【AI】（五）打印输出 Ollama 请求信息

小渣渣

需求：使用 Ollama 部署了 DeepSeek-R1 模型，想查看一些插件发出的请求信息，方便了解其中细节。例如：Open WebUI、continue、cline、Roo Code 等。

回顾：

【AI】（三）腾讯云使用 HAI 部署 DeepSeek-R1 教程
https://www.itsvse.com/thread-10931-1-1.html

【AI】（四）使用 Open WebUI 调用 DeepSeek-R1 模型
https://www.itsvse.com/thread-10934-1-1.html

为了打印出服务器端的输入请求，需要开启 Debug 模式。编辑 /etc/systemd/system/ollama.service.d/override.conf 文件，添加如下配置：

登录可见。

重新加载和启动 ollama 服务，命令如下：

登录可见。

使用 journalctl 查看服务输出日志，命令如下：

登录可见。

使用 Open WebUI 调用 ollama 进行测试，如下图：



日志如下：

Feb 06 21:25:48 VM-0-8-ubuntu ollama[13503]: [GIN] 2025/02/06 - 21:25:48 | 200 |  6.186257471s |      172.18.0.2 | POST     "/api/chat"
Feb 06 21:25:48 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:48.411+08:00 level=DEBUG source=sched.go:407 msg="context for request finished"
Feb 06 21:25:48 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:48.411+08:00 level=DEBUG source=sched.go:339 msg="runner with non-zero duration has gone idle, adding timer" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 duration=2562047h47m16.854775807s
Feb 06 21:25:48 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:48.411+08:00 level=DEBUG source=sched.go:357 msg="after processing request finished event" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 refCount=0
Feb 06 21:25:54 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:54.834+08:00 level=DEBUG source=sched.go:575 msg="evaluating already loaded" model=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
Feb 06 21:25:54 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:54.835+08:00 level=DEBUG source=routes.go:1470 msg="chat request" images=0 prompt=<｜User｜>我的名字叫小渣渣，你是谁？<｜Assistant｜>
Feb 06 21:25:54 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:25:54.836+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=0 cache=728 prompt=13 used=2 remaining=11
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: [GIN] 2025/02/06 - 21:26:02 | 200 |  7.642182053s |      172.18.0.2 | POST     "/api/chat"
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.454+08:00 level=DEBUG source=sched.go:407 msg="context for request finished"
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.454+08:00 level=DEBUG source=sched.go:339 msg="runner with non-zero duration has gone idle, adding timer" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 duration=2562047h47m16.854775807s
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.454+08:00 level=DEBUG source=sched.go:357 msg="after processing request finished event" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 refCount=0
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.491+08:00 level=DEBUG source=sched.go:575 msg="evaluating already loaded" model=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.491+08:00 level=DEBUG source=routes.go:1470 msg="chat request" images=0 prompt="<｜User｜>### Task:\nGenerate a concise, 3-5 word title with an emoji summarizing the chat history.\n### Guidelines:\n- The title should clearly represent the main theme or subject of the conversation.\n- Use emojis that enhance understanding of the topic, but avoid quotation marks or special formatting.\n- Write the title in the chat's primary language; default to English if multilingual.\n- Prioritize accuracy over excessive creativity; keep it clear and simple.\n### Output:\nJSON format: { \"title\": \"your concise title here\" }\n### Examples:\n- { \"title\": \" Stock Market Trends\" },\n- { \"title\": \" Perfect Chocolate Chip Recipe\" },\n- { \"title\": \"Evolution of Music Streaming\" },\n- { \"title\": \"Remote Work Productivity Tips\" },\n- { \"title\": \"Artificial Intelligence in Healthcare\" },\n- { \"title\": \" Video Game Development Insights\" }\n### Chat History:\n<chat_history>\nUSER: 我的名字叫小渣渣，你是谁？\nASSISTANT: 你好，小渣渣！我是DeepSeek-R1-Lite-Preview，一个由深度求索公司开发的智能助手，我会尽我所能为您提供帮助。有什么我可以为您做的吗？\n</chat_history><｜Assistant｜>"
Feb 06 21:26:02 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:02.495+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=1 cache=567 prompt=312 used=6 remaining=306
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: [GIN] 2025/02/06 - 21:26:14 | 200 | 12.263297485s |      172.18.0.2 | POST     "/api/chat"
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.731+08:00 level=DEBUG source=sched.go:407 msg="context for request finished"
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.731+08:00 level=DEBUG source=sched.go:339 msg="runner with non-zero duration has gone idle, adding timer" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 duration=2562047h47m16.854775807s
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.731+08:00 level=DEBUG source=sched.go:357 msg="after processing request finished event" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 refCount=0
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.769+08:00 level=DEBUG source=sched.go:575 msg="evaluating already loaded" model=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.769+08:00 level=DEBUG source=routes.go:1470 msg="chat request" images=0 prompt="<｜User｜>### Task:\nGenerate 1-3 broad tags categorizing the main themes of the chat history, along with 1-3 more specific subtopic tags.\n\n### Guidelines:\n- Start with high-level domains (e.g. Science, Technology, Philosophy, Arts, Politics, Business, Health, Sports, Entertainment, Education)\n- Consider including relevant subfields/subdomains if they are strongly represented throughout the conversation\n- If content is too short (less than 3 messages) or too diverse, use only [\"General\"]\n- Use the chat's primary language; default to English if multilingual\n- Prioritize accuracy over specificity\n\n### Output:\nJSON format: { \"tags\": [\"tag1\", \"tag2\", \"tag3\"] }\n\n### Chat History:\n<chat_history>\nUSER: 我的名字叫小渣渣，你是谁？\nASSISTANT: 你好，小渣渣！我是DeepSeek-R1-Lite-Preview，一个由深度求索公司开发的智能助手，我会尽我所能为您提供帮助。有什么我可以为您做的吗？\n</chat_history><｜Assistant｜>"
Feb 06 21:26:14 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:14.773+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=1 cache=637 prompt=249 used=7 remaining=242
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:17.717+08:00 level=DEBUG source=sched.go:575 msg="evaluating already loaded" model=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:17.718+08:00 level=DEBUG source=server.go:966 msg="new runner detected, loading model for cgo tokenization"
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from /data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   0:                       general.architecture str              = qwen2
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   1:                               general.type str              = model
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   2:                               general.name str              = DeepSeek R1 Distill Qwen 32B
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   3:                           general.basename str              = DeepSeek-R1-Distill-Qwen
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   4:                         general.size_label str              = 32B
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   5:                          qwen2.block_count u32              = 64
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   6:                       qwen2.context_length u32              = 131072
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   7:                     qwen2.embedding_length u32              = 5120
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   8:                  qwen2.feed_forward_length u32              = 27648
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv   9:                 qwen2.attention.head_count u32              = 40
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  10:              qwen2.attention.head_count_kv u32              = 8
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  11:                       qwen2.rope.freq_base f32              = 1000000.000000
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  12:     qwen2.attention.layer_norm_rms_epsilon f32              = 0.000010
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  13:                          general.file_type u32              = 15
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  14:                       tokenizer.ggml.model str              = gpt2
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  15:                         tokenizer.ggml.pre str              = deepseek-r1-qwen
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  16:                      tokenizer.ggml.tokens arr[str,152064]  = ["!", "\"", "#", "$", "%", "&", "'", ...
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  17:                  tokenizer.ggml.token_type arr[i32,152064]  = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  18:                      tokenizer.ggml.merges arr[str,151387]  = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  19:                tokenizer.ggml.bos_token_id u32              = 151646
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  20:                tokenizer.ggml.eos_token_id u32              = 151643
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  21:            tokenizer.ggml.padding_token_id u32              = 151643
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  22:               tokenizer.ggml.add_bos_token bool             = true
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  23:               tokenizer.ggml.add_eos_token bool             = false
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  24:                    tokenizer.chat_template str              = {% if not add_generation_prompt is de...
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - kv  25:               general.quantization_version u32              = 2
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - type  f32:  321 tensors
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - type q4_K:  385 tensors
Feb 06 21:26:17 VM-0-8-ubuntu ollama[13503]: llama_model_loader: - type q6_K:   65 tensors
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_vocab: special tokens cache size = 22
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_vocab: token to piece cache size = 0.9310 MB
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: format           = GGUF V3 (latest)
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: arch             = qwen2
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: vocab type       = BPE
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: n_vocab          = 152064
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: n_merges         = 151387
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: vocab_only       = 1
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: model type       = ?B
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: model ftype      = all F32
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: model params     = 32.76 B
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: model size       = 18.48 GiB (4.85 BPW)
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: general.name     = DeepSeek R1 Distill Qwen 32B
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: BOS token        = 151646 '<｜begin▁of▁sentence｜>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOS token        = 151643 '<｜end▁of▁sentence｜>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOT token        = 151643 '<｜end▁of▁sentence｜>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: PAD token        = 151643 '<｜end▁of▁sentence｜>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: LF token         = 148848 'ÄĬ'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM PRE token    = 151659 '<|fim_prefix|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM SUF token    = 151661 '<|fim_suffix|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM MID token    = 151660 '<|fim_middle|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM PAD token    = 151662 '<|fim_pad|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM REP token    = 151663 '<|repo_name|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: FIM SEP token    = 151664 '<|file_sep|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOG token        = 151643 '<｜end▁of▁sentence｜>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOG token        = 151662 '<|fim_pad|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOG token        = 151663 '<|repo_name|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: EOG token        = 151664 '<|file_sep|>'
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llm_load_print_meta: max token length = 256
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: llama_model_load: vocab only - skipping tensors
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:18.440+08:00 level=DEBUG source=routes.go:1470 msg="chat request" images=0 prompt="<｜User｜>我的名字叫小渣渣，你是谁？<｜Assistant｜>\n你好，小渣渣！我是DeepSeek-R1-Lite-Preview，一个由深度求索公司开发的智能助手，我会尽我所能为您提供帮助。有什么我可以为您做的吗？<｜end▁of▁sentence｜><｜User｜>你好 DeepSeek-R1<｜Assistant｜>"
Feb 06 21:26:18 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:18.491+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=0 cache=223 prompt=64 used=13 remaining=51
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: [GIN] 2025/02/06 - 21:26:24 | 200 |  6.737131375s |      172.18.0.2 | POST     "/api/chat"
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:24.426+08:00 level=DEBUG source=sched.go:407 msg="context for request finished"
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:24.426+08:00 level=DEBUG source=sched.go:357 msg="after processing request finished event" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 refCount=1
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: [GIN] 2025/02/06 - 21:26:24 | 200 | 10.172441322s |      172.18.0.2 | POST     "/api/chat"
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:24.918+08:00 level=DEBUG source=sched.go:407 msg="context for request finished"
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:24.918+08:00 level=DEBUG source=sched.go:339 msg="runner with non-zero duration has gone idle, adding timer" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 duration=2562047h47m16.854775807s
Feb 06 21:26:24 VM-0-8-ubuntu ollama[13503]: time=2025-02-06T21:26:24.918+08:00 level=DEBUG source=sched.go:357 msg="after processing request finished event" modelPath=/data/ollama/models/blobs/sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 refCount=0

参考：
https://stackoverflow.com/questions/78609187/how-to-print-input-requests-and-output-responses-in-ollama-server
https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md

小渣渣

使用如下命令也可以编辑 ollama 服务文件：

登录可见。

当你运行这个命令时，系统会打开一个文本编辑器（通常是 vi 或 nano），让你编辑 /etc/systemd/system/ollama.service 文件。

小渣渣

Linux 查看 Systemd 服务的日志输出
https://www.itsvse.com/thread-10154-1-1.html