github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 13:19:40 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #3476] Vision models (bak/llava) crash and exit Ollama docker container #2141

New Issue
Closed
opened 2026-04-12 12:22:28 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-12 12:22:28 -05:00
Owner
Copy Link

Originally created by @RELNO on GitHub (Apr 3, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3476

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

When running vision models such as bakllava or llava (all sizes), the docker container is crashing, returning the following error: RemoteProtocolError: Server disconnected without sending a response.

Note: llama2 or other llms run well, i.e

completion = ollama.chat(
    model='llama2', messages=[{"role": "user", "content": 'What is the capital of France?'}])
print(completion)

{'model': 'llama2', 'created_at': '2024-04-03T15:30:45.152975648Z', 'message': {'role': 'assistant', 'content': '\nThe capital of France is Paris.'}, 'done': True, 'total_duration': 4938479062, 'load_duration': 2702775270, 'prompt_eval_count': 27, 'prompt_eval_duration': 1461042000, 'eval_count': 9, 'eval_duration': 773529000}

The docker logs [shortened the 100+ goroutines]


goroutine 100 gp=0xc000702fc0 m=34 mp=0xc000b01008 [syscall]:
runtime.cgocall(0xebebf0, 0xc0000485e0)
        runtime/cgocall.go:157 +0x4b fp=0xc0000485b8 sp=0xc000048580 pc=0x40a72b
github.com/ollama/ollama/llm._Cfunc_dyn_llama_server_init({0x7f8bdc003f00, 0x7f8bce0a5450, 0x7f8bce09c720, 0x7f8bce0a48a0, 0x7f8bce0abb40, 0x7f8bce0a3250, 0x7f8bce09e840, 0x7f8bce09c7a0, 0x7f8bce0ac580, 0x7f8bce0ab6e0, ...}, ...)
        _cgo_gotypes.go:286 +0x45 fp=0xc0000485e0 sp=0xc0000485b8 pc=0xce9345
github.com/ollama/ollama/llm.newDynExtServer.func7(0x200?, 0xc000458000, 0xc000a69ce0)
        github.com/ollama/ollama/llm/dyn_ext_server.go:151 +0x112 fp=0xc000048720 sp=0xc0000485e0 pc=0xceaaf2
github.com/ollama/ollama/llm.newDynExtServer({0xc00062c000, 0x36}, {0xc00044a2a0, _}, {_, _, _}, {0xc000931910, 0x1, 0x1}, ...)
        github.com/ollama/ollama/llm/dyn_ext_server.go:151 +0xb58 fp=0xc000048a80 sp=0xc000048720 pc=0xcea638
github.com/ollama/ollama/llm.newLlmServer({{_, _, _}, {_, _}, {_, _}}, {_, _}, {0x0, ...}, ...)
        github.com/ollama/ollama/llm/llm.go:166 +0x4c5 fp=0xc000048c40 sp=0xc000048a80 pc=0xce6b25
github.com/ollama/ollama/llm.New({0xc00044a2a0, 0x62}, {0x0, 0x0, 0x0}, {0xc000931910, _, _}, {{0x0, 0x1000, ...}, ...})
        github.com/ollama/ollama/llm/llm.go:131 +0x90e fp=0xc000048ed8 sp=0xc000048c40 pc=0xce64ce
github.com/ollama/ollama/server.load(0xc000003080?, 0xc000003080, {{0x0, 0x1000, 0x200, 0x1, 0xffffffffffffffff, 0x0, 0x0, 0x1, ...}, ...}, ...)
        github.com/ollama/ollama/server/routes.go:86 +0x325 fp=0xc000049028 sp=0xc000048ed8 pc=0xe973a5
github.com/ollama/ollama/server.ChatHandler(0xc00023a100)
        github.com/ollama/ollama/server/routes.go:1272 +0xa4b fp=0xc000049730 sp=0xc000049028 pc=0xea310b
github.com/gin-gonic/gin.(*Context).Next(0xc00023a100)
        github.com/gin-gonic/gin@v1.9.1/context.go:174 +0x2b fp=0xc000049750 sp=0xc000049730 pc=0xe6ab8b
github.com/ollama/ollama/server.(*Server).GenerateRoutes.allowedHostsMiddleware.func3(0xc00023a100)
        github.com/ollama/ollama/server/routes.go:990 +0x15d fp=0xc0000497a8 sp=0xc000049750 pc=0xea187d
github.com/gin-gonic/gin.(*Context).Next(...)
        github.com/gin-gonic/gin@v1.9.1/context.go:174
github.com/gin-gonic/gin.CustomRecoveryWithWriter.func1(0xc00023a100)
        github.com/gin-gonic/gin@v1.9.1/recovery.go:102 +0x7a fp=0xc0000497f8 sp=0xc0000497a8 pc=0xe77a7a
github.com/gin-gonic/gin.(*Context).Next(...)
        github.com/gin-gonic/gin@v1.9.1/context.go:174
github.com/gin-gonic/gin.LoggerWithConfig.func1(0xc00023a100)
        github.com/gin-gonic/gin@v1.9.1/logger.go:240 +0xdd fp=0xc0000499a8 sp=0xc0000497f8 pc=0xe76bbd
github.com/gin-gonic/gin.(*Context).Next(...)
        github.com/gin-gonic/gin@v1.9.1/context.go:174
github.com/gin-gonic/gin.(*Engine).handleHTTPRequest(0xc0001c24e0, 0xc00023a100)
        github.com/gin-gonic/gin@v1.9.1/gin.go:620 +0x66e fp=0xc000049b28 sp=0xc0000499a8 pc=0xe760ae
github.com/gin-gonic/gin.(*Engine).ServeHTTP(0xc0001c24e0, {0x11872730, 0xc0005caa80}, 0xc000613560)
        github.com/gin-gonic/gin@v1.9.1/gin.go:576 +0x1b2 fp=0xc000049b60 sp=0xc000049b28 pc=0xe75872
net/http.serverHandler.ServeHTTP({0x11870610?}, {0x11872730?, 0xc0005caa80?}, 0x6?)
        net/http/server.go:3137 +0x8e fp=0xc000049b90 sp=0xc000049b60 pc=0x70028e
net/http.(*conn).serve(0xc0001fcc60, {0x11874ae8, 0xc00060e570})
        net/http/server.go:2039 +0x5e8 fp=0xc000049fb8 sp=0xc000049b90 pc=0x6fb648
net/http.(*Server).Serve.gowrap3()
        net/http/server.go:3285 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x700aa8
runtime.goexit({})
        runtime/asm_amd64.s:1695 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x4742e1
created by net/http.(*Server).Serve in goroutine 1
        net/http/server.go:3285 +0x4b4

goroutine 1 gp=0xc0000061c0 m=nil [IO wait, 1 minutes]:
runtime.gopark(0xc00004e008?, 0x0?, 0xc0?, 0x61?, 0xc0007d9868?)
        runtime/proc.go:402 +0xce fp=0xc0007d9830 sp=0xc0007d9810 pc=0x44160e
runtime.netpollblock(0xc0007d98c8?, 0x409ec6?, 0x0?)
        runtime/netpoll.go:573 +0xf7 fp=0xc0007d9868 sp=0xc0007d9830 pc=0x43a377
internal/poll.runtime_pollWait(0x7f8cbbfc66d0, 0x72)
        runtime/netpoll.go:345 +0x85 fp=0xc0007d9888 sp=0xc0007d9868 pc=0x46e9e5
internal/poll.(*pollDesc).wait(0x3?, 0x1?, 0x0)
        internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0007d98b0 sp=0xc0007d9888 pc=0x503707
internal/poll.(*pollDesc).waitRead(...)
        internal/poll/fd_poll_runtime.go:89
internal/poll.(*FD).Accept(0xc0005b4600)
        internal/poll/fd_unix.go:611 +0x2ac fp=0xc0007d9958 sp=0xc0007d98b0 pc=0x508aac
net.(*netFD).accept(0xc0005b4600)
        net/fd_unix.go:172 +0x29 fp=0xc0007d9a10 sp=0xc0007d9958 pc=0x598f89
net.(*TCPListener).accept(0xc00049a820)
        net/tcpsock_posix.go:159 +0x1e fp=0xc0007d9a38 sp=0xc0007d9a10 pc=0x5aeebe
net.(*TCPListener).Accept(0xc00049a820)
        net/tcpsock.go:327 +0x30 fp=0xc0007d9a68 sp=0xc0007d9a38 pc=0x5ae0b0
net/http.(*onceCloseListener).Accept(0xc0001fcc60?)
        <autogenerated>:1 +0x24 fp=0xc0007d9a80 sp=0xc0007d9a68 pc=0x722ce4
net/http.(*Server).Serve(0xc0005cc690, {0x118724c0, 0xc00049a820})
        net/http/server.go:3255 +0x33e fp=0xc0007d9bb0 sp=0xc0007d9a80 pc=0x7006be
github.com/ollama/ollama/server.Serve({0x118724c0, 0xc00049a820})
        github.com/ollama/ollama/server/routes.go:1145 +0x4eb fp=0xc0007d9cc0 sp=0xc0007d9bb0 pc=0xea1dab
github.com/ollama/ollama/cmd.RunServer(0xc00023ae00?, {0x11fc38c0?, 0x4?, 0x1055b75?})
        github.com/ollama/ollama/cmd/cmd.go:787 +0x1b9 fp=0xc0007d9d58 sp=0xc0007d9cc0 pc=0xeb5b59
github.com/spf13/cobra.(*Command).execute(0xc0004e0f08, {0x11fc38c0, 0x0, 0x0})
        github.com/spf13/cobra@v1.7.0/command.go:940 +0x882 fp=0xc0007d9e78 sp=0xc0007d9d58 pc=0x795c62
github.com/spf13/cobra.(*Command).ExecuteC(0xc0004e0308)
        github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc0007d9f30 sp=0xc0007d9e78 pc=0x7964a5
github.com/spf13/cobra.(*Command).Execute(...)
        github.com/spf13/cobra@v1.7.0/command.go:992
github.com/spf13/cobra.(*Command).ExecuteContext(...)
        github.com/spf13/cobra@v1.7.0/command.go:985
main.main()
        github.com/ollama/ollama/main.go:11 +0x4d fp=0xc0007d9f50 sp=0xc0007d9f30 pc=0xebdd0d
runtime.main()
        runtime/proc.go:271 +0x29d fp=0xc0007d9fe0 sp=0xc0007d9f50 pc=0x4411dd
runtime.goexit({})
        runtime/asm_amd64.s:1695 +0x1 fp=0xc0007d9fe8 sp=0xc0007d9fe0 pc=0x4742e1
...

goroutine 101 gp=0xc001284700 m=nil [IO wait, 1 minutes]:
runtime.gopark(0x40?, 0x3f8ccccd3f4ccccd?, 0x0?, 0x0?, 0xb?)
        runtime/proc.go:402 +0xce fp=0xc0004cb5a8 sp=0xc0004cb588 pc=0x44160e
runtime.netpollblock(0x4866f8?, 0x409ec6?, 0x0?)
        runtime/netpoll.go:573 +0xf7 fp=0xc0004cb5e0 sp=0xc0004cb5a8 pc=0x43a377
internal/poll.runtime_pollWait(0x7f8cbbfc65d8, 0x72)
        runtime/netpoll.go:345 +0x85 fp=0xc0004cb600 sp=0xc0004cb5e0 pc=0x46e9e5
internal/poll.(*pollDesc).wait(0xc0005b4180?, 0xc000acca61?, 0x0)
        internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0004cb628 sp=0xc0004cb600 pc=0x503707
internal/poll.(*pollDesc).waitRead(...)
        internal/poll/fd_poll_runtime.go:89
internal/poll.(*FD).Read(0xc0005b4180, {0xc000acca61, 0x1, 0x1})
        internal/poll/fd_unix.go:164 +0x27a fp=0xc0004cb6c0 sp=0xc0004cb628 pc=0x5049fa
net.(*netFD).Read(0xc0005b4180, {0xc000acca61?, 0x40?, 0x3f8ccccd3f4ccccd?})
        net/fd_posix.go:55 +0x25 fp=0xc0004cb708 sp=0xc0004cb6c0 pc=0x596fa5
net.(*conn).Read(0xc0000c6738, {0xc000acca61?, 0x0?, 0xc0004cb7d0?})
        net/net.go:179 +0x45 fp=0xc0004cb750 sp=0xc0004cb708 pc=0x5a5e05
net.(*TCPConn).Read(0x0?, {0xc000acca61?, 0x11f5d540?, 0xea4be0?})
        <autogenerated>:1 +0x25 fp=0xc0004cb780 sp=0xc0004cb750 pc=0x5b7485
net/http.(*connReader).backgroundRead(0xc000acca50)
        net/http/server.go:681 +0x37 fp=0xc0004cb7c8 sp=0xc0004cb780 pc=0x6f55b7
net/http.(*connReader).startBackgroundRead.gowrap2()
        net/http/server.go:677 +0x25 fp=0xc0004cb7e0 sp=0xc0004cb7c8 pc=0x6f54e5
runtime.goexit({})
        runtime/asm_amd64.s:1695 +0x1 fp=0xc0004cb7e8 sp=0xc0004cb7e0 pc=0x4742e1
created by net/http.(*connReader).startBackgroundRead in goroutine 100
        net/http/server.go:677 +0xba

What did you expect to see?

Vision model should work as expected.

Steps to reproduce

docker pull ollama/ollama
docker run -d --gpus=all -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama

docker exec -it ollama ollama run llama2
docker exec -it ollama ollama run bakllava  < crashes the container here

-- or in python --

ollama.chat(
	model="bakllava",
    
	messages=[
		{
			'role': 'user',
			'content': 'how to make a cake',
			'images': []
		}
	]
)

Are there any recent changes that introduced the issue?

no, was running well a couple of days ago. no update to any of the pieces.

OS

Linux

Architecture

No response

Platform

Docker

Ollama version

ollama version is 0.1.30

GPU

Nvidia

GPU info

4 X NVIDIA GeForce GTX TITAN X @ 12gb

CPU

Intel

Other software

No response

Originally created by @RELNO on GitHub (Apr 3, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3476 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? When running vision models such as bakllava or llava (all sizes), the docker container is crashing, returning the following error: `RemoteProtocolError: Server disconnected without sending a response.` Note: llama2 or other llms run well, i.e ``` completion = ollama.chat( model='llama2', messages=[{"role": "user", "content": 'What is the capital of France?'}]) print(completion) {'model': 'llama2', 'created_at': '2024-04-03T15:30:45.152975648Z', 'message': {'role': 'assistant', 'content': '\nThe capital of France is Paris.'}, 'done': True, 'total_duration': 4938479062, 'load_duration': 2702775270, 'prompt_eval_count': 27, 'prompt_eval_duration': 1461042000, 'eval_count': 9, 'eval_duration': 773529000} ``` The docker logs [shortened the 100+ goroutines] ``` goroutine 100 gp=0xc000702fc0 m=34 mp=0xc000b01008 [syscall]: runtime.cgocall(0xebebf0, 0xc0000485e0) runtime/cgocall.go:157 +0x4b fp=0xc0000485b8 sp=0xc000048580 pc=0x40a72b github.com/ollama/ollama/llm._Cfunc_dyn_llama_server_init({0x7f8bdc003f00, 0x7f8bce0a5450, 0x7f8bce09c720, 0x7f8bce0a48a0, 0x7f8bce0abb40, 0x7f8bce0a3250, 0x7f8bce09e840, 0x7f8bce09c7a0, 0x7f8bce0ac580, 0x7f8bce0ab6e0, ...}, ...) _cgo_gotypes.go:286 +0x45 fp=0xc0000485e0 sp=0xc0000485b8 pc=0xce9345 github.com/ollama/ollama/llm.newDynExtServer.func7(0x200?, 0xc000458000, 0xc000a69ce0) github.com/ollama/ollama/llm/dyn_ext_server.go:151 +0x112 fp=0xc000048720 sp=0xc0000485e0 pc=0xceaaf2 github.com/ollama/ollama/llm.newDynExtServer({0xc00062c000, 0x36}, {0xc00044a2a0, _}, {_, _, _}, {0xc000931910, 0x1, 0x1}, ...) github.com/ollama/ollama/llm/dyn_ext_server.go:151 +0xb58 fp=0xc000048a80 sp=0xc000048720 pc=0xcea638 github.com/ollama/ollama/llm.newLlmServer({{_, _, _}, {_, _}, {_, _}}, {_, _}, {0x0, ...}, ...) github.com/ollama/ollama/llm/llm.go:166 +0x4c5 fp=0xc000048c40 sp=0xc000048a80 pc=0xce6b25 github.com/ollama/ollama/llm.New({0xc00044a2a0, 0x62}, {0x0, 0x0, 0x0}, {0xc000931910, _, _}, {{0x0, 0x1000, ...}, ...}) github.com/ollama/ollama/llm/llm.go:131 +0x90e fp=0xc000048ed8 sp=0xc000048c40 pc=0xce64ce github.com/ollama/ollama/server.load(0xc000003080?, 0xc000003080, {{0x0, 0x1000, 0x200, 0x1, 0xffffffffffffffff, 0x0, 0x0, 0x1, ...}, ...}, ...) github.com/ollama/ollama/server/routes.go:86 +0x325 fp=0xc000049028 sp=0xc000048ed8 pc=0xe973a5 github.com/ollama/ollama/server.ChatHandler(0xc00023a100) github.com/ollama/ollama/server/routes.go:1272 +0xa4b fp=0xc000049730 sp=0xc000049028 pc=0xea310b github.com/gin-gonic/gin.(*Context).Next(0xc00023a100) github.com/gin-gonic/gin@v1.9.1/context.go:174 +0x2b fp=0xc000049750 sp=0xc000049730 pc=0xe6ab8b github.com/ollama/ollama/server.(*Server).GenerateRoutes.allowedHostsMiddleware.func3(0xc00023a100) github.com/ollama/ollama/server/routes.go:990 +0x15d fp=0xc0000497a8 sp=0xc000049750 pc=0xea187d github.com/gin-gonic/gin.(*Context).Next(...) github.com/gin-gonic/gin@v1.9.1/context.go:174 github.com/gin-gonic/gin.CustomRecoveryWithWriter.func1(0xc00023a100) github.com/gin-gonic/gin@v1.9.1/recovery.go:102 +0x7a fp=0xc0000497f8 sp=0xc0000497a8 pc=0xe77a7a github.com/gin-gonic/gin.(*Context).Next(...) github.com/gin-gonic/gin@v1.9.1/context.go:174 github.com/gin-gonic/gin.LoggerWithConfig.func1(0xc00023a100) github.com/gin-gonic/gin@v1.9.1/logger.go:240 +0xdd fp=0xc0000499a8 sp=0xc0000497f8 pc=0xe76bbd github.com/gin-gonic/gin.(*Context).Next(...) github.com/gin-gonic/gin@v1.9.1/context.go:174 github.com/gin-gonic/gin.(*Engine).handleHTTPRequest(0xc0001c24e0, 0xc00023a100) github.com/gin-gonic/gin@v1.9.1/gin.go:620 +0x66e fp=0xc000049b28 sp=0xc0000499a8 pc=0xe760ae github.com/gin-gonic/gin.(*Engine).ServeHTTP(0xc0001c24e0, {0x11872730, 0xc0005caa80}, 0xc000613560) github.com/gin-gonic/gin@v1.9.1/gin.go:576 +0x1b2 fp=0xc000049b60 sp=0xc000049b28 pc=0xe75872 net/http.serverHandler.ServeHTTP({0x11870610?}, {0x11872730?, 0xc0005caa80?}, 0x6?) net/http/server.go:3137 +0x8e fp=0xc000049b90 sp=0xc000049b60 pc=0x70028e net/http.(*conn).serve(0xc0001fcc60, {0x11874ae8, 0xc00060e570}) net/http/server.go:2039 +0x5e8 fp=0xc000049fb8 sp=0xc000049b90 pc=0x6fb648 net/http.(*Server).Serve.gowrap3() net/http/server.go:3285 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x700aa8 runtime.goexit({}) runtime/asm_amd64.s:1695 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x4742e1 created by net/http.(*Server).Serve in goroutine 1 net/http/server.go:3285 +0x4b4 goroutine 1 gp=0xc0000061c0 m=nil [IO wait, 1 minutes]: runtime.gopark(0xc00004e008?, 0x0?, 0xc0?, 0x61?, 0xc0007d9868?) runtime/proc.go:402 +0xce fp=0xc0007d9830 sp=0xc0007d9810 pc=0x44160e runtime.netpollblock(0xc0007d98c8?, 0x409ec6?, 0x0?) runtime/netpoll.go:573 +0xf7 fp=0xc0007d9868 sp=0xc0007d9830 pc=0x43a377 internal/poll.runtime_pollWait(0x7f8cbbfc66d0, 0x72) runtime/netpoll.go:345 +0x85 fp=0xc0007d9888 sp=0xc0007d9868 pc=0x46e9e5 internal/poll.(*pollDesc).wait(0x3?, 0x1?, 0x0) internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0007d98b0 sp=0xc0007d9888 pc=0x503707 internal/poll.(*pollDesc).waitRead(...) internal/poll/fd_poll_runtime.go:89 internal/poll.(*FD).Accept(0xc0005b4600) internal/poll/fd_unix.go:611 +0x2ac fp=0xc0007d9958 sp=0xc0007d98b0 pc=0x508aac net.(*netFD).accept(0xc0005b4600) net/fd_unix.go:172 +0x29 fp=0xc0007d9a10 sp=0xc0007d9958 pc=0x598f89 net.(*TCPListener).accept(0xc00049a820) net/tcpsock_posix.go:159 +0x1e fp=0xc0007d9a38 sp=0xc0007d9a10 pc=0x5aeebe net.(*TCPListener).Accept(0xc00049a820) net/tcpsock.go:327 +0x30 fp=0xc0007d9a68 sp=0xc0007d9a38 pc=0x5ae0b0 net/http.(*onceCloseListener).Accept(0xc0001fcc60?) <autogenerated>:1 +0x24 fp=0xc0007d9a80 sp=0xc0007d9a68 pc=0x722ce4 net/http.(*Server).Serve(0xc0005cc690, {0x118724c0, 0xc00049a820}) net/http/server.go:3255 +0x33e fp=0xc0007d9bb0 sp=0xc0007d9a80 pc=0x7006be github.com/ollama/ollama/server.Serve({0x118724c0, 0xc00049a820}) github.com/ollama/ollama/server/routes.go:1145 +0x4eb fp=0xc0007d9cc0 sp=0xc0007d9bb0 pc=0xea1dab github.com/ollama/ollama/cmd.RunServer(0xc00023ae00?, {0x11fc38c0?, 0x4?, 0x1055b75?}) github.com/ollama/ollama/cmd/cmd.go:787 +0x1b9 fp=0xc0007d9d58 sp=0xc0007d9cc0 pc=0xeb5b59 github.com/spf13/cobra.(*Command).execute(0xc0004e0f08, {0x11fc38c0, 0x0, 0x0}) github.com/spf13/cobra@v1.7.0/command.go:940 +0x882 fp=0xc0007d9e78 sp=0xc0007d9d58 pc=0x795c62 github.com/spf13/cobra.(*Command).ExecuteC(0xc0004e0308) github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc0007d9f30 sp=0xc0007d9e78 pc=0x7964a5 github.com/spf13/cobra.(*Command).Execute(...) github.com/spf13/cobra@v1.7.0/command.go:992 github.com/spf13/cobra.(*Command).ExecuteContext(...) github.com/spf13/cobra@v1.7.0/command.go:985 main.main() github.com/ollama/ollama/main.go:11 +0x4d fp=0xc0007d9f50 sp=0xc0007d9f30 pc=0xebdd0d runtime.main() runtime/proc.go:271 +0x29d fp=0xc0007d9fe0 sp=0xc0007d9f50 pc=0x4411dd runtime.goexit({}) runtime/asm_amd64.s:1695 +0x1 fp=0xc0007d9fe8 sp=0xc0007d9fe0 pc=0x4742e1 ... goroutine 101 gp=0xc001284700 m=nil [IO wait, 1 minutes]: runtime.gopark(0x40?, 0x3f8ccccd3f4ccccd?, 0x0?, 0x0?, 0xb?) runtime/proc.go:402 +0xce fp=0xc0004cb5a8 sp=0xc0004cb588 pc=0x44160e runtime.netpollblock(0x4866f8?, 0x409ec6?, 0x0?) runtime/netpoll.go:573 +0xf7 fp=0xc0004cb5e0 sp=0xc0004cb5a8 pc=0x43a377 internal/poll.runtime_pollWait(0x7f8cbbfc65d8, 0x72) runtime/netpoll.go:345 +0x85 fp=0xc0004cb600 sp=0xc0004cb5e0 pc=0x46e9e5 internal/poll.(*pollDesc).wait(0xc0005b4180?, 0xc000acca61?, 0x0) internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0004cb628 sp=0xc0004cb600 pc=0x503707 internal/poll.(*pollDesc).waitRead(...) internal/poll/fd_poll_runtime.go:89 internal/poll.(*FD).Read(0xc0005b4180, {0xc000acca61, 0x1, 0x1}) internal/poll/fd_unix.go:164 +0x27a fp=0xc0004cb6c0 sp=0xc0004cb628 pc=0x5049fa net.(*netFD).Read(0xc0005b4180, {0xc000acca61?, 0x40?, 0x3f8ccccd3f4ccccd?}) net/fd_posix.go:55 +0x25 fp=0xc0004cb708 sp=0xc0004cb6c0 pc=0x596fa5 net.(*conn).Read(0xc0000c6738, {0xc000acca61?, 0x0?, 0xc0004cb7d0?}) net/net.go:179 +0x45 fp=0xc0004cb750 sp=0xc0004cb708 pc=0x5a5e05 net.(*TCPConn).Read(0x0?, {0xc000acca61?, 0x11f5d540?, 0xea4be0?}) <autogenerated>:1 +0x25 fp=0xc0004cb780 sp=0xc0004cb750 pc=0x5b7485 net/http.(*connReader).backgroundRead(0xc000acca50) net/http/server.go:681 +0x37 fp=0xc0004cb7c8 sp=0xc0004cb780 pc=0x6f55b7 net/http.(*connReader).startBackgroundRead.gowrap2() net/http/server.go:677 +0x25 fp=0xc0004cb7e0 sp=0xc0004cb7c8 pc=0x6f54e5 runtime.goexit({}) runtime/asm_amd64.s:1695 +0x1 fp=0xc0004cb7e8 sp=0xc0004cb7e0 pc=0x4742e1 created by net/http.(*connReader).startBackgroundRead in goroutine 100 net/http/server.go:677 +0xba ``` ### What did you expect to see? Vision model should work as expected. ### Steps to reproduce ``` docker pull ollama/ollama docker run -d --gpus=all -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama docker exec -it ollama ollama run llama2 docker exec -it ollama ollama run bakllava < crashes the container here ``` -- or in python -- ``` ollama.chat( model="bakllava", messages=[ { 'role': 'user', 'content': 'how to make a cake', 'images': [] } ] ) ``` ### Are there any recent changes that introduced the issue? no, was running well a couple of days ago. no update to any of the pieces. ### OS Linux ### Architecture _No response_ ### Platform Docker ### Ollama version ollama version is 0.1.30 ### GPU Nvidia ### GPU info 4 X NVIDIA GeForce GTX TITAN X @ 12gb ### CPU Intel ### Other software _No response_
GiteaMirror added the bugnvidia labels 2026-04-12 12:22:28 -05:00
GiteaMirror commented 2026-04-12 12:22:29 -05:00
Author
Owner
Copy Link

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

I'm pretty sure this has been fixed for a while. I'm able to load both of these models on a 2x cuda setup (I don't have a 4x setup to test)

If you're still having problems on the latest version, please share an updated server log and I'll reopen the issue.

<!-- gh-comment-id:2249108620 --> @dhiltgen commented on GitHub (Jul 25, 2024): I'm pretty sure this has been fixed for a while. I'm able to load both of these models on a 2x cuda setup (I don't have a 4x setup to test) If you're still having problems on the latest version, please share an updated server log and I'll reopen the issue.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug nvidia
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#2141
Delete Branch "%!s()"

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