huangfu/DronePlanning
2
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

修改为东南天坐标系

Browse Source
This commit is contained in:
huangfu
2026-01-20 09:49:52 +08:00
parent 9538757047
commit 333fad40ac
7201 changed files with 1030888 additions and 85410 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
logs/embedding_model.log
View File

@@ -2,16 +2,16 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA GeForce RTX 4060 Ti, compute capability 8.9, VMM: yes
build: 6912 (961660b8c) with cc (Ubuntu 11.4.0-1ubuntu1~22.04.2) 11.4.0 for x86_64-linux-gnu
system info: n_threads = 6, n_threads_batch = 6, total_threads = 16
build: 6097 (9515c613) with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu
system info: n_threads = 8, n_threads_batch = 8, total_threads = 16
system_info: n_threads = 6 (n_threads_batch = 6) / 16 | CUDA : ARCHS = 890 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX_VNNI = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
system_info: n_threads = 8 (n_threads_batch = 8) / 16 | CUDA : ARCHS = 500,610,700,750,800,860,890 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX_VNNI = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
main: binding port with default address family
main: HTTP server is listening, hostname: 0.0.0.0, port: 8090, http threads: 15
main: loading model
srv load_model: loading model '/home/huangfukk/models/gguf/Qwen3/Qwen3-Embedding-4B/Qwen3-Embedding-4B-Q5_K_M.gguf'
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 4060 Ti) (0000:01:00.0) - 13681 MiB free
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 4060 Ti) - 2767 MiB free
llama_model_loader: loaded meta data with 36 key-value pairs and 398 tensors from /home/huangfukk/models/gguf/Qwen3/Qwen3-Embedding-4B/Qwen3-Embedding-4B-Q5_K_M.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen3
@@ -88,8 +88,6 @@ print_info: f_attn_scale = 0.0e+00
print_info: n_ff = 9728
print_info: n_expert = 0
print_info: n_expert_used = 0
print_info: n_expert_groups = 0
print_info: n_group_used = 0
print_info: causal attn = 1
print_info: pooling type = 3
print_info: rope type = 2
@@ -124,8 +122,8 @@ print_info: max token length = 256
load_tensors: loading model tensors, this can take a while... (mmap = true)
load_tensors: offloading 36 repeating layers to GPU
load_tensors: offloaded 36/37 layers to GPU
load_tensors: CPU_Mapped model buffer size = 303.75 MiB
load_tensors: CUDA0 model buffer size = 2445.68 MiB
load_tensors: CPU_Mapped model buffer size = 303.75 MiB
..........................................................................................
llama_context: constructing llama_context
llama_context: n_seq_max = 1
@@ -134,18 +132,17 @@ llama_context: n_ctx_per_seq = 4096
llama_context: n_batch = 2048
llama_context: n_ubatch = 512
llama_context: causal_attn = 1
llama_context: flash_attn = auto
llama_context: flash_attn = 0
llama_context: kv_unified = false
llama_context: freq_base = 1000000.0
llama_context: freq_scale = 1
llama_context: n_ctx_per_seq (4096) < n_ctx_train (40960) -- the full capacity of the model will not be utilized
llama_context: CPU output buffer size = 0.59 MiB
llama_kv_cache: CUDA0 KV buffer size = 576.00 MiB
llama_kv_cache: size = 576.00 MiB ( 4096 cells, 36 layers, 1/1 seqs), K (f16): 288.00 MiB, V (f16): 288.00 MiB
llama_context: Flash Attention was auto, set to enabled
llama_kv_cache_unified: CUDA0 KV buffer size = 576.00 MiB
llama_kv_cache_unified: size = 576.00 MiB ( 4096 cells, 36 layers, 1/1 seqs), K (f16): 288.00 MiB, V (f16): 288.00 MiB
llama_context: CUDA0 compute buffer size = 604.96 MiB
llama_context: CUDA_Host compute buffer size = 13.01 MiB
llama_context: graph nodes = 1268
llama_context: CUDA_Host compute buffer size = 17.01 MiB
llama_context: graph nodes = 1411
llama_context: graph splits = 4 (with bs=512), 3 (with bs=1)
common_init_from_params: added <|endoftext|> logit bias = -inf
common_init_from_params: added <|im_end|> logit bias = -inf
@@ -156,10 +153,6 @@ common_init_from_params: setting dry_penalty_last_n to ctx_size = 4096
common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
srv init: initializing slots, n_slots = 1
slot init: id 0 | task -1 | new slot n_ctx_slot = 4096
srv init: prompt cache is enabled, size limit: 8192 MiB
srv init: use `--cache-ram 0` to disable the prompt cache
srv init: for more info see https://github.com/ggml-org/llama.cpp/pull/16391
srv init: thinking = 0
main: model loaded
main: chat template, chat_template: {%- if tools %}
{{- '<|im_start|>system\n' }}
@@ -227,29 +220,7 @@ How are you?<|im_end|>
'
main: server is listening on http://0.0.0.0:8090 - starting the main loop
srv update_slots: all slots are idle
srv log_server_r: request: GET /health 127.0.0.1 200
slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = -1
slot launch_slot_: id 0 | task 0 | processing task
slot update_slots: id 0 | task 0 | new prompt, n_ctx_slot = 4096, n_keep = 0, task.n_tokens = 41
slot update_slots: id 0 | task 0 | n_tokens = 0, memory_seq_rm [0, end)
slot update_slots: id 0 | task 0 | prompt processing progress, n_tokens = 41, batch.n_tokens = 41, progress = 1.000000
slot update_slots: id 0 | task 0 | prompt done, n_tokens = 41, batch.n_tokens = 41
slot release: id 0 | task 0 | stop processing: n_tokens = 41, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/embeddings 127.0.0.1 200
slot get_availabl: id 0 | task -1 | selected slot by LCP similarity, sim_best = 1.000 (> 0.100 thold), f_keep = 1.000
slot launch_slot_: id 0 | task 2 | processing task
slot update_slots: id 0 | task 2 | new prompt, n_ctx_slot = 4096, n_keep = 0, task.n_tokens = 41
slot update_slots: id 0 | task 2 | need to evaluate at least 1 token for each active slot (n_past = 41, task.n_tokens() = 41)
slot update_slots: id 0 | task 2 | n_past was set to 40
slot update_slots: id 0 | task 2 | n_tokens = 40, memory_seq_rm [40, end)
slot update_slots: id 0 | task 2 | prompt processing progress, n_tokens = 41, batch.n_tokens = 1, progress = 1.000000
slot update_slots: id 0 | task 2 | prompt done, n_tokens = 41, batch.n_tokens = 1
slot release: id 0 | task 2 | stop processing: n_tokens = 41, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/embeddings 127.0.0.1 200
srv operator(): operator(): cleaning up before exit...
llama_memory_breakdown_print: | memory breakdown [MiB] | total free self model context compute unaccounted |
llama_memory_breakdown_print: | - CUDA0 (RTX 4060 Ti) | 15944 = 5959 + (3626 = 2445 + 576 + 604) + 6358 |
llama_memory_breakdown_print: | - Host | 316 = 303 + 0 + 13 |
Received second interrupt, terminating immediately.
srv log_server_r: request: GET /health 127.0.0.1 200
srv operator(): operator(): cleaning up before exit...

58
logs/fastapi.log
View File

@@ -1,55 +1,15 @@
2025-12-06 13:42:09,376 - INFO - Anonymized telemetry enabled. See https://docs.trychroma.com/telemetry for more information.
2025-12-06 13:42:09,460 - ERROR - 在系统提示词中未找到'可用节点定义'部分的JSON代码块。
2025-12-06 13:42:09,460 - WARNING - 使用备用方案解析节点定义...
2025-12-06 13:42:09,460 - INFO - 从第1个JSON块中成功解析出节点定义
2025-12-06 13:42:09,461 - INFO - 动作节点: ['battle_damage_assessment', 'deliver_payload', 'fly_to_waypoint', 'land', 'loiter', 'move_direction', 'object_detect', 'orbit_around_point', 'orbit_around_target', 'preflight_checks', 'search_pattern', 'strike_target', 'take_picture', 'takeoff', 'track_object']
2025-12-06 13:42:09,461 - INFO - 条件节点: ['at_waypoint', 'object_detected', 'target_destroyed', 'time_elapsed']
INFO: Started server process [3031182]
2026-01-20 09:41:43,111 - INFO - Anonymized telemetry enabled. See https://docs.trychroma.com/telemetry for more information.
2026-01-20 09:41:43,260 - INFO - 成功找到节点定义JSON代码块
2026-01-20 09:41:43,260 - INFO - 成功解析出动作节点: ['approach_target', 'deliver_payload', 'fly_sequence', 'fly_to_waypoint', 'land', 'loiter', 'manual_confirmation', 'move_direction', 'object_detect', 'return_emergency', 'rotate', 'rotate_search', 'search_pattern', 'system_checks', 'take_photos', 'takeoff', 'track_object']
2026-01-20 09:41:43,260 - INFO - 成功解析出条件节点: ['at_waypoint', 'object_detected']
INFO: Started server process [34239]
INFO: Waiting for application startup.
2025-12-06 13:42:09,463 - INFO - WebSocket event loop configured.
2026-01-20 09:41:43,263 - INFO - WebSocket event loop configured.
INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
2025-12-06 13:42:10,978 - INFO - 接收到用户请求: 无人机起飞到80米高度后，先移动至搜索区，搜索并锁定任一红色车辆，跟踪接近距离目标车辆10m后进行拍照，完成拍照后返回。
2025-12-06 13:42:11,392 - INFO - HTTP Request: POST http://localhost:8081/v1/chat/completions "HTTP/1.1 200 OK"
2025-12-06 13:42:11,395 - INFO - 分类结果: complex
2025-12-06 13:42:11,395 - INFO - --- 开始从向量数据库检索上下文 ---
2025-12-06 13:42:11,467 - INFO - --- 成功检索到上下文信息 ---
2025-12-06 13:42:11,467 - INFO - 📚 检索到的上下文内容:
在地图上有一个名为 '跷跷板' 的地点或区域，它的leisure是'playground'，其中心位置坐标大约在 (x:15, y:-8.5, z:1.2)。
在地图上有一个名为 'A地' 的地点或区域，它的building是'commercial'，其中心位置坐标大约在 (x:10, y:-10, z:2)。
在地图上有一个名为 '学生宿舍' 的地点或区域，它的building是'dormitory'，其中心位置坐标大约在 (x:5, y:3, z:2)。
2025-12-06 13:42:11,467 - INFO - --- 第 1/3 次尝试生成Pytree ---
2025-12-06 13:42:21,443 - INFO - HTTP Request: POST http://localhost:8081/v1/chat/completions "HTTP/1.1 200 OK"
2025-12-06 13:42:21,447 - INFO - ✅ JSON Schema验证成功
2025-12-06 13:42:21,447 - INFO - ✅ 成功生成并验证了Pytree
2025-12-06 13:42:21,507 - INFO - ✅ 任务树可视化成功
2025-12-06 13:42:21,507 - INFO - 图形已保存到: /home/huangfukk/DronePlanning/backend_service/generated_visualizations/py_tree.png
2025-12-06 13:42:21,507 - INFO - 未在模型输出中发现 <think> 推理链片段。若需捕获，请设置 ENABLE_REASONING_CAPTURE=true 以放宽JSON强制格式。
INFO: 127.0.0.1:38420 - "POST /generate_plan HTTP/1.1" 200 OK
INFO: 127.0.0.1:38430 - "GET /docs HTTP/1.1" 200 OK
2025-12-06 13:42:24,416 - INFO - 接收到用户请求: 无人机起飞到80米高度后，先移动至搜索区，搜索并锁定任一红色车辆，跟踪接近距离目标车辆10m后进行拍照，完成拍照后返回。
2025-12-06 13:42:24,809 - INFO - HTTP Request: POST http://localhost:8081/v1/chat/completions "HTTP/1.1 200 OK"
2025-12-06 13:42:24,809 - INFO - 分类结果: complex
2025-12-06 13:42:24,809 - INFO - --- 开始从向量数据库检索上下文 ---
2025-12-06 13:42:24,842 - INFO - --- 成功检索到上下文信息 ---
2025-12-06 13:42:24,842 - INFO - 📚 检索到的上下文内容:
在地图上有一个名为 '跷跷板' 的地点或区域，它的leisure是'playground'，其中心位置坐标大约在 (x:15, y:-8.5, z:1.2)。
在地图上有一个名为 'A地' 的地点或区域，它的building是'commercial'，其中心位置坐标大约在 (x:10, y:-10, z:2)。
在地图上有一个名为 '学生宿舍' 的地点或区域，它的building是'dormitory'，其中心位置坐标大约在 (x:5, y:3, z:2)。
2025-12-06 13:42:24,842 - INFO - --- 第 1/3 次尝试生成Pytree ---
2025-12-06 13:42:34,645 - INFO - HTTP Request: POST http://localhost:8081/v1/chat/completions "HTTP/1.1 200 OK"
2025-12-06 13:42:34,647 - INFO - ✅ JSON Schema验证成功
2025-12-06 13:42:34,647 - INFO - ✅ 成功生成并验证了Pytree
2025-12-06 13:42:34,695 - INFO - ✅ 任务树可视化成功
2025-12-06 13:42:34,695 - INFO - 图形已保存到: /home/huangfukk/DronePlanning/backend_service/generated_visualizations/py_tree.png
2025-12-06 13:42:34,695 - INFO - 未在模型输出中发现 <think> 推理链片段。若需捕获，请设置 ENABLE_REASONING_CAPTURE=true 以放宽JSON强制格式。
INFO: 127.0.0.1:51574 - "POST /generate_plan HTTP/1.1" 200 OK
INFO: 127.0.0.1:44013 - "GET /docs HTTP/1.1" 200 OK
INFO: Shutting down
INFO: Waiting for application shutdown.
2025-12-06 13:46:44,939 - INFO - Backend service shutting down.
2026-01-20 09:42:11,885 - INFO - Backend service shutting down.
INFO: Application shutdown complete.
INFO: Finished server process [3031182]
INFO: Finished server process [34239]

118
logs/inference_model.log
View File

@@ -2,16 +2,16 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA GeForce RTX 4060 Ti, compute capability 8.9, VMM: yes
build: 6912 (961660b8c) with cc (Ubuntu 11.4.0-1ubuntu1~22.04.2) 11.4.0 for x86_64-linux-gnu
system info: n_threads = 6, n_threads_batch = 6, total_threads = 16
build: 6097 (9515c613) with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu
system info: n_threads = 8, n_threads_batch = 8, total_threads = 16
system_info: n_threads = 6 (n_threads_batch = 6) / 16 | CUDA : ARCHS = 890 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX_VNNI = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
system_info: n_threads = 8 (n_threads_batch = 8) / 16 | CUDA : ARCHS = 500,610,700,750,800,860,890 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX_VNNI = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
main: binding port with default address family
main: HTTP server is listening, hostname: 0.0.0.0, port: 8081, http threads: 15
main: loading model
srv load_model: loading model '/home/huangfukk/models/gguf/Qwen3/Qwen3-4B/Qwen3-4B-Q5_K_M.gguf'
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 4060 Ti) (0000:01:00.0) - 13681 MiB free
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 4060 Ti) - 2789 MiB free
llama_model_loader: loaded meta data with 28 key-value pairs and 398 tensors from /home/huangfukk/models/gguf/Qwen3/Qwen3-4B/Qwen3-4B-Q5_K_M.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen3
@@ -80,8 +80,6 @@ print_info: f_attn_scale = 0.0e+00
print_info: n_ff = 9728
print_info: n_expert = 0
print_info: n_expert_used = 0
print_info: n_expert_groups = 0
print_info: n_group_used = 0
print_info: causal attn = 1
print_info: pooling type = -1
print_info: rope type = 2
@@ -116,8 +114,8 @@ print_info: max token length = 256
load_tensors: loading model tensors, this can take a while... (mmap = true)
load_tensors: offloading 36 repeating layers to GPU
load_tensors: offloaded 36/37 layers to GPU
load_tensors: CPU_Mapped model buffer size = 304.29 MiB
load_tensors: CUDA0 model buffer size = 2445.68 MiB
load_tensors: CPU_Mapped model buffer size = 304.29 MiB
..........................................................................................
llama_context: constructing llama_context
llama_context: n_seq_max = 1
@@ -126,18 +124,17 @@ llama_context: n_ctx_per_seq = 8192
llama_context: n_batch = 2048
llama_context: n_ubatch = 512
llama_context: causal_attn = 1
llama_context: flash_attn = auto
llama_context: flash_attn = 0
llama_context: kv_unified = false
llama_context: freq_base = 1000000.0
llama_context: freq_scale = 1
llama_context: n_ctx_per_seq (8192) < n_ctx_train (40960) -- the full capacity of the model will not be utilized
llama_context: CPU output buffer size = 0.58 MiB
llama_kv_cache: CUDA0 KV buffer size = 1152.00 MiB
llama_kv_cache: size = 1152.00 MiB ( 8192 cells, 36 layers, 1/1 seqs), K (f16): 576.00 MiB, V (f16): 576.00 MiB
llama_context: Flash Attention was auto, set to enabled
llama_kv_cache_unified: CUDA0 KV buffer size = 1152.00 MiB
llama_kv_cache_unified: size = 1152.00 MiB ( 8192 cells, 36 layers, 1/1 seqs), K (f16): 576.00 MiB, V (f16): 576.00 MiB
llama_context: CUDA0 compute buffer size = 606.03 MiB
llama_context: CUDA_Host compute buffer size = 21.01 MiB
llama_context: graph nodes = 1267
llama_context: CUDA_Host compute buffer size = 25.01 MiB
llama_context: graph nodes = 1410
llama_context: graph splits = 4 (with bs=512), 3 (with bs=1)
common_init_from_params: added <|endoftext|> logit bias = -inf
common_init_from_params: added <|im_end|> logit bias = -inf
@@ -148,10 +145,6 @@ common_init_from_params: setting dry_penalty_last_n to ctx_size = 8192
common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
srv init: initializing slots, n_slots = 1
slot init: id 0 | task -1 | new slot n_ctx_slot = 8192
srv init: prompt cache is enabled, size limit: 8192 MiB
srv init: use `--cache-ram 0` to disable the prompt cache
srv init: for more info see https://github.com/ggml-org/llama.cpp/pull/16391
srv init: thinking = 0
main: model loaded
main: chat template, chat_template: {%- if tools %}
{{- '<|im_start|>system\n' }}
@@ -249,91 +242,8 @@ How are you?<|im_end|>
'
main: server is listening on http://0.0.0.0:8081 - starting the main loop
srv update_slots: all slots are idle
srv log_server_r: request: GET /health 127.0.0.1 200
srv params_from_: Chat format: Content-only
slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = -1
slot launch_slot_: id 0 | task 0 | processing task
slot update_slots: id 0 | task 0 | new prompt, n_ctx_slot = 8192, n_keep = 0, task.n_tokens = 375
slot update_slots: id 0 | task 0 | n_tokens = 0, memory_seq_rm [0, end)
slot update_slots: id 0 | task 0 | prompt processing progress, n_tokens = 375, batch.n_tokens = 375, progress = 1.000000
slot update_slots: id 0 | task 0 | prompt done, n_tokens = 375, batch.n_tokens = 375
slot print_timing: id 0 | task 0 |
prompt eval time = 111.12 ms / 375 tokens ( 0.30 ms per token, 3374.64 tokens per second)
eval time = 207.35 ms / 10 tokens ( 20.73 ms per token, 48.23 tokens per second)
total time = 318.47 ms / 385 tokens
slot release: id 0 | task 0 | stop processing: n_tokens = 384, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
srv params_from_: Chat format: Content-only
slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 237472165130
srv get_availabl: updating prompt cache
srv prompt_save: - saving prompt with length 384, total state size = 54.005 MiB
srv load: - looking for better prompt, base f_keep = 0.008, sim = 0.001
srv update: - cache state: 1 prompts, 54.005 MiB (limits: 8192.000 MiB, 8192 tokens, 58248 est)
srv update: - prompt 0x63c2bba5e140: 384 tokens, checkpoints: 0, 54.005 MiB
srv get_availabl: prompt cache update took 22.38 ms
slot launch_slot_: id 0 | task 11 | processing task
slot update_slots: id 0 | task 11 | new prompt, n_ctx_slot = 8192, n_keep = 0, task.n_tokens = 2554
slot update_slots: id 0 | task 11 | n_tokens = 3, memory_seq_rm [3, end)
slot update_slots: id 0 | task 11 | prompt processing progress, n_tokens = 2051, batch.n_tokens = 2048, progress = 0.803054
slot update_slots: id 0 | task 11 | n_tokens = 2051, memory_seq_rm [2051, end)
slot update_slots: id 0 | task 11 | prompt processing progress, n_tokens = 2554, batch.n_tokens = 503, progress = 1.000000
slot update_slots: id 0 | task 11 | prompt done, n_tokens = 2554, batch.n_tokens = 503
slot print_timing: id 0 | task 11 |
prompt eval time = 608.76 ms / 2551 tokens ( 0.24 ms per token, 4190.45 tokens per second)
eval time = 9337.00 ms / 409 tokens ( 22.83 ms per token, 43.80 tokens per second)
total time = 9945.76 ms / 2960 tokens
slot release: id 0 | task 11 | stop processing: n_tokens = 2962, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
srv params_from_: Chat format: Content-only
slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 237482215959
srv get_availabl: updating prompt cache
srv prompt_save: - saving prompt with length 2962, total state size = 416.566 MiB
srv load: - looking for better prompt, base f_keep = 0.001, sim = 0.008
srv load: - found better prompt with f_keep = 0.977, sim = 1.000
srv update: - cache state: 1 prompts, 416.566 MiB (limits: 8192.000 MiB, 8192 tokens, 58249 est)
srv update: - prompt 0x63c2bba9e270: 2962 tokens, checkpoints: 0, 416.566 MiB
srv get_availabl: prompt cache update took 162.15 ms
slot launch_slot_: id 0 | task 422 | processing task
slot update_slots: id 0 | task 422 | new prompt, n_ctx_slot = 8192, n_keep = 0, task.n_tokens = 375
slot update_slots: id 0 | task 422 | need to evaluate at least 1 token for each active slot (n_past = 375, task.n_tokens() = 375)
slot update_slots: id 0 | task 422 | n_past was set to 374
slot update_slots: id 0 | task 422 | n_tokens = 374, memory_seq_rm [374, end)
slot update_slots: id 0 | task 422 | prompt processing progress, n_tokens = 375, batch.n_tokens = 1, progress = 1.000000
slot update_slots: id 0 | task 422 | prompt done, n_tokens = 375, batch.n_tokens = 1
slot print_timing: id 0 | task 422 |
prompt eval time = 30.72 ms / 1 tokens ( 30.72 ms per token, 32.55 tokens per second)
eval time = 196.03 ms / 10 tokens ( 19.60 ms per token, 51.01 tokens per second)
total time = 226.75 ms / 11 tokens
slot release: id 0 | task 422 | stop processing: n_tokens = 384, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
srv params_from_: Chat format: Content-only
slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 237485582368
srv get_availabl: updating prompt cache
srv prompt_save: - saving prompt with length 384, total state size = 54.005 MiB
srv load: - looking for better prompt, base f_keep = 0.008, sim = 0.001
srv load: - found better prompt with f_keep = 0.862, sim = 1.000
srv update: - cache state: 1 prompts, 54.005 MiB (limits: 8192.000 MiB, 8192 tokens, 58248 est)
srv update: - prompt 0x63c2bf15ea60: 384 tokens, checkpoints: 0, 54.005 MiB
srv get_availabl: prompt cache update took 76.62 ms
slot launch_slot_: id 0 | task 433 | processing task
slot update_slots: id 0 | task 433 | new prompt, n_ctx_slot = 8192, n_keep = 0, task.n_tokens = 2554
slot update_slots: id 0 | task 433 | need to evaluate at least 1 token for each active slot (n_past = 2554, task.n_tokens() = 2554)
slot update_slots: id 0 | task 433 | n_past was set to 2553
slot update_slots: id 0 | task 433 | n_tokens = 2553, memory_seq_rm [2553, end)
slot update_slots: id 0 | task 433 | prompt processing progress, n_tokens = 2554, batch.n_tokens = 1, progress = 1.000000
slot update_slots: id 0 | task 433 | prompt done, n_tokens = 2554, batch.n_tokens = 1
slot print_timing: id 0 | task 433 |
prompt eval time = 37.62 ms / 1 tokens ( 37.62 ms per token, 26.58 tokens per second)
eval time = 9680.67 ms / 409 tokens ( 23.67 ms per token, 42.25 tokens per second)
total time = 9718.29 ms / 410 tokens
slot release: id 0 | task 433 | stop processing: n_tokens = 2962, truncated = 0
srv update_slots: all slots are idle
srv log_server_r: request: POST /v1/chat/completions 127.0.0.1 200
srv operator(): operator(): cleaning up before exit...
llama_memory_breakdown_print: | memory breakdown [MiB] | total free self model context compute unaccounted |
llama_memory_breakdown_print: | - CUDA0 (RTX 4060 Ti) | 15944 = 5959 + (4203 = 2445 + 1152 + 606) + 5781 |
llama_memory_breakdown_print: | - Host | 325 = 304 + 0 + 21 |
Received second interrupt, terminating immediately.
ed second interrupt, terminating immediately.
srv operator(): operator(): cleaning up before exit...
srv log_server_r: request: GET /health 127.0.0.1 200
srv operator(): operator(): cleaning up before exit...