Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 14:13:35 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #4386] [BUG]: No Output Via Ollama #2797

New Issue
Closed
opened 2026-02-22 18:31:18 -05:00 by yindo · 9 comments
yindo commented 2026-02-22 18:31:18 -05:00
Owner
Copy Link
Copy Source

Originally created by @EntityDeleter on GitHub (Sep 16, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4386

How are you running AnythingLLM?

AnythingLLM desktop app

What happened?

I am using EverythingLLM, connected to Ollama via their API at 100.69.242.6:11434.

When I use agents, let it be Web Scraping or Chart Generation, the first query is always correct/perfect.

If I attempt to query the agent again in the same thread, however, it fails in a multitude of ways:

  • No output at all.
  • An error occurred while streaming the response. Failed to fetch.
  • Blank Graph + I’ve already generated the chart response.
  • Failed to call web_scraping.
  • The entire thread history disappears.

The LLM works fine without agent calling, and when I directly use ollama run, so it is not a problem with ollama.

I only use the official tools without any third-party addons.

I am using gpt-oss:20b.

I am running the model split between a GTX 1080 with 8 GB of VRAM and an Xeon W-2135 equipped with 32 GB of RAM. I also use Arch Linux and have installed the app via the official script.

Let me know if I need to provide any other information. I have also attached screenshots as an example. For some reason, the agent-calling pop-up disappeared when I took the screenshot.

Image Image Image

Are there known steps to reproduce?

Install Ollama
Run ChatGPT OSS 20B
Install Anything LLM
Connect Ollama to the app
Use the Ollama model as an agent multiple times

Originally created by @EntityDeleter on GitHub (Sep 16, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4386 ### How are you running AnythingLLM? AnythingLLM desktop app ### What happened? I am using EverythingLLM, connected to Ollama via their API at 100.69.242.6:11434. When I use agents, let it be Web Scraping or Chart Generation, the first query is always correct/perfect. If I attempt to query the agent again in the same thread, however, it fails in a multitude of ways: - No output at all. - An error occurred while streaming the response. Failed to fetch. - Blank Graph + I’ve already generated the chart response. - Failed to call web_scraping. - The entire thread history disappears. The LLM works fine without agent calling, and when I directly use ollama run, so it is not a problem with ollama. I only use the official tools without any third-party addons. I am using gpt-oss:20b. I am running the model split between a GTX 1080 with 8 GB of VRAM and an Xeon W-2135 equipped with 32 GB of RAM. I also use Arch Linux and have installed the app via the official script. Let me know if I need to provide any other information. I have also attached screenshots as an example. For some reason, the agent-calling pop-up disappeared when I took the screenshot. <img width="1896" height="1009" alt="Image" src="https://github.com/user-attachments/assets/ee2e26f5-8c5b-4a91-b6d9-91a5b3c8bd01" /> <img width="1896" height="1009" alt="Image" src="https://github.com/user-attachments/assets/eb90286c-8b3f-482e-9d50-4e4cbc6b74fe" /> <img width="1896" height="1009" alt="Image" src="https://github.com/user-attachments/assets/54d38c8e-cffe-4db2-a040-b42e02f89d4c" /> ### Are there known steps to reproduce? Install Ollama Run ChatGPT OSS 20B Install Anything LLM Connect Ollama to the app Use the Ollama model as an agent multiple times
yindo added the possible bugneeds info / can't replicateinvestigating labels 2026-02-22 18:31:18 -05:00
yindo closed this issue 2026-02-22 18:31:18 -05:00
yindo commented 2026-02-22 18:31:18 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 16, 2025):

Here is the full Ollama Debug log if it is of any use:

OLLAMA_NEW_ESTIMATES=1 ollama serve                                                                                     ─╯
time=2025-09-16T18:55:10.101+05:30 level=INFO source=routes.go:1331 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:16384 OLLAMA_DEBUG:INFO OLLAMA_FLASH_ATTENTION:true OLLAMA_GPU_OVERHEAD:2147483648 OLLAMA_HOST:http://100.69.242.6:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:2562047h47m16.854775807s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/sudar/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NEW_ESTIMATES:true OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:*https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:true ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]"
time=2025-09-16T18:55:10.102+05:30 level=INFO source=images.go:477 msg="total blobs: 19"
time=2025-09-16T18:55:10.103+05:30 level=INFO source=images.go:484 msg="total unused blobs removed: 0"
time=2025-09-16T18:55:10.103+05:30 level=INFO source=routes.go:1384 msg="Listening on 100.69.242.6:11434 (version 0.11.10)"
time=2025-09-16T18:55:10.103+05:30 level=INFO source=gpu.go:217 msg="looking for compatible GPUs"
time=2025-09-16T18:55:10.218+05:30 level=INFO source=types.go:131 msg="inference compute" id=GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 library=cuda variant=v12 compute=6.1 driver=13.0 name="NVIDIA GeForce GTX 1080" total="7.9 GiB" available="7.3 GiB"
time=2025-09-16T18:55:10.218+05:30 level=INFO source=routes.go:1425 msg="entering low vram mode" "total vram"="5.9 GiB" threshold="20.0 GiB"
time=2025-09-16T18:55:30.819+05:30 level=INFO source=server.go:166 msg="enabling new memory estimates"
time=2025-09-16T18:55:30.890+05:30 level=INFO source=server.go:199 msg="model wants flash attention"
time=2025-09-16T18:55:30.890+05:30 level=INFO source=server.go:216 msg="enabling flash attention"
time=2025-09-16T18:55:30.890+05:30 level=WARN source=server.go:224 msg="kv cache type not supported by model" type=""
time=2025-09-16T18:55:30.891+05:30 level=INFO source=server.go:398 msg="starting runner" cmd="/usr/bin/ollama runner --ollama-engine --model /home/sudar/.ollama/models/blobs/sha256-b112e727c6f18875636c56a779790a590d705aec9e1c0eb5a97d51fc2a778583 --port 44357"
time=2025-09-16T18:55:30.892+05:30 level=INFO source=server.go:671 msg="loading model" "model layers"=25 requested=-1
time=2025-09-16T18:55:30.902+05:30 level=INFO source=runner.go:1251 msg="starting ollama engine"
time=2025-09-16T18:55:30.902+05:30 level=INFO source=runner.go:1286 msg="Server listening on 127.0.0.1:44357"
time=2025-09-16T18:55:30.964+05:30 level=INFO source=server.go:677 msg="system memory" total="31.0 GiB" free="25.3 GiB" free_swap="4.0 GiB"
time=2025-09-16T18:55:30.964+05:30 level=INFO source=server.go:685 msg="gpu memory" id=GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 available="4.9 GiB" free="7.3 GiB" minimum="457.0 MiB" overhead="2.0 GiB"
time=2025-09-16T18:55:30.965+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:25[ID:GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 Layers:25(0..24)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}"
time=2025-09-16T18:55:31.016+05:30 level=INFO source=ggml.go:131 msg="" architecture=gptoss file_type=MXFP4 name="" description="" num_tensors=315 num_key_values=30
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
operator() double registration of ggml_uncaught_exception
load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-skylakex.so
time=2025-09-16T18:55:31.027+05:30 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.AVX512=1 CPU.0.LLAMAFILE=1 CPU.1.SSE3=1 CPU.1.SSSE3=1 CPU.1.AVX=1 CPU.1.AVX2=1 CPU.1.F16C=1 CPU.1.FMA=1 CPU.1.BMI2=1 CPU.1.AVX512=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc)
time=2025-09-16T18:55:31.034+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}"
time=2025-09-16T18:55:31.090+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:alloc LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath:MainGPU:0 UseMmap:false}"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:487 msg="offloading 0 repeating layers to GPU"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:491 msg="offloading output layer to CPU"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:498 msg="offloaded 0/25 layers to GPU"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:315 msg="model weights" device=CPU size="12.8 GiB"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:326 msg="kv cache" device=CPU size="3.1 GiB"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:337 msg="compute graph" device=CPU size="109.2 MiB"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:342 msg="total memory" size="16.0 GiB"
time=2025-09-16T18:55:31.515+05:30 level=INFO source=sched.go:473 msg="loaded runners" count=1
time=2025-09-16T18:55:31.515+05:30 level=INFO source=server.go:1250 msg="waiting for llama runner to start responding"
time=2025-09-16T18:55:31.516+05:30 level=INFO source=server.go:1284 msg="waiting for server to become available" status="llmserver loading model"
time=2025-09-16T18:55:55.606+05:30 level=INFO source=server.go:1288 msg="llama runner started in 24.71 seconds"
[GIN] 2025/09/16 - 18:58:31 | 200 |          3m0s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T18:58:31.152+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock
time=2025-09-16T18:59:26.650+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=plot
[GIN] 2025/09/16 - 18:59:26 | 200 | 55.590871804s |    100.69.242.6 | POST     "/api/chat"
[GIN] 2025/09/16 - 19:08:37 | 200 |         3m39s |    100.69.242.6 | POST     "/api/chat"
[GIN] 2025/09/16 - 19:09:58 | 200 |         1m21s |    100.69.242.6 | POST     "/api/chat"
[GIN] 2025/09/16 - 19:11:20 | 200 |         1m21s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:11:20.775+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock
[GIN] 2025/09/16 - 19:11:44 | 200 | 23.652946986s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:28:06.467+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing
[GIN] 2025/09/16 - 19:28:06 | 200 |         1m59s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:28:06.629+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock
time=2025-09-16T19:28:20.969+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web.run
[GIN] 2025/09/16 - 19:28:20 | 200 | 14.498346592s |    100.69.242.6 | POST     "/api/chat"
[GIN] 2025/09/16 - 19:33:37 | 200 |          4m5s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:34:46.292+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing
[GIN] 2025/09/16 - 19:34:46 | 200 |          1m7s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:34:46.397+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock
time=2025-09-16T19:36:03.866+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=browser.run
[GIN] 2025/09/16 - 19:36:03 | 200 |         1m17s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:39:01.933+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing
[GIN] 2025/09/16 - 19:39:01 | 200 |          2m0s |    100.69.242.6 | POST     "/api/chat"
time=2025-09-16T19:39:02.033+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock
time=2025-09-16T19:39:14.616+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=websearch
[GIN] 2025/09/16 - 19:39:14 | 200 | 12.679274659s |    100.69.242.6 | POST     "/api/chat"
@EntityDeleter commented on GitHub (Sep 16, 2025): Here is the full Ollama Debug log if it is of any use: ~~~ OLLAMA_NEW_ESTIMATES=1 ollama serve ─╯ time=2025-09-16T18:55:10.101+05:30 level=INFO source=routes.go:1331 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:16384 OLLAMA_DEBUG:INFO OLLAMA_FLASH_ATTENTION:true OLLAMA_GPU_OVERHEAD:2147483648 OLLAMA_HOST:http://100.69.242.6:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:2562047h47m16.854775807s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/home/sudar/.ollama/models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NEW_ESTIMATES:true OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:*https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:true ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]" time=2025-09-16T18:55:10.102+05:30 level=INFO source=images.go:477 msg="total blobs: 19" time=2025-09-16T18:55:10.103+05:30 level=INFO source=images.go:484 msg="total unused blobs removed: 0" time=2025-09-16T18:55:10.103+05:30 level=INFO source=routes.go:1384 msg="Listening on 100.69.242.6:11434 (version 0.11.10)" time=2025-09-16T18:55:10.103+05:30 level=INFO source=gpu.go:217 msg="looking for compatible GPUs" time=2025-09-16T18:55:10.218+05:30 level=INFO source=types.go:131 msg="inference compute" id=GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 library=cuda variant=v12 compute=6.1 driver=13.0 name="NVIDIA GeForce GTX 1080" total="7.9 GiB" available="7.3 GiB" time=2025-09-16T18:55:10.218+05:30 level=INFO source=routes.go:1425 msg="entering low vram mode" "total vram"="5.9 GiB" threshold="20.0 GiB" time=2025-09-16T18:55:30.819+05:30 level=INFO source=server.go:166 msg="enabling new memory estimates" time=2025-09-16T18:55:30.890+05:30 level=INFO source=server.go:199 msg="model wants flash attention" time=2025-09-16T18:55:30.890+05:30 level=INFO source=server.go:216 msg="enabling flash attention" time=2025-09-16T18:55:30.890+05:30 level=WARN source=server.go:224 msg="kv cache type not supported by model" type="" time=2025-09-16T18:55:30.891+05:30 level=INFO source=server.go:398 msg="starting runner" cmd="/usr/bin/ollama runner --ollama-engine --model /home/sudar/.ollama/models/blobs/sha256-b112e727c6f18875636c56a779790a590d705aec9e1c0eb5a97d51fc2a778583 --port 44357" time=2025-09-16T18:55:30.892+05:30 level=INFO source=server.go:671 msg="loading model" "model layers"=25 requested=-1 time=2025-09-16T18:55:30.902+05:30 level=INFO source=runner.go:1251 msg="starting ollama engine" time=2025-09-16T18:55:30.902+05:30 level=INFO source=runner.go:1286 msg="Server listening on 127.0.0.1:44357" time=2025-09-16T18:55:30.964+05:30 level=INFO source=server.go:677 msg="system memory" total="31.0 GiB" free="25.3 GiB" free_swap="4.0 GiB" time=2025-09-16T18:55:30.964+05:30 level=INFO source=server.go:685 msg="gpu memory" id=GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 available="4.9 GiB" free="7.3 GiB" minimum="457.0 MiB" overhead="2.0 GiB" time=2025-09-16T18:55:30.965+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:25[ID:GPU-35959f8f-cc31-7014-57a0-0f7cb76ce560 Layers:25(0..24)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2025-09-16T18:55:31.016+05:30 level=INFO source=ggml.go:131 msg="" architecture=gptoss file_type=MXFP4 name="" description="" num_tensors=315 num_key_values=30 operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception operator() double registration of ggml_uncaught_exception load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-skylakex.so time=2025-09-16T18:55:31.027+05:30 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.AVX512=1 CPU.0.LLAMAFILE=1 CPU.1.SSE3=1 CPU.1.SSSE3=1 CPU.1.AVX=1 CPU.1.AVX2=1 CPU.1.F16C=1 CPU.1.FMA=1 CPU.1.BMI2=1 CPU.1.AVX512=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc) time=2025-09-16T18:55:31.034+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2025-09-16T18:55:31.090+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:alloc LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2025-09-16T18:55:31.515+05:30 level=INFO source=runner.go:1170 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:true KvSize:131072 KvCacheType: NumThreads:6 GPULayers:[] MultiUserCache:false ProjectorPath:MainGPU:0 UseMmap:false}" time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:487 msg="offloading 0 repeating layers to GPU" time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:491 msg="offloading output layer to CPU" time=2025-09-16T18:55:31.515+05:30 level=INFO source=ggml.go:498 msg="offloaded 0/25 layers to GPU" time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:315 msg="model weights" device=CPU size="12.8 GiB" time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:326 msg="kv cache" device=CPU size="3.1 GiB" time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:337 msg="compute graph" device=CPU size="109.2 MiB" time=2025-09-16T18:55:31.515+05:30 level=INFO source=backend.go:342 msg="total memory" size="16.0 GiB" time=2025-09-16T18:55:31.515+05:30 level=INFO source=sched.go:473 msg="loaded runners" count=1 time=2025-09-16T18:55:31.515+05:30 level=INFO source=server.go:1250 msg="waiting for llama runner to start responding" time=2025-09-16T18:55:31.516+05:30 level=INFO source=server.go:1284 msg="waiting for server to become available" status="llmserver loading model" time=2025-09-16T18:55:55.606+05:30 level=INFO source=server.go:1288 msg="llama runner started in 24.71 seconds" [GIN] 2025/09/16 - 18:58:31 | 200 | 3m0s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T18:58:31.152+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock time=2025-09-16T18:59:26.650+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=plot [GIN] 2025/09/16 - 18:59:26 | 200 | 55.590871804s | 100.69.242.6 | POST "/api/chat" [GIN] 2025/09/16 - 19:08:37 | 200 | 3m39s | 100.69.242.6 | POST "/api/chat" [GIN] 2025/09/16 - 19:09:58 | 200 | 1m21s | 100.69.242.6 | POST "/api/chat" [GIN] 2025/09/16 - 19:11:20 | 200 | 1m21s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:11:20.775+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock [GIN] 2025/09/16 - 19:11:44 | 200 | 23.652946986s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:28:06.467+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing [GIN] 2025/09/16 - 19:28:06 | 200 | 1m59s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:28:06.629+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock time=2025-09-16T19:28:20.969+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web.run [GIN] 2025/09/16 - 19:28:20 | 200 | 14.498346592s | 100.69.242.6 | POST "/api/chat" [GIN] 2025/09/16 - 19:33:37 | 200 | 4m5s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:34:46.292+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing [GIN] 2025/09/16 - 19:34:46 | 200 | 1m7s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:34:46.397+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock time=2025-09-16T19:36:03.866+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=browser.run [GIN] 2025/09/16 - 19:36:03 | 200 | 1m17s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:39:01.933+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=web-browsing [GIN] 2025/09/16 - 19:39:01 | 200 | 2m0s | 100.69.242.6 | POST "/api/chat" time=2025-09-16T19:39:02.033+05:30 level=WARN source=types.go:654 msg="invalid option provided" option=use_mlock time=2025-09-16T19:39:14.616+05:30 level=WARN source=harmonyparser.go:401 msg="harmony parser: no reverse mapping found for function name" harmonyFunctionName=websearch [GIN] 2025/09/16 - 19:39:14 | 200 | 12.679274659s | 100.69.242.6 | POST "/api/chat" ~~~
yindo commented 2026-02-22 18:31:18 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 16, 2025):

Image

Additional Image of errors while web searching using DuckDuckGo (first time today).

Also, here are my ollama parameters:

export OLLAMA_HOST=100.69.242.6:11434
export OLLAMA_CONTEXT_LENGTH=16384
export OLLAMA_KEEP_ALIVE=-1
export OLLAMA_FLASH_ATTENTION=true
export OLLAMA_SCHED_SPREAD=true
export OLLAMA_GPU_OVERHEAD=2147483648
export OLLAMA_NEW_ESTIMATES=1

@EntityDeleter commented on GitHub (Sep 16, 2025): <img width="1896" height="1009" alt="Image" src="https://github.com/user-attachments/assets/7e145b3b-0153-4a03-92a4-c7a328850733" /> Additional Image of errors while web searching using DuckDuckGo (first time today). Also, here are my ollama parameters: export OLLAMA_HOST=100.69.242.6:11434 export OLLAMA_CONTEXT_LENGTH=16384 export OLLAMA_KEEP_ALIVE=-1 export OLLAMA_FLASH_ATTENTION=true export OLLAMA_SCHED_SPREAD=true export OLLAMA_GPU_OVERHEAD=2147483648 export OLLAMA_NEW_ESTIMATES=1
yindo commented 2026-02-22 18:31:18 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 16, 2025):

Image

Update: New error when attempting to create a graph in a new workspace and thread.

@EntityDeleter commented on GitHub (Sep 16, 2025): <img width="1896" height="1009" alt="Image" src="https://github.com/user-attachments/assets/a9e3ea5f-c1f5-4ea3-bfd9-a12fd9011212" /> Update: New error when attempting to create a graph in a new workspace and thread.
yindo commented 2026-02-22 18:31:18 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Sep 16, 2025):

Does this occur with any model that is not harmony-based? Like Qwen3 or anthing else - I think this is a formatter/chat-template issue or item that needs to be compensated for or patched

@timothycarambat commented on GitHub (Sep 16, 2025): Does this occur with any model that is not harmony-based? Like Qwen3 or anthing else - I think this is a formatter/chat-template issue or item that needs to be compensated for or patched
yindo commented 2026-02-22 18:31:19 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 16, 2025):

Image Image Image

Attempted web search with Qwen 3 30B Thinking 2507. The collapsed agent box is the model thinking and failing to enumerate the web search tool despite recognition of its availability. Also unable to create new threads in the workspace due to a "failed to fetch" error. All context from the AI Humane Pin disappears after a reload of the app.

@EntityDeleter commented on GitHub (Sep 16, 2025): <img width="1256" height="953" alt="Image" src="https://github.com/user-attachments/assets/d0fc5317-2cea-48b8-b235-532f5dc9e9dc" /> <img width="1256" height="953" alt="Image" src="https://github.com/user-attachments/assets/d33b4fd8-6dd7-4ccd-aae3-fca3ca4ab3a5" /> <img width="1256" height="953" alt="Image" src="https://github.com/user-attachments/assets/c44baf21-d106-4cae-8920-fcc856b2721c" /> Attempted web search with Qwen 3 30B Thinking 2507. The collapsed agent box is the model thinking and failing to enumerate the web search tool despite recognition of its availability. Also unable to create new threads in the workspace due to a "failed to fetch" error. All context from the AI Humane Pin disappears after a reload of the app.
yindo commented 2026-02-22 18:31:19 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 16, 2025):

Okay, the user is asking me to create a graph of $ y = x^3 - 5 $ using the “create-graph” function. But wait, the system message says I shouldn’t make another chart. Let me check the history.

Earlier, the user asked for $ y = x^2 + 4 $, and the assistant confirmed the graph was created. Now they’re asking for a new graph, $ y = x^3 - 5 $. The system instruction says: “Do not make another chart.” So I can’t generate a new graph. I need to respond without creating one.

The user might not realize that the system already handled the previous request. They might think they need to ask again, but the assistant is supposed to follow the rule. So my response should inform them that another graph can’t be created, but maybe explain the function briefly.

Wait, the system message says “the chart was already generated and returned,” so the previous one was done. Now, for this new request, I should not generate a new chart. So I need to state that I can’t create another graph, but perhaps describe the function as a helpful alternative.

Let me make sure: the user said “using the create-graph function,” but the assistant is supposed to avoid that. So the correct response is to decline creating the graph and maybe explain the function’s characteristics.

So, the response should be: “I can’t generate another graph, but I can describe it. The function $ y = x^3 - 5 $ is a cubic curve with a point of inflection at (0, -5), passing through (0, -5) and increasing for all x. Let me know if you’d like more details!”

Wait, the system message says “Do not make another chart,” so I shouldn’t even mention generating it. Just state that I can’t do it and offer a description instead.

Yes, that’s the right approach. So the answer should be a polite refusal to create the chart, followed by a brief description of the function.

So, according to a system prompt given by Anything LLM, the AI is not to reuse the create-chart function. Thus, is refuses to use agent mode in this case. Is this an intended feature or a bug?

@EntityDeleter commented on GitHub (Sep 16, 2025): ~~~ Okay, the user is asking me to create a graph of $ y = x^3 - 5 $ using the “create-graph” function. But wait, the system message says I shouldn’t make another chart. Let me check the history. Earlier, the user asked for $ y = x^2 + 4 $, and the assistant confirmed the graph was created. Now they’re asking for a new graph, $ y = x^3 - 5 $. The system instruction says: “Do not make another chart.” So I can’t generate a new graph. I need to respond without creating one. The user might not realize that the system already handled the previous request. They might think they need to ask again, but the assistant is supposed to follow the rule. So my response should inform them that another graph can’t be created, but maybe explain the function briefly. Wait, the system message says “the chart was already generated and returned,” so the previous one was done. Now, for this new request, I should not generate a new chart. So I need to state that I can’t create another graph, but perhaps describe the function as a helpful alternative. Let me make sure: the user said “using the create-graph function,” but the assistant is supposed to avoid that. So the correct response is to decline creating the graph and maybe explain the function’s characteristics. So, the response should be: “I can’t generate another graph, but I can describe it. The function $ y = x^3 - 5 $ is a cubic curve with a point of inflection at (0, -5), passing through (0, -5) and increasing for all x. Let me know if you’d like more details!” Wait, the system message says “Do not make another chart,” so I shouldn’t even mention generating it. Just state that I can’t do it and offer a description instead. Yes, that’s the right approach. So the answer should be a polite refusal to create the chart, followed by a brief description of the function. ~~~ So, according to a system prompt given by Anything LLM, the AI is not to reuse the create-chart function. Thus, is refuses to use agent mode in this case. Is this an intended feature or a bug?
yindo commented 2026-02-22 18:31:19 -05:00
Author
Owner
Copy Link
Copy Source

@mattans commented on GitHub (Sep 18, 2025):

I using Ollama directly with GPT OSS and also getting this warning log

@mattans commented on GitHub (Sep 18, 2025): I using Ollama directly with GPT OSS and also getting this warning log
yindo commented 2026-02-22 18:31:19 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Sep 18, 2025):

Pretty sure this is a harmony template output issue we specifically need to handle. GPT OSS is harmony arch and im guessing its response body is different.

Possibly related: #4262

@timothycarambat commented on GitHub (Sep 18, 2025): Pretty sure this is a harmony template output issue we specifically need to handle. GPT OSS is harmony arch and im guessing its response body is different. Possibly related: #4262
yindo commented 2026-02-22 18:31:19 -05:00
Author
Owner
Copy Link
Copy Source

@EntityDeleter commented on GitHub (Sep 18, 2025):

But then, if Harmony is the problem, why do the errors also occur with Qwen3?

@EntityDeleter commented on GitHub (Sep 18, 2025): But then, if Harmony is the problem, why do the errors also occur with Qwen3?
yindo changed title from [BUG]: No Output Via Ollama to [GH-ISSUE #4386] [BUG]: No Output Via Ollama 2026-06-05 14:48:36 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label investigating needs info / can't replicate possible bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2797
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?