50.23.792.913 I srv proxy_reques: proxying request to model gemma-4-26B-A4B on port 46193

[46193] 50.23.619.515 I srv params_from_: Chat format: peg-gemma4

[46193] 50.23.619.637 I slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 2289971110874

[46193] 50.23.619.638 I srv get_availabl: updating prompt cache

[46193] 50.23.619.725 W srv prompt_save: - saving prompt with length 701, total state size = 90.225 MiB (draft: 10.199 MiB)

[46193] 50.23.642.965 I srv load: - looking for better prompt, base f_keep = 0.020, sim = 0.011

[46193] 50.23.642.971 I srv update: - cache state: 11 prompts, 1561.227 MiB (limits: 0.000 MiB, 32768 tokens, 32768 est)

[46193] 50.23.642.972 I srv update: - prompt 0x58211674b490: 3794 tokens, checkpoints: 0, 166.351 MiB

[46193] 50.23.642.975 I srv update: - prompt 0x5820ff86f9c0: 7463 tokens, checkpoints: 0, 212.118 MiB

[46193] 50.23.642.975 I srv update: - prompt 0x58211670c5a0: 1591 tokens, checkpoints: 0, 138.871 MiB

[46193] 50.23.642.975 I srv update: - prompt 0x58211670b4c0: 1038 tokens, checkpoints: 0, 131.972 MiB

[46193] 50.23.642.976 I srv update: - prompt 0x582107c81e60: 1949 tokens, checkpoints: 0, 143.336 MiB

[46193] 50.23.642.977 I srv update: - prompt 0x582107b4c330: 1804 tokens, checkpoints: 0, 141.527 MiB

[46193] 50.23.642.978 I srv update: - prompt 0x5821166fd260: 3100 tokens, checkpoints: 0, 136.539 MiB

[46193] 50.23.642.978 I srv update: - prompt 0x582115b02450: 4129 tokens, checkpoints: 0, 170.530 MiB

[46193] 50.23.642.978 I srv update: - prompt 0x5821166b92c0: 1675 tokens, checkpoints: 0, 139.918 MiB

[46193] 50.23.642.978 I srv update: - prompt 0x582116364800: 698 tokens, checkpoints: 0, 89.839 MiB

[46193] 50.23.642.978 I srv update: - prompt 0x5821168ff050: 701 tokens, checkpoints: 0, 90.225 MiB

[46193] 50.23.642.979 I srv get_availabl: prompt cache update took 23.34 ms

[46193] 50.23.643.036 I slot launch_slot_: id 0 | task 8298 | processing task, is_child = 0

[46193] 50.24.441.751 I slot print_timing: id 0 | task 8298 | n_decoded = 100, tg = 163.00 t/s

[46193] 50.27.455.059 I slot print_timing: id 0 | task 8298 | n_decoded = 616, tg = 169.85 t/s

[46193] 50.30.467.849 I slot print_timing: id 0 | task 8298 | n_decoded = 1096, tg = 165.07 t/s

[46193] 50.33.477.927 I slot print_timing: id 0 | task 8298 | n_decoded = 1575, tg = 163.22 t/s

[46193] 50.36.398.689 I slot print_timing: id 0 | task 8298 | prompt eval time = 185.18 ms / 1272 tokens ( 0.15 ms per token, 6869.10 tokens per second)

[46193] 50.36.398.692 I slot print_timing: id 0 | task 8298 | eval time = 12570.44 ms / 2031 tokens ( 6.19 ms per token, 161.57 tokens per second)

[46193] 50.36.398.692 I slot print_timing: id 0 | task 8298 | total time = 12755.61 ms / 3303 tokens

[46193] 50.36.398.693 I slot print_timing: id 0 | task 8298 | graphs reused = 8994

[46193] 50.36.398.694 I slot print_timing: id 0 | task 8298 | draft acceptance = 0.73329 ( 1207 accepted / 1646 generated)

[46193] 50.36.398.706 I statistics draft-mtp: #calls(b,g,a) = 13 9089 9089, #gen drafts = 9089, #acc drafts = 6584, #gen tokens = 18178, #acc tokens = 11527, dur(b,g,a) = 0.018, 60880.802, 1.926 ms

[46193] 50.36.398.764 I slot release: id 0 | task 8298 | stop processing: n_tokens = 3316, truncated = 0

[46193] 50.36.398.777 I srv update_slots: all slots are idle

make me feel alright?

50.36.973.390 I srv proxy_reques: proxying request to model gemma-4-26B-A4B on port 46193

[46193] 50.36.799.172 I srv params_from_: Chat format: peg-gemma4

[46193] 50.36.799.292 I slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 2289984639550

[46193] 50.36.799.294 I srv get_availabl: updating prompt cache

[46193] 50.36.799.451 W srv prompt_save: - saving prompt with length 3316, total state size = 160.388 MiB (draft: 19.681 MiB)

[46193] 50.36.838.940 I srv load: - looking for better prompt, base f_keep = 0.004, sim = 0.018

[46193] 50.36.838.946 I srv update: - cache state: 12 prompts, 1721.615 MiB (limits: 0.000 MiB, 32768 tokens, 32768 est)

[46193] 50.36.838.948 I srv update: - prompt 0x58211674b490: 3794 tokens, checkpoints: 0, 166.351 MiB

[46193] 50.36.838.948 I srv update: - prompt 0x5820ff86f9c0: 7463 tokens, checkpoints: 0, 212.118 MiB

[46193] 50.36.838.949 I srv update: - prompt 0x58211670c5a0: 1591 tokens, checkpoints: 0, 138.871 MiB

[46193] 50.36.838.949 I srv update: - prompt 0x58211670b4c0: 1038 tokens, checkpoints: 0, 131.972 MiB

[46193] 50.36.838.949 I srv update: - prompt 0x582107c81e60: 1949 tokens, checkpoints: 0, 143.336 MiB

[46193] 50.36.838.949 I srv update: - prompt 0x582107b4c330: 1804 tokens, checkpoints: 0, 141.527 MiB

[46193] 50.36.838.949 I srv update: - prompt 0x5821166fd260: 3100 tokens, checkpoints: 0, 136.539 MiB

[46193] 50.36.838.950 I srv update: - prompt 0x582115b02450: 4129 tokens, checkpoints: 0, 170.530 MiB

[46193] 50.36.838.950 I srv update: - prompt 0x5821166b92c0: 1675 tokens, checkpoints: 0, 139.918 MiB

[46193] 50.36.838.950 I srv update: - prompt 0x582116364800: 698 tokens, checkpoints: 0, 89.839 MiB

[46193] 50.36.838.950 I srv update: - prompt 0x5821168ff050: 701 tokens, checkpoints: 0, 90.225 MiB

[46193] 50.36.838.951 I srv update: - prompt 0x582108a24b60: 3316 tokens, checkpoints: 0, 160.388 MiB

[46193] 50.36.838.951 I srv get_availabl: prompt cache update took 39.66 ms

[46193] 50.36.839.018 I slot launch_slot_: id 0 | task 9123 | processing task, is_child = 0

[46193] 50.37.594.196 I slot print_timing: id 0 | task 9123 | n_decoded = 100, tg = 155.70 t/s

[46193] 50.40.009.510 I slot print_timing: id 0 | task 9123 | prompt eval time = 112.87 ms / 721 tokens ( 0.16 ms per token, 6388.11 tokens per second)

[46193] 50.40.009.513 I slot print_timing: id 0 | task 9123 | eval time = 3057.58 ms / 477 tokens ( 6.41 ms per token, 156.01 tokens per second)

[46193] 50.40.009.514 I slot print_timing: id 0 | task 9123 | total time = 3170.45 ms / 1198 tokens

[46193] 50.40.009.514 I slot print_timing: id 0 | task 9123 | graphs reused = 9199

[46193] 50.40.009.515 I slot print_timing: id 0 | task 9123 | draft acceptance = 0.64904 ( 270 accepted / 416 generated)

[46193] 50.40.009.527 I statistics draft-mtp: #calls(b,g,a) = 14 9297 9297, #gen drafts = 9297, #acc drafts = 6735, #gen tokens = 18594, #acc tokens = 11797, dur(b,g,a) = 0.020, 62279.211, 1.981 ms

[46193] 50.40.009.562 I slot release: id 0 | task 9123 | stop processing: n_tokens = 1212, truncated = 0

[46193] 50.40.009.574 I srv update_slots: all slots are idle

Have you ever seen the lain?

188.10.762.594 I srv proxy_reques: proxying request to model gemma-4-26B-A4B on port 46193

[46193] 188.10.589.671 I srv params_from_: Chat format: peg-gemma4

[46193] 188.10.590.018 I slot get_availabl: id 0 | task -1 | selected slot by LRU, t_last = 2298237243280

[46193] 188.10.590.021 I srv get_availabl: updating prompt cache

[46193] 188.10.590.544 W srv prompt_save: - saving prompt with length 5507, total state size = 187.719 MiB (draft: 24.253 MiB)

[46193] 188.10.641.749 I srv load: - looking for better prompt, base f_keep = 0.002, sim = 0.029

[46193] 188.10.641.756 W srv update: - cache token limit (32768, est: 32768) reached, removing oldest entry (size = 84.562 MiB)

[46193] 188.10.643.670 W srv update: - cache token limit (32768, est: 32768) reached, removing oldest entry (size = 156.060 MiB)

[46193] 188.10.647.701 W srv update: - cache token limit (32768, est: 32768) reached, removing oldest entry (size = 137.336 MiB)

[46193] 188.10.651.894 W srv update: - cache token limit (32768, est: 32768) reached, removing oldest entry (size = 184.588 MiB)

[46193] 188.10.658.388 I srv update: - cache state: 7 prompts, 1177.541 MiB (limits: 0.000 MiB, 32768 tokens, 32768 est)

[46193] 188.10.658.390 I srv update: - prompt 0x5821073809d0: 8138 tokens, checkpoints: 0, 220.538 MiB

[46193] 188.10.658.391 I srv update: - prompt 0x5821226fc270: 2539 tokens, checkpoints: 0, 150.696 MiB

[46193] 188.10.658.391 I srv update: - prompt 0x5821226f6e80: 4720 tokens, checkpoints: 0, 177.902 MiB

[46193] 188.10.658.391 I srv update: - prompt 0x5821225c4820: 1327 tokens, checkpoints: 0, 135.577 MiB

[46193] 188.10.658.392 I srv update: - prompt 0x5821162f3480: 2178 tokens, checkpoints: 0, 146.193 MiB

[46193] 188.10.658.392 I srv update: - prompt 0x5821190fa1b0: 3198 tokens, checkpoints: 0, 158.916 MiB

[46193] 188.10.658.392 I srv update: - prompt 0x5821190c61a0: 5507 tokens, checkpoints: 0, 187.719 MiB

[46193] 188.10.658.393 I srv get_availabl: prompt cache update took 68.37 ms

[46193] 188.10.658.460 I slot launch_slot_: id 0 | task 22113 | processing task, is_child = 0

[46193] 188.11.844.921 I slot print_timing: id 0 | task 22113 | n_decoded = 169, tg = 152.56 t/s

[46193] 188.14.855.903 I slot print_timing: id 0 | task 22113 | n_decoded = 610, tg = 148.10 t/s

[46193] 188.15.733.298 I slot print_timing: id 0 | task 22113 | prompt eval time = 78.62 ms / 363 tokens ( 0.22 ms per token, 4617.09 tokens per second)

[46193] 188.15.733.302 I slot print_timing: id 0 | task 22113 | eval time = 4996.17 ms / 752 tokens ( 6.64 ms per token, 150.52 tokens per second)

[46193] 188.15.733.302 I slot print_timing: id 0 | task 22113 | total time = 5074.80 ms / 1115 tokens

[46193] 188.15.733.303 I slot print_timing: id 0 | task 22113 | graphs reused = 22147

[46193] 188.15.733.303 I slot print_timing: id 0 | task 22113 | draft acceptance = 0.58526 ( 405 accepted / 692 generated)

[46193] 188.15.733.316 I statistics draft-mtp: #calls(b,g,a) = 30 22380 22380, #gen drafts = 22380, #acc drafts = 16116, #gen tokens = 44760, #acc tokens = 28094, dur(b,g,a) = 0.039, 149913.895, 4.856 ms

[46193] 188.15.733.347 I slot release: id 0 | task 22113 | stop processing: n_tokens = 1125, truncated = 0

[46193] 188.15.733.358 I srv update_slots: all slots are idle

Do you remember how it felt? Let's try this again later.