Need help with running with Lemonade server, I keep getting no results #942

gerroon · 2025-12-08T04:20:51Z

gerroon
Dec 8, 2025

Hi

I am using Lemonade on Win11, it works with GAIA UI and OpenWebUi fine, all local. I added it to Perplexica the way I use in OpenwebUI but no good result.

When the query runs, I see Lemonade starts acitivity. Log below

=================
[Server] POST /api/v0/chat/completions - 200
[Server PRE-ROUTE] POST /api/v0/chat/completions
[Server DEBUG] No tools in request
[Server] Model already loaded: gpt-oss-20b-mxfp4-GGUF
[Server] POST /api/v1/chat/completions - srv  params_from_: Chat format: GPT-OSS
slot get_availabl: id  1 | task -1 | selected slot by LCP similarity, sim_best = 0.113 (> 0.100 thold), f_keep = 0.155
srv  get_availabl: updating prompt cache
srv   prompt_save:  - saving prompt with length 438, total state size = 13.624 MiB
srv          load:  - looking for better prompt, base f_keep = 0.155, sim = 0.113
srv          load:  - found better prompt with f_keep = 0.857, sim = 1.000
srv        update:  - cache state: 24 prompts, 1511.476 MiB (limits: 8192.000 MiB, 4096 tokens, 169988 est)
srv        update:    - prompt 000001AC303FD780:    2051 tokens, checkpoints:  1,    82.846 MiB
srv        update:    - prompt 000001B1D6C2A9C0:    1507 tokens, checkpoints:  1,    83.362 MiB
srv        update:    - prompt 000001B1D6CC49E0:    1957 tokens, checkpoints:  3,   106.342 MiB
srv        update:    - prompt 000001B1D6CC4890:    2467 tokens, checkpoints:  2,   119.591 MiB
srv        update:    - prompt 000001B1D6CC47B0:     972 tokens, checkpoints:  1,    59.303 MiB
srv        update:    - prompt 000001AC309B8320:     814 tokens, checkpoints:  3,    62.634 MiB
srv        update:    - prompt 000001AC309B80F0:    1369 tokens, checkpoints:  1,    40.474 MiB
srv        update:    - prompt 000001AC309B86A0:    3195 tokens, checkpoints:  3,   135.372 MiB
srv        update:    - prompt 000001AC309B8630:    1131 tokens, checkpoints:  3,    74.030 MiB
srv        update:    - prompt 000001AC309B8080:     505 tokens, checkpoints:  2,    32.009 MiB
srv        update:    - prompt 000001AC309B8550:     355 tokens, checkpoints:  2,    25.959 MiB
srv        update:    - prompt 000001AC309B8710:    1071 tokens, checkpoints:  3,    61.719 MiB
srv        update:    - prompt 000001AC309B8780:    3438 tokens, checkpoints:  2,   134.035 MiB
srv        update:    - prompt 000001AC309B8160:    1473 tokens, checkpoints:  2,    82.002 MiB
srv        update:    - prompt 000001AC309B84E0:     942 tokens, checkpoints:  1,    30.461 MiB
srv        update:    - prompt 000001AC309B8240:    1345 tokens, checkpoints:  1,    39.911 MiB
srv        update:    - prompt 000001B1D6CC4A50:    1430 tokens, checkpoints:  1,    56.982 MiB
srv        update:    - prompt 000001AC309B8390:    1148 tokens, checkpoints:  1,    53.464 MiB
srv        update:    - prompt 000001AC309B82B0:     851 tokens, checkpoints:  2,    35.714 MiB
srv        update:    - prompt 000001B1D6CC4900:    1343 tokens, checkpoints:  1,    52.527 MiB
srv        update:    - prompt 000001AC308E50F0:     533 tokens, checkpoints:  3,    47.017 MiB
srv        update:    - prompt 000001AC308E5240:     451 tokens, checkpoints:  3,    46.477 MiB
srv        update:    - prompt 000001AC308E51D0:     578 tokens, checkpoints:  1,    22.840 MiB
srv        update:    - prompt 000001AC308E5160:     438 tokens, checkpoints:  2,    26.405 MiB
srv  get_availabl: prompt cache update took 697.83 ms
slot launch_slot_: id  1 | task -1 | sampler chain: logits -> logit-bias -> penalties -> dry -> top-n-sigma -> top-k -> typical -> top-p -> min-p -> xtc -> temp-ext -> dist
slot launch_slot_: id  1 | task 38486 | processing task
slot update_slots: id  1 | task 38486 | new prompt, n_ctx_slot = 4096, n_keep = 16, task.n_tokens = 600
slot update_slots: id  1 | task 38486 | n_past = 600, slot.prompt.tokens.size() = 700, seq_id = 1, pos_min = 573, n_swa = 128
state_read_meta: failed to find available cells in kv cache
state_seq_set_data: error loading state: failed to restore kv cache
slot update_slots: id  1 | task 38486 | failed to restore context checkpoint (pos_min = 0, pos_max = 534, size = 12.545 MiB)
slot update_slots: id  1 | task 38486 | forcing full prompt re-processing due to lack of cache data (likely due to SWA or hybrid/recurrent memory, see https://github.com/ggml-org/llama.cpp/pull/13194#issuecomment-2868343055)
slot update_slots: id  1 | task 38486 | n_tokens = 0, memory_seq_rm [0, end)
slot update_slots: id  1 | task 38486 | prompt processing progress, n_tokens = 536, batch.n_tokens = 536, progress = 0.893333
slot update_slots: id  1 | task 38486 | n_tokens = 536, memory_seq_rm [536, end)
slot update_slots: id  1 | task 38486 | prompt processing progress, n_tokens = 600, batch.n_tokens = 64, progress = 1.000000
slot update_slots: id  1 | task 38486 | prompt done, n_tokens = 600, batch.n_tokens = 64
slot print_timing: id  1 | task 38486 |
prompt eval time =     232.69 ms /   600 tokens (    0.39 ms per token,  2578.52 tokens per second)
       eval time =     301.42 ms /   100 tokens (    3.01 ms per token,   331.77 tokens per second)
      total time =     534.11 ms /   700 tokens
slot      release: id  1 | task 38486 | stop processing: n_tokens = 699, truncated = 0
srv  update_slots: all slots are idle
srv  log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
200 OK
[Server DEBUG] Response message does NOT contain tool_calls
[Server DEBUG] Message content: <question>
Vision Pro prices
</question>

=== Telemetry ===
Input tokens:  600
Output tokens: 100
TTFT (s):      0.23
TPS:           331.77
=================
[Server] POST /api/v0/chat/completions - 200
[Server PRE-ROUTE] POST /api/v0/chat/completions
[Server DEBUG] No tools in request
[Server] Model already loaded: gpt-oss-20b-mxfp4-GGUF
[Server] POST /api/v1/chat/completions - Streaming
srv  params_from_: Chat format: GPT-OSS
slot get_availabl: id  3 | task -1 | selected slot by LCP similarity, sim_best = 0.980 (> 0.100 thold), f_keep = 0.761
slot launch_slot_: id  3 | task -1 | sampler chain: logits -> logit-bias -> penalties -> dry -> top-n-sigma -> top-k -> typical -> top-p -> min-p -> xtc -> temp-ext -> dist
slot launch_slot_: id  3 | task 38588 | processing task
slot update_slots: id  3 | task 38588 | new prompt, n_ctx_slot = 4096, n_keep = 16, task.n_tokens = 933
slot update_slots: id  3 | task 38588 | n_past = 914, slot.prompt.tokens.size() = 1201, seq_id = 3, pos_min = 1074, n_swa = 128
state_read_meta: failed to find available cells in kv cache
state_seq_set_data: error loading state: failed to restore kv cache
slot update_slots: id  3 | task 38588 | failed to restore context checkpoint (pos_min = 98, pos_max = 867, size = 18.056 MiB)
slot update_slots: id  3 | task 38588 | forcing full prompt re-processing due to lack of cache data (likely due to SWA or hybrid/recurrent memory, see https://github.com/ggml-org/llama.cpp/pull/13194#issuecomment-2868343055)
slot update_slots: id  3 | task 38588 | n_tokens = 0, memory_seq_rm [0, end)
slot update_slots: id  3 | task 38588 | prompt processing progress, n_tokens = 869, batch.n_tokens = 869, progress = 0.931404
slot update_slots: id  3 | task 38588 | n_tokens = 869, memory_seq_rm [869, end)
slot update_slots: id  3 | task 38588 | prompt processing progress, n_tokens = 933, batch.n_tokens = 64, progress = 1.000000
slot update_slots: id  3 | task 38588 | prompt done, n_tokens = 933, batch.n_tokens = 64
slot print_timing: id  3 | task 38588 |
prompt eval time =     146.58 ms /   933 tokens (    0.16 ms per token,  6364.95 tokens per second)
       eval time =     766.39 ms /   244 tokens (    3.14 ms per token,   318.38 tokens per second)
      total time =     912.97 ms /  1177 tokens
slot      release: id  3 | task 38588 | stop processing: n_tokens = 1176, truncated = 0
srv  update_slots: all slots are idle
[Server] Streaming completed - 200 OK
srv  log_server_r: request: POST /v1/chat/completions 127.0.0.1 200

=== Telemetry ===
Input tokens:  933
Output tokens: 244
TTFT (s):      0.147
TPS:           318.38

ItzCrazyKns · 2025-12-08T04:34:09Z

ItzCrazyKns
Dec 8, 2025
Maintainer

Hey! Can you please try rephrasing your question to something like "What are the prices of vision pro?" and see if that works? We're working on a new release which should fix these issues.

7 replies

gerroon Dec 8, 2025
Author

Qwen3-4B-Instruct-2507-GGUF
"What are the prices of vision pro?"
"Answer

As of now, there is no publicly available information in the provided context regarding the prices of the Apple Vision Pro. Therefore, I cannot provide specific pricing details based on the given data .

For up-to-date and accurate pricing information on the Apple Vision Pro, it is recommended to consult official Apple website or authorized retailers.

If you have additional context or a source where pricing information might be found, please share it so I can assist further."

ItzCrazyKns Dec 8, 2025
Maintainer

Can you see any sources coming up in the UI? If not, please share the logs from Perplexica container.

gerroon Dec 8, 2025
Author

I guess I should mention that my docker is behind proxy, this proxy used to work last year when I was working with ollama, I am not sure it isnot passing this proxy to searx or perplexica?

Lemonade runs on a machine that has direct access.

docker-compose.yaml
" environment:
https_proxy: http://192.168.2.3:8118
http_proxy: http://192.168.2.3:8118
"

gerroon Dec 8, 2025
Author

I tried it a custom searxng that I use daily, that doesnt work either
(Qwen3-4B-Instruct-2507-GGUF)

"
What are the prices of vision pro?
Answer

As of now, there is no publicly available information in the provided context regarding the price of the Apple Vision Pro. Therefore, I cannot provide a specific price or pricing details based on the given data. If you are looking for up-to-date pricing, it may be necessary to consult official Apple product pages or recent news reports on Apple’s Vision Pro release .

Please note: The context does not contain any details about the Vision Pro’s cost, features, or availability.

Hmm, sorry I could not find any relevant information on this topic. Would you like me to search again or ask something else?
"

   environment:
      SEARXNG_API_URL: http://192.168.2.4:8080
      SEARXNG_UWSGI_WORKERS: 6
      SEARXNG_UWSGI_THREADS: 6
      https_proxy: http://192.168.2.3:8118
      http_proxy: http://192.168.2.3:8118

ItzCrazyKns Dec 11, 2025
Maintainer

Could you share the logs from the backend (if any)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Need help with running with Lemonade server, I keep getting no results #942

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Replies: 1 comment 7 replies

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Need help with running with Lemonade server, I keep getting no results #942

Uh oh!

Uh oh!

gerroon Dec 8, 2025

Replies: 1 comment · 7 replies

Uh oh!

ItzCrazyKns Dec 8, 2025 Maintainer

Uh oh!

gerroon Dec 8, 2025 Author

Uh oh!

ItzCrazyKns Dec 8, 2025 Maintainer

Uh oh!

Uh oh!

gerroon Dec 8, 2025 Author

Uh oh!

gerroon Dec 8, 2025 Author

Uh oh!

ItzCrazyKns Dec 11, 2025 Maintainer

gerroon
Dec 8, 2025

Replies: 1 comment 7 replies

ItzCrazyKns
Dec 8, 2025
Maintainer

gerroon Dec 8, 2025
Author

ItzCrazyKns Dec 8, 2025
Maintainer

gerroon Dec 8, 2025
Author

gerroon Dec 8, 2025
Author

ItzCrazyKns Dec 11, 2025
Maintainer