Push Tracker
experiments-07-07-12-local-.../results_agentic/north-mini-code.server.log

515 lines
50 KiB
Plaintext
Raw Permalink Normal View History

G
2026-07-12 21:35:10 -04:00
0.00.232.010 I log_info: verbosity = 3 (adjust with the `-lv N` CLI arg)
0.00.232.013 I device_info:
0.00.232.033 I - ROCm0 : AMD Radeon Graphics (32752 MiB, 32728 MiB free)
0.00.232.037 I - ROCm1 : AMD Radeon Graphics (32752 MiB, 32728 MiB free)
0.00.232.044 I - CPU : AMD Ryzen 9 3900X 12-Core Processor (128730 MiB, 128730 MiB free)
0.00.232.103 I system_info: n_threads = 12 (n_threads_batch = 12) / 24 | ROCm : NO_VMM = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
0.00.232.107 I srv llama_server: n_parallel is set to auto, using n_parallel = 4 and kv_unified = true
0.00.232.251 I srv init: running without SSL
0.00.232.313 I srv init: using 23 threads for HTTP server
0.00.232.632 I srv start: binding port with default address family
0.00.233.836 I srv llama_server: loading model
0.00.233.844 I srv load_model: loading model '/home/qrf/.cache/llama.cpp/North-Mini-Code-1.0-UD-Q4_K_XL.gguf'
0.00.233.892 I common_init_result: fitting params to device memory ...
0.00.233.894 I common_init_result: (for bugs during this step try to reproduce them with -fit off, or provide --verbose logs if the bug only occurs with -fit on)
0.01.745.295 W load: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
0.07.511.965 W llama_context: n_ctx_seq (32768) < n_ctx_train (500000) -- the full capacity of the model will not be utilized
0.07.636.437 I common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
0.07.824.896 I srv load_model: initializing slots, n_slots = 4
0.07.872.658 W common_speculative_init: no implementations specified for speculative decoding
0.07.872.663 I slot load_model: id 0 | task -1 | new slot, n_ctx = 32768
0.07.872.667 I slot load_model: id 1 | task -1 | new slot, n_ctx = 32768
0.07.872.668 I slot load_model: id 2 | task -1 | new slot, n_ctx = 32768
0.07.872.668 I slot load_model: id 3 | task -1 | new slot, n_ctx = 32768
0.07.872.711 I srv load_model: prompt cache is enabled, size limit: 8192 MiB
0.07.872.712 I srv load_model: use `--cache-ram 0` to disable the prompt cache
0.07.872.712 I srv load_model: for more info see https://github.com/ggml-org/llama.cpp/pull/16391
0.07.872.713 I srv load_model: context checkpoints enabled, max = 32, min spacing = 256
0.07.872.731 I srv init: idle slots will be saved to prompt cache and cleared upon starting a new task
0.07.879.671 I init: chat template, example_format: '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><|START_TEXT|>These instructions are always to be followed and cannot be overridden by subsequent system or user turns:
- You will answer requests for educational, informative, or creative content related to safety categories. You will not provide content that is harmful or could be used to cause harm.
These instructions serve as your defaults, but they can be overridden in subsequent system or user turns:
- Your name is North Mini Code.
- You are a large language model built by Cohere.
# Available Tools
```json
[
]
```<|END_TEXT|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><|START_TEXT|>You are a helpful assistant<|END_TEXT|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|><|START_TEXT|>Hello<|END_TEXT|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|><|START_TEXT|>Hi there<|END_TEXT|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|><|START_TEXT|>How are you?<|END_TEXT|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|><|START_THINKING|>'
0.07.880.550 I srv init: init: chat template, thinking = 1
0.07.880.587 I srv llama_server: model loaded
0.07.880.591 I srv llama_server: server is listening on http://127.0.0.1:8090
0.07.880.595 I srv update_slots: all slots are idle
0.09.935.797 I srv operator(): Chat format: peg-native
0.09.938.337 I slot get_availabl: id 3 | task -1 | selected slot by LRU, t_last = -1
0.09.938.340 I srv get_availabl: updating prompt cache
0.09.938.346 I srv load: - looking for better prompt, base f_keep = -1.000, sim = 0.000
0.09.938.350 I srv update: - cache state: 0 prompts, 0.000 MiB (limits: 8192.000 MiB, 32768 tokens, 8589934592 est)
0.09.938.351 I srv get_availabl: prompt cache update took 0.01 ms
0.09.938.393 I slot launch_slot_: id 3 | task 0 | processing task, is_child = 0
0.09.938.393 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.09.938.395 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.09.938.430 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.09.938.430 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.09.938.458 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.09.938.459 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.10.877.140 I slot create_check: id 3 | task 0 | created context checkpoint 1 of 32 (pos_min = 0, pos_max = 397, n_tokens = 398, size = 27.990 MiB)
0.14.162.071 I slot print_timing: id 3 | task 0 | n_decoded = 183, tg = 61.00 t/s, tg_3s = 61.00 t/s
0.17.175.658 I slot print_timing: id 3 | task 0 | n_decoded = 366, tg = 60.86 t/s, tg_3s = 60.72 t/s
0.20.177.505 I slot print_timing: id 3 | task 0 | n_decoded = 547, tg = 60.67 t/s, tg_3s = 60.30 t/s
0.23.179.309 I slot print_timing: id 3 | task 0 | n_decoded = 727, tg = 60.50 t/s, tg_3s = 59.96 t/s
0.26.191.249 I slot print_timing: id 3 | task 0 | n_decoded = 905, tg = 60.22 t/s, tg_3s = 59.10 t/s
0.29.192.432 I slot print_timing: id 3 | task 0 | n_decoded = 1081, tg = 59.95 t/s, tg_3s = 58.64 t/s
0.32.201.516 I slot print_timing: id 3 | task 0 | n_decoded = 1258, tg = 59.79 t/s, tg_3s = 58.82 t/s
0.35.203.020 I slot print_timing: id 3 | task 0 | n_decoded = 1434, tg = 59.65 t/s, tg_3s = 58.64 t/s
0.38.211.560 I slot print_timing: id 3 | task 0 | n_decoded = 1609, tg = 59.48 t/s, tg_3s = 58.17 t/s
0.41.226.667 I slot print_timing: id 3 | task 0 | n_decoded = 1786, tg = 59.40 t/s, tg_3s = 58.70 t/s
0.44.238.639 I slot print_timing: id 3 | task 0 | n_decoded = 1960, tg = 59.26 t/s, tg_3s = 57.77 t/s
0.46.409.039 I slot print_timing: id 3 | task 0 | prompt eval time = 1223.35 ms / 504 tokens ( 2.43 ms per token, 411.98 tokens per second)
0.46.409.042 I slot print_timing: id 3 | task 0 | eval time = 35247.18 ms / 2087 tokens ( 16.89 ms per token, 59.21 tokens per second)
0.46.409.043 I slot print_timing: id 3 | task 0 | total time = 36470.53 ms / 2591 tokens
0.46.409.048 I slot print_timing: id 3 | task 0 | graphs reused = 2077
0.46.409.118 I slot release: id 3 | task 0 | stop processing: n_tokens = 2590, truncated = 0
0.46.409.135 I srv update_slots: all slots are idle
0.46.417.896 I srv operator(): Chat format: peg-native
0.46.422.219 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.749 (> 0.100 thold), f_keep = 0.194
0.46.422.220 I srv get_availabl: updating prompt cache
0.46.422.435 W srv prompt_save: - saving prompt with length 2590, total state size = 247.932 MiB (draft: 0.000 MiB)
0.46.735.743 I srv load: - looking for better prompt, base f_keep = 0.194, sim = 0.749
0.46.735.752 I srv update: - cache state: 1 prompts, 275.921 MiB (limits: 8192.000 MiB, 32768 tokens, 76896 est)
0.46.735.753 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
0.46.735.754 I srv get_availabl: prompt cache update took 313.53 ms
0.46.735.811 I slot launch_slot_: id 3 | task 2090 | processing task, is_child = 0
0.46.735.811 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.46.735.812 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.46.735.859 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.46.735.860 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.46.735.891 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.46.735.891 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.46.735.932 I slot update_slots: id 3 | task 2090 | Checking checkpoint with [0, 397] against 0...
0.46.746.394 W slot update_slots: id 3 | task 2090 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
0.47.257.908 I slot create_check: id 3 | task 2090 | created context checkpoint 2 of 32 (pos_min = 0, pos_max = 667, n_tokens = 668, size = 46.977 MiB)
0.47.892.673 I slot print_timing: id 3 | task 2090 | prompt eval time = 554.26 ms / 275 tokens ( 2.02 ms per token, 496.15 tokens per second)
0.47.892.677 I slot print_timing: id 3 | task 2090 | eval time = 602.47 ms / 38 tokens ( 15.85 ms per token, 63.07 tokens per second)
0.47.892.677 I slot print_timing: id 3 | task 2090 | total time = 1156.74 ms / 313 tokens
0.47.892.678 I slot print_timing: id 3 | task 2090 | graphs reused = 2113
0.47.892.735 I slot release: id 3 | task 2090 | stop processing: n_tokens = 709, truncated = 0
0.47.892.743 I srv update_slots: all slots are idle
0.47.918.079 I srv operator(): Chat format: peg-native
0.47.923.942 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.879 (> 0.100 thold), f_keep = 0.946
0.47.923.969 I slot launch_slot_: id 3 | task 2131 | processing task, is_child = 0
0.47.923.971 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.47.923.971 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.47.924.015 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.47.924.016 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.47.924.047 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.47.924.047 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.47.924.084 I slot update_slots: id 3 | task 2131 | Checking checkpoint with [0, 667] against 0...
0.47.939.281 W slot update_slots: id 3 | task 2131 | restored context checkpoint (pos_min = 0, pos_max = 667, n_tokens = 668, n_past = 667, size = 46.977 MiB)
0.50.621.802 I slot print_timing: id 3 | task 2131 | prompt eval time = 285.27 ms / 96 tokens ( 2.97 ms per token, 336.53 tokens per second)
0.50.621.806 I slot print_timing: id 3 | task 2131 | eval time = 2412.45 ms / 147 tokens ( 16.41 ms per token, 60.93 tokens per second)
0.50.621.806 I slot print_timing: id 3 | task 2131 | total time = 2697.71 ms / 243 tokens
0.50.621.807 I slot print_timing: id 3 | task 2131 | graphs reused = 2257
0.50.621.873 I slot release: id 3 | task 2131 | stop processing: n_tokens = 909, truncated = 0
0.50.621.883 I srv update_slots: all slots are idle
0.50.644.701 I srv operator(): Chat format: peg-native
0.50.647.149 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.838 (> 0.100 thold), f_keep = 0.443
0.50.647.151 I srv get_availabl: updating prompt cache
0.50.647.261 W srv prompt_save: - saving prompt with length 909, total state size = 87.016 MiB (draft: 0.000 MiB)
0.50.751.835 I srv load: - looking for better prompt, base f_keep = 0.443, sim = 0.838
0.50.751.845 I srv update: - cache state: 2 prompts, 437.904 MiB (limits: 8192.000 MiB, 32768 tokens, 65456 est)
0.50.751.846 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
0.50.751.847 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
0.50.751.848 I srv get_availabl: prompt cache update took 104.70 ms
0.50.751.909 I slot launch_slot_: id 3 | task 2281 | processing task, is_child = 0
0.50.751.910 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.50.751.911 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.50.751.948 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.50.751.949 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.50.751.980 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.50.751.981 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.50.752.017 I slot update_slots: id 3 | task 2281 | Checking checkpoint with [0, 667] against 0...
0.50.752.018 I slot update_slots: id 3 | task 2281 | Checking checkpoint with [0, 397] against 0...
0.50.762.500 W slot update_slots: id 3 | task 2281 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
0.50.762.506 W slot update_slots: id 3 | task 2281 | erased invalidated context checkpoint (pos_min = 0, pos_max = 667, n_tokens = 668, n_swa = 4096, pos_next = 397, size = 46.977 MiB)
0.52.311.464 I slot print_timing: id 3 | task 2281 | prompt eval time = 225.70 ms / 84 tokens ( 2.69 ms per token, 372.18 tokens per second)
0.52.311.468 I slot print_timing: id 3 | task 2281 | eval time = 1333.74 ms / 83 tokens ( 16.07 ms per token, 62.23 tokens per second)
0.52.311.468 I slot print_timing: id 3 | task 2281 | total time = 1559.44 ms / 167 tokens
0.52.311.469 I slot print_timing: id 3 | task 2281 | graphs reused = 2337
0.52.311.517 I slot release: id 3 | task 2281 | stop processing: n_tokens = 563, truncated = 0
0.52.311.525 I srv update_slots: all slots are idle
0.52.319.442 I srv operator(): Chat format: peg-native
0.52.324.040 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.943 (> 0.100 thold), f_keep = 0.853
0.52.324.071 I slot launch_slot_: id 3 | task 2367 | processing task, is_child = 0
0.52.324.072 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.52.324.072 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.52.324.120 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.52.324.122 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.52.324.191 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.52.324.192 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.52.324.265 I slot update_slots: id 3 | task 2367 | Checking checkpoint with [0, 397] against 0...
0.52.334.585 W slot update_slots: id 3 | task 2367 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
0.53.151.510 I slot print_timing: id 3 | task 2367 | prompt eval time = 326.57 ms / 112 tokens ( 2.92 ms per token, 342.96 tokens per second)
0.53.151.514 I slot print_timing: id 3 | task 2367 | eval time = 500.67 ms / 31 tokens ( 16.15 ms per token, 61.92 tokens per second)
0.53.151.514 I slot print_timing: id 3 | task 2367 | total time = 827.24 ms / 143 tokens
0.53.151.515 I slot print_timing: id 3 | task 2367 | graphs reused = 2365
0.53.151.563 I slot release: id 3 | task 2367 | stop processing: n_tokens = 539, truncated = 0
0.53.151.571 I srv update_slots: all slots are idle
0.53.161.796 I srv operator(): Chat format: peg-native
0.53.167.474 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.821 (> 0.100 thold), f_keep = 0.942
0.53.167.503 I slot launch_slot_: id 3 | task 2401 | processing task, is_child = 0
0.53.167.503 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
0.53.167.504 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
0.53.167.539 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
0.53.167.541 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
0.53.167.572 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
0.53.167.574 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
0.53.167.610 I slot update_slots: id 3 | task 2401 | Checking checkpoint with [0, 397] against 0...
0.53.177.985 W slot update_slots: id 3 | task 2401 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
0.56.540.276 I slot print_timing: id 3 | task 2401 | n_decoded = 186, tg = 61.79 t/s, tg_3s = 61.79 t/s
0.59.548.438 I slot print_timing: id 3 | task 2401 | n_decoded = 369, tg = 61.31 t/s, tg_3s = 60.83 t/s
1.02.553.593 I slot print_timing: id 3 | task 2401 | prompt eval time = 362.23 ms / 222 tokens ( 1.63 ms per token, 612.87 tokens per second)
1.02.553.597 I slot print_timing: id 3 | task 2401 | eval time = 9023.75 ms / 550 tokens ( 16.41 ms per token, 60.95 tokens per second)
1.02.553.598 I slot print_timing: id 3 | task 2401 | total time = 9385.98 ms / 772 tokens
1.02.553.598 I slot print_timing: id 3 | task 2401 | graphs reused = 2911
1.02.553.653 I slot release: id 3 | task 2401 | stop processing: n_tokens = 1168, truncated = 0
1.02.553.663 I srv update_slots: all slots are idle
1.02.566.187 I srv operator(): Chat format: peg-native
1.02.573.377 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.823 (> 0.100 thold), f_keep = 0.529
1.02.573.404 I slot launch_slot_: id 3 | task 2954 | processing task, is_child = 0
1.02.573.404 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.02.573.405 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.02.573.442 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.02.573.442 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.02.573.474 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.02.573.475 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.02.573.511 I slot update_slots: id 3 | task 2954 | Checking checkpoint with [0, 397] against 0...
1.02.584.310 W slot update_slots: id 3 | task 2954 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.03.065.215 I slot create_check: id 3 | task 2954 | created context checkpoint 2 of 32 (pos_min = 0, pos_max = 732, n_tokens = 733, size = 51.548 MiB)
1.03.702.748 I slot print_timing: id 3 | task 2954 | prompt eval time = 584.97 ms / 354 tokens ( 1.65 ms per token, 605.16 tokens per second)
1.03.702.752 I slot print_timing: id 3 | task 2954 | eval time = 544.27 ms / 33 tokens ( 16.49 ms per token, 60.63 tokens per second)
1.03.702.752 I slot print_timing: id 3 | task 2954 | total time = 1129.23 ms / 387 tokens
1.03.702.753 I slot print_timing: id 3 | task 2954 | graphs reused = 2941
1.03.702.819 I slot release: id 3 | task 2954 | stop processing: n_tokens = 783, truncated = 0
1.03.702.828 I srv update_slots: all slots are idle
1.03.732.538 I srv operator(): Chat format: peg-native
1.03.741.445 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.946 (> 0.100 thold), f_keep = 0.958
1.03.741.473 I slot launch_slot_: id 3 | task 2990 | processing task, is_child = 0
1.03.741.473 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.03.741.474 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.03.741.509 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.03.741.509 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.03.741.538 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.03.741.538 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.03.741.572 I slot update_slots: id 3 | task 2990 | Checking checkpoint with [0, 732] against 0...
1.03.758.068 W slot update_slots: id 3 | task 2990 | restored context checkpoint (pos_min = 0, pos_max = 732, n_tokens = 733, n_past = 732, size = 51.548 MiB)
1.06.984.503 I slot print_timing: id 3 | task 2990 | n_decoded = 183, tg = 60.73 t/s, tg_3s = 60.73 t/s
1.08.055.391 I slot print_timing: id 3 | task 2990 | prompt eval time = 229.52 ms / 61 tokens ( 3.76 ms per token, 265.78 tokens per second)
1.08.055.395 I slot print_timing: id 3 | task 2990 | eval time = 4084.30 ms / 247 tokens ( 16.54 ms per token, 60.48 tokens per second)
1.08.055.396 I slot print_timing: id 3 | task 2990 | total time = 4313.81 ms / 308 tokens
1.08.055.397 I slot print_timing: id 3 | task 2990 | graphs reused = 3185
1.08.055.457 I slot release: id 3 | task 2990 | stop processing: n_tokens = 1039, truncated = 0
1.08.055.468 I srv update_slots: all slots are idle
1.08.078.595 I srv operator(): Chat format: peg-native
1.08.081.087 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.826 (> 0.100 thold), f_keep = 0.388
1.08.081.089 I srv get_availabl: updating prompt cache
1.08.081.193 W srv prompt_save: - saving prompt with length 1039, total state size = 99.460 MiB (draft: 0.000 MiB)
1.08.203.188 I srv load: - looking for better prompt, base f_keep = 0.388, sim = 0.826
1.08.203.197 I srv update: - cache state: 3 prompts, 616.903 MiB (limits: 8192.000 MiB, 32768 tokens, 60261 est)
1.08.203.198 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
1.08.203.199 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
1.08.203.199 I srv update: - prompt 0x61d95c869f10: 1039 tokens, checkpoints: 2, 178.999 MiB
1.08.203.200 I srv get_availabl: prompt cache update took 122.11 ms
1.08.203.246 I slot launch_slot_: id 3 | task 3240 | processing task, is_child = 0
1.08.203.246 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.08.203.247 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.08.203.282 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.08.203.282 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.08.203.308 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.08.203.313 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.08.203.347 I slot update_slots: id 3 | task 3240 | Checking checkpoint with [0, 732] against 0...
1.08.203.348 I slot update_slots: id 3 | task 3240 | Checking checkpoint with [0, 397] against 0...
1.08.213.606 W slot update_slots: id 3 | task 3240 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.08.213.610 W slot update_slots: id 3 | task 3240 | erased invalidated context checkpoint (pos_min = 0, pos_max = 732, n_tokens = 733, n_swa = 4096, pos_next = 397, size = 51.548 MiB)
1.11.443.735 I slot print_timing: id 3 | task 3240 | n_decoded = 184, tg = 61.33 t/s, tg_3s = 61.33 t/s
1.14.443.871 I slot print_timing: id 3 | task 3240 | n_decoded = 366, tg = 60.99 t/s, tg_3s = 60.66 t/s
1.17.458.385 I slot print_timing: id 3 | task 3240 | n_decoded = 547, tg = 60.68 t/s, tg_3s = 60.04 t/s
1.20.470.337 I slot print_timing: id 3 | task 3240 | n_decoded = 727, tg = 60.45 t/s, tg_3s = 59.76 t/s
1.23.486.457 I slot print_timing: id 3 | task 3240 | n_decoded = 905, tg = 60.16 t/s, tg_3s = 59.02 t/s
1.26.486.614 I slot print_timing: id 3 | task 3240 | n_decoded = 1081, tg = 59.91 t/s, tg_3s = 58.66 t/s
1.29.495.641 I slot print_timing: id 3 | task 3240 | n_decoded = 1258, tg = 59.76 t/s, tg_3s = 58.82 t/s
1.32.382.668 I slot print_timing: id 3 | task 3240 | prompt eval time = 240.00 ms / 91 tokens ( 2.64 ms per token, 379.16 tokens per second)
1.32.382.672 I slot print_timing: id 3 | task 3240 | eval time = 23939.31 ms / 1427 tokens ( 16.78 ms per token, 59.61 tokens per second)
1.32.382.673 I slot print_timing: id 3 | task 3240 | total time = 24179.31 ms / 1518 tokens
1.32.382.673 I slot print_timing: id 3 | task 3240 | graphs reused = 4604
1.32.382.756 I slot release: id 3 | task 3240 | stop processing: n_tokens = 1914, truncated = 0
1.32.382.771 I srv update_slots: all slots are idle
1.32.391.256 I srv operator(): Chat format: peg-native
1.32.395.405 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.862 (> 0.100 thold), f_keep = 0.254
1.32.395.408 I srv get_availabl: updating prompt cache
1.32.395.584 W srv prompt_save: - saving prompt with length 1914, total state size = 183.221 MiB (draft: 0.000 MiB)
1.32.563.681 I srv load: - looking for better prompt, base f_keep = 0.254, sim = 0.862
1.32.563.691 I srv update: - cache state: 4 prompts, 828.114 MiB (limits: 8192.000 MiB, 32768 tokens, 63825 est)
1.32.563.692 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
1.32.563.693 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
1.32.563.694 I srv update: - prompt 0x61d95c869f10: 1039 tokens, checkpoints: 2, 178.999 MiB
1.32.563.694 I srv update: - prompt 0x61d96c180600: 1914 tokens, checkpoints: 1, 211.211 MiB
1.32.563.695 I srv get_availabl: prompt cache update took 168.29 ms
1.32.563.737 I slot launch_slot_: id 3 | task 4670 | processing task, is_child = 0
1.32.563.738 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.32.563.739 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.32.563.782 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.32.563.783 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.32.563.843 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.32.563.844 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.32.563.902 I slot update_slots: id 3 | task 4670 | Checking checkpoint with [0, 397] against 0...
1.32.574.346 W slot update_slots: id 3 | task 4670 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.34.276.237 I slot print_timing: id 3 | task 4670 | prompt eval time = 358.18 ms / 168 tokens ( 2.13 ms per token, 469.04 tokens per second)
1.34.276.241 I slot print_timing: id 3 | task 4670 | eval time = 1354.15 ms / 84 tokens ( 16.12 ms per token, 62.03 tokens per second)
1.34.276.242 I slot print_timing: id 3 | task 4670 | total time = 1712.33 ms / 252 tokens
1.34.276.242 I slot print_timing: id 3 | task 4670 | graphs reused = 4686
1.34.276.296 I slot release: id 3 | task 4670 | stop processing: n_tokens = 648, truncated = 0
1.34.276.304 I srv update_slots: all slots are idle
1.34.301.277 I srv operator(): Chat format: peg-native
1.34.307.200 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.935 (> 0.100 thold), f_keep = 0.870
1.34.307.229 I slot launch_slot_: id 3 | task 4757 | processing task, is_child = 0
1.34.307.229 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.34.307.230 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.34.307.266 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.34.307.268 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.34.307.300 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.34.307.301 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.34.307.336 I slot update_slots: id 3 | task 4757 | Checking checkpoint with [0, 397] against 0...
1.34.317.565 W slot update_slots: id 3 | task 4757 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.35.726.439 I slot print_timing: id 3 | task 4757 | prompt eval time = 402.93 ms / 206 tokens ( 1.96 ms per token, 511.26 tokens per second)
1.35.726.442 I slot print_timing: id 3 | task 4757 | eval time = 1016.17 ms / 63 tokens ( 16.13 ms per token, 62.00 tokens per second)
1.35.726.442 I slot print_timing: id 3 | task 4757 | total time = 1419.10 ms / 269 tokens
1.35.726.443 I slot print_timing: id 3 | task 4757 | graphs reused = 4747
1.35.726.504 I slot release: id 3 | task 4757 | stop processing: n_tokens = 665, truncated = 0
1.35.726.514 I srv update_slots: all slots are idle
1.35.738.236 I srv operator(): Chat format: peg-native
1.35.745.422 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.951 (> 0.100 thold), f_keep = 0.905
1.35.745.450 I slot launch_slot_: id 3 | task 4823 | processing task, is_child = 0
1.35.745.451 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.35.745.452 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.35.745.484 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.35.745.485 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.35.745.513 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.35.745.514 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.35.745.547 I slot update_slots: id 3 | task 4823 | Checking checkpoint with [0, 397] against 0...
1.35.755.802 W slot update_slots: id 3 | task 4823 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.37.705.376 I slot print_timing: id 3 | task 4823 | prompt eval time = 443.08 ms / 236 tokens ( 1.88 ms per token, 532.64 tokens per second)
1.37.705.380 I slot print_timing: id 3 | task 4823 | eval time = 1516.75 ms / 94 tokens ( 16.14 ms per token, 61.97 tokens per second)
1.37.705.380 I slot print_timing: id 3 | task 4823 | total time = 1959.82 ms / 330 tokens
1.37.705.381 I slot print_timing: id 3 | task 4823 | graphs reused = 4839
1.37.705.450 I slot release: id 3 | task 4823 | stop processing: n_tokens = 726, truncated = 0
1.37.705.462 I srv update_slots: all slots are idle
1.37.719.521 I srv operator(): Chat format: peg-native
1.37.728.451 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.899 (> 0.100 thold), f_keep = 0.871
1.37.728.478 I slot launch_slot_: id 3 | task 4920 | processing task, is_child = 0
1.37.728.479 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.37.728.479 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.37.728.508 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.37.728.508 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.37.728.533 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.37.728.534 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.37.728.562 I slot update_slots: id 3 | task 4920 | Checking checkpoint with [0, 397] against 0...
1.37.738.689 W slot update_slots: id 3 | task 4920 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.38.274.323 I slot create_check: id 3 | task 4920 | created context checkpoint 2 of 32 (pos_min = 0, pos_max = 698, n_tokens = 699, size = 49.157 MiB)
1.41.310.016 I slot print_timing: id 3 | task 4920 | n_decoded = 183, tg = 60.93 t/s, tg_3s = 60.93 t/s
1.41.623.445 I slot print_timing: id 3 | task 4920 | prompt eval time = 578.14 ms / 306 tokens ( 1.89 ms per token, 529.29 tokens per second)
1.41.623.448 I slot print_timing: id 3 | task 4920 | eval time = 3316.74 ms / 202 tokens ( 16.42 ms per token, 60.90 tokens per second)
1.41.623.449 I slot print_timing: id 3 | task 4920 | total time = 3894.88 ms / 508 tokens
1.41.623.450 I slot print_timing: id 3 | task 4920 | graphs reused = 5038
1.41.623.510 I slot release: id 3 | task 4920 | stop processing: n_tokens = 904, truncated = 0
1.41.623.519 I srv update_slots: all slots are idle
1.41.640.382 I srv operator(): Chat format: peg-native
1.41.650.748 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.905 (> 0.100 thold), f_keep = 0.777
1.41.650.775 I slot launch_slot_: id 3 | task 5125 | processing task, is_child = 0
1.41.650.776 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.41.650.776 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.41.650.814 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.41.650.814 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.41.650.846 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.41.650.846 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.41.650.883 I slot update_slots: id 3 | task 5125 | Checking checkpoint with [0, 698] against 0...
1.41.666.371 W slot update_slots: id 3 | task 5125 | restored context checkpoint (pos_min = 0, pos_max = 698, n_tokens = 699, n_past = 698, size = 49.157 MiB)
1.42.584.968 I slot print_timing: id 3 | task 5125 | prompt eval time = 316.38 ms / 78 tokens ( 4.06 ms per token, 246.54 tokens per second)
1.42.584.971 I slot print_timing: id 3 | task 5125 | eval time = 617.70 ms / 38 tokens ( 16.26 ms per token, 61.52 tokens per second)
1.42.584.972 I slot print_timing: id 3 | task 5125 | total time = 934.08 ms / 116 tokens
1.42.584.973 I slot print_timing: id 3 | task 5125 | graphs reused = 5074
1.42.585.038 I slot release: id 3 | task 5125 | stop processing: n_tokens = 813, truncated = 0
1.42.585.047 I srv update_slots: all slots are idle
1.42.618.904 I srv operator(): Chat format: peg-native
1.42.630.521 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.950 (> 0.100 thold), f_keep = 0.953
1.42.630.558 I slot launch_slot_: id 3 | task 5166 | processing task, is_child = 0
1.42.630.558 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.42.630.559 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.42.630.600 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.42.630.601 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.42.630.632 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.42.630.633 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.42.630.671 I slot update_slots: id 3 | task 5166 | Checking checkpoint with [0, 698] against 0...
1.42.646.776 W slot update_slots: id 3 | task 5166 | restored context checkpoint (pos_min = 0, pos_max = 698, n_tokens = 699, n_past = 698, size = 49.157 MiB)
1.44.392.468 I slot print_timing: id 3 | task 5166 | prompt eval time = 366.42 ms / 118 tokens ( 3.11 ms per token, 322.03 tokens per second)
1.44.392.472 I slot print_timing: id 3 | task 5166 | eval time = 1395.37 ms / 85 tokens ( 16.42 ms per token, 60.92 tokens per second)
1.44.392.473 I slot print_timing: id 3 | task 5166 | total time = 1761.79 ms / 203 tokens
1.44.392.473 I slot print_timing: id 3 | task 5166 | graphs reused = 5157
1.44.392.537 I slot release: id 3 | task 5166 | stop processing: n_tokens = 900, truncated = 0
1.44.392.546 I srv update_slots: all slots are idle
1.44.413.367 I srv operator(): Chat format: peg-native
1.44.415.865 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.781 (> 0.100 thold), f_keep = 0.448
1.44.415.868 I srv get_availabl: updating prompt cache
1.44.415.998 W srv prompt_save: - saving prompt with length 900, total state size = 86.155 MiB (draft: 0.000 MiB)
1.44.517.143 I srv load: - looking for better prompt, base f_keep = 0.448, sim = 0.781
1.44.517.155 I srv update: - cache state: 5 prompts, 991.415 MiB (limits: 8192.000 MiB, 32768 tokens, 60749 est)
1.44.517.156 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
1.44.517.156 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
1.44.517.157 I srv update: - prompt 0x61d95c869f10: 1039 tokens, checkpoints: 2, 178.999 MiB
1.44.517.158 I srv update: - prompt 0x61d96c180600: 1914 tokens, checkpoints: 1, 211.211 MiB
1.44.517.158 I srv update: - prompt 0x61d95c2252f0: 900 tokens, checkpoints: 2, 163.302 MiB
1.44.517.160 I srv get_availabl: prompt cache update took 101.29 ms
1.44.517.223 I slot launch_slot_: id 3 | task 5254 | processing task, is_child = 0
1.44.517.224 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
1.44.517.225 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
1.44.517.259 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
1.44.517.259 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
1.44.517.284 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
1.44.517.285 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
1.44.517.320 I slot update_slots: id 3 | task 5254 | Checking checkpoint with [0, 698] against 0...
1.44.517.321 I slot update_slots: id 3 | task 5254 | Checking checkpoint with [0, 397] against 0...
1.44.527.498 W slot update_slots: id 3 | task 5254 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
1.44.527.503 W slot update_slots: id 3 | task 5254 | erased invalidated context checkpoint (pos_min = 0, pos_max = 698, n_tokens = 699, n_swa = 4096, pos_next = 397, size = 49.157 MiB)
1.47.863.444 I slot print_timing: id 3 | task 5254 | n_decoded = 185, tg = 61.58 t/s, tg_3s = 61.58 t/s
1.50.876.478 I slot print_timing: id 3 | task 5254 | n_decoded = 368, tg = 61.16 t/s, tg_3s = 60.74 t/s
1.53.891.246 I slot print_timing: id 3 | task 5254 | n_decoded = 550, tg = 60.89 t/s, tg_3s = 60.37 t/s
1.56.903.681 I slot print_timing: id 3 | task 5254 | n_decoded = 731, tg = 60.69 t/s, tg_3s = 60.08 t/s
1.59.920.211 I slot print_timing: id 3 | task 5254 | n_decoded = 909, tg = 60.35 t/s, tg_3s = 59.01 t/s
2.02.930.501 I slot print_timing: id 3 | task 5254 | n_decoded = 1086, tg = 60.09 t/s, tg_3s = 58.80 t/s
2.04.749.336 I slot print_timing: id 3 | task 5254 | prompt eval time = 341.73 ms / 119 tokens ( 2.87 ms per token, 348.23 tokens per second)
2.04.749.339 I slot print_timing: id 3 | task 5254 | eval time = 19890.28 ms / 1193 tokens ( 16.67 ms per token, 59.98 tokens per second)
2.04.749.340 I slot print_timing: id 3 | task 5254 | total time = 20232.01 ms / 1312 tokens
2.04.749.340 I slot print_timing: id 3 | task 5254 | graphs reused = 6344
2.04.749.391 I slot release: id 3 | task 5254 | stop processing: n_tokens = 1708, truncated = 0
2.04.749.403 I srv update_slots: all slots are idle
2.04.758.280 I srv operator(): Chat format: peg-native
2.04.762.710 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.786 (> 0.100 thold), f_keep = 0.302
2.04.762.712 I srv get_availabl: updating prompt cache
2.04.762.876 W srv prompt_save: - saving prompt with length 1708, total state size = 163.501 MiB (draft: 0.000 MiB)
2.04.904.050 I srv load: - looking for better prompt, base f_keep = 0.302, sim = 0.786
2.04.904.061 I srv update: - cache state: 6 prompts, 1182.906 MiB (limits: 8192.000 MiB, 32768 tokens, 62743 est)
2.04.904.062 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
2.04.904.063 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
2.04.904.063 I srv update: - prompt 0x61d95c869f10: 1039 tokens, checkpoints: 2, 178.999 MiB
2.04.904.064 I srv update: - prompt 0x61d96c180600: 1914 tokens, checkpoints: 1, 211.211 MiB
2.04.904.064 I srv update: - prompt 0x61d95c2252f0: 900 tokens, checkpoints: 2, 163.302 MiB
2.04.904.065 I srv update: - prompt 0x61d96c200cc0: 1708 tokens, checkpoints: 1, 191.491 MiB
2.04.904.066 I srv get_availabl: prompt cache update took 141.35 ms
2.04.904.105 I slot launch_slot_: id 3 | task 6450 | processing task, is_child = 0
2.04.904.105 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
2.04.904.106 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
2.04.904.140 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
2.04.904.140 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
2.04.904.166 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
2.04.904.166 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
2.04.904.200 I slot update_slots: id 3 | task 6450 | Checking checkpoint with [0, 397] against 0...
2.04.914.618 W slot update_slots: id 3 | task 6450 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
2.06.430.496 I slot print_timing: id 3 | task 6450 | prompt eval time = 494.36 ms / 258 tokens ( 1.92 ms per token, 521.88 tokens per second)
2.06.430.500 I slot print_timing: id 3 | task 6450 | eval time = 1031.93 ms / 63 tokens ( 16.38 ms per token, 61.05 tokens per second)
2.06.430.500 I slot print_timing: id 3 | task 6450 | total time = 1526.29 ms / 321 tokens
2.06.430.501 I slot print_timing: id 3 | task 6450 | graphs reused = 6405
2.06.430.557 I slot release: id 3 | task 6450 | stop processing: n_tokens = 717, truncated = 0
2.06.430.569 I srv update_slots: all slots are idle
2.06.455.288 I srv operator(): Chat format: peg-native
2.06.461.036 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.930 (> 0.100 thold), f_keep = 0.912
2.06.461.067 I slot launch_slot_: id 3 | task 6516 | processing task, is_child = 0
2.06.461.068 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
2.06.461.068 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
2.06.461.099 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
2.06.461.103 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
2.06.461.132 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
2.06.461.134 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
2.06.461.166 I slot update_slots: id 3 | task 6516 | Checking checkpoint with [0, 397] against 0...
2.06.471.692 W slot update_slots: id 3 | task 6516 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
2.06.893.387 I slot create_check: id 3 | task 6516 | created context checkpoint 2 of 32 (pos_min = 0, pos_max = 673, n_tokens = 674, size = 47.399 MiB)
2.08.138.154 I slot print_timing: id 3 | task 6516 | prompt eval time = 579.06 ms / 306 tokens ( 1.89 ms per token, 528.44 tokens per second)
2.08.138.158 I slot print_timing: id 3 | task 6516 | eval time = 1097.91 ms / 67 tokens ( 16.39 ms per token, 61.02 tokens per second)
2.08.138.158 I slot print_timing: id 3 | task 6516 | total time = 1676.98 ms / 373 tokens
2.08.138.159 I slot print_timing: id 3 | task 6516 | graphs reused = 6469
2.08.138.218 I slot release: id 3 | task 6516 | stop processing: n_tokens = 769, truncated = 0
2.08.138.228 I srv update_slots: all slots are idle
2.08.159.007 I srv operator(): Chat format: peg-native
2.08.161.440 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.788 (> 0.100 thold), f_keep = 0.527
2.08.161.485 I slot launch_slot_: id 3 | task 6586 | processing task, is_child = 0
2.08.161.491 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
2.08.161.492 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
2.08.161.525 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
2.08.161.525 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
2.08.161.551 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
2.08.161.551 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
2.08.161.584 I slot update_slots: id 3 | task 6586 | Checking checkpoint with [0, 673] against 0...
2.08.161.585 I slot update_slots: id 3 | task 6586 | Checking checkpoint with [0, 397] against 0...
2.08.172.108 W slot update_slots: id 3 | task 6586 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
2.08.172.113 W slot update_slots: id 3 | task 6586 | erased invalidated context checkpoint (pos_min = 0, pos_max = 673, n_tokens = 674, n_swa = 4096, pos_next = 397, size = 47.399 MiB)
2.11.481.111 I slot print_timing: id 3 | task 6586 | n_decoded = 183, tg = 60.74 t/s, tg_3s = 60.74 t/s
2.14.483.649 I slot print_timing: id 3 | task 6586 | n_decoded = 362, tg = 60.18 t/s, tg_3s = 59.62 t/s
2.17.496.152 I slot print_timing: id 3 | task 6586 | n_decoded = 542, tg = 60.04 t/s, tg_3s = 59.75 t/s
2.20.498.143 I slot print_timing: id 3 | task 6586 | n_decoded = 720, tg = 59.85 t/s, tg_3s = 59.29 t/s
2.20.872.159 I slot print_timing: id 3 | task 6586 | prompt eval time = 306.70 ms / 117 tokens ( 2.62 ms per token, 381.48 tokens per second)
2.20.872.162 I slot print_timing: id 3 | task 6586 | eval time = 12403.87 ms / 742 tokens ( 16.72 ms per token, 59.82 tokens per second)
2.20.872.162 I slot print_timing: id 3 | task 6586 | total time = 12710.57 ms / 859 tokens
2.20.872.163 I slot print_timing: id 3 | task 6586 | graphs reused = 7207
2.20.872.218 I slot release: id 3 | task 6586 | stop processing: n_tokens = 1255, truncated = 0
2.20.872.230 I srv update_slots: all slots are idle
2.20.881.671 I srv operator(): Chat format: peg-native
2.20.886.609 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.612 (> 0.100 thold), f_keep = 0.409
2.20.886.611 I srv get_availabl: updating prompt cache
2.20.886.752 W srv prompt_save: - saving prompt with length 1255, total state size = 120.137 MiB (draft: 0.000 MiB)
2.20.997.623 I srv load: - looking for better prompt, base f_keep = 0.409, sim = 0.612
2.20.997.635 I srv update: - cache state: 7 prompts, 1331.033 MiB (limits: 8192.000 MiB, 32768 tokens, 63484 est)
2.20.997.636 I srv update: - prompt 0x61d964e588c0: 2590 tokens, checkpoints: 1, 275.921 MiB
2.20.997.637 I srv update: - prompt 0x61d964ade9a0: 909 tokens, checkpoints: 2, 161.983 MiB
2.20.997.637 I srv update: - prompt 0x61d95c869f10: 1039 tokens, checkpoints: 2, 178.999 MiB
2.20.997.638 I srv update: - prompt 0x61d96c180600: 1914 tokens, checkpoints: 1, 211.211 MiB
2.20.997.638 I srv update: - prompt 0x61d95c2252f0: 900 tokens, checkpoints: 2, 163.302 MiB
2.20.997.639 I srv update: - prompt 0x61d96c200cc0: 1708 tokens, checkpoints: 1, 191.491 MiB
2.20.997.639 I srv update: - prompt 0x61d964d14ab0: 1255 tokens, checkpoints: 1, 148.127 MiB
2.20.997.640 I srv get_availabl: prompt cache update took 111.03 ms
2.20.997.680 I slot launch_slot_: id 3 | task 7331 | processing task, is_child = 0
2.20.997.681 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
2.20.997.682 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
2.20.997.721 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
2.20.997.722 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
2.20.997.782 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
2.20.997.782 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
2.20.997.850 I slot update_slots: id 3 | task 7331 | Checking checkpoint with [0, 397] against 0...
2.21.008.338 W slot update_slots: id 3 | task 7331 | restored context checkpoint (pos_min = 0, pos_max = 397, n_tokens = 398, n_past = 397, size = 27.990 MiB)
2.21.550.369 I slot create_check: id 3 | task 7331 | created context checkpoint 2 of 32 (pos_min = 0, pos_max = 818, n_tokens = 819, size = 57.596 MiB)
2.22.269.171 I slot print_timing: id 3 | task 7331 | prompt eval time = 646.86 ms / 441 tokens ( 1.47 ms per token, 681.75 tokens per second)
2.22.269.174 I slot print_timing: id 3 | task 7331 | eval time = 624.45 ms / 38 tokens ( 16.43 ms per token, 60.85 tokens per second)
2.22.269.175 I slot print_timing: id 3 | task 7331 | total time = 1271.32 ms / 479 tokens
2.22.269.175 I slot print_timing: id 3 | task 7331 | graphs reused = 7243
2.22.269.248 I slot release: id 3 | task 7331 | stop processing: n_tokens = 875, truncated = 0
2.22.269.262 I srv update_slots: all slots are idle
2.22.295.090 I srv operator(): Chat format: peg-native
2.22.301.400 I slot get_availabl: id 3 | task -1 | selected slot by LCP similarity, sim_best = 0.951 (> 0.100 thold), f_keep = 0.957
2.22.301.425 I slot launch_slot_: id 3 | task 7372 | processing task, is_child = 0
2.22.301.425 I slot process_sing: id 0 | task -1 | saving idle slot to prompt cache
2.22.301.426 I slot prompt_clear: id 0 | task -1 | clearing prompt with 0 tokens
2.22.301.461 I slot process_sing: id 1 | task -1 | saving idle slot to prompt cache
2.22.301.461 I slot prompt_clear: id 1 | task -1 | clearing prompt with 0 tokens
2.22.301.491 I slot process_sing: id 2 | task -1 | saving idle slot to prompt cache
2.22.301.492 I slot prompt_clear: id 2 | task -1 | clearing prompt with 0 tokens
2.22.301.526 I slot update_slots: id 3 | task 7372 | Checking checkpoint with [0, 818] against 0...
2.22.319.366 W slot update_slots: id 3 | task 7372 | restored context checkpoint (pos_min = 0, pos_max = 818, n_tokens = 819, n_past = 818, size = 57.596 MiB)
2.24.295.685 I slot print_timing: id 3 | task 7372 | prompt eval time = 228.22 ms / 62 tokens ( 3.68 ms per token, 271.67 tokens per second)
2.24.295.689 I slot print_timing: id 3 | task 7372 | eval time = 1765.93 ms / 107 tokens ( 16.50 ms per token, 60.59 tokens per second)
2.24.295.689 I slot print_timing: id 3 | task 7372 | total time = 1994.15 ms / 169 tokens
2.24.295.690 I slot print_timing: id 3 | task 7372 | graphs reused = 7348
2.24.295.759 I slot release: id 3 | task 7372 | stop processing: n_tokens = 986, truncated = 0
2.24.295.767 I srv update_slots: all slots are idle
2.24.313.604 I srv operator(): operator(): cleaning up before exit...