using Ollama to serve inference workloads in prod.

@wlhee commented on GitHub (Apr 10, 2024): > @wlhee see my [comment above](https://github.com/ollama/ollama/issues/1736#issuecomment-1919592282). I think it's a problem we're having w/ Cloudflare. The only way to really get around it is to hit Ctrl+C and retry; you won't lose any progress. Thanks for the tip. ctrl+c and retry does resume the progress, but I still see the extreme slowness issue. More importantly, I'd like to have the root cause to this problem as it would affect using Ollama to serve inference workloads in prod.

GiteaMirror commented

2026-04-22 03:16:47 -05:00

@sammcj commented on GitHub (Apr 10, 2024):

Seems to be a constant issue for me as well, also in Australia

@sammcj commented on GitHub (Apr 10, 2024): Seems to be a constant issue for me as well, also in Australia

GiteaMirror commented

2026-04-22 03:16:48 -05:00

@frosty1492 commented on GitHub (Apr 11, 2024):

So I tried several times to download any of the mods. No success. I am
fully aware that I have a slow internet link, but it is all I have and
works fine for other stuff. The OP was, why can't we download the mods as a
regular file from Git Hub, etc. then have Ollama pull them from our drives
and install the mods. I already do this on many hubs and sites for many
applications.

On Tue, Apr 9, 2024 at 10:54 PM Sam @.***> wrote:

Seems to be a constant issue for me as well, also in Australia

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2046495362,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6D26MGD6VMZGQJFINDY4SZYHAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANBWGQ4TKMZWGI
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Apr 11, 2024): So I tried several times to download any of the mods. No success. I am fully aware that I have a slow internet link, but it is all I have and works fine for other stuff. The OP was, why can't we download the mods as a regular file from Git Hub, etc. then have Ollama pull them from our drives and install the mods. I already do this on many hubs and sites for many applications. On Tue, Apr 9, 2024 at 10:54 PM Sam ***@***.***> wrote: > Seems to be a constant issue for me as well, also in Australia > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2046495362>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6D26MGD6VMZGQJFINDY4SZYHAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANBWGQ4TKMZWGI> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:16:48 -05:00

@asitjain commented on GitHub (Apr 12, 2024):

On Windows 11 here stuck at 99%, I stopped the transfer using CTRL+C, exited the power shell (x86) and then started Power Shell (No x86 so that must be 64 bit) as administrator. Restarted again and it worked like a charm and finished in 5 seconds.

@asitjain commented on GitHub (Apr 12, 2024): On Windows 11 here stuck at 99%, I stopped the transfer using CTRL+C, exited the power shell (x86) and then started Power Shell (No x86 so that must be 64 bit) as administrator. Restarted again and it worked like a charm and finished in 5 seconds.

GiteaMirror commented

2026-04-22 03:16:49 -05:00

@LagSlug commented on GitHub (Apr 13, 2024):

Been having the same problem, but I think the problem is with the models/blobs not updating unless either:

the request is manually cancelled, at which point the files are updated
the model download succeeds

So the status objects aren't being sent until the very end, and it makes it appear that the system has halted. If you look over at the stats window in docker you can see, however, that the download speed remains constant.

@LagSlug commented on GitHub (Apr 13, 2024): Been having the same problem, but I think the problem is with the models/blobs not updating unless either: 1. the request is manually cancelled, at which point the files are updated 2. the model download succeeds So the status objects aren't being sent until the very end, and it makes it appear that the system has halted. If you look over at the stats window in docker you can see, however, that the download speed remains constant. ![image](https://github.com/ollama/ollama/assets/96115715/a30b343f-9077-4665-9e65-e44aeee5701b)

GiteaMirror commented

2026-04-22 03:16:49 -05:00

@matheusnienow commented on GitHub (Apr 20, 2024):

Same issue. Restarting the download fixed it.

@matheusnienow commented on GitHub (Apr 20, 2024): Same issue. Restarting the download fixed it.

GiteaMirror commented

2026-04-22 03:16:50 -05:00

@nielpattin commented on GitHub (Apr 20, 2024):

same issue. Version 0.1.32 when trying to ollama run llama3 for example. It just randomly slowed down at 88% I Ctrl+C quit it and re-run again it started at full speed and then the same thing happened after at 96% it slowed down again. OMG

@nielpattin commented on GitHub (Apr 20, 2024): same issue. Version 0.1.32 when trying to ` ollama run llama3` for example. It just randomly slowed down at 88% I `Ctrl+C` quit it and re-run again it started at full speed and then the same thing happened after at 96% it slowed down again. OMG

GiteaMirror commented

2026-04-22 03:16:51 -05:00

@gauravagnihotristla commented on GitHub (Apr 24, 2024):

Having same issue ollama version is 0.1.32

ollama run codegemma
pulling manifest
pulling 7cb8cb805c53...  99% ▕██████████████████████████████ ▏ 5.0 GB/5.0 GB   55 KB/s  11m21s

re-running the command did not fix the issue

@gauravagnihotristla commented on GitHub (Apr 24, 2024): Having same issue `ollama version is 0.1.32` ``` ollama run codegemma pulling manifest pulling 7cb8cb805c53... 99% ▕██████████████████████████████ ▏ 5.0 GB/5.0 GB 55 KB/s 11m21s ``` re-running the command did not fix the issue

GiteaMirror commented

2026-04-22 03:16:51 -05:00

@nielpattin commented on GitHub (Apr 24, 2024):

Having same issue ollama version is 0.1.32

ollama run codegemma
pulling manifest
pulling 7cb8cb805c53...  99% ▕██████████████████████████████ ▏ 5.0 GB/5.0 GB   55 KB/s  11m21s

re-running the command did not fix the issue

sOmE TiMe tHeY did, you have to keep trying... (Your luck is 50/50)

@nielpattin commented on GitHub (Apr 24, 2024): > Having same issue `ollama version is 0.1.32` > > ``` > ollama run codegemma > pulling manifest > pulling 7cb8cb805c53... 99% ▕██████████████████████████████ ▏ 5.0 GB/5.0 GB 55 KB/s 11m21s > ``` > > re-running the command did not fix the issue sOmE TiMe tHeY did, you have to keep trying... (Your luck is 50/50)

GiteaMirror commented

2026-04-22 03:16:52 -05:00

@Gurpreethgnis commented on GitHub (Apr 25, 2024):

This not fixed. I have a 1 GB Fibre network.

Apple M1 Pro

ollama -v
ollama version is 0.1.32

First stop

ollama run llama2
pulling manifest
pulling 8934d96d3f08... 12% ▕█ ▏ 454 MB/3.8 GB 3.3 MB/s 16m58s

I get this error

Error: max retries exceeded: Get "https:// .... : net/http: TLS handshake timeout

Second Stop

ollama run llama2
pulling manifest
pulling 8934d96d3f08... 94% ▕███████████████ ▏ 3.6 GB/3.8 GB 3.6 MB/s 1m1s
Error: max retries exceeded: 400: <?xml version="1.0" encoding="UTF-8"?><Error><Code>InvalidArgument</Code><Message>Invalid Argument: range must be positive.</Message></Error>

Third Stop

ollama run llama2
pulling manifest
pulling 8934d96d3f08... 100% ▕████████████████▏ 3.8 GB
Error: max retries exceeded: 400: <?xml version="1.0" encoding="UTF-8"?><Error><Code>InvalidArgument</Code><Message>Invalid Argument: range must be positive.</Message></Error>

Fourth time

ollama run llama2
pulling manifest
pulling 8934d96d3f08... 100% ▕███████████████████████████████████████████████████████▏ 3.8 GB

It has been stuck for the past half an hour and nothing is happening!

@Gurpreethgnis commented on GitHub (Apr 25, 2024): This not fixed. I have a 1 GB Fibre network. Apple M1 Pro `ollama -v` `ollama version is 0.1.32` **First stop** `ollama run llama2` `pulling manifest ` `pulling 8934d96d3f08... 12% ▕█ ▏ 454 MB/3.8 GB 3.3 MB/s 16m58s` I get this error `Error: max retries exceeded: Get "https:// .... : net/http: TLS handshake timeout` **Second Stop** `ollama run llama2` `pulling manifest` `pulling 8934d96d3f08... 94% ▕███████████████ ▏ 3.6 GB/3.8 GB 3.6 MB/s 1m1s` `Error: max retries exceeded: 400: <?xml version="1.0" encoding="UTF-8"?><Error><Code>InvalidArgument</Code><Message>Invalid Argument: range must be positive.</Message></Error>` **Third Stop** `ollama run llama2` `pulling manifest ` `pulling 8934d96d3f08... 100% ▕████████████████▏ 3.8 GB ` `Error: max retries exceeded: 400: <?xml version="1.0" encoding="UTF-8"?><Error><Code>InvalidArgument</Code><Message>Invalid Argument: range must be positive.</Message></Error>` **Fourth time** `ollama run llama2` `pulling manifest` `pulling 8934d96d3f08... 100% ▕███████████████████████████████████████████████████████▏ 3.8 GB` It has been stuck for the past half an hour and nothing is happening!

GiteaMirror commented

2026-04-22 03:16:52 -05:00

@Gurpreethgnis commented on GitHub (Apr 25, 2024):

** Update**

pulling 8934d96d3f08... 100% ▕████████████▏ 3.8 GB
Error: max retries exceeded: unexpected EOF

So, not working!

@Gurpreethgnis commented on GitHub (Apr 25, 2024): ** Update** `pulling 8934d96d3f08... 100% ▕████████████▏ 3.8 GB` `Error: max retries exceeded: unexpected EOF` So, not working!

GiteaMirror commented

2026-04-22 03:16:53 -05:00

@Gurpreethgnis commented on GitHub (Apr 25, 2024):

Tried with another model. All models when they reach 100% start giving this error infinitely

Error: max retries exceeded: Get ... : net/http: TLS handshake timeout

@Gurpreethgnis commented on GitHub (Apr 25, 2024): Tried with another model. All models when they reach 100% start giving this error infinitely `Error: max retries exceeded: Get ... : net/http: TLS handshake timeout`

GiteaMirror commented

2026-04-22 03:16:53 -05:00

@gauravagnihotristla commented on GitHub (Apr 25, 2024):

Tried with another model. All models when they reach 100% start giving this error infinitely

Error: max retries exceeded: Get ... : net/http: TLS handshake timeout

Are you on a VPN?

@gauravagnihotristla commented on GitHub (Apr 25, 2024): > Tried with another model. All models when they reach 100% start giving this error infinitely > > > > `Error: max retries exceeded: Get ... : net/http: TLS handshake timeout` Are you on a VPN?

GiteaMirror commented

2026-04-22 03:16:54 -05:00

@Gurpreethgnis commented on GitHub (Apr 25, 2024):

Tried with another model. All models when they reach 100% start giving this error infinitely
Error: max retries exceeded: Get ... : net/http: TLS handshake timeout

Are you on a VPN?

no

@Gurpreethgnis commented on GitHub (Apr 25, 2024): > > Tried with another model. All models when they reach 100% start giving this error infinitely > > `Error: max retries exceeded: Get ... : net/http: TLS handshake timeout` > > Are you on a VPN? no

GiteaMirror commented

2026-04-22 03:16:55 -05:00

@taozhiyuai commented on GitHub (May 7, 2024):

just ctrl+c,and restart pulling

@taozhiyuai commented on GitHub (May 7, 2024): just ctrl+c,and restart pulling

GiteaMirror commented

2026-04-22 03:16:55 -05:00

@agomezh commented on GitHub (May 8, 2024):

I can also say, that disconnecting & reconnecting to the wifi network did the trick.

Thanks!

@agomezh commented on GitHub (May 8, 2024): I can also say, that disconnecting & reconnecting to the wifi network did the trick. Thanks!

GiteaMirror commented

2026-04-22 03:16:56 -05:00

@sychedelix commented on GitHub (May 8, 2024):

+1 and stopping the command and re-running it works.

@sychedelix commented on GitHub (May 8, 2024): +1 and stopping the command and re-running it works.

GiteaMirror commented

2026-04-22 03:16:57 -05:00

@bmizerany commented on GitHub (May 9, 2024):

Hello, Everyone!

At Ollama we're working on a solution to this issue, and have been seeing some positive results!

Now we need your help testing in your enviroments as well!

How to help:

Run a test pull through our staging server

From the list below, pick one (or many) of the models that you have not pulled already, and perform a pull.

ollama pull issue1736.ollama.dev/library/llama3:8b
ollama pull issue1736.ollama.dev/library/gemma:2b
ollama pull issue1736.ollama.dev/library/mistral
ollama pull issue1736.ollama.dev/library/dolphin-mistral
ollama pull issue1736.ollama.dev/library/wizardlm2
ollama pull issue1736.ollama.dev/library/llava-phi3
ollama pull issue1736.ollama.dev/library/llava-llama3
ollama pull issue1736.ollama.dev/library/dolphin-phi
ollama pull issue1736.ollama.dev/library/nomic-embed-text
ollama pull issue1736.ollama.dev/library/phi3
ollama pull issue1736.ollama.dev/library/orca-mini

Remove and retry 2 or 3 more times

ollama rm issue1736.ollama.dev/library/<model>[:<tag>]
ollama pull issue1736.ollama.dev/library/<model>[:<tag>]

Report back!

Please respond here answering these questions to the best of your ability:
- What was the full ollama pull command you ran including model?
- What OS are you running the ollama server on?
- What speed range did you see? (e.g. 30-50 MB/s)
- What version of Ollama are you using?
- What region of the world is your ollama running?
- What is the top speed of your internet connection?
- Was it faster, slower, the same as a normal ollama pull <model> for the same model(s)?

Thank you all so much in advance. We look forward to hearing back from you.

@bmizerany commented on GitHub (May 9, 2024): Hello, Everyone! At Ollama we're working on a solution to this issue, and have been seeing some positive results! Now we need your help testing in your enviroments as well! How to help: 1. Run a test pull through our staging server From the list below, pick one (or many) of the models that you *have not* pulled already, and perform a pull. ``` ollama pull issue1736.ollama.dev/library/llama3:8b ollama pull issue1736.ollama.dev/library/gemma:2b ollama pull issue1736.ollama.dev/library/mistral ollama pull issue1736.ollama.dev/library/dolphin-mistral ollama pull issue1736.ollama.dev/library/wizardlm2 ollama pull issue1736.ollama.dev/library/llava-phi3 ollama pull issue1736.ollama.dev/library/llava-llama3 ollama pull issue1736.ollama.dev/library/dolphin-phi ollama pull issue1736.ollama.dev/library/nomic-embed-text ollama pull issue1736.ollama.dev/library/phi3 ollama pull issue1736.ollama.dev/library/orca-mini ``` 1. Remove and retry 2 or 3 more times ``` ollama rm issue1736.ollama.dev/library/<model>[:<tag>] ollama pull issue1736.ollama.dev/library/<model>[:<tag>] ``` 1. Report back! Please respond here answering these questions to the best of your ability: - What was the full `ollama pull` command you ran including model? - What OS are you running the ollama server on? - What speed range did you see? (e.g. `30-50 MB/s`) - What version of Ollama are you using? - What region of the world is your ollama running? - What is the top speed of your internet connection? - Was it faster, slower, the same as a normal `ollama pull <model>` for the same model(s)? Thank you all so much in advance. We look forward to hearing back from you.

GiteaMirror commented

2026-04-22 03:16:58 -05:00

@mchiang0610 commented on GitHub (May 9, 2024):

Thank you @bmizerany!

ollama pull issue1736.ollama.dev/library/llama3:8b

MacOS Sonoma 14.4.1
Speed starts at around 58MB/s; after 1% it reduces to 25-30MB/s; after about 10% the speed increases to 58MB/s and fluctuates between 45-60MB/s. Same results all 3 times
ollama v0.1.34
Canada, Toronto (East)
Top speed is around 600Mbps (75MB/s) down on my wifi
Slightly faster than normal pull

@mchiang0610 commented on GitHub (May 9, 2024): Thank you @bmizerany! `ollama pull issue1736.ollama.dev/library/llama3:8b` - MacOS Sonoma 14.4.1 - Speed starts at around 58MB/s; after 1% it reduces to 25-30MB/s; after about 10% the speed increases to 58MB/s and fluctuates between 45-60MB/s. Same results all 3 times - ollama v0.1.34 - Canada, Toronto (East) - Top speed is around 600Mbps (75MB/s) down on my wifi - Slightly faster than normal pull

GiteaMirror commented

2026-04-22 03:16:58 -05:00

@adriens commented on GitHub (May 9, 2024):

ℹ️ Settings

Key	value
command	`ollama pull issue1736.ollama.dev/library/gemma:2b`
OS	`Linux Mint 21.3 Virginia`
ollama version	`0.1.34`
Place	`Nouméa, New-Caledonia`
Top speed	`9.7 MB/s` down on my wifi`
📈 Effect	Much much faster than usual
Avg. Speed trial 1️⃣	`3.7 MB/s`
Avg. Speed trial 2️⃣	`9.4 MB/s`
Avg. Speed trial 3️⃣	`9.4 MB/s`

📷 Details

1️⃣ Try

2️⃣ Try

3️⃣ Try

@adriens commented on GitHub (May 9, 2024): ## :information_source: Settings | Key | value | | --- | --- | | command | `ollama pull issue1736.ollama.dev/library/gemma:2b`| | OS | `Linux Mint 21.3 Virginia`| |ollama version | `0.1.34`| | Place | [`Nouméa, New-Caledonia`](https://en.wikipedia.org/wiki/Noum%C3%A9a#/media/File:Locator_map_of_Noum%C3%A9a_2018.png) | | Top speed | `9.7 MB/s` down on my wifi` | :chart_with_upwards_trend: Effect | Much much faster than usual | | Avg. Speed trial :one: | `3.7 MB/s`| | Avg. Speed trial :two: | `9.4 MB/s`| | Avg. Speed trial :three: | `9.4 MB/s`| ## :camera: Details ### :one: Try ![image](https://github.com/ollama/ollama/assets/5235127/30804799-1d94-4e69-8842-93626087fd63) ### :two: Try ![image](https://github.com/ollama/ollama/assets/5235127/59fa0f5d-3299-4a0d-888f-87559ef0b756) ### :three: Try ![image](https://github.com/ollama/ollama/assets/5235127/43580d4d-860c-4507-b8ab-395bb8ec17b3)

GiteaMirror commented

2026-04-22 03:16:59 -05:00

@MaxJa4 commented on GitHub (May 9, 2024):

Command: ollama pull issue1736.ollama.dev/library/dolphin-phi (1.6 GB)
OS: Win 11 Pro
Speed range: 28-30 MB/s
Version: v0.1.34
Region: Europe, Germany / domestic internet
Max Line Speed: ~30-31 MB/s
Speed compared to normal pull: Same

Command: ollama pull issue1736.ollama.dev/library/dolphin-phi (1.6 GB)
OS: Ubuntu 22 LTS Server
Speed range: 57-77 MB/s
Version: v0.1.34
Region: Europe, Germany / cloud server infrastructure
Max Line Speed: ~1000 MB/s at time of testing (varies, usually 600-1200)
Speed compared to normal pull: Slower, normal pull had 100-250 MB/s

Command: ollama pull issue1736.ollama.dev/library/llama3:8b (4.7 GB)
OS: Ubuntu 22 LTS Server
Speed range: 68-80 MB/s
Version: v0.1.34
Region: Europe, Germany / cloud server infrastructure
Max Line Speed: ~1000 MB/s at time of testing (varies, usually 600-1200)
Speed compared to normal pull: Slower, normal pull had 200-420 MB/s

It used to be super slow after ~95%, but since a few days or so it got a lot better!
Hope this helps. Thanks for working on Ollama!

@MaxJa4 commented on GitHub (May 9, 2024): Command: `ollama pull issue1736.ollama.dev/library/dolphin-phi` (1.6 GB) OS: `Win 11 Pro` Speed range: `28-30 MB/s` Version: `v0.1.34` Region: `Europe, Germany / domestic internet` Max Line Speed: `~30-31 MB/s` Speed compared to normal pull: `Same` --- Command: `ollama pull issue1736.ollama.dev/library/dolphin-phi` (1.6 GB) OS: `Ubuntu 22 LTS Server` Speed range: `57-77 MB/s` Version: `v0.1.34` Region: `Europe, Germany / cloud server infrastructure` Max Line Speed: `~1000 MB/s at time of testing (varies, usually 600-1200)` Speed compared to normal pull: `Slower, normal pull had 100-250 MB/s` Command: `ollama pull issue1736.ollama.dev/library/llama3:8b` (4.7 GB) OS: `Ubuntu 22 LTS Server` Speed range: `68-80 MB/s` Version: `v0.1.34` Region: `Europe, Germany / cloud server infrastructure` Max Line Speed: `~1000 MB/s at time of testing (varies, usually 600-1200)` Speed compared to normal pull: `Slower, normal pull had 200-420 MB/s` --- It used to be super slow after ~95%, but since a few days or so it got a lot better! Hope this helps. Thanks for working on Ollama!

GiteaMirror commented

2026-04-22 03:17:00 -05:00

@bmizerany commented on GitHub (May 9, 2024):

Command: ollama pull issue1736.ollama.dev/library/llama3:8b (4.7 GB) OS: Ubuntu 22 LTS Server Speed range: 68-80 MB/s Version: v0.1.34 Region: Europe, Germany / cloud server infrastructure Max Line Speed: ~1000 MB/s at time of testing (varies, usually 600-1200) Speed compared to normal pull: Slower, normal pull had 200-420 MB/s

@MaxJa4 Do you mind trying again? It's likely the cache wasn't warm for your region when you tried it. Hopefully it's faster now?

@bmizerany commented on GitHub (May 9, 2024): > Command: `ollama pull issue1736.ollama.dev/library/llama3:8b` (4.7 GB) OS: `Ubuntu 22 LTS Server` Speed range: `68-80 MB/s` Version: `v0.1.34` Region: `Europe, Germany / cloud server infrastructure` Max Line Speed: `~1000 MB/s at time of testing (varies, usually 600-1200)` Speed compared to normal pull: `Slower, normal pull had 200-420 MB/s` @MaxJa4 Do you mind trying again? It's likely the cache wasn't warm for your region when you tried it. Hopefully it's faster now?

GiteaMirror commented

2026-04-22 03:17:00 -05:00

@MaxJa4 commented on GitHub (May 9, 2024):

@MaxJa4 Do you mind trying again? It's likely the cache wasn't warm for your region when you tried it. Hopefully it's faster now?

Sure, tried it again three times for the llama3 model (staging). This time it was 320-340 MB/s and pretty stable throughout the whole download, including the very last percentages.

@MaxJa4 commented on GitHub (May 9, 2024): > @MaxJa4 Do you mind trying again? It's likely the cache wasn't warm for your region when you tried it. Hopefully it's faster now? Sure, tried it again three times for the llama3 model (staging). This time it was 320-340 MB/s and pretty stable throughout the whole download, including the very last percentages.

GiteaMirror commented

2026-04-22 03:17:00 -05:00

@bmizerany commented on GitHub (May 9, 2024):

@MaxJa4 thank you so much for the extra help!

@bmizerany commented on GitHub (May 9, 2024): @MaxJa4 thank you so much for the extra help!

GiteaMirror commented

2026-04-22 03:17:01 -05:00

@alwqx commented on GitHub (May 10, 2024):

Key	value
command	ollama pull issue1736.ollama.dev/library/gemma:2b
OS	MacOS Sonoma 14.4
ollama version	0.1.34
Place	Shanghai, China
Top speed	4.8 MB/s down on my wifi
Effect	same as usual
Speed trial 1️⃣	Avg: 4.6 MB/s, Max: 4.8 MB/s
Speed trial 2️⃣	Avg: 4.6 MB/s, Max: 4.8 MB/s
Speed trial 3️⃣	Avg: 4.6 MB/s, Max: 4.9 MB/s

$ ollama pull issue1736.ollama.dev/library/gemma:2b
pulling manifest
pulling c1864a5eb193...  35% ▕████████                          ▏ 586 MB/1.7 GB  4.6 MB/s   3m56s

start with 150KB/s -> 1MB/s -> 2MB/s -> 4.6MB/s in 3 second and remain this speed until pull finished.

GiteaMirror commented

2026-04-22 03:17:01 -05:00

@taozhiyuai commented on GitHub (May 10, 2024):

◦	What was the full ollama pull command you ran including model? , YES
◦	What OS are you running the ollama server on? , MAC OS 14.4
◦	What speed range did you see? (e.g. 30-50 MB/s) , 25MB/S
◦	What version of Ollama are you using? v.0.1.34
◦	What region of the world is your ollama running? ; HUBEI, P.R.China
◦	What is the top speed of your internet connection? 100MB/S WHEN I DOWNLOAD MODELS ON OLLAMA.COM
◦	Was it faster, slower, the same as a normal ollama pull <model> for the same model(s)? ; SLOWER

taozhiyu@603e5f4a42f1 ~ % ollama pull issue1736.ollama.dev/library/llama3:8b
pulling manifest
Error: pull model manifest: Get "https://issue1736.ollama.dev/v2/library/llama3/manifests/8b": net/http: TLS handshake timeout

ollama pull issue1736.ollama.dev/library/llama3:8b ; FAIL,net/http: TLS handshake timeout
ollama pull issue1736.ollama.dev/library/gemma:2b ;CAN DOWNLOAD AT 25MB/S MAX
ollama pull issue1736.ollama.dev/library/mistral ; CAN DOWNLOAD AT 25MB/S MAX, SOMETIMES DROP TO 0MB/S, MUST CTRL+C AND RESTART WHEN 98%
ollama pull issue1736.ollama.dev/library/dolphin-mistral ; CAN DOWNLOAD AT 31MB/S MAX;SOMETIMES DROP TO 0MB/S,
ollama pull issue1736.ollama.dev/library/wizardlm2
ollama pull issue1736.ollama.dev/library/llava-phi3 ; CAN DOWNLOAD AT 29MB/S MAX
ollama pull issue1736.ollama.dev/library/llava-llama3
ollama pull issue1736.ollama.dev/library/dolphin-phi
ollama pull issue1736.ollama.dev/library/nomic-embed-text ; CAN DOWNLOAD AT 19MB/S MAX
ollama pull issue1736.ollama.dev/library/phi3
ollama pull issue1736.ollama.dev/library/orca-mini ; CAN DOWNLOAD AT 29MB/S MAX

taozhiyu@603e5f4a42f1 ~ % ping issue1736.ollama.dev
PING issue1736.ollama.dev (66.241.124.100): 56 data bytes
Request timeout for icmp_seq 0
64 bytes from 66.241.124.100: icmp_seq=1 ttl=50 time=180.223 ms
64 bytes from 66.241.124.100: icmp_seq=2 ttl=50 time=185.270 ms
64 bytes from 66.241.124.100: icmp_seq=3 ttl=50 time=191.586 ms
64 bytes from 66.241.124.100: icmp_seq=4 ttl=50 time=193.099 ms
64 bytes from 66.241.124.100: icmp_seq=5 ttl=50 time=169.806 ms
Request timeout for icmp_seq 6
64 bytes from 66.241.124.100: icmp_seq=7 ttl=50 time=136.826 ms
64 bytes from 66.241.124.100: icmp_seq=8 ttl=50 time=167.838 ms
64 bytes from 66.241.124.100: icmp_seq=9 ttl=50 time=171.574 ms
Request timeout for icmp_seq 10
64 bytes from 66.241.124.100: icmp_seq=11 ttl=50 time=136.509 ms
64 bytes from 66.241.124.100: icmp_seq=12 ttl=50 time=165.411 ms
Request timeout for icmp_seq 13
64 bytes from 66.241.124.100: icmp_seq=14 ttl=50 time=166.876 ms
64 bytes from 66.241.124.100: icmp_seq=15 ttl=50 time=190.036 ms
Request timeout for icmp_seq 16
64 bytes from 66.241.124.100: icmp_seq=17 ttl=50 time=166.609 ms
Request timeout for icmp_seq 18
Request timeout for icmp_seq 19
64 bytes from 66.241.124.100: icmp_seq=20 ttl=50 time=175.115 ms
64 bytes from 66.241.124.100: icmp_seq=21 ttl=50 time=178.726 ms
64 bytes from 66.241.124.100: icmp_seq=22 ttl=50 time=175.443 ms
^C
--- issue1736.ollama.dev ping statistics ---
23 packets transmitted, 16 packets received, 30.4% packet loss
round-trip min/avg/max/stddev = 136.509/171.934/193.099/15.973 ms

taozhiyu@603e5f4a42f1 ~ % networkQuality
==== SUMMARY ====
Uplink capacity: 432.862 Kbps
Downlink capacity: 795.363 Mbps
Responsiveness: Low (714.286 milliseconds | 84 RPM)
Idle Latency: 20.667 milliseconds | 3000 RPM

taozhiyu@603e5f4a42f1 ~ % traceroute issue1736.ollama.dev
traceroute to issue1736.ollama.dev (66.241.124.100), 64 hops max, 40 byte packets
1 xiaoqiang (192.168.31.1) 3.657 ms 2.969 ms 2.667 ms
2 192.168.1.1 (192.168.1.1) 4.323 ms 5.977 ms 4.195 ms
3 100.64.0.1 (100.64.0.1) 8.548 ms 7.425 ms 6.664 ms
4 58.50.123.141 (58.50.123.141) 7.843 ms 6.885 ms 6.698 ms
5 * * 58.50.122.1 (58.50.122.1) 10.428 ms
6 202.97.68.101 (202.97.68.101) 27.711 ms
202.97.19.61 (202.97.19.61) 23.056 ms
202.97.29.153 (202.97.29.153) 24.625 ms
7 * * 202.97.71.34 (202.97.71.34) 23.779 ms
8 * 202.97.83.22 (202.97.83.22) 25.817 ms
202.97.39.201 (202.97.39.201) 30.011 ms
9 203.86.97.6 (203.86.97.6) 154.323 ms
xe-1-5-2-3.a03.tokyjp05.jp.bb.gin.ntt.net (129.250.9.125) 139.994 ms
203.86.97.6 (203.86.97.6) 187.262 ms
10 ae-3.r33.tokyjp05.jp.bb.gin.ntt.net (129.250.5.95) 123.159 ms 119.312 ms
ae-5.r32.tokyjp05.jp.bb.gin.ntt.net (129.250.5.93) 145.196 ms
11 * ae-0.a00.tokyjp09.jp.bb.gin.ntt.net (129.250.7.14) 128.363 ms 156.647 ms
12 * * *
13 * * *
14 * 103.84.154.10 (103.84.154.10) 134.021 ms *
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 * *^C

@bmizerany

@taozhiyuai commented on GitHub (May 10, 2024): ◦ What was the full ollama pull command you ran including model? , YES ◦ What OS are you running the ollama server on? , MAC OS 14.4 ◦ What speed range did you see? (e.g. 30-50 MB/s) , 25MB/S ◦ What version of Ollama are you using? v.0.1.34 ◦ What region of the world is your ollama running? ; HUBEI, P.R.China ◦ What is the top speed of your internet connection? 100MB/S WHEN I DOWNLOAD MODELS ON OLLAMA.COM ◦ Was it faster, slower, the same as a normal ollama pull <model> for the same model(s)? ; SLOWER <img width="809" alt="截屏2024-05-10 13 54 11" src="https://github.com/ollama/ollama/assets/146583103/8495f493-f4bd-4420-8ceb-525fbd3c7562"> taozhiyu@603e5f4a42f1 ~ % ollama pull issue1736.ollama.dev/library/llama3:8b pulling manifest Error: pull model manifest: Get "https://issue1736.ollama.dev/v2/library/llama3/manifests/8b": net/http: TLS handshake timeout ollama pull issue1736.ollama.dev/library/llama3:8b ; FAIL,net/http: TLS handshake timeout ollama pull issue1736.ollama.dev/library/gemma:2b ;CAN DOWNLOAD AT 25MB/S MAX ollama pull issue1736.ollama.dev/library/mistral ; CAN DOWNLOAD AT 25MB/S MAX, SOMETIMES DROP TO 0MB/S, MUST CTRL+C AND RESTART WHEN 98% ollama pull issue1736.ollama.dev/library/dolphin-mistral ; CAN DOWNLOAD AT 31MB/S MAX;SOMETIMES DROP TO 0MB/S, ollama pull issue1736.ollama.dev/library/wizardlm2 ollama pull issue1736.ollama.dev/library/llava-phi3 ; CAN DOWNLOAD AT 29MB/S MAX ollama pull issue1736.ollama.dev/library/llava-llama3 ollama pull issue1736.ollama.dev/library/dolphin-phi ollama pull issue1736.ollama.dev/library/nomic-embed-text ; CAN DOWNLOAD AT 19MB/S MAX ollama pull issue1736.ollama.dev/library/phi3 ollama pull issue1736.ollama.dev/library/orca-mini ; CAN DOWNLOAD AT 29MB/S MAX taozhiyu@603e5f4a42f1 ~ % ping issue1736.ollama.dev PING issue1736.ollama.dev (66.241.124.100): 56 data bytes Request timeout for icmp_seq 0 64 bytes from 66.241.124.100: icmp_seq=1 ttl=50 time=180.223 ms 64 bytes from 66.241.124.100: icmp_seq=2 ttl=50 time=185.270 ms 64 bytes from 66.241.124.100: icmp_seq=3 ttl=50 time=191.586 ms 64 bytes from 66.241.124.100: icmp_seq=4 ttl=50 time=193.099 ms 64 bytes from 66.241.124.100: icmp_seq=5 ttl=50 time=169.806 ms Request timeout for icmp_seq 6 64 bytes from 66.241.124.100: icmp_seq=7 ttl=50 time=136.826 ms 64 bytes from 66.241.124.100: icmp_seq=8 ttl=50 time=167.838 ms 64 bytes from 66.241.124.100: icmp_seq=9 ttl=50 time=171.574 ms Request timeout for icmp_seq 10 64 bytes from 66.241.124.100: icmp_seq=11 ttl=50 time=136.509 ms 64 bytes from 66.241.124.100: icmp_seq=12 ttl=50 time=165.411 ms Request timeout for icmp_seq 13 64 bytes from 66.241.124.100: icmp_seq=14 ttl=50 time=166.876 ms 64 bytes from 66.241.124.100: icmp_seq=15 ttl=50 time=190.036 ms Request timeout for icmp_seq 16 64 bytes from 66.241.124.100: icmp_seq=17 ttl=50 time=166.609 ms Request timeout for icmp_seq 18 Request timeout for icmp_seq 19 64 bytes from 66.241.124.100: icmp_seq=20 ttl=50 time=175.115 ms 64 bytes from 66.241.124.100: icmp_seq=21 ttl=50 time=178.726 ms 64 bytes from 66.241.124.100: icmp_seq=22 ttl=50 time=175.443 ms ^C --- issue1736.ollama.dev ping statistics --- 23 packets transmitted, 16 packets received, 30.4% packet loss round-trip min/avg/max/stddev = 136.509/171.934/193.099/15.973 ms taozhiyu@603e5f4a42f1 ~ % networkQuality ==== SUMMARY ==== Uplink capacity: 432.862 Kbps Downlink capacity: 795.363 Mbps Responsiveness: Low (714.286 milliseconds | 84 RPM) Idle Latency: 20.667 milliseconds | 3000 RPM taozhiyu@603e5f4a42f1 ~ % traceroute issue1736.ollama.dev traceroute to issue1736.ollama.dev (66.241.124.100), 64 hops max, 40 byte packets 1 xiaoqiang (192.168.31.1) 3.657 ms 2.969 ms 2.667 ms 2 192.168.1.1 (192.168.1.1) 4.323 ms 5.977 ms 4.195 ms 3 100.64.0.1 (100.64.0.1) 8.548 ms 7.425 ms 6.664 ms 4 58.50.123.141 (58.50.123.141) 7.843 ms 6.885 ms 6.698 ms 5 * * 58.50.122.1 (58.50.122.1) 10.428 ms 6 202.97.68.101 (202.97.68.101) 27.711 ms 202.97.19.61 (202.97.19.61) 23.056 ms 202.97.29.153 (202.97.29.153) 24.625 ms 7 * * 202.97.71.34 (202.97.71.34) 23.779 ms 8 * 202.97.83.22 (202.97.83.22) 25.817 ms 202.97.39.201 (202.97.39.201) 30.011 ms 9 203.86.97.6 (203.86.97.6) 154.323 ms xe-1-5-2-3.a03.tokyjp05.jp.bb.gin.ntt.net (129.250.9.125) 139.994 ms 203.86.97.6 (203.86.97.6) 187.262 ms 10 ae-3.r33.tokyjp05.jp.bb.gin.ntt.net (129.250.5.95) 123.159 ms 119.312 ms ae-5.r32.tokyjp05.jp.bb.gin.ntt.net (129.250.5.93) 145.196 ms 11 * ae-0.a00.tokyjp09.jp.bb.gin.ntt.net (129.250.7.14) 128.363 ms 156.647 ms 12 * * * 13 * * * 14 * 103.84.154.10 (103.84.154.10) 134.021 ms * 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 * *^C @bmizerany

GiteaMirror commented

2026-04-22 03:17:02 -05:00

@tao12345666333 commented on GitHub (May 10, 2024):

Command	ollama pull issue1736.ollama.dev/library/gemma:2b
OS	Fedora Linux
ollama	0.1.34
Seep	3M/s (average)
Effect	Slower than the existing repo.
Region	Beijing, China

Command	ollama pull issue1736.ollama.dev/library/phi3
OS	Fedora Linux
ollama	0.1.34
Seep	203 KB/s (max)
Effect	Slower than the existing repo. The gap is very obvious.
Region	Beijing, China

@tao12345666333 commented on GitHub (May 10, 2024): ![Screenshot from 2024-05-10 12-03-15](https://github.com/ollama/ollama/assets/3264292/daca6b7d-bd5a-4f99-bfa0-090f6368cdfd) ![Screenshot from 2024-05-10 12-07-43](https://github.com/ollama/ollama/assets/3264292/65472a0f-6067-46b9-bd83-388ca12bec09) <table> <tr> <td> Command <td> ollama pull issue1736.ollama.dev/library/gemma:2b <tr> <td>OS <td> Fedora Linux <tr> <td>ollama <td>0.1.34 <tr> <td>Seep <td> 3M/s (average) <tr> <td>Effect <td> <b>Slower than the existing repo.</b> <tr> <td>Region <td> Beijing, China </table> ![Screenshot from 2024-05-10 12-08-51](https://github.com/ollama/ollama/assets/3264292/a1eb6629-6c44-45d4-8b66-96e5aec3cbb9) ![Screenshot from 2024-05-10 13-03-20](https://github.com/ollama/ollama/assets/3264292/a8b265a0-8ccd-4a04-8c34-d36d79edf627) <table> <tr> <td> Command <td> ollama pull issue1736.ollama.dev/library/phi3 <tr> <td>OS <td> Fedora Linux <tr> <td>ollama <td>0.1.34 <tr> <td>Seep <td> 203 KB/s (max) <tr> <td>Effect <td> <b>Slower than the existing repo.</b> The gap is very obvious. <tr> <td>Region <td> Beijing, China </table>

GiteaMirror commented

2026-04-22 03:17:03 -05:00

@bmizerany commented on GitHub (May 10, 2024):

@tao12345666333 By any chance did you try pulling one of those models through the staging server after waiting maybe 10 minutes or more? You may have not hit the cache on the first pull, which we have been seeing is slower than pulling from ollama.com instead, but the pulls through the staging server after the first are much faster than the first.

@bmizerany commented on GitHub (May 10, 2024): @tao12345666333 By any chance did you try pulling one of those models through the staging server after waiting maybe 10 minutes or more? You may have not hit the cache on the first pull, which we have been seeing is slower than pulling from ollama.com instead, but the pulls through the staging server *after* the first are much faster than the first.

GiteaMirror commented

2026-04-22 03:17:03 -05:00

@tao12345666333 commented on GitHub (May 11, 2024):

@bmizerany After waiting for a while, I tried again and it was indeed much faster than the first time.

However, based on my current test results, they are consistently slower than the existing repo (ollama.com).

@tao12345666333 commented on GitHub (May 11, 2024): @bmizerany After waiting for a while, I tried again and it was indeed much faster than the first time. However, based on my current test results, they are consistently slower than the existing repo (ollama.com). ![image](https://github.com/ollama/ollama/assets/3264292/1ff0c093-6a69-4b53-96cf-808763dbdf9c)

GiteaMirror commented

2026-04-22 03:17:04 -05:00

@sammcj commented on GitHub (May 19, 2024):

I've been hitting this problem for a few months now, it used to be I could just cmd+c and resume the download to fix it - but something interesting changed over the last week.

Now instead of stalling at 95-99%, now pulling models runs until about 95%~ and then the progress starts jumping around, the speed bounces around and the percentage complete even goes backwards at times!

Then after a minute or so it crashes with the error: Error: max retries exceeded: EOF

ollama pull llama3:8b-text-q6_K
pulling manifest
pulling ce446d4caf83...  99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████  ▏ 6.5 GB/6.6 GB
Error: max retries exceeded: EOF

It looks like the Ollama is pulling models from 2606:4700::6812:85a or for IPv4 172.67.182.229 which appears to be a Cloudflare IP and I'm not seeing any packet loss or high latency with mtr/ping.

 Host                                        Loss%   Snt   Last   Avg  Best  Wrst StDev
...(truncated)
 5. be400.lsr2.200bou.nsw.aussiebb.net        0.0%     8   32.9  34.8  30.0  47.4   5.9
 6. 172.67.182.229                            0.0%     8   33.5  31.5  24.7  35.9   3.8

I've tested on:

2020 M1 Macbook Air
2023 M2 Macbook Pro
Linux Server (Fedora 40/Docker)
Two different internet connections, (1000/50, 25/10 Mbit), both in Australia
With Ollama compiled with LLAMA_CURL=0 and LLAMA_CURL=1 (not sure if they're actually used by Ollama or not?)

Here's a screen recording of the progress bar at the end of a pull: https://github.com/ollama/ollama/assets/862951/fdf830d0-c122-444f-95c4-300c5058193c

Edit: I just saw mention of issue1736.ollama.dev/library/, and tried it - it appeared to work, then got to 99% and slowed to a crawl, cmd+c and started it again and it did actually finish downloading the image - but the SHA is corrupt:

ollama pull issue1736.ollama.dev/library/llama3:8b-text-q6_K
pulling manifest
pulling ce446d4caf83...  35% ▕███████████████████████████████████████                                                                           ▏ 2.3 GB/6.6 GB   75 MB/s     57s
..
pulling ce446d4caf83...  99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████  ▏ 6.5 GB/6.6 GB   63 MB/s      1s
pulling ce446d4caf83...  99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████  ▏ 6.5 GB/6.6 GB   65 KB/s  17m55s
pulling ce446d4caf83...  99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████  ▏ 6.5 GB/6.6 GB   92 KB/s  12m32s
pulling ce446d4caf83... 100% ▕███████████████████████████████████████████████████���█████████████████████████████████████████████████████████████ ▏ 6.6 GB/6.6 GB  6.7 MB/s      3s
pulling ce446d4caf83... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ 6.6 GB
pulling 4fa551d4f938... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏  12 KB
pulling a2b54344edad... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏  335 B
verifying sha256 digest
Error: digest mismatch, file must be downloaded again: want sha256:ce446d4caf83e013841ba4ed4ae7da3c4319178743a25e463b7e04d3371cc624, got sha256:efc4303d77fe236b49446044c948d8ddc9f30bb4a8323d85d815f517d9ce9afc

See also

*Edit 2024-05-25: This PR seems to fix the broken pulls for me: https://github.com/ollama/ollama/pull/4619

@sammcj commented on GitHub (May 19, 2024): I've been hitting this problem for a few months now, it used to be I could just cmd+c and resume the download to fix it - but something interesting changed over the last week. Now instead of stalling at 95-99%, now pulling models runs until about 95%~ and then the progress starts jumping around, the speed bounces around and the percentage complete even goes backwards at times! Then after a minute or so it crashes with the error: `Error: max retries exceeded: EOF` ``` ollama pull llama3:8b-text-q6_K pulling manifest pulling ce446d4caf83... 99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 6.5 GB/6.6 GB Error: max retries exceeded: EOF ``` It looks like the Ollama is pulling models from `2606:4700::6812:85a` or for IPv4 `172.67.182.229` which appears to be a Cloudflare IP and I'm not seeing any packet loss or high latency with mtr/ping. ``` Host Loss% Snt Last Avg Best Wrst StDev ...(truncated) 5. be400.lsr2.200bou.nsw.aussiebb.net 0.0% 8 32.9 34.8 30.0 47.4 5.9 6. 172.67.182.229 0.0% 8 33.5 31.5 24.7 35.9 3.8 ``` I've tested on: - 2020 M1 Macbook Air - 2023 M2 Macbook Pro - Linux Server (Fedora 40/Docker) - Two different internet connections, (1000/50, 25/10 Mbit), both in Australia - With Ollama compiled with LLAMA_CURL=0 and LLAMA_CURL=1 (not sure if they're actually used by Ollama or not?) Here's a screen recording of the progress bar at the end of a pull: https://github.com/ollama/ollama/assets/862951/fdf830d0-c122-444f-95c4-300c5058193c --- *Edit: I just saw mention of issue1736.ollama.dev/library/*, and tried it - it appeared to work, then got to 99% and slowed to a crawl, cmd+c and started it again and it did actually finish downloading the image - but the SHA is corrupt: ``` ollama pull issue1736.ollama.dev/library/llama3:8b-text-q6_K pulling manifest pulling ce446d4caf83... 35% ▕███████████████████████████████████████ ▏ 2.3 GB/6.6 GB 75 MB/s 57s .. pulling ce446d4caf83... 99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 6.5 GB/6.6 GB 63 MB/s 1s pulling ce446d4caf83... 99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 6.5 GB/6.6 GB 65 KB/s 17m55s pulling ce446d4caf83... 99% ▕████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ▏ 6.5 GB/6.6 GB 92 KB/s 12m32s pulling ce446d4caf83... 100% ▕███████████████████████████████████████████████████��█████████████████████████████████████████████████████████████ ▏ 6.6 GB/6.6 GB 6.7 MB/s 3s pulling ce446d4caf83... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ 6.6 GB pulling 4fa551d4f938... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ 12 KB pulling a2b54344edad... 100% ▕██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ 335 B verifying sha256 digest Error: digest mismatch, file must be downloaded again: want sha256:ce446d4caf83e013841ba4ed4ae7da3c4319178743a25e463b7e04d3371cc624, got sha256:efc4303d77fe236b49446044c948d8ddc9f30bb4a8323d85d815f517d9ce9afc ``` See also - https://github.com/ollama/ollama/issues/1036 - https://github.com/ollama/ollama/issues/4520 *Edit 2024-05-25: This PR seems to fix the broken pulls for me: https://github.com/ollama/ollama/pull/4619

GiteaMirror commented

2026-04-22 03:17:05 -05:00

@hangxingliu commented on GitHub (May 25, 2024):

I believe that this issue has been fixed by the patch https://github.com/ollama/ollama/issues/4520, which was merged just now.
I had the same issue (The download progress is always stuck at the last 1 gb) today. Fortunately, @noxer just fixed this issue, the download process became available when I used ollama compiled from the latest code.

@hangxingliu commented on GitHub (May 25, 2024): I believe that this issue has been fixed by the patch https://github.com/ollama/ollama/issues/4520, which was merged just now. I had the same issue (The download progress is always stuck at the last 1 gb) today. Fortunately, @noxer just fixed this issue, the download process became available when I used ollama compiled from the latest code.

GiteaMirror commented

2026-04-22 03:17:05 -05:00

@davidtranjs commented on GitHub (May 30, 2024):

This issue still happens after I updated to latest version. Start fast untill the last G, I have to cancel and resume it about 10 times to finish the download

@davidtranjs commented on GitHub (May 30, 2024): This issue still happens after I updated to latest version. Start fast untill the last G, I have to cancel and resume it about 10 times to finish the download <img width="1198" alt="image" src="https://github.com/ollama/ollama/assets/16059174/ee1f697e-4121-4963-be2b-3b9bb81bc477">

GiteaMirror commented

2026-04-22 03:17:06 -05:00

@metamec commented on GitHub (May 30, 2024):

Same problem with the latest version. In addition, by downloading so many chunks simultaneously, it's overwhelming my connection. Everything else grinds to a halt. I had to ctrl+c just to load this page. (And again just to submit this comment. 😅)

@metamec commented on GitHub (May 30, 2024): Same problem with the latest version. In addition, by downloading so many chunks simultaneously, it's overwhelming my connection. Everything else grinds to a halt. I had to ctrl+c just to load this page. (And again just to submit this comment. 😅)

GiteaMirror commented

2026-04-22 03:17:07 -05:00

@MaxJa4 commented on GitHub (May 30, 2024):

In addition, by downloading so many chunks simultaneously, it's overwhelming my connection. Everything else grinds to a halt. I had to ctrl+c just to load this page. (And again just to submit this comment. 😅)

Same here, it also consumes a rather high amount of CPU for "just" a download.
Tried it on a GPU cloud server... it reports 20 Gbit/s uplink, but is throttled to ~1 GBit/s due to 100% CPU usage just from model downloading.

@MaxJa4 commented on GitHub (May 30, 2024): > In addition, by downloading so many chunks simultaneously, it's overwhelming my connection. Everything else grinds to a halt. I had to ctrl+c just to load this page. (And again just to submit this comment. 😅) Same here, it also consumes a rather high amount of CPU for "just" a download. Tried it on a GPU cloud server... it reports 20 Gbit/s uplink, but is throttled to ~1 GBit/s due to 100% CPU usage just from model downloading.

GiteaMirror commented

2026-04-22 03:17:08 -05:00

@dhaneshdutta commented on GitHub (Jun 22, 2024):

same problem with verision 0.1.45, download speed slows down when approaching the end of downloading a model

@dhaneshdutta commented on GitHub (Jun 22, 2024): same problem with verision 0.1.45, download speed slows down when approaching the end of downloading a model

GiteaMirror commented

2026-04-22 03:17:08 -05:00

@IAkumaI commented on GitHub (Jun 27, 2024):

Same problem on 0.1.47. Retry it does not change anything.

Also there are logs:

time=2024-06-27T11:58:26.831Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 40 attempt 0 failed: read tcp 172.17.0.3:43764->188.114.98.233:443: read: connection reset by peer, retrying in 1s"
time=2024-06-27T11:58:26.888Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 44 attempt 0 failed: read tcp 172.17.0.3:42344->188.114.99.233:443: read: connection reset by peer, retrying in 1s"
time=2024-06-27T11:58:26.942Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 45 attempt 1 failed: read tcp 172.17.0.3:54754->188.114.99.233:443: read: connection reset by peer, retrying in 2s"
time=2024-06-27T11:58:26.955Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 5 attempt 0 failed: read tcp 172.17.0.3:42334->188.114.99.233:443: read: connection reset by peer, retrying in 1s"
time=2024-06-27T11:58:27.136Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 38 attempt 0 failed: read tcp 172.17.0.3:32950->188.114.99.233:443: read: connection reset by peer, retrying in 1s"
time=2024-06-27T11:58:27.348Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 32 attempt 0 failed: read tcp 172.17.0.3:32954->188.114.99.233:443: read: connection reset by peer, retrying in 1s"
time=2024-06-27T11:58:31.702Z level=INFO source=download.go:251 msg="6a0746a1ec1a part 17 stalled; retrying. If this persists, press ctrl-c to exit, then 'ollama pull' to find a faster connection."
time=2024-06-27T11:58:31.955Z level=INFO source=download.go:251 msg="6a0746a1ec1a part 5 stalled; retrying. If this persists, press ctrl-c to exit, then 'ollama pull' to find a faster connection."

@IAkumaI commented on GitHub (Jun 27, 2024): Same problem on 0.1.47. Retry it does not change anything. Also there are logs: ``` time=2024-06-27T11:58:26.831Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 40 attempt 0 failed: read tcp 172.17.0.3:43764->188.114.98.233:443: read: connection reset by peer, retrying in 1s" time=2024-06-27T11:58:26.888Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 44 attempt 0 failed: read tcp 172.17.0.3:42344->188.114.99.233:443: read: connection reset by peer, retrying in 1s" time=2024-06-27T11:58:26.942Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 45 attempt 1 failed: read tcp 172.17.0.3:54754->188.114.99.233:443: read: connection reset by peer, retrying in 2s" time=2024-06-27T11:58:26.955Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 5 attempt 0 failed: read tcp 172.17.0.3:42334->188.114.99.233:443: read: connection reset by peer, retrying in 1s" time=2024-06-27T11:58:27.136Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 38 attempt 0 failed: read tcp 172.17.0.3:32950->188.114.99.233:443: read: connection reset by peer, retrying in 1s" time=2024-06-27T11:58:27.348Z level=INFO source=download.go:178 msg="6a0746a1ec1a part 32 attempt 0 failed: read tcp 172.17.0.3:32954->188.114.99.233:443: read: connection reset by peer, retrying in 1s" time=2024-06-27T11:58:31.702Z level=INFO source=download.go:251 msg="6a0746a1ec1a part 17 stalled; retrying. If this persists, press ctrl-c to exit, then 'ollama pull' to find a faster connection." time=2024-06-27T11:58:31.955Z level=INFO source=download.go:251 msg="6a0746a1ec1a part 5 stalled; retrying. If this persists, press ctrl-c to exit, then 'ollama pull' to find a faster connection." ```

GiteaMirror commented

2026-04-22 03:17:09 -05:00

@frosty1492 commented on GitHub (Jun 27, 2024):

Perhaps the downloads need to be on a different server with a proper
connection.

On Thu, Jun 27, 2024 at 7:11 AM Valery Ozarnichuk @.***>
wrote:

Same problem on 0.1.47. Retry it does not change anything

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2194522284,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6BNUZ4JOAB4F5NXNKDZJP6NZAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCOJUGUZDEMRYGQ
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Jun 27, 2024): Perhaps the downloads need to be on a different server with a proper connection. On Thu, Jun 27, 2024 at 7:11 AM Valery Ozarnichuk ***@***.***> wrote: > Same problem on 0.1.47. Retry it does not change anything > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2194522284>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6BNUZ4JOAB4F5NXNKDZJP6NZAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCOJUGUZDEMRYGQ> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:17:09 -05:00

@ahaggard2013 commented on GitHub (Jul 7, 2024):

I was only experiencing this on windows. worked on linux/mac fine. I bypassed this for windows by running ollama in docker and mounting the dockers model directory to the windows model directory for downloads. Worked as a temporary work around for anyone seeing this.

edit: this worked once. seemed to be able to download via docker, but had the same issue after one successful model download.

@ahaggard2013 commented on GitHub (Jul 7, 2024): I was only experiencing this on windows. worked on linux/mac fine. I bypassed this for windows by running ollama in docker and mounting the dockers model directory to the windows model directory for downloads. Worked as a temporary work around for anyone seeing this. edit: this worked once. seemed to be able to download via docker, but had the same issue after one successful model download.

GiteaMirror commented

2026-04-22 03:17:09 -05:00

@metamec commented on GitHub (Jul 11, 2024):

It's not hard to understand why it's saturating home networks when according to netstat -aon, it's establishing 63 sockets to download a single file.

@metamec commented on GitHub (Jul 11, 2024): It's not hard to understand why it's saturating home networks when according to `netstat -aon`, it's establishing 63 sockets to download a single file.

GiteaMirror commented

2026-04-22 03:17:10 -05:00

@frosty1492 commented on GitHub (Jul 11, 2024):

I do not understand why anyone would set up a server for software downloads
and not have the DL files to be in a compressed format to be
de-compressed and installed after a single file download. Pardon me but
that causes concern about the efficiency of the software. On a second note,
the actual Ollama files were downloaded and installed in less than a minute
on my system. So is there another explanation? I'm just VERY curious....

On Thu, Jul 11, 2024 at 12:41 PM metamec @.***> wrote:

It's not hard to understand why it's saturating home networks when
according to netstat -aon, it's establishing 63 sockets to download a
single file.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2223519520,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6AQ3MYFEESB32RJ4P3ZL27VVAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRTGUYTSNJSGA
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Jul 11, 2024): I do not understand why anyone would set up a server for software downloads and not have the DL files to be in a compressed format to be de-compressed and installed after a single file download. Pardon me but that causes concern about the efficiency of the software. On a second note, the actual Ollama files were downloaded and installed in less than a minute on my system. So is there another explanation? I'm just VERY curious.... On Thu, Jul 11, 2024 at 12:41 PM metamec ***@***.***> wrote: > It's not hard to understand why it's saturating home networks when > according to netstat -aon, it's establishing 63 sockets to download a > single file. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2223519520>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6AQ3MYFEESB32RJ4P3ZL27VVAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRTGUYTSNJSGA> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:17:11 -05:00

@goodie1974 commented on GitHub (Jul 20, 2024):

it stinks! i'm on DSL and before I got to bed I start downloading. But it's constantly giving me this shit:
wsarecv: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.

@goodie1974 commented on GitHub (Jul 20, 2024): it stinks! i'm on DSL and before I got to bed I start downloading. But it's constantly giving me this shit: wsarecv: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.

GiteaMirror commented

2026-04-22 03:17:11 -05:00

@frosty1492 commented on GitHub (Jul 23, 2024):

So I'm on a very slow DSL hard line, but just because I have it on my
T-Mobil service. I decided to tether my system to my cell phone wifi. It
took 2 attempts but I got all of llama3 and llava installed along with
Docker and WebUI. Now experimenting but it all works. Got to say Llama3 is
impressive but VERY SLOW to respond. Llava is much faster but seems less
astute. Any tips ore comments?

On Sat, Jul 20, 2024 at 12:03 PM goodie1974 @.***>
wrote:

it stinks! i'm on DSL and before I got to bed I start downloading. But
it's constantly giving me this shit:
wsarecv: A connection attempt failed because the connected party did not
properly respond after a period of time, or established connection failed
because connected host has failed to respond.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2241211492,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6HANLMZHA3S2X5EIX3ZNKJ4LAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENBRGIYTCNBZGI
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Jul 23, 2024): So I'm on a very slow DSL hard line, but just because I have it on my T-Mobil service. I decided to tether my system to my cell phone wifi. It took 2 attempts but I got all of llama3 and llava installed along with Docker and WebUI. Now experimenting but it all works. Got to say Llama3 is impressive but VERY SLOW to respond. Llava is much faster but seems less astute. Any tips ore comments? On Sat, Jul 20, 2024 at 12:03 PM goodie1974 ***@***.***> wrote: > it stinks! i'm on DSL and before I got to bed I start downloading. But > it's constantly giving me this shit: > wsarecv: A connection attempt failed because the connected party did not > properly respond after a period of time, or established connection failed > because connected host has failed to respond. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2241211492>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6HANLMZHA3S2X5EIX3ZNKJ4LAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENBRGIYTCNBZGI> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:17:12 -05:00

@vamsi-panchada commented on GitHub (Jul 25, 2024):

I've facing the issue while pulling the model, the model download speed is suddenly dropping from 300 Mbps to 10 Mbps. when the model is at 40-50% downloaded, along with that it's gradually going down to 1 Mbps duing the last 10-5%, could someone help me in resolving this issue.

thank you

@vamsi-panchada commented on GitHub (Jul 25, 2024): I've facing the issue while pulling the model, the model download speed is suddenly dropping from 300 Mbps to 10 Mbps. when the model is at 40-50% downloaded, along with that it's gradually going down to 1 Mbps duing the last 10-5%, could someone help me in resolving this issue. thank you

GiteaMirror commented

2026-04-22 03:17:13 -05:00

@vineetgupta0 commented on GitHub (Jul 25, 2024):

I was facing the same issue after 3 retries and turning off windows firewall it worked. I don't face any similar problem with my macbook m2 so can say network is not at the fault and my windows machine is powerful enough, we can see the speeds were in kb/s before and later it was higher, even after getting the high speed it suddenly drips back to kb's

@vineetgupta0 commented on GitHub (Jul 25, 2024): I was facing the same issue after 3 retries and turning off windows firewall it worked. I don't face any similar problem with my macbook m2 so can say network is not at the fault and my windows machine is powerful enough, we can see the speeds were in kb/s before and later it was higher, even after getting the high speed it suddenly drips back to kb's ![image](https://github.com/user-attachments/assets/e27306fb-3db2-4bce-aac3-7b93356cca60)

GiteaMirror commented

2026-04-22 03:17:13 -05:00

@rampageservices commented on GitHub (Aug 2, 2024):

*Edit 2024-05-25: This PR seems to fix the broken pulls for me: #4619

As mentioned in the quoted message above from earlier in this thread. The PR #4625 that is yet to be pulled resolves the remaining issue for some users.

@rampageservices commented on GitHub (Aug 2, 2024): > *Edit 2024-05-25: This PR seems to fix the broken pulls for me: #4619 As mentioned in the quoted message above from earlier in this thread. The PR #4625 that is yet to be pulled resolves the remaining issue for some users.

GiteaMirror commented

2026-04-22 03:17:14 -05:00

@sammcj commented on GitHub (Sep 11, 2024):

Looks like the stalled at 90%~ bug is back!

@sammcj commented on GitHub (Sep 11, 2024): Looks like the stalled at 90%~ bug is back! <img width="763" alt="image" src="https://github.com/user-attachments/assets/6b7a6820-cb4e-4da1-a620-6d18bd9c29e2">

GiteaMirror commented

2026-04-22 03:17:14 -05:00

@2102819372 commented on GitHub (Sep 20, 2024):

i don't know how to describe the problem, i follow the log recommendations retried so many times,but is useless

@2102819372 commented on GitHub (Sep 20, 2024): i don't know how to describe the problem, i follow the log recommendations retried so many times,but is useless ![image](https://github.com/user-attachments/assets/752b93ae-2876-451e-8107-f0a08ca268c1) ![image](https://github.com/user-attachments/assets/3bf0d41c-e981-4069-b9e8-27f6e6283cd5)

GiteaMirror commented

2026-04-22 03:17:15 -05:00

@sammcj commented on GitHub (Oct 29, 2024):

FYI this is still a an issue as of 2024-10-29, currently causing the following issues (other than generally being annoying 😅):

When trying to drive adoption of Ollama, users experience this issue and either think they've done something wrong or label it as 'unreliable' (overheard this one today).
When trying to use Ollama in an automated environment such as deploying to AWS this bug causes other dependant services/tooling to be blocked until the slow portion of the download is completed.
- Current workarounds include dodgy shell scripts that cancel the pull after a given time period and start it again.

@sammcj commented on GitHub (Oct 29, 2024): FYI this is still a an issue as of 2024-10-29, currently causing the following issues (other than generally being annoying 😅): - When trying to drive adoption of Ollama, users experience this issue and either think they've done something wrong or label it as 'unreliable' (overheard this one today). - When trying to use Ollama in an automated environment such as deploying to AWS this bug causes other dependant services/tooling to be blocked until the slow portion of the download is completed. - Current workarounds include dodgy shell scripts that cancel the pull after a given time period and start it again.

GiteaMirror commented

2026-04-22 03:17:15 -05:00

@FreshLucas-git commented on GitHub (Nov 8, 2024):

Same issue here. This issue still exists on v0.4.0.

@FreshLucas-git commented on GitHub (Nov 8, 2024): Same issue here. This issue still exists on v0.4.0.

GiteaMirror commented

2026-04-22 03:17:17 -05:00

@sammcj commented on GitHub (Nov 10, 2024):

This is impacting the new create model from Huggingface feature as well:

# ollama create rombos-llm-v2.5-qwen-32b:q6_k -f modelfiles/Modelfile-qwen2.5
transferring model data
pulling manifest
pulling b7dbfcbd6d48...  93% ▕████   ▏  25 GB/ 26 GB  713 KB/s  43m52s ^C

# ollama create rombos-llm-v2.5-qwen-32b:q6_k -f modelfiles/Modelfile-qwen2.5
transferring model data
pulling manifest
pulling b7dbfcbd6d48...  95% ▕█████   ▏  25 GB/ 26 GB   52 MB/s     26s

@sammcj commented on GitHub (Nov 10, 2024): This is impacting the new create model from Huggingface feature as well: ``` # ollama create rombos-llm-v2.5-qwen-32b:q6_k -f modelfiles/Modelfile-qwen2.5 transferring model data pulling manifest pulling b7dbfcbd6d48... 93% ▕████ ▏ 25 GB/ 26 GB 713 KB/s 43m52s ^C # ollama create rombos-llm-v2.5-qwen-32b:q6_k -f modelfiles/Modelfile-qwen2.5 transferring model data pulling manifest pulling b7dbfcbd6d48... 95% ▕█████ ▏ 25 GB/ 26 GB 52 MB/s 26s ```

GiteaMirror commented

2026-04-22 03:17:17 -05:00

@charlieamer commented on GitHub (Dec 25, 2024):

Hello all. One of the 2 things solved this issue for me (probably not both though):

I used to save my models to HDD instead of SSD. After I switched to SSD this issue was gone.
(while I used HDD for ollama) I've restarted my computer and made sure I am not using ollama for anything else other then downloading. Once it's done I begin using ollama.

I am not sure which of the 2 things above helped me, but i don't have the issuse anymore

@charlieamer commented on GitHub (Dec 25, 2024): Hello all. One of the 2 things solved this issue for me (probably not both though): 1. I used to save my models to HDD instead of SSD. After I switched to SSD this issue was gone. 2. (while I used HDD for ollama) I've restarted my computer and made sure I am not using ollama for anything else other then downloading. Once it's done I begin using ollama. I am not sure which of the 2 things above helped me, but i don't have the issuse anymore

GiteaMirror commented

2026-04-22 03:17:17 -05:00

@rampageservices commented on GitHub (Dec 25, 2024):

That is helpful feedback Amer. It could mean that some sort of checking
needs to be done on the I/O side or check to ensure the download function
be done in a separate thread if not done already. (It's been quite a few
months since I've seen the codebase.)

On Wed, Dec 25, 2024, 2:48 PM Amer @.***> wrote:

Hello all. One of the 2 things solved this issue for me (probably not both
though):

I used to save my models to HDD instead of SSD. After I switched to
SSD this issue was gone.

I've restarted my computer and made sure I am not using ollama for
anything else other then downloading. Once it's done I begin using ollama.

I am not sure which of the 2 things above helped me, but i don't have the
issuse anymore

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2561660738,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCSFSHC26D4CDPPTS632HJINBAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNRRGY3DANZTHA
.
You are receiving this because you commented.Message ID:
@.***>

@rampageservices commented on GitHub (Dec 25, 2024): That is helpful feedback Amer. It could mean that some sort of checking needs to be done on the I/O side or check to ensure the download function be done in a separate thread if not done already. (It's been quite a few months since I've seen the codebase.) On Wed, Dec 25, 2024, 2:48 PM Amer ***@***.***> wrote: > Hello all. One of the 2 things solved this issue for me (probably not both > though): > > 1. I used to save my models to HDD instead of SSD. After I switched to > SSD this issue was gone. > 2. I've restarted my computer and made sure I am not using ollama for > anything else other then downloading. Once it's done I begin using ollama. > > I am not sure which of the 2 things above helped me, but i don't have the > issuse anymore > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2561660738>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AE32PCSFSHC26D4CDPPTS632HJINBAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNRRGY3DANZTHA> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

GiteaMirror commented

2026-04-22 03:17:18 -05:00

@ngocdd commented on GitHub (Dec 29, 2024):

the same issue fro my side the speed too slow

@ngocdd commented on GitHub (Dec 29, 2024): the same issue fro my side the speed too slow ![image](https://github.com/user-attachments/assets/60c1d6f5-acdb-4f2d-b96c-63192712ea05)

GiteaMirror commented

2026-04-22 03:17:19 -05:00

@oyaro-tech commented on GitHub (Jan 7, 2025):

I had the same issue on MacOS only when using an external SSD drive to store models. After formatting it to APFS, everything works fine

@oyaro-tech commented on GitHub (Jan 7, 2025): I had the same issue on MacOS only when using an external SSD drive to store models. After formatting it to APFS, everything works fine

GiteaMirror commented

2026-04-22 03:17:20 -05:00

@eduanmoldeep commented on GitHub (Jan 12, 2025):

I found a much simple solution that worked for me

Close the terminal
Restart download with a new one

from stalling 200 kbps, jumped back to 5Mbps

Hope it helps you, who had to scroll this deep here

@eduanmoldeep commented on GitHub (Jan 12, 2025): I found a much simple solution that worked for me 1. Close the terminal 2. Restart download with a new one from stalling 200 kbps, jumped back to 5Mbps Hope it helps you, who had to scroll this deep here

GiteaMirror commented

2026-04-22 03:17:20 -05:00

@Nitish-bot commented on GitHub (Jan 20, 2025):

I found a much simple solution that worked for me
1. Close the terminal

2. Restart download with a new one
from stalling 200 kbps, jumped back to 5Mbps

Hope it helps you, who had to scroll this deep here

@eduanmoldeep it is the first comment

@Nitish-bot commented on GitHub (Jan 20, 2025): > I found a much simple solution that worked for me > > 1. Close the terminal > > 2. Restart download with a new one > > > from stalling 200 kbps, jumped back to 5Mbps > > Hope it helps you, who had to scroll this deep here @eduanmoldeep it is the first comment

GiteaMirror commented

2026-04-22 03:17:22 -05:00

@harisiqbal12 commented on GitHub (Jan 25, 2025):

Still facing this issue, my mistral model was downloaded of 50% then my internet fluctuated and it's back to 10% i'm on latest version. 0.5.7

@harisiqbal12 commented on GitHub (Jan 25, 2025): Still facing this issue, my mistral model was downloaded of 50% then my internet fluctuated and it's back to 10% i'm on latest version. 0.5.7

GiteaMirror commented

2026-04-22 03:17:22 -05:00

@eduanmoldeep commented on GitHub (Jan 25, 2025):

I found a much simple solution that worked for me
1. Close the terminal

2. Restart download with a new one
from stalling 200 kbps, jumped back to 5Mbps
Hope it helps you, who had to scroll this deep here
@eduanmoldeep it is the first comment

@Nitish-bot
might sound odd, but stopping download didn't help,
I had to close the whole terminal and in a fresh instance only was able to get the throttling solved.

Thus wrote a seperate comment

@eduanmoldeep commented on GitHub (Jan 25, 2025): > > I found a much simple solution that worked for me > > ``` > > 1. Close the terminal > > > > 2. Restart download with a new one > > ``` > > > > > > > > > > > > > > > > > > > > > > > > from stalling 200 kbps, jumped back to 5Mbps > > Hope it helps you, who had to scroll this deep here > > [@eduanmoldeep](https://github.com/eduanmoldeep) it is the first comment @Nitish-bot might sound odd, but stopping download didn't help, I had to close the whole terminal and in a fresh instance only was able to get the throttling solved. Thus wrote a seperate comment

GiteaMirror commented

2026-04-22 03:17:23 -05:00

@krecco commented on GitHub (Jan 30, 2025):

Is there any other simple way to download a model? Downloading mixtral, downloaded 100+ GB with restarts (a lot is lost after restart) but still not finished....

@krecco commented on GitHub (Jan 30, 2025): Is there any other simple way to download a model? Downloading mixtral, downloaded 100+ GB with restarts (a lot is lost after restart) but still not finished....

GiteaMirror commented

2026-04-22 03:17:25 -05:00

@HildaJ commented on GitHub (Jan 30, 2025):

I closed the terminal, and entered ollama pull [your model name] instead of ollama run [your model name]. It works for me.

@HildaJ commented on GitHub (Jan 30, 2025): I closed the terminal, and entered `ollama pull [your model name]` instead of `ollama run [your model name]`. It works for me.

GiteaMirror commented

2026-04-22 03:17:26 -05:00

@pdevine commented on GitHub (Jan 30, 2025):

@HildaJ The mechanism for pulling is the same for both ollama pull and ollama run, so I think you just got lucky.

@pdevine commented on GitHub (Jan 30, 2025): @HildaJ The mechanism for pulling is the same for both `ollama pull` and `ollama run`, so I think you just got lucky.

GiteaMirror commented

2026-04-22 03:17:27 -05:00

@krecco commented on GitHub (Jan 30, 2025):

agreed, i always use pull and its not working ... normally i try to load instruct model, not the default one

@krecco commented on GitHub (Jan 30, 2025): agreed, i always use pull and its not working ... normally i try to load instruct model, not the default one

GiteaMirror commented

2026-04-22 03:17:28 -05:00

@rampageservices commented on GitHub (Jan 31, 2025):

I am going to take a look at this over the next week and suggest a PR. This
issue continues to persist after such a long time. To ease initial use for
users who are new to the environment of which I am positive there will be
thousands more, this will continue to be a wall that may delay future AI
innovators from creating the next big thing.

On Thu, Jan 30, 2025, 11:18 PM BoRut @.***> wrote:

agreed, i always use pull and its not working ...

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2624785412,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCQEIFILCN2RKMFZOKL2NI7FHAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRUG44DKNBRGI
.
You are receiving this because you commented.Message ID:
@.***>

@rampageservices commented on GitHub (Jan 31, 2025): I am going to take a look at this over the next week and suggest a PR. This issue continues to persist after such a long time. To ease initial use for users who are new to the environment of which I am positive there will be thousands more, this will continue to be a wall that may delay future AI innovators from creating the next big thing. On Thu, Jan 30, 2025, 11:18 PM BoRut ***@***.***> wrote: > agreed, i always use pull and its not working ... > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2624785412>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AE32PCQEIFILCN2RKMFZOKL2NI7FHAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRUG44DKNBRGI> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

GiteaMirror commented

2026-04-22 03:17:30 -05:00

@flyfox666 commented on GitHub (Feb 1, 2025):

I've run into the same problem, every time I reopen and start it the speed is really fast, but after a while it slows down and the drop is visible to the naked eye.The download speed goes down in steps.

@flyfox666 commented on GitHub (Feb 1, 2025): I've run into the same problem, every time I reopen and start it the speed is really fast, but after a while it slows down and the drop is visible to the naked eye.The download speed goes down in steps.

GiteaMirror commented

2026-04-22 03:17:30 -05:00

@rampageservices commented on GitHub (Feb 1, 2025):

The slow down may actually be a symptom of your ISP performing QoS on large
downloads to ensure bandwidth for other users on your network. This is
outside of Ollama's control and not a result of anything in particular
wrong with the software itself.

On Sat, Feb 1, 2025, 3:00 PM bbw1986 @.***> wrote:

I've run into the same problem, every time I reopen and start it the speed
is really fast, but after a while it slows down and the drop is visible to
the naked eye.The download speed goes down in steps.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2628818157,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCROFO2DX5WYT76KGHT2NRWJ5AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRYHAYTQMJVG4
.
You are receiving this because you commented.Message ID:
@.***>

@rampageservices commented on GitHub (Feb 1, 2025): The slow down may actually be a symptom of your ISP performing QoS on large downloads to ensure bandwidth for other users on your network. This is outside of Ollama's control and not a result of anything in particular wrong with the software itself. On Sat, Feb 1, 2025, 3:00 PM bbw1986 ***@***.***> wrote: > I've run into the same problem, every time I reopen and start it the speed > is really fast, but after a while it slows down and the drop is visible to > the naked eye.The download speed goes down in steps. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2628818157>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AE32PCROFO2DX5WYT76KGHT2NRWJ5AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRYHAYTQMJVG4> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

GiteaMirror commented

2026-04-22 03:17:32 -05:00

@arunofhyd commented on GitHub (Feb 1, 2025):

I'm also experiencing this exact issue.

The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed.

This is the ultimate fix!

@arunofhyd commented on GitHub (Feb 1, 2025): > I'm also experiencing this exact issue. > > The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed. This is the ultimate fix!

GiteaMirror commented

2026-04-22 03:17:33 -05:00

@krecco commented on GitHub (Feb 2, 2025):

I have tested it on ubuntu server, cli only and I havent had any problems since.

@krecco commented on GitHub (Feb 2, 2025): I have tested it on ubuntu server, cli only and I havent had any problems since.

GiteaMirror commented

2026-04-22 03:17:35 -05:00

@imyhxy commented on GitHub (Feb 11, 2025):

It's frustrated that I have to stop and start the pull command over and over again once the download speed is lower than 1MB/s. 😢

@imyhxy commented on GitHub (Feb 11, 2025): It's frustrated that I have to stop and start the pull command over and over again once the download speed is lower than 1MB/s. 😢

GiteaMirror commented

2026-04-22 03:17:36 -05:00

@arunofhyd commented on GitHub (Feb 11, 2025):

It's frustrated that I have to stop and start the pull command over and over again once the download speed is lower than 1MB/s. 😢

You can just stop and restart, it works!

@arunofhyd commented on GitHub (Feb 11, 2025): > It's frustrated that I have to stop and start the pull command over and over again once the download speed is lower than 1MB/s. 😢 You can just stop and restart, it works!

GiteaMirror commented

2026-04-22 03:17:38 -05:00

@imyhxy commented on GitHub (Feb 11, 2025):

@IngeniousThomas Yes, but it will slow down again very soon.

@imyhxy commented on GitHub (Feb 11, 2025): @IngeniousThomas Yes, but it will slow down again very soon.

GiteaMirror commented

2026-04-22 03:17:38 -05:00

@shrijayan commented on GitHub (Feb 17, 2025):

I'm also experiencing this exact issue.

The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed.

Now this work around is also not working.

@shrijayan commented on GitHub (Feb 17, 2025): > I'm also experiencing this exact issue. > > The workaround I've found is to stop the download when it slows down and then restart it - this resumes the download at full speed. Now this work around is also not working.

GiteaMirror commented

2026-04-22 03:17:39 -05:00

@frosty1492 commented on GitHub (Feb 18, 2025):

Can anyone explain why we can not simply download this as an installable
file? This process is not good at all.

On Mon, Feb 17, 2025 at 6:09 AM Shrijayan @.***> wrote:

I'm also experiencing this exact issue.

The workaround I've found is to stop the download when it slows down and
then restart it - this resumes the download at full speed.

Now this work around is also not working.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6F2GW6CKEJDFZJLXGL2QHGONAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRSHEZDSNJXG4
.
You are receiving this because you commented.Message ID:
@.***>
[image: shrijayan]shrijayan left a comment (ollama/ollama#1736)
https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577

I'm also experiencing this exact issue.

The workaround I've found is to stop the download when it slows down and
then restart it - this resumes the download at full speed.

Now this work around is also not working.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABT3R6F2GW6CKEJDFZJLXGL2QHGONAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRSHEZDSNJXG4
.
You are receiving this because you commented.Message ID:
@.***>

--
John Foster

@frosty1492 commented on GitHub (Feb 18, 2025): Can anyone explain why we can not simply download this as an installable file? This process is not good at all. On Mon, Feb 17, 2025 at 6:09 AM Shrijayan ***@***.***> wrote: > I'm also experiencing this exact issue. > > The workaround I've found is to stop the download when it slows down and > then restart it - this resumes the download at full speed. > > Now this work around is also not working. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6F2GW6CKEJDFZJLXGL2QHGONAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRSHEZDSNJXG4> > . > You are receiving this because you commented.Message ID: > ***@***.***> > [image: shrijayan]*shrijayan* left a comment (ollama/ollama#1736) > <https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577> > > I'm also experiencing this exact issue. > > The workaround I've found is to stop the download when it slows down and > then restart it - this resumes the download at full speed. > > Now this work around is also not working. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2662929577>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABT3R6F2GW6CKEJDFZJLXGL2QHGONAVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNRSHEZDSNJXG4> > . > You are receiving this because you commented.Message ID: > ***@***.***> > -- John Foster

GiteaMirror commented

2026-04-22 03:17:40 -05:00

@dmilneconcordinfiniti commented on GitHub (Mar 6, 2025):

Ctrl + C then closing the terminal before reapplying "ollama pull [model]" worked for me.

I had also added the "OLLAMA_NOPRUNE=true" environment variable to enable (encourage?) resumable downloads.

I'm also downloading from Australia and I don't think it's QoS - as it happens consistently on the last gigabyte of an Ollama download regardless of the size of the model. 9GB model throttles to dial-up speed at 8GB downloaded; while a 19GB model throttles to dial-up speed at 18GB downloaded.

@dmilneconcordinfiniti commented on GitHub (Mar 6, 2025): Ctrl + C then closing the terminal before reapplying "ollama pull [model]" worked for me. I had also added the "OLLAMA_NOPRUNE=true" environment variable to enable (encourage?) resumable downloads. I'm also downloading from Australia and I don't think it's QoS - as it happens consistently on the last gigabyte of an Ollama download regardless of the size of the model. 9GB model throttles to dial-up speed at 8GB downloaded; while a 19GB model throttles to dial-up speed at 18GB downloaded.

GiteaMirror commented

2026-04-22 03:17:41 -05:00

@seeadoog commented on GitHub (Mar 10, 2025):

just like playing monkey

@seeadoog commented on GitHub (Mar 10, 2025): just like playing monkey

GiteaMirror commented

2026-04-22 03:17:42 -05:00

@imyhxy commented on GitHub (Mar 11, 2025):

Try this:

while true; do timeout 600s ollama pull deepseek-r1:7b && break; sleep 10; done

Restart the download process every 600s until the download is finished.

@imyhxy commented on GitHub (Mar 11, 2025): Try this: ```bash while true; do timeout 600s ollama pull deepseek-r1:7b && break; sleep 10; done ``` Restart the download process every 600s until the download is finished.

GiteaMirror commented

2026-04-22 03:17:43 -05:00

@sandeep-tukaram commented on GitHub (Mar 15, 2025):

Don't understand why 15MB/s , order of 10MBps, limit on the download speed when I do ollama pull on a model.. Is it because of the registry host - Docker hub?

$ ollama pull llama3.3 
pulling manifest 
pulling 4824460d29f2...  73% ▕████████████████████████████████████████████████████                    ▏  31 GB/ 42 GB   11 MB/s  16m26s

Tops this my ISP speed drops from about 100MBps to 15MBps (speedtest ookla) when downloading from Ollama. Fail to wrap my head around this crawling speed.

@sandeep-tukaram commented on GitHub (Mar 15, 2025): Don't understand why 15MB/s , order of 10MBps, limit on the download speed when I do ollama pull on a model.. Is it because of the registry host - Docker hub? ``` $ ollama pull llama3.3 pulling manifest pulling 4824460d29f2... 73% ▕████████████████████████████████████████████████████ ▏ 31 GB/ 42 GB 11 MB/s 16m26s ``` Tops this my ISP speed drops from about 100MBps to 15MBps (speedtest ookla) when downloading from Ollama. Fail to wrap my head around this crawling speed.

GiteaMirror commented

2026-04-22 03:17:44 -05:00

@sandeep-tukaram commented on GitHub (Mar 15, 2025):

The slow down may actually be a symptom of your ISP performing QoS on large
downloads to ensure bandwidth for other users on your network. This is
outside of Ollama's control and not a result of anything in particular
wrong with the software itself.
…

Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds unrealistic.

@sandeep-tukaram commented on GitHub (Mar 15, 2025): > The slow down may actually be a symptom of your ISP performing QoS on large > downloads to ensure bandwidth for other users on your network. This is > outside of Ollama's control and not a result of anything in particular > wrong with the software itself. > […](#) Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds unrealistic.

GiteaMirror commented

2026-04-22 03:17:45 -05:00

@rampageservices commented on GitHub (Mar 15, 2025):

I have achieved 60MB/s on every recent model pull so... not sure what to
tell you. It was a suggestion that I was providing. No need to directly
attack me. If you have this issue, please look into it on your end as well.

On Sat, Mar 15, 2025, 6:41 PM Sandeep T @.***> wrote:

The slow down may actually be a symptom of your ISP performing QoS on large
downloads to ensure bandwidth for other users on your network. This is
outside of Ollama's control and not a result of anything in particular
wrong with the software itself.
… <#m_6001758609550271610_>

Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds
unrealistic.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCVWTGK2ES5P4B7UVTT2UP7V3AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMRWGQZTCOBQHA
.
You are receiving this because you commented.Message ID:
@.***>
[image: sandeep-tukaram]sandeep-tukaram left a comment
(ollama/ollama#1736)
https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808

The slow down may actually be a symptom of your ISP performing QoS on large
downloads to ensure bandwidth for other users on your network. This is
outside of Ollama's control and not a result of anything in particular
wrong with the software itself.
… <#m_6001758609550271610_>

Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds
unrealistic.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCVWTGK2ES5P4B7UVTT2UP7V3AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMRWGQZTCOBQHA
.
You are receiving this because you commented.Message ID:
@.***>

@rampageservices commented on GitHub (Mar 15, 2025): I have achieved 60MB/s on every recent model pull so... not sure what to tell you. It was a suggestion that I was providing. No need to directly attack me. If you have this issue, please look into it on your end as well. On Sat, Mar 15, 2025, 6:41 PM Sandeep T ***@***.***> wrote: > The slow down may actually be a symptom of your ISP performing QoS on large > downloads to ensure bandwidth for other users on your network. This is > outside of Ollama's control and not a result of anything in particular > wrong with the software itself. > … <#m_6001758609550271610_> > > Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds > unrealistic. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AE32PCVWTGK2ES5P4B7UVTT2UP7V3AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMRWGQZTCOBQHA> > . > You are receiving this because you commented.Message ID: > ***@***.***> > [image: sandeep-tukaram]*sandeep-tukaram* left a comment > (ollama/ollama#1736) > <https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808> > > The slow down may actually be a symptom of your ISP performing QoS on large > downloads to ensure bandwidth for other users on your network. This is > outside of Ollama's control and not a result of anything in particular > wrong with the software itself. > … <#m_6001758609550271610_> > > Every ISP caping downloads at 15 MB/s and crawling at 99%. Sounds > unrealistic. > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/1736#issuecomment-2726431808>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AE32PCVWTGK2ES5P4B7UVTT2UP7V3AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMRWGQZTCOBQHA> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

GiteaMirror commented

2026-04-22 03:17:46 -05:00

@pdevine commented on GitHub (Mar 16, 2025):

Hey guys, using ollama 0.6.1 can you set the OLLAMA_EXPERIMENT env to client2 and try pulling? Instructions for setting the environment variables are here.

You'll get something that looks like this:

When it's finished you won't see the verifying sha256 digest message as the checksums are validated as the data is streamed.

If you want to set it back to the default client you can set OLLAMA_EXPERIMENT to client1 or remove the environment variable.

@pdevine commented on GitHub (Mar 16, 2025): Hey guys, using ollama 0.6.1 can you set the `OLLAMA_EXPERIMENT` env to `client2` and try pulling? Instructions for setting the environment variables are [here](https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-configure-ollama-server). You'll get something that looks like this: <img width="1652" alt="Image" src="https://github.com/user-attachments/assets/b5d9ec5a-e1ab-47e6-bb5c-d62c4deb6b02" /> When it's finished you won't see the `verifying sha256 digest` message as the checksums are validated as the data is streamed. If you want to set it back to the default client you can set `OLLAMA_EXPERIMENT` to `client1` or remove the environment variable.

GiteaMirror commented

2026-04-22 03:17:49 -05:00

@ViZAlice commented on GitHub (Jun 6, 2025):

Yeah, it works! And when I repeat the last pull command, it will keep processing.

@EmmaWebGH 很抱歉你遇到了这个问题。@mxyng之前提到过，其中一个数据流挂了，但报告显示一切正常。你应该可以重新加载Ctrl + c镜像，然后重新拉取，它应该会从上次中断的地方继续运行。你能再确认一下你正在运行吗0.1.22？

@ViZAlice commented on GitHub (Jun 6, 2025): Yeah, it works! And when I repeat the last pull command, it will keep processing. > @EmmaWebGH 很抱歉你遇到了这个问题。[@mxyng](https://github.com/mxyng)之前提到过，其中一个数据流挂了，但报告显示一切正常。你应该可以重新加载`Ctrl + c`镜像，然后重新拉取，它应该会从上次中断的地方继续运行。你能再确认一下你正在运行吗`0.1.22`？

GiteaMirror commented

2026-04-22 03:17:51 -05:00

@SPOOKEXE commented on GitHub (Jun 11, 2025):

setting the OLLAMA_EXPERIMENT to client2 fixed my issue it seems!

@SPOOKEXE commented on GitHub (Jun 11, 2025): setting the OLLAMA_EXPERIMENT to client2 fixed my issue it seems!

GiteaMirror commented

2026-04-22 03:17:53 -05:00

@curlydingo commented on GitHub (Jul 19, 2025):

client2 has fixed the issue for me too, thanks @pdevine !

@curlydingo commented on GitHub (Jul 19, 2025): client2 has fixed the issue for me too, thanks @pdevine !

GiteaMirror commented

2026-04-22 03:17:55 -05:00

@ferrouswheel commented on GitHub (Aug 3, 2025):

client2 fixed mine too. Downloaded most of qwen3-coder, but it would get stuck at 80kb/s and 30 minutes remaining despite almost being complete. Restarting ollama and/or the download with ollama pull didn't change anything.

But adding -e OLLAMA_EXPERIMENT=client2 to docker run makes the download complete instantly (suggesting almost all parts were already downloaded)

This seems like a high prority bug since the ease of downloading models is the primary reason I use ollama.

@ferrouswheel commented on GitHub (Aug 3, 2025): `client2` fixed mine too. Downloaded most of qwen3-coder, but it would get stuck at 80kb/s and 30 minutes remaining despite almost being complete. Restarting ollama and/or the download with `ollama pull` didn't change anything. But adding `-e OLLAMA_EXPERIMENT=client2` to docker run makes the download complete instantly (suggesting almost all parts were already downloaded) This seems like a high prority bug since the ease of downloading models is the primary reason I use ollama.

GiteaMirror commented

2026-04-22 03:17:56 -05:00

@div002123 commented on GitHub (Aug 6, 2025):

➜  ~ export OLLAMA_EXPERIMENT=client2
➜  ~ ollama run gpt-oss:20b

not working...

0%-95% 30min
95%~ 30min~1hour

It turns out I was being throttled; switching networks solved the problem. Just ignore me.😂

@div002123 commented on GitHub (Aug 6, 2025): ``` ➜ ~ export OLLAMA_EXPERIMENT=client2 ➜ ~ ollama run gpt-oss:20b ``` not working... 0%-95% 30min 95%~ 30min~1hour -------- It turns out I was being throttled; switching networks solved the problem. Just ignore me.😂

GiteaMirror commented

2026-04-22 03:17:57 -05:00

@Rizmi commented on GitHub (Aug 28, 2025):

The slow down may actually be a symptom of your ISP performing QoS on large
downloads to ensure bandwidth for other users on your network. This is
outside of Ollama's control and not a result of anything in particular
wrong with the software itself.

On Sat, Feb 1, 2025, 3:00 PM bbw1986 @.***> wrote:

I've run into the same problem, every time I reopen and start it the speed
is really fast, but after a while it slows down and the drop is visible to
the naked eye.The download speed goes down in steps.

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/1736#issuecomment-2628818157,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AE32PCROFO2DX5WYT76KGHT2NRWJ5AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRYHAYTQMJVG4
.
You are receiving this because you commented.Message ID:
@.***>

Oh yeah? U genuinely think tht isp gonna block extractly at 98% of a 5gb model? This issue old af and still not fixed, instead blame the isp 😆

@Rizmi commented on GitHub (Aug 28, 2025): > The slow down may actually be a symptom of your ISP performing QoS on large > downloads to ensure bandwidth for other users on your network. This is > outside of Ollama's control and not a result of anything in particular > wrong with the software itself. > > On Sat, Feb 1, 2025, 3:00 PM bbw1986 ***@***.***> wrote: > > > I've run into the same problem, every time I reopen and start it the speed > > is really fast, but after a while it slows down and the drop is visible to > > the naked eye.The download speed goes down in steps. > > > > — > > Reply to this email directly, view it on GitHub > > <https://github.com/ollama/ollama/issues/1736#issuecomment-2628818157>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/AE32PCROFO2DX5WYT76KGHT2NRWJ5AVCNFSM6AAAAABBGGQZPKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRYHAYTQMJVG4> > > . > > You are receiving this because you commented.Message ID: > > ***@***.***> > > > Oh yeah? U genuinely think tht isp gonna block extractly at 98% of a 5gb model? This issue old af and still not fixed, instead blame the isp 😆

GiteaMirror commented

2026-04-22 03:17:59 -05:00

@DaRacci commented on GitHub (Sep 10, 2025):

Still relevant, downloading any model seems to result in a crawl past ~96% unless its a super small model.

Its not an ISP issue, DNS issue or network related, this is something directly related to ollama.

Ollama version: 0.11.7

@DaRacci commented on GitHub (Sep 10, 2025): Still relevant, downloading any model seems to result in a crawl past ~96% unless its a super small model. Its not an ISP issue, DNS issue or network related, this is something directly related to ollama. Ollama version: 0.11.7

GiteaMirror commented

2026-04-22 03:18:00 -05:00

@ferrouswheel commented on GitHub (Sep 14, 2025):

After updating ollama in docker OLLAMA_EXPERIMENT=client2 results in a stream of "context cancelled" log errors and model downloading doesn't occur.

model name: deepseek-r1:8b
ollama version: 0.11.10

time=2025-09-14T20:15:18.602Z level=ERROR source=server.go:304 msg=pulling model=deepseek-r1:8b error="Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/e6/e6a7edc1a4d7d9b2de136a221a57336b76316cfe53a252aeba814496c5ae439d/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&[REMOVED CREDENTIALS]\": context canceled"
time=2025-09-14T20:15:18.603Z level=INFO source=server.go:164 msg=http error="context canceled" status=200 method=POST path=/api/pull content-length=71 remote=172.17.0.1:60820 proto=HTTP/1.1 query=""

Not sure if any of the tokens/signatures in the query string are non-public so have stripped them.

Removing the OLLAMA_EXPERIMENT environment variable allows downloading to occur again but exhibits the old behaviour of slowing to a halt at ~95%. Here is a network graph of the hosts traffic to show how extreme the slow down is.

(the first two traffic peaks are docker pulls, the last is the ollama pull stuck at 95%)

Edit: after 40 minutes it said: time=2025-09-14T21:03:30.184Z level=INFO source=download.go:295 msg="e6a7edc1a4d7 part 12 attempt 0 failed: unexpected EOF, retrying in 1s" and appears to have now completed the download.

I also tried pulling the smallest model I could quickly find and embeddinggemma:300m downloaded with stalling.

@ferrouswheel commented on GitHub (Sep 14, 2025): After updating ollama in docker `OLLAMA_EXPERIMENT=client2` results in a stream of "context cancelled" log errors and model downloading doesn't occur. model name: deepseek-r1:8b ollama version: 0.11.10 ``` time=2025-09-14T20:15:18.602Z level=ERROR source=server.go:304 msg=pulling model=deepseek-r1:8b error="Get \"https://dd20bb891979d25aebc8bec07b2b3bbc.r2.cloudflarestorage.com/ollama/docker/registry/v2/blobs/sha256/e6/e6a7edc1a4d7d9b2de136a221a57336b76316cfe53a252aeba814496c5ae439d/data?X-Amz-Algorithm=AWS4-HMAC-SHA256&[REMOVED CREDENTIALS]\": context canceled" time=2025-09-14T20:15:18.603Z level=INFO source=server.go:164 msg=http error="context canceled" status=200 method=POST path=/api/pull content-length=71 remote=172.17.0.1:60820 proto=HTTP/1.1 query="" ``` Not sure if any of the tokens/signatures in the query string are non-public so have stripped them. Removing the `OLLAMA_EXPERIMENT` environment variable allows downloading to occur again but exhibits the old behaviour of slowing to a halt at ~95%. Here is a network graph of the hosts traffic to show how extreme the slow down is. <img width="1241" height="492" alt="Image" src="https://github.com/user-attachments/assets/5468bd3d-d58f-47a4-a077-9afa312083b9" /> (the first two traffic peaks are docker pulls, the last is the ollama pull stuck at 95%) Edit: after 40 minutes it said: `time=2025-09-14T21:03:30.184Z level=INFO source=download.go:295 msg="e6a7edc1a4d7 part 12 attempt 0 failed: unexpected EOF, retrying in 1s"` and appears to have now completed the download. I also tried pulling the smallest model I could quickly find and `embeddinggemma:300m` downloaded with stalling.

GiteaMirror commented

2026-04-22 03:18:02 -05:00

@barbuz commented on GitHub (Sep 15, 2025):

Same here (version 0.11.10), download speed pulls the brakes hard after 90% on a 3GB model (gemma3). OLLAMA_EXPERIMENT=client2 as suggested above did not fix it, the only valid workaround seems to be stopping the download and restarting it.

@barbuz commented on GitHub (Sep 15, 2025): Same here (version 0.11.10), download speed pulls the brakes hard after 90% on a 3GB model (gemma3). `OLLAMA_EXPERIMENT=client2` as suggested above did not fix it, the only valid workaround seems to be stopping the download and restarting it.

GiteaMirror commented

2026-04-22 03:18:02 -05:00

@adriens commented on GitHub (Sep 15, 2025):

Same here

@adriens commented on GitHub (Sep 15, 2025): Same here

GiteaMirror commented

2026-04-22 03:18:03 -05:00

@omalagouzNW commented on GitHub (Sep 17, 2025):

I tried all the methods above and was stuck at 95% at 155 KB/s downloading a 5GB model, the only thing that actually fixed the issue was switching from Wi-Fi to my phone's 5G personal hotspot, went back to at least 50 MB/s and finished downloading in 30 seconds.

@omalagouzNW commented on GitHub (Sep 17, 2025): I tried all the methods above and was stuck at 95% at 155 KB/s downloading a 5GB model, the only thing that actually fixed the issue was switching from Wi-Fi to my phone's 5G personal hotspot, went back to at least 50 MB/s and finished downloading in 30 seconds.

GiteaMirror commented

2026-04-22 03:18:04 -05:00

@Tarun047 commented on GitHub (Oct 18, 2025):

Wow its full 2 years and this problem is not resolved yet!
So much for running models locally!

@Tarun047 commented on GitHub (Oct 18, 2025): Wow its full 2 years and this problem is not resolved yet! So much for running models locally!

GiteaMirror commented

2026-04-22 03:18:04 -05:00

@FakeUser2121 commented on GitHub (Oct 19, 2025):

This problem is still not fixed and its been almost 2 years. Lm studio downloads 100s of Gb of models just fine. I dont know why isnt this fixed yet.
PS C:\Users\user1> ollama pull llama3.2-vision
pulling manifest
pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB
PS C:\Users\user1> ollama pull llama3.2-vision
pulling manifest
pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB 17 KB/s 57m49s
PS C:\Users\user1> ollama pull llama3.2-vision
pulling manifest
pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB
PS C:\Users\user1> ollama pull llama3.2-vision
pulling manifest
pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB
PS C:\Users\user1> ollama pull llama3.2-vision
pulling manifest
pulling 9999d473417a: 100% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB 116 KB/s 2m17s

@FakeUser2121 commented on GitHub (Oct 19, 2025): This problem is still not fixed and its been almost 2 years. Lm studio downloads 100s of Gb of models just fine. I dont know why isnt this fixed yet. PS C:\Users\user1> ollama pull llama3.2-vision pulling manifest pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB PS C:\Users\user1> ollama pull llama3.2-vision pulling manifest pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB 17 KB/s 57m49s PS C:\Users\user1> ollama pull llama3.2-vision pulling manifest pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB PS C:\Users\user1> ollama pull llama3.2-vision pulling manifest pulling 9999d473417a: 99% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB PS C:\Users\user1> ollama pull llama3.2-vision pulling manifest pulling 9999d473417a: 100% ▕█████████████████████████████████████████████████████████ ▏ 7.8 GB/7.8 GB 116 KB/s 2m17s

GiteaMirror commented

2026-04-22 03:18:05 -05:00

@rampageservices commented on GitHub (Oct 20, 2025):

This problem is still not fixed and its been almost 2 years. Lm studio downloads 100s of Gb of models just fine. I dont know why isnt this fixed yet.

I just want to point out that this is not the way to ask for help with Open Source software. You are not entitled to free fixes. To others reading this, don't be this person.

@rampageservices commented on GitHub (Oct 20, 2025): > This problem is still not fixed and its been almost 2 years. Lm studio downloads 100s of Gb of models just fine. I dont know why isnt this fixed yet. I just want to point out that this is not the way to ask for help with Open Source software. You are not entitled to free fixes. To others reading this, don't be this person.

GiteaMirror commented

2026-04-22 03:18:06 -05:00

@mendi80 commented on GitHub (Oct 23, 2025):

working solution. after the slow down start, Ctrl+C, switch the network to your phone hotspot (or other network with different public ip address), continue the downloading (ollama pull mymodel) for some seconds, Ctrl+C again. back to you home router, continue downloading. now it will continue at max speed.

very strange bug. but i guess it is easy to fix.

@mendi80 commented on GitHub (Oct 23, 2025): working solution. after the slow down start, Ctrl+C, switch the network to your phone hotspot (or other network with different public ip address), continue the downloading (ollama pull mymodel) for some seconds, Ctrl+C again. back to you home router, continue downloading. now it will continue at max speed. very strange bug. but i guess it is easy to fix.

GiteaMirror commented

2026-04-22 03:18:07 -05:00

@Phizicks commented on GitHub (Oct 24, 2025):

instead of going through cloudflare etc, could think the possibility in future to allow torrenting of the models.

@Phizicks commented on GitHub (Oct 24, 2025): instead of going through cloudflare etc, could think the possibility in future to allow torrenting of the models.

GiteaMirror commented

2026-04-22 03:18:09 -05:00

@sudhir-asuracore commented on GitHub (Oct 26, 2025):

I just switch on my vpn, connect to a zone in USA and the download completes the last few percent at full bandwidth. Still not sure what the cause is, but switching to a vpn spot helps.

@sudhir-asuracore commented on GitHub (Oct 26, 2025): I just switch on my vpn, connect to a zone in USA and the download completes the last few percent at full bandwidth. Still not sure what the cause is, but switching to a vpn spot helps.

GiteaMirror commented

2026-04-22 03:18:09 -05:00

@AbhishekRH commented on GitHub (Oct 27, 2025):

I just switch on my vpn, connect to a zone in USA and the download completes the last few percent at full bandwidth. Still not sure what the cause is, but switching to a vpn spot helps.

This ACTUALLY WORKS!! 🔥

Thanks man!

@AbhishekRH commented on GitHub (Oct 27, 2025): > I just switch on my vpn, connect to a zone in USA and the download completes the last few percent at full bandwidth. Still not sure what the cause is, but switching to a vpn spot helps. This ACTUALLY WORKS!! 🔥 Thanks man!

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#26751