diff --git "a/results/tool_use_levels.json" "b/results/tool_use_levels.json" new file mode 100644--- /dev/null +++ "b/results/tool_use_levels.json" @@ -0,0 +1,57643 @@ +{ + "host": { + "cpu_model": "Intel(R) Xeon(R) Platinum 8559C", + "cpu_threads": 192, + "cgroup": { + "memory.max": "104000000000", + "memory.swap.max": "2936848384", + "memory.high": "max" + } + }, + "config": { + "model": "models/Ling-3.0-tiny-Q4_K_M.gguf", + "bin": "llama.cpp/build/bin/sdq-chat", + "fixture": "tools/tiny_fixture", + "levels_mb": [ + 1024, + 2048, + 4096, + 8192, + 16384 + ], + "repeat": 10, + "ctx": 256, + "ubatch": 32, + "threads": 4, + "max_tokens": 80, + "timeout": 1800, + "tool_timeout": 120, + "evict_helper": "./pagecache_resident", + "reserve_map": { + "1024": 700, + "2048": 700, + "4096": 700, + "8192": 1000, + "16384": 1400 + }, + "output_dir": "results/tool_use_runs", + "out_json": "results/tool_use_levels.json", + "retry_failed": true + }, + "runs": [ + { + "level_mb": 1024, + "repeat": 1, + "status": "OK", + "wall_s": 34.49, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.556, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.5, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.45, + "tok_s": 7.6, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8779, + "expert_hits": 526, + "expert_misses": 8253, + "hit_rate": 5.99157, + "ssb_demand_bytes": 11810433024, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11810433024, + "ssd_bytes_per_token": 71147200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8028, + "demand_wait_us": 14703653, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954748928, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 44, + "miss": 342, + "hit_rate": 11.399, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 32, + "miss": 374, + "hit_rate": 7.88177, + "demand_bytes": 496336896, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 0, + "miss": 274, + "hit_rate": 0, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r1_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.1, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.95, + "tok_s": 8.1, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1598, + "expert_misses": 8999, + "hit_rate": 15.0797, + "ssb_demand_bytes": 12979279872, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12979279872, + "ssd_bytes_per_token": 76348700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8774, + "demand_wait_us": 16021196, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953245696, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 129, + "miss": 325, + "hit_rate": 28.4141, + "demand_bytes": 431308800, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 117, + "miss": 359, + "hit_rate": 24.5798, + "demand_bytes": 476430336, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 163, + "miss": 281, + "hit_rate": 36.7117, + "demand_bytes": 372916224, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r1_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.561, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.6, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.73, + "tok_s": 7.52, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9185, + "expert_hits": 665, + "expert_misses": 8520, + "hit_rate": 7.24007, + "ssb_demand_bytes": 12210186240, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12210186240, + "ssd_bytes_per_token": 70173500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8295, + "demand_wait_us": 15004124, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 961163264, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 52, + "miss": 363, + "hit_rate": 12.5301, + "demand_bytes": 481738752, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 340, + "hit": 0, + "miss": 340, + "hit_rate": 0, + "demand_bytes": 520151040, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r1_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 915.8, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.92, + "tok_s": 7.81, + "ssd_mb": 3079.7, + "ssd_mb_per_token": 205.3 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7470, + "expert_hits": 747, + "expert_misses": 6723, + "hit_rate": 10, + "ssb_demand_bytes": 9663584256, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9663584256, + "ssd_bytes_per_token": 73767800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6497, + "demand_wait_us": 11844457, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 960315392, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 58, + "miss": 295, + "hit_rate": 16.4306, + "demand_bytes": 391495680, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 58, + "miss": 283, + "hit_rate": 17.0088, + "demand_bytes": 375570432, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 58, + "miss": 239, + "hit_rate": 19.5286, + "demand_bytes": 317177856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 63, + "miss": 264, + "hit_rate": 19.2661, + "demand_bytes": 350355456, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 61, + "miss": 253, + "hit_rate": 19.4268, + "demand_bytes": 335757312, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 71, + "miss": 222, + "hit_rate": 24.2321, + "demand_bytes": 294617088, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 299, + "hit": 0, + "miss": 299, + "hit_rate": 0, + "demand_bytes": 457426944, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r1_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.6, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.45, + "tok_s": 7.6, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + { + "tokens": 32, + "seconds": 3.95, + "tok_s": 8.1, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + { + "tokens": 13, + "seconds": 1.73, + "tok_s": 7.52, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + { + "tokens": 15, + "seconds": 1.92, + "tok_s": 7.81, + "ssd_mb": 3079.7, + "ssd_mb_per_token": 205.3 + } + ], + "mean_decode_tok_s": 7.757499999999999, + "mean_ssd_mb_per_token": 205.375, + "mean_hot_rate_pct": 9.577835, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 2, + "status": "OK", + "wall_s": 34.85, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.548, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 907.9, + "peak_anon_mb": 423.5, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.61, + "tok_s": 6.85, + "ssd_mb": 2348.7, + "ssd_mb_per_token": 213.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8807, + "expert_hits": 529, + "expert_misses": 8278, + "hit_rate": 6.00659, + "ssb_demand_bytes": 11849287680, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11849287680, + "ssd_bytes_per_token": 71381300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8052, + "demand_wait_us": 14810009, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 952025088, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 39, + "miss": 371, + "hit_rate": 9.5122, + "demand_bytes": 492355584, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 44, + "miss": 342, + "hit_rate": 11.399, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 33, + "miss": 373, + "hit_rate": 8.12808, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 302, + "hit": 0, + "miss": 302, + "hit_rate": 0, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r2_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 908.9, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.9, + "tok_s": 8.21, + "ssd_mb": 6275.9, + "ssd_mb_per_token": 196.1 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1599, + "expert_misses": 8998, + "hit_rate": 15.0892, + "ssb_demand_bytes": 12977952768, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12977952768, + "ssd_bytes_per_token": 76340900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8773, + "demand_wait_us": 16035629, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953077760, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 127, + "miss": 341, + "hit_rate": 27.1368, + "demand_bytes": 452542464, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 116, + "miss": 360, + "hit_rate": 24.3697, + "demand_bytes": 477757440, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 163, + "miss": 281, + "hit_rate": 36.7117, + "demand_bytes": 372916224, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r2_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.521, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.2, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.69, + "tok_s": 7.68, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9149, + "expert_hits": 665, + "expert_misses": 8484, + "hit_rate": 7.26855, + "ssb_demand_bytes": 12155111424, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12155111424, + "ssd_bytes_per_token": 69857000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8258, + "demand_wait_us": 15198503, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 960745472, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 52, + "miss": 363, + "hit_rate": 12.5301, + "demand_bytes": 481738752, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 304, + "hit": 0, + "miss": 304, + "hit_rate": 0, + "demand_bytes": 465076224, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r2_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 909.1, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 2.07, + "tok_s": 7.26, + "ssd_mb": 3084.8, + "ssd_mb_per_token": 205.7 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 743, + "expert_misses": 6702, + "hit_rate": 9.97985, + "ssb_demand_bytes": 9630646272, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9630646272, + "ssd_bytes_per_token": 73516400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6477, + "demand_wait_us": 11965604, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953282560, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 57, + "miss": 296, + "hit_rate": 16.1473, + "demand_bytes": 392822784, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 57, + "miss": 284, + "hit_rate": 16.7155, + "demand_bytes": 376897536, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 58, + "miss": 239, + "hit_rate": 19.5286, + "demand_bytes": 317177856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 63, + "miss": 264, + "hit_rate": 19.2661, + "demand_bytes": 350355456, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 0, + "miss": 274, + "hit_rate": 0, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r2_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.2, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.61, + "tok_s": 6.85, + "ssd_mb": 2348.7, + "ssd_mb_per_token": 213.5 + }, + { + "tokens": 32, + "seconds": 3.9, + "tok_s": 8.21, + "ssd_mb": 6275.9, + "ssd_mb_per_token": 196.1 + }, + { + "tokens": 13, + "seconds": 1.69, + "tok_s": 7.68, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + { + "tokens": 15, + "seconds": 2.07, + "tok_s": 7.26, + "ssd_mb": 3084.8, + "ssd_mb_per_token": 205.7 + } + ], + "mean_decode_tok_s": 7.5, + "mean_ssd_mb_per_token": 205.35, + "mean_hot_rate_pct": 9.5860475, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 3, + "status": "OK", + "wall_s": 34.26, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.566, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.2, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.46, + "tok_s": 7.55, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 526, + "expert_misses": 8295, + "hit_rate": 5.96304, + "ssb_demand_bytes": 11874686976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11874686976, + "ssd_bytes_per_token": 71534300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8070, + "demand_wait_us": 14599767, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954421248, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 44, + "miss": 342, + "hit_rate": 11.399, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 32, + "miss": 374, + "hit_rate": 7.88177, + "demand_bytes": 496336896, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r3_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 908.4, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 4.03, + "tok_s": 7.95, + "ssd_mb": 6273.3, + "ssd_mb_per_token": 196.0 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1601, + "expert_misses": 8996, + "hit_rate": 15.108, + "ssb_demand_bytes": 12975298560, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12975298560, + "ssd_bytes_per_token": 76325300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8771, + "demand_wait_us": 16007273, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 952504320, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 118, + "miss": 358, + "hit_rate": 24.7899, + "demand_bytes": 475103232, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 164, + "miss": 280, + "hit_rate": 36.9369, + "demand_bytes": 371589120, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r3_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.525, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.6, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.64, + "tok_s": 7.91, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9185, + "expert_hits": 665, + "expert_misses": 8520, + "hit_rate": 7.24007, + "ssb_demand_bytes": 12210186240, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12210186240, + "ssd_bytes_per_token": 70173500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8295, + "demand_wait_us": 15122207, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 961134592, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 51, + "miss": 329, + "hit_rate": 13.4211, + "demand_bytes": 436617216, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 340, + "hit": 0, + "miss": 340, + "hit_rate": 0, + "demand_bytes": 520151040, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r3_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 913.6, + "peak_anon_mb": 429.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.82, + "tok_s": 8.24, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7444, + "expert_hits": 740, + "expert_misses": 6704, + "hit_rate": 9.94089, + "ssb_demand_bytes": 9633097728, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9633097728, + "ssd_bytes_per_token": 73535100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6479, + "demand_wait_us": 11821003, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 958009344, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 56, + "miss": 297, + "hit_rate": 15.864, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 56, + "miss": 285, + "hit_rate": 16.4223, + "demand_bytes": 378224640, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 58, + "miss": 257, + "hit_rate": 18.4127, + "demand_bytes": 341065728, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 0, + "miss": 273, + "hit_rate": 0, + "demand_bytes": 417650688, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r3_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.6, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.46, + "tok_s": 7.55, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + { + "tokens": 32, + "seconds": 4.03, + "tok_s": 7.95, + "ssd_mb": 6273.3, + "ssd_mb_per_token": 196.0 + }, + { + "tokens": 13, + "seconds": 1.64, + "tok_s": 7.91, + "ssd_mb": 2679.0, + "ssd_mb_per_token": 206.1 + }, + { + "tokens": 15, + "seconds": 1.82, + "tok_s": 8.24, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + } + ], + "mean_decode_tok_s": 7.9125, + "mean_ssd_mb_per_token": 205.475, + "mean_hot_rate_pct": 9.563, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 4, + "status": "OK", + "wall_s": 34.68, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.588, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.3, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.43, + "tok_s": 7.72, + "ssd_mb": 2355.1, + "ssd_mb_per_token": 214.1 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8820, + "expert_hits": 524, + "expert_misses": 8296, + "hit_rate": 5.94104, + "ssb_demand_bytes": 11875811328, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11875811328, + "ssd_bytes_per_token": 71541000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8071, + "demand_wait_us": 14474172, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954535936, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 42, + "miss": 344, + "hit_rate": 10.8808, + "demand_bytes": 456523776, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 33, + "miss": 373, + "hit_rate": 8.12808, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 44, + "miss": 308, + "hit_rate": 12.5, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r4_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 908.7, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.86, + "tok_s": 8.3, + "ssd_mb": 6279.7, + "ssd_mb_per_token": 196.2 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1596, + "expert_misses": 9001, + "hit_rate": 15.0609, + "ssb_demand_bytes": 12981934080, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12981934080, + "ssd_bytes_per_token": 76364300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8776, + "demand_wait_us": 15916850, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 952819712, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 128, + "miss": 326, + "hit_rate": 28.1938, + "demand_bytes": 432635904, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 117, + "miss": 359, + "hit_rate": 24.5798, + "demand_bytes": 476430336, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 162, + "miss": 282, + "hit_rate": 36.4865, + "demand_bytes": 374243328, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r4_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.541, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 917.1, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.67, + "tok_s": 7.81, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9202, + "expert_hits": 666, + "expert_misses": 8536, + "hit_rate": 7.23756, + "ssb_demand_bytes": 12234866688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12234866688, + "ssd_bytes_per_token": 70315300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8310, + "demand_wait_us": 14931210, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 961609728, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 357, + "hit": 0, + "miss": 357, + "hit_rate": 0, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r4_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 913.9, + "peak_anon_mb": 429.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 2.05, + "tok_s": 7.33, + "ssd_mb": 3087.4, + "ssd_mb_per_token": 205.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7468, + "expert_hits": 741, + "expert_misses": 6727, + "hit_rate": 9.92234, + "ssb_demand_bytes": 9668487168, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9668487168, + "ssd_bytes_per_token": 73805200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6502, + "demand_wait_us": 11930050, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 958283776, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 57, + "miss": 296, + "hit_rate": 16.1473, + "demand_bytes": 392822784, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 56, + "miss": 285, + "hit_rate": 16.4223, + "demand_bytes": 378224640, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 63, + "miss": 264, + "hit_rate": 19.2661, + "demand_bytes": 350355456, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 297, + "hit": 0, + "miss": 297, + "hit_rate": 0, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r4_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 917.1, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.43, + "tok_s": 7.72, + "ssd_mb": 2355.1, + "ssd_mb_per_token": 214.1 + }, + { + "tokens": 32, + "seconds": 3.86, + "tok_s": 8.3, + "ssd_mb": 6279.7, + "ssd_mb_per_token": 196.2 + }, + { + "tokens": 13, + "seconds": 1.67, + "tok_s": 7.81, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + { + "tokens": 15, + "seconds": 2.05, + "tok_s": 7.33, + "ssd_mb": 3087.4, + "ssd_mb_per_token": 205.8 + } + ], + "mean_decode_tok_s": 7.79, + "mean_ssd_mb_per_token": 205.525, + "mean_hot_rate_pct": 9.54046, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 5, + "status": "OK", + "wall_s": 34.57, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.568, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.8, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.56, + "tok_s": 7.07, + "ssd_mb": 2353.8, + "ssd_mb_per_token": 214.0 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 525, + "expert_misses": 8296, + "hit_rate": 5.95171, + "ssb_demand_bytes": 11876014080, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11876014080, + "ssd_bytes_per_token": 71542300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8071, + "demand_wait_us": 14694881, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 955015168, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 43, + "miss": 343, + "hit_rate": 11.1399, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 33, + "miss": 373, + "hit_rate": 8.12808, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 44, + "miss": 318, + "hit_rate": 12.1547, + "demand_bytes": 422019072, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r5_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.9, + "peak_anon_mb": 427.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 4.04, + "tok_s": 7.92, + "ssd_mb": 6268.2, + "ssd_mb_per_token": 195.9 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10568, + "expert_hits": 1605, + "expert_misses": 8963, + "hit_rate": 15.1874, + "ssb_demand_bytes": 12925624320, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12925624320, + "ssd_bytes_per_token": 76033100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8737, + "demand_wait_us": 15737980, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 954060800, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 137, + "miss": 334, + "hit_rate": 29.087, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 144, + "miss": 302, + "hit_rate": 32.287, + "demand_bytes": 400785408, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 119, + "miss": 357, + "hit_rate": 25, + "demand_bytes": 473776128, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 164, + "miss": 280, + "hit_rate": 36.9369, + "demand_bytes": 371589120, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 154, + "miss": 288, + "hit_rate": 34.8416, + "demand_bytes": 382205952, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 414, + "hit": 0, + "miss": 414, + "hit_rate": 0, + "demand_bytes": 633360384, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r5_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.521, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.4, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.6, + "tok_s": 8.13, + "ssd_mb": 2680.3, + "ssd_mb_per_token": 206.2 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9207, + "expert_hits": 664, + "expert_misses": 8543, + "hit_rate": 7.2119, + "ssb_demand_bytes": 12245170176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12245170176, + "ssd_bytes_per_token": 70374500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8318, + "demand_wait_us": 14729897, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 960892928, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 51, + "miss": 329, + "hit_rate": 13.4211, + "demand_bytes": 436617216, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 52, + "miss": 363, + "hit_rate": 12.5301, + "demand_bytes": 481738752, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 362, + "hit": 0, + "miss": 362, + "hit_rate": 0, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r5_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 913.7, + "peak_anon_mb": 429.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.96, + "tok_s": 7.65, + "ssd_mb": 3087.3, + "ssd_mb_per_token": 205.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7468, + "expert_hits": 741, + "expert_misses": 6727, + "hit_rate": 9.92234, + "ssb_demand_bytes": 9668487168, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9668487168, + "ssd_bytes_per_token": 73805200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6502, + "demand_wait_us": 11921071, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 958111744, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 57, + "miss": 296, + "hit_rate": 16.1473, + "demand_bytes": 392822784, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 57, + "miss": 284, + "hit_rate": 16.7155, + "demand_bytes": 376897536, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 297, + "hit": 0, + "miss": 297, + "hit_rate": 0, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r5_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.4, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.56, + "tok_s": 7.07, + "ssd_mb": 2353.8, + "ssd_mb_per_token": 214.0 + }, + { + "tokens": 32, + "seconds": 4.04, + "tok_s": 7.92, + "ssd_mb": 6268.2, + "ssd_mb_per_token": 195.9 + }, + { + "tokens": 13, + "seconds": 1.6, + "tok_s": 8.13, + "ssd_mb": 2680.3, + "ssd_mb_per_token": 206.2 + }, + { + "tokens": 15, + "seconds": 1.96, + "tok_s": 7.65, + "ssd_mb": 3087.3, + "ssd_mb_per_token": 205.8 + } + ], + "mean_decode_tok_s": 7.692500000000001, + "mean_ssd_mb_per_token": 205.475, + "mean_hot_rate_pct": 9.5683375, + "max_expert_requests": 10568 + }, + { + "level_mb": 1024, + "repeat": 6, + "status": "OK", + "wall_s": 34.76, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.559, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 915.5, + "peak_anon_mb": 430.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.44, + "tok_s": 7.64, + "ssd_mb": 2357.6, + "ssd_mb_per_token": 214.3 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8848, + "expert_hits": 522, + "expert_misses": 8326, + "hit_rate": 5.89964, + "ssb_demand_bytes": 11921301504, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11921301504, + "ssd_bytes_per_token": 71815100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8101, + "demand_wait_us": 14680769, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 959926272, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 42, + "miss": 344, + "hit_rate": 10.8808, + "demand_bytes": 456523776, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 31, + "miss": 375, + "hit_rate": 7.63547, + "demand_bytes": 497664000, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 44, + "miss": 318, + "hit_rate": 12.1547, + "demand_bytes": 422019072, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 343, + "hit": 0, + "miss": 343, + "hit_rate": 0, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r6_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.0, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.89, + "tok_s": 8.23, + "ssd_mb": 6273.3, + "ssd_mb_per_token": 196.0 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1601, + "expert_misses": 8996, + "hit_rate": 15.108, + "ssb_demand_bytes": 12975298560, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12975298560, + "ssd_bytes_per_token": 76325300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8771, + "demand_wait_us": 16033369, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953163776, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 118, + "miss": 358, + "hit_rate": 24.7899, + "demand_bytes": 475103232, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 164, + "miss": 280, + "hit_rate": 36.9369, + "demand_bytes": 371589120, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r6_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.54, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.7, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.7, + "tok_s": 7.67, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9185, + "expert_hits": 666, + "expert_misses": 8519, + "hit_rate": 7.25095, + "ssb_demand_bytes": 12208859136, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12208859136, + "ssd_bytes_per_token": 70165900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8294, + "demand_wait_us": 14995429, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 961204224, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 340, + "hit": 0, + "miss": 340, + "hit_rate": 0, + "demand_bytes": 520151040, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r6_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 915.2, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 2.09, + "tok_s": 7.18, + "ssd_mb": 3081.0, + "ssd_mb_per_token": 205.4 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7427, + "expert_hits": 746, + "expert_misses": 6681, + "hit_rate": 10.0444, + "ssb_demand_bytes": 9599127552, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9599127552, + "ssd_bytes_per_token": 73275800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6455, + "demand_wait_us": 11945786, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 959709184, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 51, + "miss": 293, + "hit_rate": 14.8256, + "demand_bytes": 388841472, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 58, + "miss": 295, + "hit_rate": 16.4306, + "demand_bytes": 391495680, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 67, + "miss": 257, + "hit_rate": 20.679, + "demand_bytes": 341065728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 56, + "miss": 285, + "hit_rate": 16.4223, + "demand_bytes": 378224640, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 59, + "miss": 238, + "hit_rate": 19.8653, + "demand_bytes": 315850752, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 61, + "miss": 253, + "hit_rate": 19.4268, + "demand_bytes": 335757312, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 256, + "hit": 0, + "miss": 256, + "hit_rate": 0, + "demand_bytes": 391643136, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r6_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.7, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.44, + "tok_s": 7.64, + "ssd_mb": 2357.6, + "ssd_mb_per_token": 214.3 + }, + { + "tokens": 32, + "seconds": 3.89, + "tok_s": 8.23, + "ssd_mb": 6273.3, + "ssd_mb_per_token": 196.0 + }, + { + "tokens": 13, + "seconds": 1.7, + "tok_s": 7.67, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + { + "tokens": 15, + "seconds": 2.09, + "tok_s": 7.18, + "ssd_mb": 3081.0, + "ssd_mb_per_token": 205.4 + } + ], + "mean_decode_tok_s": 7.68, + "mean_ssd_mb_per_token": 205.425, + "mean_hot_rate_pct": 9.5757475, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 7, + "status": "OK", + "wall_s": 34.72, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.564, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 911.6, + "peak_anon_mb": 427.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.48, + "tok_s": 7.44, + "ssd_mb": 2346.1, + "ssd_mb_per_token": 213.3 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 531, + "expert_misses": 8290, + "hit_rate": 6.01973, + "ssb_demand_bytes": 11868051456, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11868051456, + "ssd_bytes_per_token": 71494300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8064, + "demand_wait_us": 14435108, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 955830272, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 38, + "miss": 372, + "hit_rate": 9.26829, + "demand_bytes": 493682688, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 46, + "miss": 340, + "hit_rate": 11.9171, + "demand_bytes": 451215360, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 34, + "miss": 372, + "hit_rate": 8.37438, + "demand_bytes": 493682688, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r7_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.3, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 4.01, + "tok_s": 7.97, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1598, + "expert_misses": 8999, + "hit_rate": 15.0797, + "ssb_demand_bytes": 12979279872, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12979279872, + "ssd_bytes_per_token": 76348700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8774, + "demand_wait_us": 15671748, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953425920, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 129, + "miss": 325, + "hit_rate": 28.4141, + "demand_bytes": 431308800, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 156, + "miss": 279, + "hit_rate": 35.8621, + "demand_bytes": 370262016, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 118, + "miss": 358, + "hit_rate": 24.7899, + "demand_bytes": 475103232, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 163, + "miss": 281, + "hit_rate": 36.7117, + "demand_bytes": 372916224, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r7_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.573, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.3, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.68, + "tok_s": 7.75, + "ssd_mb": 2676.4, + "ssd_mb_per_token": 205.9 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9197, + "expert_hits": 667, + "expert_misses": 8530, + "hit_rate": 7.25236, + "ssb_demand_bytes": 12225890304, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12225890304, + "ssd_bytes_per_token": 70263700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8305, + "demand_wait_us": 14857215, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 960843776, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 67, + "miss": 320, + "hit_rate": 17.3127, + "demand_bytes": 424673280, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 352, + "hit": 0, + "miss": 352, + "hit_rate": 0, + "demand_bytes": 538509312, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r7_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 913.6, + "peak_anon_mb": 429.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.97, + "tok_s": 7.63, + "ssd_mb": 3086.1, + "ssd_mb_per_token": 205.7 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7444, + "expert_hits": 742, + "expert_misses": 6702, + "hit_rate": 9.96776, + "ssb_demand_bytes": 9630443520, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9630443520, + "ssd_bytes_per_token": 73514800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6477, + "demand_wait_us": 11893030, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 957931520, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 51, + "miss": 293, + "hit_rate": 14.8256, + "demand_bytes": 388841472, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 57, + "miss": 296, + "hit_rate": 16.1473, + "demand_bytes": 392822784, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 57, + "miss": 284, + "hit_rate": 16.7155, + "demand_bytes": 376897536, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 0, + "miss": 273, + "hit_rate": 0, + "demand_bytes": 417650688, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r7_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.3, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.48, + "tok_s": 7.44, + "ssd_mb": 2346.1, + "ssd_mb_per_token": 213.3 + }, + { + "tokens": 32, + "seconds": 4.01, + "tok_s": 7.97, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + { + "tokens": 13, + "seconds": 1.68, + "tok_s": 7.75, + "ssd_mb": 2676.4, + "ssd_mb_per_token": 205.9 + }, + { + "tokens": 15, + "seconds": 1.97, + "tok_s": 7.63, + "ssd_mb": 3086.1, + "ssd_mb_per_token": 205.7 + } + ], + "mean_decode_tok_s": 7.6975, + "mean_ssd_mb_per_token": 205.275, + "mean_hot_rate_pct": 9.5798875, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 8, + "status": "OK", + "wall_s": 35.08, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.545, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 909.8, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.49, + "tok_s": 7.4, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 526, + "expert_misses": 8295, + "hit_rate": 5.96304, + "ssb_demand_bytes": 11874686976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11874686976, + "ssd_bytes_per_token": 71534300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8070, + "demand_wait_us": 14860121, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954028032, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 44, + "miss": 342, + "hit_rate": 11.399, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 32, + "miss": 374, + "hit_rate": 7.88177, + "demand_bytes": 496336896, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r8_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.8, + "peak_anon_mb": 427.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 4.11, + "tok_s": 7.79, + "ssd_mb": 6268.2, + "ssd_mb_per_token": 195.9 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10568, + "expert_hits": 1605, + "expert_misses": 8963, + "hit_rate": 15.1874, + "ssb_demand_bytes": 12925624320, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12925624320, + "ssd_bytes_per_token": 76033100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8737, + "demand_wait_us": 16129697, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 953995264, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 112, + "miss": 341, + "hit_rate": 24.7241, + "demand_bytes": 452542464, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 144, + "miss": 302, + "hit_rate": 32.287, + "demand_bytes": 400785408, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 118, + "miss": 358, + "hit_rate": 24.7899, + "demand_bytes": 475103232, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 165, + "miss": 279, + "hit_rate": 37.1622, + "demand_bytes": 370262016, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 154, + "miss": 288, + "hit_rate": 34.8416, + "demand_bytes": 382205952, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 414, + "hit": 0, + "miss": 414, + "hit_rate": 0, + "demand_bytes": 633360384, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r8_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.536, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.1, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.7, + "tok_s": 7.67, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9185, + "expert_hits": 666, + "expert_misses": 8519, + "hit_rate": 7.25095, + "ssb_demand_bytes": 12208859136, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12208859136, + "ssd_bytes_per_token": 70165900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8294, + "demand_wait_us": 15104687, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 960651264, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 340, + "hit": 0, + "miss": 340, + "hit_rate": 0, + "demand_bytes": 520151040, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r8_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 909.3, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 2.05, + "tok_s": 7.32, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7426, + "expert_hits": 740, + "expert_misses": 6686, + "hit_rate": 9.96499, + "ssb_demand_bytes": 9605560320, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9605560320, + "ssd_bytes_per_token": 73324900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6461, + "demand_wait_us": 12215647, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 953454592, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 56, + "miss": 297, + "hit_rate": 15.864, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 57, + "miss": 284, + "hit_rate": 16.7155, + "demand_bytes": 376897536, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 255, + "hit": 0, + "miss": 255, + "hit_rate": 0, + "demand_bytes": 390113280, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r8_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.1, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.49, + "tok_s": 7.4, + "ssd_mb": 2352.5, + "ssd_mb_per_token": 213.9 + }, + { + "tokens": 32, + "seconds": 4.11, + "tok_s": 7.79, + "ssd_mb": 6268.2, + "ssd_mb_per_token": 195.9 + }, + { + "tokens": 13, + "seconds": 1.7, + "tok_s": 7.67, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + { + "tokens": 15, + "seconds": 2.05, + "tok_s": 7.32, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + } + ], + "mean_decode_tok_s": 7.545, + "mean_ssd_mb_per_token": 205.425, + "mean_hot_rate_pct": 9.591595, + "max_expert_requests": 10568 + }, + { + "level_mb": 1024, + "repeat": 9, + "status": "OK", + "wall_s": 34.99, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.91, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.5, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.55, + "tok_s": 7.1, + "ssd_mb": 2356.4, + "ssd_mb_per_token": 214.2 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 523, + "expert_misses": 8298, + "hit_rate": 5.92903, + "ssb_demand_bytes": 11878668288, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11878668288, + "ssd_bytes_per_token": 71558200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8073, + "demand_wait_us": 14530302, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954716160, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 44, + "miss": 342, + "hit_rate": 11.399, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 30, + "miss": 376, + "hit_rate": 7.38916, + "demand_bytes": 498991104, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 44, + "miss": 318, + "hit_rate": 12.1547, + "demand_bytes": 422019072, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 45, + "miss": 307, + "hit_rate": 12.7841, + "demand_bytes": 407420928, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r9_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 908.4, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 4.21, + "tok_s": 7.61, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10597, + "expert_hits": 1598, + "expert_misses": 8999, + "hit_rate": 15.0797, + "ssb_demand_bytes": 12979279872, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12979279872, + "ssd_bytes_per_token": 76348700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8774, + "demand_wait_us": 16362790, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 952545280, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 126, + "miss": 342, + "hit_rate": 26.9231, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 129, + "miss": 325, + "hit_rate": 28.4141, + "demand_bytes": 431308800, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 143, + "miss": 303, + "hit_rate": 32.0628, + "demand_bytes": 402112512, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 117, + "miss": 359, + "hit_rate": 24.5798, + "demand_bytes": 476430336, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 163, + "miss": 281, + "hit_rate": 36.7117, + "demand_bytes": 372916224, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 443, + "hit": 0, + "miss": 443, + "hit_rate": 0, + "demand_bytes": 677726208, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r9_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.541, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 916.8, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.64, + "tok_s": 7.92, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9184, + "expert_hits": 666, + "expert_misses": 8518, + "hit_rate": 7.25174, + "ssb_demand_bytes": 12207329280, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12207329280, + "ssd_bytes_per_token": 70157100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8293, + "demand_wait_us": 14999562, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 961298432, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 66, + "miss": 321, + "hit_rate": 17.0543, + "demand_bytes": 426000384, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 339, + "hit": 0, + "miss": 339, + "hit_rate": 0, + "demand_bytes": 518621184, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r9_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 909.8, + "peak_anon_mb": 425.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.87, + "tok_s": 8.04, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 740, + "expert_misses": 6705, + "hit_rate": 9.93956, + "ssb_demand_bytes": 9634627584, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9634627584, + "ssd_bytes_per_token": 73546800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6480, + "demand_wait_us": 11978134, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954007552, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 50, + "miss": 294, + "hit_rate": 14.5349, + "demand_bytes": 390168576, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 57, + "miss": 296, + "hit_rate": 16.1473, + "demand_bytes": 392822784, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 56, + "miss": 285, + "hit_rate": 16.4223, + "demand_bytes": 378224640, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 0, + "miss": 274, + "hit_rate": 0, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r9_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 916.8, + "peak_anon_mb": 431.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.55, + "tok_s": 7.1, + "ssd_mb": 2356.4, + "ssd_mb_per_token": 214.2 + }, + { + "tokens": 32, + "seconds": 4.21, + "tok_s": 7.61, + "ssd_mb": 6277.1, + "ssd_mb_per_token": 196.2 + }, + { + "tokens": 13, + "seconds": 1.64, + "tok_s": 7.92, + "ssd_mb": 2677.7, + "ssd_mb_per_token": 206.0 + }, + { + "tokens": 15, + "seconds": 1.87, + "tok_s": 8.04, + "ssd_mb": 3088.6, + "ssd_mb_per_token": 205.9 + } + ], + "mean_decode_tok_s": 7.6674999999999995, + "mean_ssd_mb_per_token": 205.575, + "mean_hot_rate_pct": 9.5500075, + "max_expert_requests": 10597 + }, + { + "level_mb": 1024, + "repeat": 10, + "status": "OK", + "wall_s": 34.53, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.925, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 910.1, + "peak_anon_mb": 426.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.43, + "tok_s": 7.68, + "ssd_mb": 2356.4, + "ssd_mb_per_token": 214.2 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 523, + "expert_misses": 8298, + "hit_rate": 5.92903, + "ssb_demand_bytes": 11878668288, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 11878668288, + "ssd_bytes_per_token": 71558200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8073, + "demand_wait_us": 14564053, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 954310656, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 0, + "miss": 451, + "hit_rate": 0, + "demand_bytes": 689965056, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 0, + "miss": 444, + "hit_rate": 0, + "demand_bytes": 679256064, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 40, + "miss": 391, + "hit_rate": 9.28074, + "demand_bytes": 518897664, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 37, + "miss": 373, + "hit_rate": 9.02439, + "demand_bytes": 495009792, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 0, + "miss": 416, + "hit_rate": 0, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 43, + "miss": 343, + "hit_rate": 11.1399, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 31, + "miss": 375, + "hit_rate": 7.63547, + "demand_bytes": 497664000, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 45, + "miss": 317, + "hit_rate": 12.4309, + "demand_bytes": 420691968, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 48, + "miss": 289, + "hit_rate": 14.2433, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 44, + "miss": 308, + "hit_rate": 12.5, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 44, + "miss": 310, + "hit_rate": 12.4294, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 0, + "miss": 354, + "hit_rate": 0, + "demand_bytes": 541569024, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 42, + "miss": 330, + "hit_rate": 11.2903, + "demand_bytes": 437944320, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 41, + "miss": 349, + "hit_rate": 10.5128, + "demand_bytes": 463159296, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 0, + "miss": 386, + "hit_rate": 0, + "demand_bytes": 590524416, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 48, + "miss": 343, + "hit_rate": 12.2762, + "demand_bytes": 455196672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 60, + "miss": 297, + "hit_rate": 16.8067, + "demand_bytes": 394149888, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 0, + "miss": 401, + "hit_rate": 0, + "demand_bytes": 613472256, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 0, + "miss": 389, + "hit_rate": 0, + "demand_bytes": 595113984, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 0, + "miss": 390, + "hit_rate": 0, + "demand_bytes": 596643840, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 0, + "miss": 316, + "hit_rate": 0, + "demand_bytes": 483434496, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r10_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 909.5, + "peak_anon_mb": 427.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.78, + "tok_s": 8.47, + "ssd_mb": 6269.5, + "ssd_mb_per_token": 195.9 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10568, + "expert_hits": 1604, + "expert_misses": 8964, + "hit_rate": 15.1779, + "ssb_demand_bytes": 12926951424, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12926951424, + "ssd_bytes_per_token": 76040900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8738, + "demand_wait_us": 15880954, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 953692160, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 0, + "miss": 535, + "hit_rate": 0, + "demand_bytes": 818472960, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 0, + "miss": 477, + "hit_rate": 0, + "demand_bytes": 729741312, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 113, + "miss": 370, + "hit_rate": 23.3954, + "demand_bytes": 491028480, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 127, + "miss": 341, + "hit_rate": 27.1368, + "demand_bytes": 452542464, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 0, + "miss": 453, + "hit_rate": 0, + "demand_bytes": 693024768, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 111, + "miss": 342, + "hit_rate": 24.5033, + "demand_bytes": 453869568, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 136, + "miss": 335, + "hit_rate": 28.8747, + "demand_bytes": 444579840, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 0, + "miss": 474, + "hit_rate": 0, + "demand_bytes": 725151744, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 130, + "miss": 324, + "hit_rate": 28.6344, + "demand_bytes": 429981696, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 157, + "miss": 278, + "hit_rate": 36.092, + "demand_bytes": 368934912, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 0, + "miss": 445, + "hit_rate": 0, + "demand_bytes": 680785920, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 116, + "miss": 334, + "hit_rate": 25.7778, + "demand_bytes": 443252736, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 144, + "miss": 302, + "hit_rate": 32.287, + "demand_bytes": 400785408, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 0, + "miss": 452, + "hit_rate": 0, + "demand_bytes": 691494912, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 134, + "miss": 328, + "hit_rate": 29.0043, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 118, + "miss": 358, + "hit_rate": 24.7899, + "demand_bytes": 475103232, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 0, + "miss": 448, + "hit_rate": 0, + "demand_bytes": 685375488, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 165, + "miss": 279, + "hit_rate": 37.1622, + "demand_bytes": 370262016, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 153, + "miss": 289, + "hit_rate": 34.6154, + "demand_bytes": 383533056, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 0, + "miss": 459, + "hit_rate": 0, + "demand_bytes": 702203904, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 0, + "miss": 457, + "hit_rate": 0, + "demand_bytes": 699144192, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 0, + "miss": 470, + "hit_rate": 0, + "demand_bytes": 719032320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 414, + "hit": 0, + "miss": 414, + "hit_rate": 0, + "demand_bytes": 633360384, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r10_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.516, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 906.4, + "peak_anon_mb": 422.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.76, + "tok_s": 7.41, + "ssd_mb": 2676.4, + "ssd_mb_per_token": 205.9 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 667, + "expert_misses": 8451, + "hit_rate": 7.3152, + "ssb_demand_bytes": 12105031680, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 12105031680, + "ssd_bytes_per_token": 69569100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 8225, + "demand_wait_us": 14931039, + "prefetch_wait_us": 0, + "arena_bytes": 311394304, + "arena_used_bytes": 311394304, + "rss_bytes": 950460416, + "n_slots": 226, + "n_small_slots": 170, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 0, + "miss": 476, + "hit_rate": 0, + "demand_bytes": 728211456, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 0, + "miss": 462, + "hit_rate": 0, + "demand_bytes": 706793472, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 48, + "miss": 403, + "hit_rate": 10.643, + "demand_bytes": 534822912, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 45, + "miss": 385, + "hit_rate": 10.4651, + "demand_bytes": 510935040, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 0, + "miss": 428, + "hit_rate": 0, + "demand_bytes": 654778368, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 54, + "miss": 346, + "hit_rate": 13.5, + "demand_bytes": 459177984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 53, + "miss": 361, + "hit_rate": 12.8019, + "demand_bytes": 479084544, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 0, + "miss": 430, + "hit_rate": 0, + "demand_bytes": 657838080, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 55, + "miss": 310, + "hit_rate": 15.0685, + "demand_bytes": 411402240, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 61, + "miss": 287, + "hit_rate": 17.5287, + "demand_bytes": 380878848, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 0, + "miss": 332, + "hit_rate": 0, + "demand_bytes": 507912192, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 54, + "miss": 308, + "hit_rate": 14.9171, + "demand_bytes": 408748032, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 56, + "miss": 315, + "hit_rate": 15.0943, + "demand_bytes": 418037760, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 0, + "miss": 353, + "hit_rate": 0, + "demand_bytes": 540039168, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 52, + "miss": 328, + "hit_rate": 13.6842, + "demand_bytes": 435290112, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 53, + "miss": 362, + "hit_rate": 12.7711, + "demand_bytes": 480411648, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 0, + "miss": 411, + "hit_rate": 0, + "demand_bytes": 628770816, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 69, + "miss": 338, + "hit_rate": 16.9533, + "demand_bytes": 448561152, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 67, + "miss": 320, + "hit_rate": 17.3127, + "demand_bytes": 424673280, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 0, + "miss": 406, + "hit_rate": 0, + "demand_bytes": 621121536, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 0, + "miss": 412, + "hit_rate": 0, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 0, + "miss": 405, + "hit_rate": 0, + "demand_bytes": 619591680, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 0, + "miss": 273, + "hit_rate": 0, + "demand_bytes": 417650688, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r10_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 915.9, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.94, + "tok_s": 7.71, + "ssd_mb": 3083.5, + "ssd_mb_per_token": 205.6 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 744, + "expert_misses": 6727, + "hit_rate": 9.95851, + "ssb_demand_bytes": 9669095424, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 9669095424, + "ssd_bytes_per_token": 73809900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 6502, + "demand_wait_us": 11963427, + "prefetch_wait_us": 0, + "arena_bytes": 310067200, + "arena_used_bytes": 310067200, + "rss_bytes": 960385024, + "n_slots": 225, + "n_small_slots": 169, + "n_big_slots": 56, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 0, + "miss": 407, + "hit_rate": 0, + "demand_bytes": 622651392, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 0, + "miss": 367, + "hit_rate": 0, + "demand_bytes": 561457152, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 52, + "miss": 292, + "hit_rate": 15.1163, + "demand_bytes": 387514368, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 58, + "miss": 295, + "hit_rate": 16.4306, + "demand_bytes": 391495680, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 0, + "miss": 345, + "hit_rate": 0, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 66, + "miss": 258, + "hit_rate": 20.3704, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 57, + "miss": 284, + "hit_rate": 16.7155, + "demand_bytes": 376897536, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 0, + "miss": 348, + "hit_rate": 0, + "demand_bytes": 532389888, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 57, + "miss": 240, + "hit_rate": 19.1919, + "demand_bytes": 318504960, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 78, + "miss": 206, + "hit_rate": 27.4648, + "demand_bytes": 273383424, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 57, + "miss": 258, + "hit_rate": 18.0952, + "demand_bytes": 342392832, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 62, + "miss": 265, + "hit_rate": 18.9602, + "demand_bytes": 351682560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 0, + "miss": 318, + "hit_rate": 0, + "demand_bytes": 486494208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 60, + "miss": 254, + "hit_rate": 19.1083, + "demand_bytes": 337084416, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 53, + "miss": 270, + "hit_rate": 16.4087, + "demand_bytes": 358318080, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 0, + "miss": 324, + "hit_rate": 0, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 70, + "miss": 223, + "hit_rate": 23.8908, + "demand_bytes": 295944192, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 74, + "miss": 224, + "hit_rate": 24.8322, + "demand_bytes": 297271296, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 0, + "miss": 315, + "hit_rate": 0, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 0, + "miss": 323, + "hit_rate": 0, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 0, + "miss": 311, + "hit_rate": 0, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 0, + "miss": 300, + "hit_rate": 0, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_1024_r10_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 915.9, + "peak_anon_mb": 432.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.43, + "tok_s": 7.68, + "ssd_mb": 2356.4, + "ssd_mb_per_token": 214.2 + }, + { + "tokens": 32, + "seconds": 3.78, + "tok_s": 8.47, + "ssd_mb": 6269.5, + "ssd_mb_per_token": 195.9 + }, + { + "tokens": 13, + "seconds": 1.76, + "tok_s": 7.41, + "ssd_mb": 2676.4, + "ssd_mb_per_token": 205.9 + }, + { + "tokens": 15, + "seconds": 1.94, + "tok_s": 7.71, + "ssd_mb": 3083.5, + "ssd_mb_per_token": 205.6 + } + ], + "mean_decode_tok_s": 7.8175, + "mean_ssd_mb_per_token": 205.4, + "mean_hot_rate_pct": 9.59516, + "max_expert_requests": 10568 + }, + { + "level_mb": 2048, + "repeat": 1, + "status": "OK", + "wall_s": 33.19, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.938, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1899.0, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.94, + "tok_s": 11.68, + "ssd_mb": 789.2, + "ssd_mb_per_token": 71.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3391, + "expert_misses": 5421, + "hit_rate": 38.4816, + "ssb_demand_bytes": 7939547136, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7939547136, + "ssd_bytes_per_token": 47828600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4442, + "demand_wait_us": 9503869, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991217152, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 40, + "miss": 376, + "hit_rate": 9.61538, + "demand_bytes": 575225856, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 220, + "miss": 117, + "hit_rate": 65.2819, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 232, + "miss": 120, + "hit_rate": 65.9091, + "demand_bytes": 159252480, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r1_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.4, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.23, + "tok_s": 9.92, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5398, + "expert_misses": 5202, + "hit_rate": 50.9245, + "ssb_demand_bytes": 7587680256, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7587680256, + "ssd_bytes_per_token": 44633400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4223, + "demand_wait_us": 9184415, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1996869632, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 119, + "miss": 416, + "hit_rate": 22.243, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 273, + "miss": 195, + "hit_rate": 58.3333, + "demand_bytes": 258785280, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 166, + "miss": 280, + "hit_rate": 37.2197, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r1_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.494, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1898.5, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.13, + "tok_s": 11.47, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4069, + "expert_misses": 5101, + "hit_rate": 44.373, + "ssb_demand_bytes": 7515076608, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7515076608, + "ssd_bytes_per_token": 43190100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4122, + "demand_wait_us": 9059242, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1990746112, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 287, + "miss": 127, + "hit_rate": 69.3237, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 286, + "miss": 129, + "hit_rate": 68.9157, + "demand_bytes": 171196416, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 77, + "miss": 334, + "hit_rate": 18.7348, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r1_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1907.3, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.48, + "tok_s": 10.13, + "ssd_mb": 1274.5, + "ssd_mb_per_token": 85.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3051, + "expert_misses": 4420, + "hit_rate": 40.8379, + "ssb_demand_bytes": 6448508928, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6448508928, + "ssd_bytes_per_token": 49225300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3441, + "demand_wait_us": 7773044, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999933440, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 163, + "miss": 190, + "hit_rate": 46.1756, + "demand_bytes": 252149760, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 176, + "miss": 148, + "hit_rate": 54.321, + "demand_bytes": 196411392, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r1_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1907.3, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.94, + "tok_s": 11.68, + "ssd_mb": 789.2, + "ssd_mb_per_token": 71.7 + }, + { + "tokens": 32, + "seconds": 3.23, + "tok_s": 9.92, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 1.13, + "tok_s": 11.47, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + { + "tokens": 15, + "seconds": 1.48, + "tok_s": 10.13, + "ssd_mb": 1274.5, + "ssd_mb_per_token": 85.0 + } + ], + "mean_decode_tok_s": 10.8, + "mean_ssd_mb_per_token": 71.375, + "mean_hot_rate_pct": 43.65425, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 2, + "status": "OK", + "wall_s": 33.61, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.917, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1895.8, + "peak_anon_mb": 1411.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.1, + "tok_s": 10.03, + "ssd_mb": 786.4, + "ssd_mb_per_token": 71.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8755, + "expert_hits": 3394, + "expert_misses": 5361, + "hit_rate": 38.7664, + "ssb_demand_bytes": 7848161280, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7848161280, + "ssd_bytes_per_token": 47278100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4382, + "demand_wait_us": 9657461, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1987866624, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 247, + "miss": 184, + "hit_rate": 57.3086, + "demand_bytes": 244187136, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 266, + "miss": 144, + "hit_rate": 64.878, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 40, + "miss": 376, + "hit_rate": 9.61538, + "demand_bytes": 575225856, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 233, + "miss": 119, + "hit_rate": 66.1932, + "demand_bytes": 157925376, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 55, + "miss": 334, + "hit_rate": 14.1388, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 250, + "hit": 66, + "miss": 184, + "hit_rate": 26.4, + "demand_bytes": 281493504, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r2_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.3, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.11, + "tok_s": 10.27, + "ssd_mb": 2258.4, + "ssd_mb_per_token": 70.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5396, + "expert_misses": 5204, + "hit_rate": 50.9057, + "ssb_demand_bytes": 7590739968, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7590739968, + "ssd_bytes_per_token": 44651400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4225, + "demand_wait_us": 9257606, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1996804096, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 273, + "miss": 195, + "hit_rate": 58.3333, + "demand_bytes": 258785280, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 155, + "miss": 298, + "hit_rate": 34.2163, + "demand_bytes": 455897088, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 166, + "miss": 280, + "hit_rate": 37.2197, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r2_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.525, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1899.3, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.21, + "tok_s": 10.73, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4069, + "expert_misses": 5101, + "hit_rate": 44.373, + "ssb_demand_bytes": 7515076608, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7515076608, + "ssd_bytes_per_token": 43190100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4122, + "demand_wait_us": 9056811, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991524352, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 287, + "miss": 127, + "hit_rate": 69.3237, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 286, + "miss": 129, + "hit_rate": 68.9157, + "demand_bytes": 171196416, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 77, + "miss": 334, + "hit_rate": 18.7348, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r2_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1907.1, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.41, + "tok_s": 10.6, + "ssd_mb": 1275.5, + "ssd_mb_per_token": 85.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3046, + "expert_misses": 4425, + "hit_rate": 40.771, + "ssb_demand_bytes": 6454941696, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6454941696, + "ssd_bytes_per_token": 49274400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3446, + "demand_wait_us": 8042808, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999712256, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 63, + "miss": 344, + "hit_rate": 15.4791, + "demand_bytes": 526270464, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 172, + "miss": 152, + "hit_rate": 53.0864, + "demand_bytes": 201719808, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 208, + "miss": 133, + "hit_rate": 60.9971, + "demand_bytes": 176504832, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r2_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1907.1, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.1, + "tok_s": 10.03, + "ssd_mb": 786.4, + "ssd_mb_per_token": 71.5 + }, + { + "tokens": 32, + "seconds": 3.11, + "tok_s": 10.27, + "ssd_mb": 2258.4, + "ssd_mb_per_token": 70.6 + }, + { + "tokens": 13, + "seconds": 1.21, + "tok_s": 10.73, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + { + "tokens": 15, + "seconds": 1.41, + "tok_s": 10.6, + "ssd_mb": 1275.5, + "ssd_mb_per_token": 85.0 + } + ], + "mean_decode_tok_s": 10.407499999999999, + "mean_ssd_mb_per_token": 71.35, + "mean_hot_rate_pct": 43.704025, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 3, + "status": "OK", + "wall_s": 32.99, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.922, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1898.9, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.98, + "tok_s": 11.22, + "ssd_mb": 784.9, + "ssd_mb_per_token": 71.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8810, + "expert_hits": 3394, + "expert_misses": 5416, + "hit_rate": 38.5244, + "ssb_demand_bytes": 7932100608, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7932100608, + "ssd_bytes_per_token": 47783700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4437, + "demand_wait_us": 9836107, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991151616, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 232, + "miss": 120, + "hit_rate": 65.9091, + "demand_bytes": 159252480, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 55, + "miss": 334, + "hit_rate": 14.1388, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 305, + "hit": 66, + "miss": 239, + "hit_rate": 21.6393, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r3_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 1904.6, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.91, + "tok_s": 11.0, + "ssd_mb": 2227.9, + "ssd_mb_per_token": 69.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5419, + "expert_misses": 5181, + "hit_rate": 51.1226, + "ssb_demand_bytes": 7556972544, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7556972544, + "ssd_bytes_per_token": 44452800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4202, + "demand_wait_us": 9526373, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997135872, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 122, + "miss": 412, + "hit_rate": 22.8464, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 130, + "miss": 349, + "hit_rate": 27.1399, + "demand_bytes": 533919744, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 277, + "miss": 191, + "hit_rate": 59.188, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 157, + "miss": 296, + "hit_rate": 34.6578, + "demand_bytes": 452837376, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 284, + "miss": 167, + "hit_rate": 62.9712, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 308, + "miss": 164, + "hit_rate": 65.2542, + "demand_bytes": 217645056, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 143, + "miss": 330, + "hit_rate": 30.2326, + "demand_bytes": 504852480, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 306, + "miss": 149, + "hit_rate": 67.2527, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 304, + "miss": 127, + "hit_rate": 70.5336, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 170, + "miss": 274, + "hit_rate": 38.2883, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 306, + "miss": 141, + "hit_rate": 68.4564, + "demand_bytes": 187121664, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 307, + "miss": 139, + "hit_rate": 68.8341, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 174, + "miss": 276, + "hit_rate": 38.6667, + "demand_bytes": 422240256, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 337, + "miss": 130, + "hit_rate": 72.1627, + "demand_bytes": 172523520, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 326, + "miss": 146, + "hit_rate": 69.0678, + "demand_bytes": 193757184, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 175, + "miss": 274, + "hit_rate": 38.9755, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 318, + "miss": 129, + "hit_rate": 71.1409, + "demand_bytes": 171196416, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 317, + "miss": 125, + "hit_rate": 71.7195, + "demand_bytes": 165888000, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 175, + "miss": 283, + "hit_rate": 38.2096, + "demand_bytes": 432949248, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 179, + "miss": 281, + "hit_rate": 38.913, + "demand_bytes": 429889536, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 168, + "miss": 304, + "hit_rate": 35.5932, + "demand_bytes": 465076224, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 447, + "hit": 166, + "miss": 281, + "hit_rate": 37.1365, + "demand_bytes": 429889536, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r3_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.552, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1898.8, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.03, + "tok_s": 12.67, + "ssd_mb": 758.7, + "ssd_mb_per_token": 58.4 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4068, + "expert_misses": 5102, + "hit_rate": 44.3621, + "ssb_demand_bytes": 7516403712, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7516403712, + "ssd_bytes_per_token": 43197700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4123, + "demand_wait_us": 9241011, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991045120, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 286, + "miss": 128, + "hit_rate": 69.0821, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 286, + "miss": 129, + "hit_rate": 68.9157, + "demand_bytes": 171196416, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 77, + "miss": 334, + "hit_rate": 18.7348, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r3_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1906.8, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.39, + "tok_s": 10.8, + "ssd_mb": 1275.7, + "ssd_mb_per_token": 85.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3049, + "expert_misses": 4422, + "hit_rate": 40.8111, + "ssb_demand_bytes": 6451163136, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6451163136, + "ssd_bytes_per_token": 49245500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3443, + "demand_wait_us": 7751026, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999372288, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 163, + "miss": 190, + "hit_rate": 46.1756, + "demand_bytes": 252149760, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 175, + "miss": 149, + "hit_rate": 54.0123, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 208, + "miss": 133, + "hit_rate": 60.9971, + "demand_bytes": 176504832, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r3_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1906.8, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.98, + "tok_s": 11.22, + "ssd_mb": 784.9, + "ssd_mb_per_token": 71.4 + }, + { + "tokens": 32, + "seconds": 2.91, + "tok_s": 11.0, + "ssd_mb": 2227.9, + "ssd_mb_per_token": 69.6 + }, + { + "tokens": 13, + "seconds": 1.03, + "tok_s": 12.67, + "ssd_mb": 758.7, + "ssd_mb_per_token": 58.4 + }, + { + "tokens": 15, + "seconds": 1.39, + "tok_s": 10.8, + "ssd_mb": 1275.7, + "ssd_mb_per_token": 85.0 + } + ], + "mean_decode_tok_s": 11.4225, + "mean_ssd_mb_per_token": 71.1, + "mean_hot_rate_pct": 43.70505, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 4, + "status": "OK", + "wall_s": 33.43, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.929, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1899.1, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.02, + "tok_s": 10.81, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3394, + "expert_misses": 5418, + "hit_rate": 38.5157, + "ssb_demand_bytes": 7935363072, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7935363072, + "ssd_bytes_per_token": 47803400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4439, + "demand_wait_us": 9576097, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991319552, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 247, + "miss": 184, + "hit_rate": 57.3086, + "demand_bytes": 244187136, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 267, + "miss": 143, + "hit_rate": 65.122, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 232, + "miss": 120, + "hit_rate": 65.9091, + "demand_bytes": 159252480, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r4_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1905.2, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 3.05, + "tok_s": 10.48, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5397, + "expert_misses": 5203, + "hit_rate": 50.9151, + "ssb_demand_bytes": 7589007360, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7589007360, + "ssd_bytes_per_token": 44641200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4224, + "demand_wait_us": 9473030, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997713408, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 272, + "miss": 196, + "hit_rate": 58.1197, + "demand_bytes": 260112384, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 167, + "miss": 279, + "hit_rate": 37.4439, + "demand_bytes": 426829824, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r4_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.542, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1901.9, + "peak_anon_mb": 1417.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.07, + "tok_s": 12.15, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9168, + "expert_hits": 4068, + "expert_misses": 5100, + "hit_rate": 44.3717, + "ssb_demand_bytes": 7513546752, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7513546752, + "ssd_bytes_per_token": 43181300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4121, + "demand_wait_us": 9112705, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1993715712, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 286, + "miss": 128, + "hit_rate": 69.0821, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 287, + "miss": 128, + "hit_rate": 69.1566, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 76, + "miss": 335, + "hit_rate": 18.4915, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 323, + "hit": 79, + "miss": 244, + "hit_rate": 24.4582, + "demand_bytes": 373284864, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r4_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1900.3, + "peak_anon_mb": 1416.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.45, + "tok_s": 10.33, + "ssd_mb": 1277.2, + "ssd_mb_per_token": 85.1 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7426, + "expert_hits": 3049, + "expert_misses": 4377, + "hit_rate": 41.0584, + "ssb_demand_bytes": 6382116864, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6382116864, + "ssd_bytes_per_token": 48718400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3398, + "demand_wait_us": 7925202, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1992601600, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 175, + "miss": 149, + "hit_rate": 54.0123, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 208, + "miss": 133, + "hit_rate": 60.9971, + "demand_bytes": 176504832, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 79, + "miss": 244, + "hit_rate": 24.4582, + "demand_bytes": 373284864, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 90, + "miss": 221, + "hit_rate": 28.9389, + "demand_bytes": 338098176, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 255, + "hit": 72, + "miss": 183, + "hit_rate": 28.2353, + "demand_bytes": 279963648, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r4_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1905.2, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.02, + "tok_s": 10.81, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + { + "tokens": 32, + "seconds": 3.05, + "tok_s": 10.48, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 1.07, + "tok_s": 12.15, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + { + "tokens": 15, + "seconds": 1.45, + "tok_s": 10.33, + "ssd_mb": 1277.2, + "ssd_mb_per_token": 85.1 + } + ], + "mean_decode_tok_s": 10.9425, + "mean_ssd_mb_per_token": 71.375, + "mean_hot_rate_pct": 43.715225000000004, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 5, + "status": "OK", + "wall_s": 33.51, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.922, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1899.1, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.09, + "tok_s": 10.05, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3399, + "expert_misses": 5413, + "hit_rate": 38.5724, + "ssb_demand_bytes": 7928727552, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7928727552, + "ssd_bytes_per_token": 47763400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4434, + "demand_wait_us": 9656362, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991303168, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 238, + "miss": 114, + "hit_rate": 67.6136, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r5_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.9, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.82, + "tok_s": 11.33, + "ssd_mb": 2255.6, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5398, + "expert_misses": 5202, + "hit_rate": 50.9245, + "ssb_demand_bytes": 7587883008, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7587883008, + "ssd_bytes_per_token": 44634600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4223, + "demand_wait_us": 9851942, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997438976, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 273, + "miss": 195, + "hit_rate": 58.3333, + "demand_bytes": 258785280, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 310, + "miss": 161, + "hit_rate": 65.8174, + "demand_bytes": 213663744, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 166, + "miss": 280, + "hit_rate": 37.2197, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r5_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.54, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1899.1, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.99, + "tok_s": 13.07, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4071, + "expert_misses": 5099, + "hit_rate": 44.3948, + "ssb_demand_bytes": 7512422400, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7512422400, + "ssd_bytes_per_token": 43174800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4120, + "demand_wait_us": 9518911, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991380992, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 287, + "miss": 127, + "hit_rate": 69.3237, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 288, + "miss": 127, + "hit_rate": 69.3976, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 77, + "miss": 334, + "hit_rate": 18.7348, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r5_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1907.0, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.39, + "tok_s": 10.77, + "ssd_mb": 1274.3, + "ssd_mb_per_token": 85.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3051, + "expert_misses": 4420, + "hit_rate": 40.8379, + "ssb_demand_bytes": 6448306176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6448306176, + "ssd_bytes_per_token": 49223700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3441, + "demand_wait_us": 8242622, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999650816, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 63, + "miss": 344, + "hit_rate": 15.4791, + "demand_bytes": 526270464, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 163, + "miss": 190, + "hit_rate": 46.1756, + "demand_bytes": 252149760, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 175, + "miss": 149, + "hit_rate": 54.0123, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r5_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1907.0, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.09, + "tok_s": 10.05, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + { + "tokens": 32, + "seconds": 2.82, + "tok_s": 11.33, + "ssd_mb": 2255.6, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 0.99, + "tok_s": 13.07, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + { + "tokens": 15, + "seconds": 1.39, + "tok_s": 10.77, + "ssd_mb": 1274.3, + "ssd_mb_per_token": 85.0 + } + ], + "mean_decode_tok_s": 11.305, + "mean_ssd_mb_per_token": 71.3, + "mean_hot_rate_pct": 43.6824, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 6, + "status": "OK", + "wall_s": 32.87, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.925, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1899.9, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 1.01, + "tok_s": 10.91, + "ssd_mb": 786.2, + "ssd_mb_per_token": 71.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3394, + "expert_misses": 5418, + "hit_rate": 38.5157, + "ssb_demand_bytes": 7935160320, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7935160320, + "ssd_bytes_per_token": 47802200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4439, + "demand_wait_us": 9615610, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1992232960, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 247, + "miss": 184, + "hit_rate": 57.3086, + "demand_bytes": 244187136, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 266, + "miss": 144, + "hit_rate": 64.878, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 220, + "miss": 117, + "hit_rate": 65.2819, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 233, + "miss": 119, + "hit_rate": 66.1932, + "demand_bytes": 157925376, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 55, + "miss": 346, + "hit_rate": 13.7157, + "demand_bytes": 529330176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r6_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.8, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.82, + "tok_s": 11.36, + "ssd_mb": 2256.9, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10587, + "expert_hits": 5397, + "expert_misses": 5190, + "hit_rate": 50.9776, + "ssb_demand_bytes": 7569321984, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7569321984, + "ssd_bytes_per_token": 44525400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4211, + "demand_wait_us": 9060363, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997176832, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 273, + "miss": 195, + "hit_rate": 58.3333, + "demand_bytes": 258785280, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 155, + "miss": 298, + "hit_rate": 34.2163, + "demand_bytes": 455897088, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 433, + "hit": 167, + "miss": 266, + "hit_rate": 38.5681, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r6_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.576, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1900.1, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.27, + "tok_s": 10.26, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4067, + "expert_misses": 5103, + "hit_rate": 44.3511, + "ssb_demand_bytes": 7517730816, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7517730816, + "ssd_bytes_per_token": 43205300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4124, + "demand_wait_us": 9082524, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991364608, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 287, + "miss": 127, + "hit_rate": 69.3237, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 284, + "miss": 131, + "hit_rate": 68.4337, + "demand_bytes": 173850624, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 77, + "miss": 334, + "hit_rate": 18.7348, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r6_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1901.0, + "peak_anon_mb": 1416.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.42, + "tok_s": 10.53, + "ssd_mb": 1277.2, + "ssd_mb_per_token": 85.1 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 3052, + "expert_misses": 4393, + "hit_rate": 40.994, + "ssb_demand_bytes": 6407202816, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6407202816, + "ssd_bytes_per_token": 48909900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3414, + "demand_wait_us": 7709378, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1993297920, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 164, + "miss": 189, + "hit_rate": 46.4589, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 175, + "miss": 149, + "hit_rate": 54.0123, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 90, + "miss": 221, + "hit_rate": 28.9389, + "demand_bytes": 338098176, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 75, + "miss": 199, + "hit_rate": 27.3723, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r6_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1904.8, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 1.01, + "tok_s": 10.91, + "ssd_mb": 786.2, + "ssd_mb_per_token": 71.5 + }, + { + "tokens": 32, + "seconds": 2.82, + "tok_s": 11.36, + "ssd_mb": 2256.9, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 1.27, + "tok_s": 10.26, + "ssd_mb": 757.4, + "ssd_mb_per_token": 58.3 + }, + { + "tokens": 15, + "seconds": 1.42, + "tok_s": 10.53, + "ssd_mb": 1277.2, + "ssd_mb_per_token": 85.1 + } + ], + "mean_decode_tok_s": 10.765, + "mean_ssd_mb_per_token": 71.35, + "mean_hot_rate_pct": 43.7096, + "max_expert_requests": 10587 + }, + { + "level_mb": 2048, + "repeat": 7, + "status": "OK", + "wall_s": 32.43, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.91, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1906.5, + "peak_anon_mb": 1422.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.92, + "tok_s": 11.9, + "ssd_mb": 787.7, + "ssd_mb_per_token": 71.6 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8835, + "expert_hits": 3393, + "expert_misses": 5442, + "hit_rate": 38.4041, + "ssb_demand_bytes": 7971876864, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7971876864, + "ssd_bytes_per_token": 48023400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4463, + "demand_wait_us": 9758991, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999118336, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 220, + "miss": 117, + "hit_rate": 65.2819, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 233, + "miss": 119, + "hit_rate": 66.1932, + "demand_bytes": 157925376, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 330, + "hit": 66, + "miss": 264, + "hit_rate": 20, + "demand_bytes": 403881984, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r7_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.4, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.99, + "tok_s": 10.69, + "ssd_mb": 2256.9, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5397, + "expert_misses": 5203, + "hit_rate": 50.9151, + "ssb_demand_bytes": 7589210112, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7589210112, + "ssd_bytes_per_token": 44642400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4224, + "demand_wait_us": 9299012, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1996906496, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 119, + "miss": 416, + "hit_rate": 22.243, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 273, + "miss": 195, + "hit_rate": 58.3333, + "demand_bytes": 258785280, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 165, + "miss": 281, + "hit_rate": 36.9955, + "demand_bytes": 429889536, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r7_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.529, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1899.1, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.02, + "tok_s": 12.71, + "ssd_mb": 758.9, + "ssd_mb_per_token": 58.4 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4068, + "expert_misses": 5102, + "hit_rate": 44.3621, + "ssb_demand_bytes": 7516606464, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7516606464, + "ssd_bytes_per_token": 43198900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4123, + "demand_wait_us": 8958895, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991327744, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 287, + "miss": 127, + "hit_rate": 69.3237, + "demand_bytes": 168542208, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 286, + "miss": 129, + "hit_rate": 68.9157, + "demand_bytes": 171196416, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 76, + "miss": 335, + "hit_rate": 18.4915, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r7_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1906.9, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.37, + "tok_s": 10.97, + "ssd_mb": 1275.6, + "ssd_mb_per_token": 85.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3048, + "expert_misses": 4423, + "hit_rate": 40.7978, + "ssb_demand_bytes": 6452287488, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6452287488, + "ssd_bytes_per_token": 49254100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3444, + "demand_wait_us": 7691882, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999515648, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 63, + "miss": 344, + "hit_rate": 15.4791, + "demand_bytes": 526270464, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 174, + "miss": 150, + "hit_rate": 53.7037, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 208, + "miss": 133, + "hit_rate": 60.9971, + "demand_bytes": 176504832, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r7_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1906.9, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.92, + "tok_s": 11.9, + "ssd_mb": 787.7, + "ssd_mb_per_token": 71.6 + }, + { + "tokens": 32, + "seconds": 2.99, + "tok_s": 10.69, + "ssd_mb": 2256.9, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 1.02, + "tok_s": 12.71, + "ssd_mb": 758.9, + "ssd_mb_per_token": 58.4 + }, + { + "tokens": 15, + "seconds": 1.37, + "tok_s": 10.97, + "ssd_mb": 1275.6, + "ssd_mb_per_token": 85.0 + } + ], + "mean_decode_tok_s": 11.5675, + "mean_ssd_mb_per_token": 71.375, + "mean_hot_rate_pct": 43.619775000000004, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 8, + "status": "OK", + "wall_s": 32.82, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.908, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1906.0, + "peak_anon_mb": 1421.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.93, + "tok_s": 11.8, + "ssd_mb": 786.6, + "ssd_mb_per_token": 71.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8845, + "expert_hits": 3393, + "expert_misses": 5452, + "hit_rate": 38.3607, + "ssb_demand_bytes": 7987378176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7987378176, + "ssd_bytes_per_token": 48116700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4473, + "demand_wait_us": 9777401, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1998618624, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 40, + "miss": 376, + "hit_rate": 9.61538, + "demand_bytes": 575225856, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 233, + "miss": 119, + "hit_rate": 66.1932, + "demand_bytes": 157925376, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 340, + "hit": 66, + "miss": 274, + "hit_rate": 19.4118, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r8_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.6, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.88, + "tok_s": 11.13, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10606, + "expert_hits": 5396, + "expert_misses": 5210, + "hit_rate": 50.8769, + "ssb_demand_bytes": 7599513600, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7599513600, + "ssd_bytes_per_token": 44703000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4231, + "demand_wait_us": 9222473, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997103104, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 119, + "miss": 416, + "hit_rate": 22.243, + "demand_bytes": 636420096, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 271, + "miss": 197, + "hit_rate": 57.906, + "demand_bytes": 261439488, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 452, + "hit": 166, + "miss": 286, + "hit_rate": 36.7257, + "demand_bytes": 437538816, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r8_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.558, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1901.4, + "peak_anon_mb": 1417.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.09, + "tok_s": 11.92, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9180, + "expert_hits": 4070, + "expert_misses": 5110, + "hit_rate": 44.3355, + "ssb_demand_bytes": 7529250816, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7529250816, + "ssd_bytes_per_token": 43271600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4131, + "demand_wait_us": 9115522, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1993752576, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 286, + "miss": 128, + "hit_rate": 69.0821, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 289, + "miss": 126, + "hit_rate": 69.6386, + "demand_bytes": 167215104, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 76, + "miss": 335, + "hit_rate": 18.4915, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 335, + "hit": 79, + "miss": 256, + "hit_rate": 23.5821, + "demand_bytes": 391643136, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r8_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1907.6, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.5, + "tok_s": 9.98, + "ssd_mb": 1275.8, + "ssd_mb_per_token": 85.1 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3048, + "expert_misses": 4423, + "hit_rate": 40.7978, + "ssb_demand_bytes": 6452490240, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6452490240, + "ssd_bytes_per_token": 49255700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3444, + "demand_wait_us": 7818530, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 2000314368, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 174, + "miss": 150, + "hit_rate": 53.7037, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r8_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1907.6, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.93, + "tok_s": 11.8, + "ssd_mb": 786.6, + "ssd_mb_per_token": 71.5 + }, + { + "tokens": 32, + "seconds": 2.88, + "tok_s": 11.13, + "ssd_mb": 2255.5, + "ssd_mb_per_token": 70.5 + }, + { + "tokens": 13, + "seconds": 1.09, + "tok_s": 11.92, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + { + "tokens": 15, + "seconds": 1.5, + "tok_s": 9.98, + "ssd_mb": 1275.8, + "ssd_mb_per_token": 85.1 + } + ], + "mean_decode_tok_s": 11.2075, + "mean_ssd_mb_per_token": 71.4, + "mean_hot_rate_pct": 43.592725, + "max_expert_requests": 10606 + }, + { + "level_mb": 2048, + "repeat": 9, + "status": "OK", + "wall_s": 32.76, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.933, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1898.8, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.98, + "tok_s": 11.25, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3396, + "expert_misses": 5416, + "hit_rate": 38.5384, + "ssb_demand_bytes": 7932708864, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7932708864, + "ssd_bytes_per_token": 47787400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4437, + "demand_wait_us": 9553169, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1990991872, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 248, + "miss": 183, + "hit_rate": 57.5406, + "demand_bytes": 242860032, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 235, + "miss": 117, + "hit_rate": 66.7614, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r9_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.8, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.98, + "tok_s": 10.75, + "ssd_mb": 2259.9, + "ssd_mb_per_token": 70.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10600, + "expert_hits": 5393, + "expert_misses": 5207, + "hit_rate": 50.8774, + "ssb_demand_bytes": 7594924032, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7594924032, + "ssd_bytes_per_token": 44676000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4228, + "demand_wait_us": 9209046, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997373440, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 271, + "miss": 197, + "hit_rate": 57.906, + "demand_bytes": 261439488, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 155, + "miss": 298, + "hit_rate": 34.2163, + "demand_bytes": 455897088, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 169, + "miss": 290, + "hit_rate": 36.8192, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 165, + "miss": 281, + "hit_rate": 36.9955, + "demand_bytes": 429889536, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r9_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.528, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1899.7, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.05, + "tok_s": 12.34, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 4065, + "expert_misses": 5105, + "hit_rate": 44.3293, + "ssb_demand_bytes": 7520587776, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7520587776, + "ssd_bytes_per_token": 43221800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4126, + "demand_wait_us": 8997461, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1991753728, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 286, + "miss": 128, + "hit_rate": 69.0821, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 264, + "miss": 116, + "hit_rate": 69.4737, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 285, + "miss": 130, + "hit_rate": 68.6747, + "demand_bytes": 172523520, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 76, + "miss": 335, + "hit_rate": 18.4915, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 79, + "miss": 246, + "hit_rate": 24.3077, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r9_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1903.2, + "peak_anon_mb": 1419.5, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.45, + "tok_s": 10.32, + "ssd_mb": 1275.9, + "ssd_mb_per_token": 85.1 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7460, + "expert_hits": 3050, + "expert_misses": 4410, + "hit_rate": 40.8847, + "ssb_demand_bytes": 6432804864, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6432804864, + "ssd_bytes_per_token": 49105400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3431, + "demand_wait_us": 7593410, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1995698176, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 62, + "miss": 345, + "hit_rate": 15.2334, + "demand_bytes": 527800320, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 175, + "miss": 149, + "hit_rate": 54.0123, + "demand_bytes": 197738496, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 90, + "miss": 221, + "hit_rate": 28.9389, + "demand_bytes": 338098176, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 289, + "hit": 75, + "miss": 214, + "hit_rate": 25.9516, + "demand_bytes": 327389184, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r9_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1904.8, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.98, + "tok_s": 11.25, + "ssd_mb": 785.1, + "ssd_mb_per_token": 71.4 + }, + { + "tokens": 32, + "seconds": 2.98, + "tok_s": 10.75, + "ssd_mb": 2259.9, + "ssd_mb_per_token": 70.6 + }, + { + "tokens": 13, + "seconds": 1.05, + "tok_s": 12.34, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + { + "tokens": 15, + "seconds": 1.45, + "tok_s": 10.32, + "ssd_mb": 1275.9, + "ssd_mb_per_token": 85.1 + } + ], + "mean_decode_tok_s": 11.165, + "mean_ssd_mb_per_token": 71.4, + "mean_hot_rate_pct": 43.657450000000004, + "max_expert_requests": 10600 + }, + { + "level_mb": 2048, + "repeat": 10, + "status": "OK", + "wall_s": 31.94, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:15", + "interval": 900, + "temperature_2m": 20.4, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.924, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 1898.7, + "peak_anon_mb": 1414.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.97, + "tok_s": 11.34, + "ssd_mb": 786.4, + "ssd_mb_per_token": 71.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8812, + "expert_hits": 3393, + "expert_misses": 5419, + "hit_rate": 38.5043, + "ssb_demand_bytes": 7936690176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7936690176, + "ssd_bytes_per_token": 47811400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4440, + "demand_wait_us": 9518003, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1990942720, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 39, + "miss": 412, + "hit_rate": 8.64745, + "demand_bytes": 630300672, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 51, + "miss": 393, + "hit_rate": 11.4865, + "demand_bytes": 601233408, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 247, + "miss": 184, + "hit_rate": 57.3086, + "demand_bytes": 244187136, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 265, + "miss": 145, + "hit_rate": 64.6341, + "demand_bytes": 192430080, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 41, + "miss": 375, + "hit_rate": 9.85577, + "demand_bytes": 573696000, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 254, + "miss": 132, + "hit_rate": 65.8031, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 267, + "miss": 139, + "hit_rate": 65.7635, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 49, + "miss": 362, + "hit_rate": 11.9221, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 248, + "miss": 114, + "hit_rate": 68.5083, + "demand_bytes": 151289856, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 221, + "miss": 116, + "hit_rate": 65.5786, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 49, + "miss": 266, + "hit_rate": 15.5556, + "demand_bytes": 406941696, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 233, + "miss": 119, + "hit_rate": 66.1932, + "demand_bytes": 157925376, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 207, + "miss": 147, + "hit_rate": 58.4746, + "demand_bytes": 195084288, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 52, + "miss": 302, + "hit_rate": 14.6893, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 216, + "miss": 156, + "hit_rate": 58.0645, + "demand_bytes": 207028224, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 216, + "miss": 174, + "hit_rate": 55.3846, + "demand_bytes": 230916096, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 64, + "miss": 322, + "hit_rate": 16.5803, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 224, + "miss": 167, + "hit_rate": 57.289, + "demand_bytes": 221626368, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 206, + "miss": 151, + "hit_rate": 57.7031, + "demand_bytes": 200392704, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 54, + "miss": 347, + "hit_rate": 13.4663, + "demand_bytes": 530860032, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 54, + "miss": 335, + "hit_rate": 13.8817, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 70, + "miss": 320, + "hit_rate": 17.9487, + "demand_bytes": 489553920, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 307, + "hit": 66, + "miss": 241, + "hit_rate": 21.4984, + "demand_bytes": 368695296, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r10_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:15, with a temperature of 20.4°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 1904.9, + "peak_anon_mb": 1421.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 2.76, + "tok_s": 11.6, + "ssd_mb": 2254.0, + "ssd_mb_per_token": 70.4 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10595, + "expert_hits": 5397, + "expert_misses": 5198, + "hit_rate": 50.9391, + "ssb_demand_bytes": 7581155328, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7581155328, + "ssd_bytes_per_token": 44595000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4219, + "demand_wait_us": 9136694, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1997406208, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 535, + "hit": 118, + "miss": 417, + "hit_rate": 22.0561, + "demand_bytes": 637949952, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 477, + "hit": 134, + "miss": 343, + "hit_rate": 28.0922, + "demand_bytes": 524740608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 270, + "miss": 213, + "hit_rate": 55.9006, + "demand_bytes": 282673152, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 271, + "miss": 197, + "hit_rate": 57.906, + "demand_bytes": 261439488, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 156, + "miss": 297, + "hit_rate": 34.4371, + "demand_bytes": 454367232, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 453, + "hit": 285, + "miss": 168, + "hit_rate": 62.9139, + "demand_bytes": 222953472, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 471, + "hit": 309, + "miss": 162, + "hit_rate": 65.6051, + "demand_bytes": 214990848, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 474, + "hit": 140, + "miss": 334, + "hit_rate": 29.5359, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 454, + "hit": 304, + "miss": 150, + "hit_rate": 66.9604, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 435, + "hit": 307, + "miss": 128, + "hit_rate": 70.5747, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 445, + "hit": 167, + "miss": 278, + "hit_rate": 37.5281, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 450, + "hit": 311, + "miss": 139, + "hit_rate": 69.1111, + "demand_bytes": 184467456, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 306, + "miss": 140, + "hit_rate": 68.6099, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 452, + "hit": 174, + "miss": 278, + "hit_rate": 38.4956, + "demand_bytes": 425299968, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 462, + "hit": 325, + "miss": 137, + "hit_rate": 70.3463, + "demand_bytes": 181813248, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 476, + "hit": 332, + "miss": 144, + "hit_rate": 69.7479, + "demand_bytes": 191102976, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 448, + "hit": 173, + "miss": 275, + "hit_rate": 38.6161, + "demand_bytes": 420710400, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 444, + "hit": 316, + "miss": 128, + "hit_rate": 71.1712, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 318, + "miss": 124, + "hit_rate": 71.9457, + "demand_bytes": 164560896, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 459, + "hit": 170, + "miss": 289, + "hit_rate": 37.037, + "demand_bytes": 442128384, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 457, + "hit": 175, + "miss": 282, + "hit_rate": 38.2932, + "demand_bytes": 431419392, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 470, + "hit": 169, + "miss": 301, + "hit_rate": 35.9574, + "demand_bytes": 460486656, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 441, + "hit": 167, + "miss": 274, + "hit_rate": 37.8685, + "demand_bytes": 419180544, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r10_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.54, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 1906.9, + "peak_anon_mb": 1422.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 1.08, + "tok_s": 12.02, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9192, + "expert_hits": 4065, + "expert_misses": 5127, + "hit_rate": 44.2232, + "ssb_demand_bytes": 7554244608, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 7554244608, + "ssd_bytes_per_token": 43415200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 4148, + "demand_wait_us": 8891765, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999507456, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 56, + "miss": 420, + "hit_rate": 11.7647, + "demand_bytes": 642539520, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 53, + "miss": 409, + "hit_rate": 11.4719, + "demand_bytes": 625711104, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 308, + "miss": 143, + "hit_rate": 68.2927, + "demand_bytes": 189775872, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 290, + "miss": 140, + "hit_rate": 67.4419, + "demand_bytes": 185794560, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 66, + "miss": 362, + "hit_rate": 15.4206, + "demand_bytes": 553807872, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 282, + "miss": 118, + "hit_rate": 70.5, + "demand_bytes": 156598272, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 286, + "miss": 128, + "hit_rate": 69.0821, + "demand_bytes": 169869312, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 66, + "miss": 364, + "hit_rate": 15.3488, + "demand_bytes": 556867584, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 256, + "miss": 109, + "hit_rate": 70.137, + "demand_bytes": 144654336, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 244, + "miss": 104, + "hit_rate": 70.1149, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 70, + "miss": 262, + "hit_rate": 21.0843, + "demand_bytes": 400822272, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 258, + "miss": 104, + "hit_rate": 71.2707, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 264, + "miss": 107, + "hit_rate": 71.159, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 60, + "miss": 293, + "hit_rate": 16.9972, + "demand_bytes": 448247808, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 265, + "miss": 115, + "hit_rate": 69.7368, + "demand_bytes": 152616960, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 284, + "miss": 131, + "hit_rate": 68.4337, + "demand_bytes": 173850624, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 76, + "miss": 335, + "hit_rate": 18.4915, + "demand_bytes": 512501760, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 291, + "miss": 116, + "hit_rate": 71.4988, + "demand_bytes": 153944064, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 275, + "miss": 112, + "hit_rate": 71.0594, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 80, + "miss": 326, + "hit_rate": 19.7044, + "demand_bytes": 498733056, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 78, + "miss": 334, + "hit_rate": 18.932, + "demand_bytes": 510971904, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 78, + "miss": 327, + "hit_rate": 19.2593, + "demand_bytes": 500262912, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 347, + "hit": 79, + "miss": 268, + "hit_rate": 22.7666, + "demand_bytes": 410001408, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r10_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 1906.9, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.44, + "tok_s": 10.42, + "ssd_mb": 1273.0, + "ssd_mb_per_token": 84.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7471, + "expert_hits": 3049, + "expert_misses": 4422, + "hit_rate": 40.8111, + "ssb_demand_bytes": 6450960384, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6450960384, + "ssd_bytes_per_token": 49244000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 3443, + "demand_wait_us": 7698318, + "prefetch_wait_us": 0, + "arena_bytes": 1349206016, + "arena_used_bytes": 1349206016, + "rss_bytes": 1999544320, + "n_slots": 979, + "n_small_slots": 735, + "n_big_slots": 244, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 63, + "miss": 344, + "hit_rate": 15.4791, + "demand_bytes": 526270464, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 60, + "miss": 307, + "hit_rate": 16.3488, + "demand_bytes": 469665792, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 155, + "miss": 189, + "hit_rate": 45.0581, + "demand_bytes": 250822656, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 162, + "miss": 191, + "hit_rate": 45.8924, + "demand_bytes": 253476864, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 65, + "miss": 280, + "hit_rate": 18.8406, + "demand_bytes": 428359680, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 174, + "miss": 150, + "hit_rate": 53.7037, + "demand_bytes": 199065600, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 209, + "miss": 132, + "hit_rate": 61.2903, + "demand_bytes": 175177728, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 58, + "miss": 290, + "hit_rate": 16.6667, + "demand_bytes": 443658240, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 185, + "miss": 112, + "hit_rate": 62.2896, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 180, + "miss": 104, + "hit_rate": 63.3803, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 77, + "miss": 223, + "hit_rate": 25.6667, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 203, + "miss": 112, + "hit_rate": 64.4444, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 215, + "miss": 112, + "hit_rate": 65.7492, + "demand_bytes": 148635648, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 75, + "miss": 243, + "hit_rate": 23.5849, + "demand_bytes": 371755008, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 197, + "miss": 117, + "hit_rate": 62.7389, + "demand_bytes": 155271168, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 200, + "miss": 123, + "hit_rate": 61.9195, + "demand_bytes": 163233792, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 72, + "miss": 252, + "hit_rate": 22.2222, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 191, + "miss": 102, + "hit_rate": 65.1877, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 193, + "miss": 105, + "hit_rate": 64.7651, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 75, + "miss": 240, + "hit_rate": 23.8095, + "demand_bytes": 367165440, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 76, + "miss": 247, + "hit_rate": 23.5294, + "demand_bytes": 377874432, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 88, + "miss": 223, + "hit_rate": 28.2958, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 300, + "hit": 76, + "miss": 224, + "hit_rate": 25.3333, + "demand_bytes": 342687744, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_2048_r10_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 1906.9, + "peak_anon_mb": 1423.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.97, + "tok_s": 11.34, + "ssd_mb": 786.4, + "ssd_mb_per_token": 71.5 + }, + { + "tokens": 32, + "seconds": 2.76, + "tok_s": 11.6, + "ssd_mb": 2254.0, + "ssd_mb_per_token": 70.4 + }, + { + "tokens": 13, + "seconds": 1.08, + "tok_s": 12.02, + "ssd_mb": 760.1, + "ssd_mb_per_token": 58.5 + }, + { + "tokens": 15, + "seconds": 1.44, + "tok_s": 10.42, + "ssd_mb": 1273.0, + "ssd_mb_per_token": 84.9 + } + ], + "mean_decode_tok_s": 11.344999999999999, + "mean_ssd_mb_per_token": 71.325, + "mean_hot_rate_pct": 43.619425, + "max_expert_requests": 10595 + }, + { + "level_mb": 4096, + "repeat": 1, + "status": "OK", + "wall_s": 36.28, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 1.258, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.44, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4643, + "expert_misses": 4193, + "hit_rate": 52.5464, + "ssb_demand_bytes": 6191861760, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6191861760, + "ssd_bytes_per_token": 37300400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2473, + "demand_wait_us": 7742559, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077457408, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 87, + "miss": 324, + "hit_rate": 21.1679, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r1_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.6, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.82, + "tok_s": 13.7, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 6208, + "expert_misses": 3103, + "hit_rate": 66.6738, + "ssb_demand_bytes": 4506476544, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4506476544, + "ssd_bytes_per_token": 28887700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1295, + "demand_wait_us": 5865981, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074356736, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 224, + "miss": 165, + "hit_rate": 57.5835, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 257, + "miss": 157, + "hit_rate": 62.0773, + "demand_bytes": 240187392, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 256, + "miss": 134, + "hit_rate": 65.641, + "demand_bytes": 205000704, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r1_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.539, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3878.0, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.7, + "tok_s": 18.54, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5115, + "expert_misses": 4003, + "hit_rate": 56.0978, + "ssb_demand_bytes": 5905446912, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5905446912, + "ssd_bytes_per_token": 33939400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2304, + "demand_wait_us": 7513073, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4066373632, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 116, + "miss": 312, + "hit_rate": 27.1028, + "demand_bytes": 477315072, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r1_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.4, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.34, + "tok_s": 11.17, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 4252, + "expert_misses": 3193, + "hit_rate": 57.1122, + "ssb_demand_bytes": 4657747968, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4657747968, + "ssd_bytes_per_token": 35555300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1452, + "demand_wait_us": 6246495, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4069969920, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 155, + "miss": 252, + "hit_rate": 38.0835, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 156, + "miss": 189, + "hit_rate": 45.2174, + "demand_bytes": 289142784, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 150, + "miss": 150, + "hit_rate": 50, + "demand_bytes": 229478400, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 157, + "miss": 161, + "hit_rate": 49.3711, + "demand_bytes": 246306816, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 139, + "miss": 185, + "hit_rate": 42.9012, + "demand_bytes": 283023360, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 123, + "miss": 200, + "hit_rate": 38.0805, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 136, + "miss": 138, + "hit_rate": 49.635, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r1_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.44, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.82, + "tok_s": 13.7, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + { + "tokens": 13, + "seconds": 0.7, + "tok_s": 18.54, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.34, + "tok_s": 11.17, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + } + ], + "mean_decode_tok_s": 14.9625, + "mean_ssd_mb_per_token": 22.125, + "mean_hot_rate_pct": 58.10755, + "max_expert_requests": 9311 + }, + { + "level_mb": 4096, + "repeat": 2, + "status": "OK", + "wall_s": 36.34, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.9, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.2, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.74, + "tok_s": 14.84, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4643, + "expert_misses": 4193, + "hit_rate": 52.5464, + "ssb_demand_bytes": 6191861760, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6191861760, + "ssd_bytes_per_token": 37300400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2473, + "demand_wait_us": 7754607, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077117440, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 87, + "miss": 324, + "hit_rate": 21.1679, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r2_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.6, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.87, + "tok_s": 13.36, + "ssd_mb": 770.9, + "ssd_mb_per_token": 30.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9309, + "expert_hits": 6209, + "expert_misses": 3100, + "hit_rate": 66.6989, + "ssb_demand_bytes": 4501886976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4501886976, + "ssd_bytes_per_token": 28858200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1292, + "demand_wait_us": 5712885, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074389504, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 224, + "miss": 170, + "hit_rate": 56.8528, + "demand_bytes": 260075520, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 226, + "miss": 163, + "hit_rate": 58.0977, + "demand_bytes": 249366528, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 388, + "hit": 255, + "miss": 133, + "hit_rate": 65.7216, + "demand_bytes": 203470848, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r2_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.56, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.5, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.87, + "tok_s": 15.02, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5114, + "expert_misses": 4004, + "hit_rate": 56.0869, + "ssb_demand_bytes": 5906976768, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5906976768, + "ssd_bytes_per_token": 33948100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2305, + "demand_wait_us": 7467056, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065824768, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 115, + "miss": 313, + "hit_rate": 26.8692, + "demand_bytes": 478844928, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r2_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.8, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.24, + "tok_s": 12.05, + "ssd_mb": 564.6, + "ssd_mb_per_token": 37.6 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7426, + "expert_hits": 4262, + "expert_misses": 3164, + "hit_rate": 57.3929, + "ssb_demand_bytes": 4613382144, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4613382144, + "ssd_bytes_per_token": 35216700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1423, + "demand_wait_us": 6053366, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070412288, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 162, + "miss": 245, + "hit_rate": 39.8034, + "demand_bytes": 374814720, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 156, + "miss": 211, + "hit_rate": 42.5068, + "demand_bytes": 322799616, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 159, + "miss": 186, + "hit_rate": 46.087, + "demand_bytes": 284553216, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 152, + "miss": 196, + "hit_rate": 43.6782, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 151, + "miss": 149, + "hit_rate": 50.3333, + "demand_bytes": 227948544, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 160, + "miss": 158, + "hit_rate": 50.3145, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 138, + "miss": 186, + "hit_rate": 42.5926, + "demand_bytes": 284553216, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 126, + "miss": 197, + "hit_rate": 39.0093, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 120, + "miss": 191, + "hit_rate": 38.5852, + "demand_bytes": 292202496, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 255, + "hit": 126, + "miss": 129, + "hit_rate": 49.4118, + "demand_bytes": 197351424, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r2_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.2, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.74, + "tok_s": 14.84, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.87, + "tok_s": 13.36, + "ssd_mb": 770.9, + "ssd_mb_per_token": 30.8 + }, + { + "tokens": 13, + "seconds": 0.87, + "tok_s": 15.02, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.24, + "tok_s": 12.05, + "ssd_mb": 564.6, + "ssd_mb_per_token": 37.6 + } + ], + "mean_decode_tok_s": 13.817499999999999, + "mean_ssd_mb_per_token": 22.05, + "mean_hot_rate_pct": 58.181275, + "max_expert_requests": 9309 + }, + { + "level_mb": 4096, + "repeat": 3, + "status": "OK", + "wall_s": 36.38, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.929, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.8, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.66, + "tok_s": 16.76, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4644, + "expert_misses": 4192, + "hit_rate": 52.5577, + "ssb_demand_bytes": 6190331904, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6190331904, + "ssd_bytes_per_token": 37291200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2472, + "demand_wait_us": 7875339, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077666304, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 88, + "miss": 323, + "hit_rate": 21.4112, + "demand_bytes": 494143488, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r3_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.9, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.87, + "tok_s": 13.34, + "ssd_mb": 769.4, + "ssd_mb_per_token": 30.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 6209, + "expert_misses": 3102, + "hit_rate": 66.6846, + "ssb_demand_bytes": 4504946688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4504946688, + "ssd_bytes_per_token": 28877900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1294, + "demand_wait_us": 6021624, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074680320, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 224, + "miss": 170, + "hit_rate": 56.8528, + "demand_bytes": 260075520, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 223, + "miss": 166, + "hit_rate": 57.3265, + "demand_bytes": 253956096, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 257, + "miss": 157, + "hit_rate": 62.0773, + "demand_bytes": 240187392, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 257, + "miss": 133, + "hit_rate": 65.8974, + "demand_bytes": 203470848, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r3_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.517, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.8, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.76, + "tok_s": 17.12, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5115, + "expert_misses": 4003, + "hit_rate": 56.0978, + "ssb_demand_bytes": 5905446912, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5905446912, + "ssd_bytes_per_token": 33939400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2304, + "demand_wait_us": 7841923, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4066127872, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 116, + "miss": 312, + "hit_rate": 27.1028, + "demand_bytes": 477315072, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r3_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.7, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.15, + "tok_s": 13.02, + "ssd_mb": 569.1, + "ssd_mb_per_token": 37.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 4246, + "expert_misses": 3199, + "hit_rate": 57.0316, + "ssb_demand_bytes": 4666927104, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4666927104, + "ssd_bytes_per_token": 35625400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1458, + "demand_wait_us": 6172187, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070273024, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 155, + "miss": 252, + "hit_rate": 38.0835, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 155, + "miss": 190, + "hit_rate": 44.9275, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 150, + "miss": 150, + "hit_rate": 50, + "demand_bytes": 229478400, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 158, + "miss": 160, + "hit_rate": 49.6855, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 133, + "miss": 191, + "hit_rate": 41.0494, + "demand_bytes": 292202496, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 123, + "miss": 200, + "hit_rate": 38.0805, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 136, + "miss": 138, + "hit_rate": 49.635, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r3_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.8, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.66, + "tok_s": 16.76, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.87, + "tok_s": 13.34, + "ssd_mb": 769.4, + "ssd_mb_per_token": 30.8 + }, + { + "tokens": 13, + "seconds": 0.76, + "tok_s": 17.12, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.15, + "tok_s": 13.02, + "ssd_mb": 569.1, + "ssd_mb_per_token": 37.9 + } + ], + "mean_decode_tok_s": 15.06, + "mean_ssd_mb_per_token": 22.125, + "mean_hot_rate_pct": 58.092925, + "max_expert_requests": 9311 + }, + { + "level_mb": 4096, + "repeat": 4, + "status": "OK", + "wall_s": 36.04, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.922, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.35, + "ssd_mb": 135.9, + "ssd_mb_per_token": 12.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8770, + "expert_hits": 4659, + "expert_misses": 4111, + "hit_rate": 53.1243, + "ssb_demand_bytes": 6066413568, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6066413568, + "ssd_bytes_per_token": 36544700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2391, + "demand_wait_us": 7944403, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077473792, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 109, + "miss": 302, + "hit_rate": 26.5207, + "demand_bytes": 462016512, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 133, + "miss": 268, + "hit_rate": 33.1671, + "demand_bytes": 410001408, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 265, + "hit": 92, + "miss": 173, + "hit_rate": 34.717, + "demand_bytes": 264665088, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r4_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3887.3, + "peak_anon_mb": 3404.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.72, + "tok_s": 14.5, + "ssd_mb": 773.9, + "ssd_mb_per_token": 31.0 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9317, + "expert_hits": 6211, + "expert_misses": 3106, + "hit_rate": 66.6631, + "ssb_demand_bytes": 4511066112, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4511066112, + "ssd_bytes_per_token": 28917100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1298, + "demand_wait_us": 5677908, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4076138496, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 225, + "miss": 165, + "hit_rate": 57.6923, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 226, + "miss": 163, + "hit_rate": 58.0977, + "demand_bytes": 249366528, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 268, + "miss": 140, + "hit_rate": 65.6863, + "demand_bytes": 214179840, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 396, + "hit": 261, + "miss": 135, + "hit_rate": 65.9091, + "demand_bytes": 206530560, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r4_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.514, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.3, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.64, + "tok_s": 20.28, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5113, + "expert_misses": 4005, + "hit_rate": 56.0759, + "ssb_demand_bytes": 5908506624, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5908506624, + "ssd_bytes_per_token": 33956900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2306, + "demand_wait_us": 7610103, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065624064, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 114, + "miss": 314, + "hit_rate": 26.6355, + "demand_bytes": 480374784, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r4_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.5, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.3, + "tok_s": 11.55, + "ssd_mb": 570.5, + "ssd_mb_per_token": 38.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4195, + "expert_misses": 3261, + "hit_rate": 56.2634, + "ssb_demand_bytes": 4761778176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4761778176, + "ssd_bytes_per_token": 36349500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1520, + "demand_wait_us": 6248167, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070010880, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 155, + "miss": 252, + "hit_rate": 38.0835, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 154, + "miss": 191, + "hit_rate": 44.6377, + "demand_bytes": 292202496, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 134, + "miss": 166, + "hit_rate": 44.6667, + "demand_bytes": 253956096, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 125, + "miss": 193, + "hit_rate": 39.3082, + "demand_bytes": 295262208, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 118, + "miss": 197, + "hit_rate": 37.4603, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 124, + "miss": 199, + "hit_rate": 38.3901, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 143, + "miss": 142, + "hit_rate": 50.1754, + "demand_bytes": 217239552, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r4_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.35, + "ssd_mb": 135.9, + "ssd_mb_per_token": 12.4 + }, + { + "tokens": 25, + "seconds": 1.72, + "tok_s": 14.5, + "ssd_mb": 773.9, + "ssd_mb_per_token": 31.0 + }, + { + "tokens": 13, + "seconds": 0.64, + "tok_s": 20.28, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.3, + "tok_s": 11.55, + "ssd_mb": 570.5, + "ssd_mb_per_token": 38.0 + } + ], + "mean_decode_tok_s": 15.42, + "mean_ssd_mb_per_token": 22.175, + "mean_hot_rate_pct": 58.031675, + "max_expert_requests": 9317 + }, + { + "level_mb": 4096, + "repeat": 5, + "status": "OK", + "wall_s": 37.63, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.93, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.5, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.17, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8827, + "expert_hits": 4646, + "expert_misses": 4181, + "hit_rate": 52.634, + "ssb_demand_bytes": 6173503488, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6173503488, + "ssd_bytes_per_token": 37189800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2461, + "demand_wait_us": 7960097, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077379584, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 90, + "miss": 321, + "hit_rate": 21.8978, + "demand_bytes": 491083776, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 322, + "hit": 99, + "miss": 223, + "hit_rate": 30.7453, + "demand_bytes": 341157888, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r5_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.4, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.96, + "tok_s": 12.77, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 6209, + "expert_misses": 3102, + "hit_rate": 66.6846, + "ssb_demand_bytes": 4504946688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4504946688, + "ssd_bytes_per_token": 28877900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1294, + "demand_wait_us": 5959364, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074110976, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 225, + "miss": 164, + "hit_rate": 57.8406, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 257, + "miss": 133, + "hit_rate": 65.8974, + "demand_bytes": 203470848, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r5_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.568, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3876.9, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.34, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5110, + "expert_misses": 4008, + "hit_rate": 56.043, + "ssb_demand_bytes": 5913096192, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5913096192, + "ssd_bytes_per_token": 33983300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2309, + "demand_wait_us": 7549809, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065259520, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 111, + "miss": 317, + "hit_rate": 25.9346, + "demand_bytes": 484964352, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r5_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.6, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.4, + "tok_s": 10.75, + "ssd_mb": 569.0, + "ssd_mb_per_token": 37.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4194, + "expert_misses": 3262, + "hit_rate": 56.25, + "ssb_demand_bytes": 4763308032, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4763308032, + "ssd_bytes_per_token": 36361100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1521, + "demand_wait_us": 6230354, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070187008, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 156, + "miss": 251, + "hit_rate": 38.3292, + "demand_bytes": 383993856, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 155, + "miss": 190, + "hit_rate": 44.9275, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 134, + "miss": 166, + "hit_rate": 44.6667, + "demand_bytes": 253956096, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 123, + "miss": 195, + "hit_rate": 38.6792, + "demand_bytes": 298321920, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 118, + "miss": 197, + "hit_rate": 37.4603, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 124, + "miss": 199, + "hit_rate": 38.3901, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 142, + "miss": 143, + "hit_rate": 49.8246, + "demand_bytes": 218769408, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r5_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.5, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.17, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.96, + "tok_s": 12.77, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.34, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.4, + "tok_s": 10.75, + "ssd_mb": 569.0, + "ssd_mb_per_token": 37.9 + } + ], + "mean_decode_tok_s": 14.0075, + "mean_ssd_mb_per_token": 22.15, + "mean_hot_rate_pct": 57.9029, + "max_expert_requests": 9311 + }, + { + "level_mb": 4096, + "repeat": 6, + "status": "OK", + "wall_s": 37.32, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.924, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.2, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.73, + "tok_s": 15.05, + "ssd_mb": 135.9, + "ssd_mb_per_token": 12.4 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8770, + "expert_hits": 4661, + "expert_misses": 4109, + "hit_rate": 53.1471, + "ssb_demand_bytes": 6063353856, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6063353856, + "ssd_bytes_per_token": 36526200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2389, + "demand_wait_us": 8059923, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077064192, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 111, + "miss": 300, + "hit_rate": 27.0073, + "demand_bytes": 458956800, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 133, + "miss": 268, + "hit_rate": 33.1671, + "demand_bytes": 410001408, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 265, + "hit": 92, + "miss": 173, + "hit_rate": 34.717, + "demand_bytes": 264665088, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r6_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3887.5, + "peak_anon_mb": 3404.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.75, + "tok_s": 14.32, + "ssd_mb": 773.9, + "ssd_mb_per_token": 31.0 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9312, + "expert_hits": 6203, + "expert_misses": 3109, + "hit_rate": 66.613, + "ssb_demand_bytes": 4515655680, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4515655680, + "ssd_bytes_per_token": 28946500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1301, + "demand_wait_us": 5723736, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4076314624, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 227, + "miss": 162, + "hit_rate": 58.3548, + "demand_bytes": 247836672, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 268, + "miss": 140, + "hit_rate": 65.6863, + "demand_bytes": 214179840, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 263, + "miss": 138, + "hit_rate": 65.586, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 391, + "hit": 252, + "miss": 139, + "hit_rate": 64.4501, + "demand_bytes": 212649984, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r6_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.542, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3887.3, + "peak_anon_mb": 3402.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.69, + "tok_s": 18.72, + "ssd_mb": 91.4, + "ssd_mb_per_token": 7.0 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9202, + "expert_hits": 5087, + "expert_misses": 4115, + "hit_rate": 55.2815, + "ssb_demand_bytes": 6076790784, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6076790784, + "ssd_bytes_per_token": 34924100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2416, + "demand_wait_us": 7912445, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4076105728, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 109, + "miss": 319, + "hit_rate": 25.4673, + "demand_bytes": 488024064, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 120, + "miss": 310, + "hit_rate": 27.907, + "demand_bytes": 474255360, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 147, + "miss": 264, + "hit_rate": 35.7664, + "demand_bytes": 403881984, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 148, + "miss": 258, + "hit_rate": 36.4532, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 357, + "hit": 127, + "miss": 230, + "hit_rate": 35.5742, + "demand_bytes": 351866880, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r6_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3882.3, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.17, + "tok_s": 12.85, + "ssd_mb": 569.1, + "ssd_mb_per_token": 37.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4192, + "expert_misses": 3264, + "hit_rate": 56.2232, + "ssb_demand_bytes": 4766367744, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4766367744, + "ssd_bytes_per_token": 36384500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1523, + "demand_wait_us": 5955257, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070899712, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 157, + "miss": 250, + "hit_rate": 38.5749, + "demand_bytes": 382464000, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 155, + "miss": 190, + "hit_rate": 44.9275, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 132, + "miss": 168, + "hit_rate": 44, + "demand_bytes": 257015808, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 123, + "miss": 195, + "hit_rate": 38.6792, + "demand_bytes": 298321920, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 123, + "miss": 200, + "hit_rate": 38.0805, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 141, + "miss": 144, + "hit_rate": 49.4737, + "demand_bytes": 220299264, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r6_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.2, + "peak_anon_mb": 3404.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.73, + "tok_s": 15.05, + "ssd_mb": 135.9, + "ssd_mb_per_token": 12.4 + }, + { + "tokens": 25, + "seconds": 1.75, + "tok_s": 14.32, + "ssd_mb": 773.9, + "ssd_mb_per_token": 31.0 + }, + { + "tokens": 13, + "seconds": 0.69, + "tok_s": 18.72, + "ssd_mb": 91.4, + "ssd_mb_per_token": 7.0 + }, + { + "tokens": 15, + "seconds": 1.17, + "tok_s": 12.85, + "ssd_mb": 569.1, + "ssd_mb_per_token": 37.9 + } + ], + "mean_decode_tok_s": 15.235, + "mean_ssd_mb_per_token": 22.075, + "mean_hot_rate_pct": 57.8162, + "max_expert_requests": 9312 + }, + { + "level_mb": 4096, + "repeat": 7, + "status": "OK", + "wall_s": 37.46, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.922, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.6, + "tok_s": 18.25, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4642, + "expert_misses": 4194, + "hit_rate": 52.5351, + "ssb_demand_bytes": 6193391616, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6193391616, + "ssd_bytes_per_token": 37309600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2474, + "demand_wait_us": 7754190, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077445120, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 87, + "miss": 324, + "hit_rate": 21.1679, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 115, + "miss": 200, + "hit_rate": 36.5079, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r7_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3886.1, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.73, + "tok_s": 14.48, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9304, + "expert_hits": 6198, + "expert_misses": 3106, + "hit_rate": 66.6165, + "ssb_demand_bytes": 4511066112, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4511066112, + "ssd_bytes_per_token": 28917100.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1298, + "demand_wait_us": 5911472, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074848256, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 224, + "miss": 170, + "hit_rate": 56.8528, + "demand_bytes": 260075520, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 224, + "miss": 165, + "hit_rate": 57.5835, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 383, + "hit": 246, + "miss": 137, + "hit_rate": 64.2298, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r7_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.54, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.1, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.66, + "tok_s": 19.73, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5112, + "expert_misses": 4006, + "hit_rate": 56.0649, + "ssb_demand_bytes": 5910036480, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5910036480, + "ssd_bytes_per_token": 33965700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2307, + "demand_wait_us": 7553557, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065406976, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 113, + "miss": 315, + "hit_rate": 26.4019, + "demand_bytes": 481904640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r7_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.3, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.29, + "tok_s": 11.64, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4195, + "expert_misses": 3261, + "hit_rate": 56.2634, + "ssb_demand_bytes": 4761778176, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4761778176, + "ssd_bytes_per_token": 36349500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1520, + "demand_wait_us": 6767586, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4069851136, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 156, + "miss": 251, + "hit_rate": 38.3292, + "demand_bytes": 383993856, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 155, + "miss": 190, + "hit_rate": 44.9275, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 133, + "miss": 167, + "hit_rate": 44.3333, + "demand_bytes": 255485952, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 123, + "miss": 195, + "hit_rate": 38.6792, + "demand_bytes": 298321920, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 124, + "miss": 199, + "hit_rate": 38.3901, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 143, + "miss": 142, + "hit_rate": 50.1754, + "demand_bytes": 217239552, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r7_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.6, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.6, + "tok_s": 18.25, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.73, + "tok_s": 14.48, + "ssd_mb": 772.4, + "ssd_mb_per_token": 30.9 + }, + { + "tokens": 13, + "seconds": 0.66, + "tok_s": 19.73, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.29, + "tok_s": 11.64, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + } + ], + "mean_decode_tok_s": 16.025, + "mean_ssd_mb_per_token": 22.125, + "mean_hot_rate_pct": 57.869975, + "max_expert_requests": 9304 + }, + { + "level_mb": 4096, + "repeat": 8, + "status": "OK", + "wall_s": 37.82, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.925, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.4, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.66, + "tok_s": 16.7, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4644, + "expert_misses": 4192, + "hit_rate": 52.5577, + "ssb_demand_bytes": 6190331904, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6190331904, + "ssd_bytes_per_token": 37291200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2472, + "demand_wait_us": 7972631, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077297664, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 89, + "miss": 322, + "hit_rate": 21.6545, + "demand_bytes": 492613632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 115, + "miss": 200, + "hit_rate": 36.5079, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r8_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.5, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.86, + "tok_s": 13.47, + "ssd_mb": 775.3, + "ssd_mb_per_token": 31.0 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9308, + "expert_hits": 6205, + "expert_misses": 3103, + "hit_rate": 66.6631, + "ssb_demand_bytes": 4506476544, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4506476544, + "ssd_bytes_per_token": 28887700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1295, + "demand_wait_us": 6187918, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074283008, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 225, + "miss": 164, + "hit_rate": 57.8406, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 387, + "hit": 253, + "miss": 134, + "hit_rate": 65.3747, + "demand_bytes": 205000704, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r8_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.564, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.4, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.64, + "tok_s": 20.16, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5116, + "expert_misses": 4002, + "hit_rate": 56.1088, + "ssb_demand_bytes": 5903917056, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5903917056, + "ssd_bytes_per_token": 33930600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2303, + "demand_wait_us": 7463247, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065792000, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 117, + "miss": 311, + "hit_rate": 27.3364, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r8_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.5, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.05, + "tok_s": 14.35, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4193, + "expert_misses": 3263, + "hit_rate": 56.2366, + "ssb_demand_bytes": 4764837888, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4764837888, + "ssd_bytes_per_token": 36372800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1522, + "demand_wait_us": 6261870, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070055936, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 155, + "miss": 252, + "hit_rate": 38.0835, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 156, + "miss": 189, + "hit_rate": 45.2174, + "demand_bytes": 289142784, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 134, + "miss": 166, + "hit_rate": 44.6667, + "demand_bytes": 253956096, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 123, + "miss": 195, + "hit_rate": 38.6792, + "demand_bytes": 298321920, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 123, + "miss": 200, + "hit_rate": 38.0805, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 141, + "miss": 144, + "hit_rate": 49.4737, + "demand_bytes": 220299264, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r8_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.4, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.66, + "tok_s": 16.7, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.86, + "tok_s": 13.47, + "ssd_mb": 775.3, + "ssd_mb_per_token": 31.0 + }, + { + "tokens": 13, + "seconds": 0.64, + "tok_s": 20.16, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.05, + "tok_s": 14.35, + "ssd_mb": 567.6, + "ssd_mb_per_token": 37.8 + } + ], + "mean_decode_tok_s": 16.17, + "mean_ssd_mb_per_token": 22.15, + "mean_hot_rate_pct": 57.89155, + "max_expert_requests": 9308 + }, + { + "level_mb": 4096, + "repeat": 9, + "status": "OK", + "wall_s": 47.79, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.559, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3887.9, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.58, + "tok_s": 18.94, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4642, + "expert_misses": 4194, + "hit_rate": 52.5351, + "ssb_demand_bytes": 6193391616, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6193391616, + "ssd_bytes_per_token": 37309600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2474, + "demand_wait_us": 8185146, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4076716032, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 87, + "miss": 324, + "hit_rate": 21.1679, + "demand_bytes": 495673344, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 115, + "miss": 200, + "hit_rate": 36.5079, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r9_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.6, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.74, + "tok_s": 14.38, + "ssd_mb": 775.3, + "ssd_mb_per_token": 31.0 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9316, + "expert_hits": 6206, + "expert_misses": 3110, + "hit_rate": 66.6166, + "ssb_demand_bytes": 4517185536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4517185536, + "ssd_bytes_per_token": 28956300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1302, + "demand_wait_us": 6274690, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074332160, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 225, + "miss": 165, + "hit_rate": 57.6923, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 223, + "miss": 171, + "hit_rate": 56.599, + "demand_bytes": 261605376, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 223, + "miss": 166, + "hit_rate": 57.3265, + "demand_bytes": 253956096, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 267, + "miss": 141, + "hit_rate": 65.4412, + "demand_bytes": 215709696, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 395, + "hit": 260, + "miss": 135, + "hit_rate": 65.8228, + "demand_bytes": 206530560, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r9_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.566, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.2, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.7, + "tok_s": 18.67, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5116, + "expert_misses": 4002, + "hit_rate": 56.1088, + "ssb_demand_bytes": 5903917056, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5903917056, + "ssd_bytes_per_token": 33930600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2303, + "demand_wait_us": 7644152, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065566720, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 117, + "miss": 311, + "hit_rate": 27.3364, + "demand_bytes": 475785216, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r9_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.9, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.22, + "tok_s": 12.32, + "ssd_mb": 563.2, + "ssd_mb_per_token": 37.5 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7437, + "expert_hits": 4201, + "expert_misses": 3236, + "hit_rate": 56.4878, + "ssb_demand_bytes": 4723531776, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4723531776, + "ssd_bytes_per_token": 36057500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1495, + "demand_wait_us": 6313825, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070436864, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 161, + "miss": 246, + "hit_rate": 39.5577, + "demand_bytes": 376344576, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 156, + "miss": 211, + "hit_rate": 42.5068, + "demand_bytes": 322799616, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 159, + "miss": 186, + "hit_rate": 46.087, + "demand_bytes": 284553216, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 153, + "miss": 195, + "hit_rate": 43.9655, + "demand_bytes": 298321920, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 135, + "miss": 165, + "hit_rate": 45, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 124, + "miss": 194, + "hit_rate": 38.9937, + "demand_bytes": 296792064, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 119, + "miss": 196, + "hit_rate": 37.7778, + "demand_bytes": 299851776, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 126, + "miss": 197, + "hit_rate": 39.0093, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 120, + "miss": 191, + "hit_rate": 38.5852, + "demand_bytes": 292202496, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 266, + "hit": 130, + "miss": 136, + "hit_rate": 48.8722, + "demand_bytes": 208060416, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r9_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3887.9, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.58, + "tok_s": 18.94, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.74, + "tok_s": 14.38, + "ssd_mb": 775.3, + "ssd_mb_per_token": 31.0 + }, + { + "tokens": 13, + "seconds": 0.7, + "tok_s": 18.67, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.22, + "tok_s": 12.32, + "ssd_mb": 563.2, + "ssd_mb_per_token": 37.5 + } + ], + "mean_decode_tok_s": 16.0775, + "mean_ssd_mb_per_token": 22.075, + "mean_hot_rate_pct": 57.937075, + "max_expert_requests": 9316 + }, + { + "level_mb": 4096, + "repeat": 10, + "status": "OK", + "wall_s": 54.44, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.564, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 3888.5, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.79, + "tok_s": 13.85, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 4642, + "expert_misses": 4194, + "hit_rate": 52.5351, + "ssb_demand_bytes": 6193391616, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 6193391616, + "ssd_bytes_per_token": 37309600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2474, + "demand_wait_us": 8804304, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2409799680, + "rss_bytes": 4077420544, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 85, + "miss": 366, + "hit_rate": 18.847, + "demand_bytes": 559927296, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 87, + "miss": 357, + "hit_rate": 19.5946, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 85, + "miss": 331, + "hit_rate": 20.4327, + "demand_bytes": 506382336, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 86, + "miss": 325, + "hit_rate": 20.9246, + "demand_bytes": 497203200, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 116, + "miss": 199, + "hit_rate": 36.8254, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 129, + "miss": 225, + "hit_rate": 36.4407, + "demand_bytes": 344217600, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 121, + "miss": 265, + "hit_rate": 31.3472, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 132, + "miss": 269, + "hit_rate": 32.9177, + "demand_bytes": 411531264, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 124, + "miss": 265, + "hit_rate": 31.8766, + "demand_bytes": 405411840, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 129, + "miss": 261, + "hit_rate": 33.0769, + "demand_bytes": 399292416, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 99, + "miss": 232, + "hit_rate": 29.9094, + "demand_bytes": 354926592, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r10_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 3885.3, + "peak_anon_mb": 3402.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.81, + "tok_s": 13.85, + "ssd_mb": 771.0, + "ssd_mb_per_token": 30.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9309, + "expert_hits": 6207, + "expert_misses": 3102, + "hit_rate": 66.6774, + "ssb_demand_bytes": 4504946688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4504946688, + "ssd_bytes_per_token": 28877900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1294, + "demand_wait_us": 6147580, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2526584832, + "rss_bytes": 4074037248, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 240, + "miss": 239, + "hit_rate": 50.1044, + "demand_bytes": 365635584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 203, + "miss": 218, + "hit_rate": 48.2185, + "demand_bytes": 333508608, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 207, + "miss": 190, + "hit_rate": 52.1411, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 213, + "miss": 203, + "hit_rate": 51.2019, + "demand_bytes": 310560768, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 226, + "miss": 164, + "hit_rate": 57.9487, + "demand_bytes": 250896384, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 224, + "miss": 170, + "hit_rate": 56.8528, + "demand_bytes": 260075520, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 224, + "miss": 165, + "hit_rate": 57.5835, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 270, + "miss": 138, + "hit_rate": 66.1765, + "demand_bytes": 211120128, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 264, + "miss": 137, + "hit_rate": 65.8354, + "demand_bytes": 209590272, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 256, + "miss": 158, + "hit_rate": 61.8357, + "demand_bytes": 241717248, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 388, + "hit": 255, + "miss": 133, + "hit_rate": 65.7216, + "demand_bytes": 203470848, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r10_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.558, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 3877.4, + "peak_anon_mb": 3393.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.66, + "tok_s": 19.58, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 5115, + "expert_misses": 4003, + "hit_rate": 56.0978, + "ssb_demand_bytes": 5905446912, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 5905446912, + "ssd_bytes_per_token": 33939400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 2304, + "demand_wait_us": 8192870, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2381930496, + "rss_bytes": 4065193984, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 108, + "miss": 368, + "hit_rate": 22.6891, + "demand_bytes": 562987008, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 105, + "miss": 357, + "hit_rate": 22.7273, + "demand_bytes": 546158592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 116, + "miss": 312, + "hit_rate": 27.1028, + "demand_bytes": 477315072, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 153, + "miss": 277, + "hit_rate": 35.5814, + "demand_bytes": 423770112, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 133, + "miss": 199, + "hit_rate": 40.0602, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 138, + "miss": 215, + "hit_rate": 39.0935, + "demand_bytes": 328919040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 148, + "miss": 263, + "hit_rate": 36.0097, + "demand_bytes": 402352128, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 149, + "miss": 257, + "hit_rate": 36.6995, + "demand_bytes": 393172992, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 153, + "miss": 259, + "hit_rate": 37.1359, + "demand_bytes": 396232704, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 147, + "miss": 258, + "hit_rate": 36.2963, + "demand_bytes": 394702848, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 113, + "miss": 160, + "hit_rate": 41.3919, + "demand_bytes": 244776960, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r10_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 3881.8, + "peak_anon_mb": 3397.8, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.15, + "tok_s": 13.03, + "ssd_mb": 570.5, + "ssd_mb_per_token": 38.0 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7456, + "expert_hits": 4196, + "expert_misses": 3260, + "hit_rate": 56.2768, + "ssb_demand_bytes": 4760248320, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 4760248320, + "ssd_bytes_per_token": 36337800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 1519, + "demand_wait_us": 6561270, + "prefetch_wait_us": 0, + "arena_bytes": 3426361344, + "arena_used_bytes": 2437668864, + "rss_bytes": 4070363136, + "n_slots": 2486, + "n_small_slots": 1865, + "n_big_slots": 621, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 155, + "miss": 252, + "hit_rate": 38.0835, + "demand_bytes": 385523712, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 154, + "miss": 213, + "hit_rate": 41.9619, + "demand_bytes": 325859328, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 155, + "miss": 190, + "hit_rate": 44.9275, + "demand_bytes": 290672640, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 151, + "miss": 197, + "hit_rate": 43.3908, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 135, + "miss": 165, + "hit_rate": 45, + "demand_bytes": 252426240, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 126, + "miss": 192, + "hit_rate": 39.6226, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 125, + "miss": 199, + "hit_rate": 38.5802, + "demand_bytes": 304441344, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 118, + "miss": 197, + "hit_rate": 37.4603, + "demand_bytes": 301381632, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 123, + "miss": 200, + "hit_rate": 38.0805, + "demand_bytes": 305971200, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 119, + "miss": 192, + "hit_rate": 38.2637, + "demand_bytes": 293732352, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 285, + "hit": 142, + "miss": 143, + "hit_rate": 49.8246, + "demand_bytes": 218769408, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_4096_r10_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 3888.5, + "peak_anon_mb": 3404.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.79, + "tok_s": 13.85, + "ssd_mb": 137.4, + "ssd_mb_per_token": 12.5 + }, + { + "tokens": 25, + "seconds": 1.81, + "tok_s": 13.85, + "ssd_mb": 771.0, + "ssd_mb_per_token": 30.8 + }, + { + "tokens": 13, + "seconds": 0.66, + "tok_s": 19.58, + "ssd_mb": 94.4, + "ssd_mb_per_token": 7.3 + }, + { + "tokens": 15, + "seconds": 1.15, + "tok_s": 13.03, + "ssd_mb": 570.5, + "ssd_mb_per_token": 38.0 + } + ], + "mean_decode_tok_s": 15.077499999999999, + "mean_ssd_mb_per_token": 22.15, + "mean_hot_rate_pct": 57.896775000000005, + "max_expert_requests": 9309 + }, + { + "level_mb": 8192, + "repeat": 1, + "status": "OK", + "wall_s": 47.75, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.578, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.3, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.68, + "tok_s": 16.16, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8786, + "expert_hits": 6657, + "expert_misses": 2129, + "hit_rate": 75.7683, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4680188, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7926452224, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 281, + "hit": 198, + "miss": 83, + "hit_rate": 70.4626, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r1_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.0, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.49, + "tok_s": 16.79, + "ssd_mb": 243.0, + "ssd_mb_per_token": 9.7 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9316, + "expert_hits": 7028, + "expert_misses": 2288, + "hit_rate": 75.4401, + "ssb_demand_bytes": 3259643904, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3259643904, + "ssd_bytes_per_token": 20895200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5189034, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3261898752, + "rss_bytes": 7924097024, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 395, + "hit": 301, + "miss": 94, + "hit_rate": 76.2025, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r1_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.516, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.3, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.72, + "tok_s": 17.96, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4262835, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7914950656, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r1_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.91, + "tok_s": 16.54, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4576002, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926673408, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r1_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.68, + "tok_s": 16.16, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.49, + "tok_s": 16.79, + "ssd_mb": 243.0, + "ssd_mb_per_token": 9.7 + }, + { + "tokens": 13, + "seconds": 0.72, + "tok_s": 17.96, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.91, + "tok_s": 16.54, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.8625, + "mean_ssd_mb_per_token": 7.125, + "mean_hot_rate_pct": 74.774575, + "max_expert_requests": 9316 + }, + { + "level_mb": 8192, + "repeat": 2, + "status": "OK", + "wall_s": 47.78, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.565, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7548.6, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.56, + "tok_s": 19.7, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8755, + "expert_hits": 6631, + "expert_misses": 2124, + "hit_rate": 75.7396, + "ssb_demand_bytes": 3026589696, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3026589696, + "ssd_bytes_per_token": 18232500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4532642, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3028688896, + "rss_bytes": 7915315200, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 250, + "hit": 172, + "miss": 78, + "hit_rate": 68.8, + "demand_bytes": 119328768, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r2_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.1, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.65, + "tok_s": 15.18, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9291, + "expert_hits": 7004, + "expert_misses": 2287, + "hit_rate": 75.3848, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4818002, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7924211712, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 370, + "hit": 277, + "miss": 93, + "hit_rate": 74.8649, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r2_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.541, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.6, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.74, + "tok_s": 17.58, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4954408, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915282432, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r2_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.2, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.13, + "tok_s": 13.27, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5673431, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7925800960, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r2_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.2, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.56, + "tok_s": 19.7, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.65, + "tok_s": 15.18, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.74, + "tok_s": 17.58, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 1.13, + "tok_s": 13.27, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.432499999999997, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.753575, + "max_expert_requests": 9291 + }, + { + "level_mb": 8192, + "repeat": 3, + "status": "OK", + "wall_s": 57.55, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.552, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.8, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.56, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8786, + "expert_hits": 6657, + "expert_misses": 2129, + "hit_rate": 75.7683, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5598312, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7926980608, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 281, + "hit": 198, + "miss": 83, + "hit_rate": 70.4626, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r3_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.1, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.79, + "tok_s": 13.93, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6361288, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7924207616, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r3_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.557, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.8, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.86, + "tok_s": 15.14, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5225311, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915495424, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r3_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7552.7, + "peak_anon_mb": 7069.2, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.98, + "tok_s": 15.37, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 5262, + "expert_misses": 2183, + "hit_rate": 70.6783, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5117861, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7919611904, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 190, + "miss": 84, + "hit_rate": 69.3431, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r3_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.8, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.56, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.79, + "tok_s": 13.93, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.86, + "tok_s": 15.14, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.98, + "tok_s": 15.37, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 15.75, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.768075, + "max_expert_requests": 9311 + }, + { + "level_mb": 8192, + "repeat": 4, + "status": "OK", + "wall_s": 46.41, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.538, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.57, + "tok_s": 19.17, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8786, + "expert_hits": 6657, + "expert_misses": 2129, + "hit_rate": 75.7683, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4372841, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7926751232, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 281, + "hit": 198, + "miss": 83, + "hit_rate": 70.4626, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r4_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7556.6, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.35, + "tok_s": 18.5, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4888097, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7923724288, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r4_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.539, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.8, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.72, + "tok_s": 17.95, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4844541, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915520000, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r4_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.2, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.05, + "tok_s": 14.23, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4616671, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926386688, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r4_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.57, + "tok_s": 19.17, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.35, + "tok_s": 18.5, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.72, + "tok_s": 17.95, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 1.05, + "tok_s": 14.23, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.4625, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.77397500000001, + "max_expert_requests": 9311 + }, + { + "level_mb": 8192, + "repeat": 5, + "status": "OK", + "wall_s": 52.21, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.575, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7563.5, + "peak_anon_mb": 7079.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.64, + "tok_s": 17.25, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8855, + "expert_hits": 6721, + "expert_misses": 2134, + "hit_rate": 75.9006, + "ssb_demand_bytes": 3041888256, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3041888256, + "ssd_bytes_per_token": 18324600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5187131, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3044007936, + "rss_bytes": 7930920960, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 350, + "hit": 262, + "miss": 88, + "hit_rate": 74.8571, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r5_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.1, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.85, + "tok_s": 13.49, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5686009, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7923273728, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r5_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.553, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7549.0, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.62, + "tok_s": 21.09, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4989696, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915610112, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r5_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7558.7, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.19, + "tok_s": 12.65, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5656434, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7925841920, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r5_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7563.5, + "peak_anon_mb": 7079.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.64, + "tok_s": 17.25, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.85, + "tok_s": 13.49, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.62, + "tok_s": 21.09, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 1.19, + "tok_s": 12.65, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.12, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.80705, + "max_expert_requests": 9311 + }, + { + "level_mb": 8192, + "repeat": 6, + "status": "OK", + "wall_s": 52.41, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.556, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.9, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.53, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8786, + "expert_hits": 6657, + "expert_misses": 2129, + "hit_rate": 75.7683, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5894649, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7927140352, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 281, + "hit": 198, + "miss": 83, + "hit_rate": 70.4626, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r6_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.0, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.75, + "tok_s": 14.25, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5919799, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7924137984, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r6_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.533, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.7, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.61, + "tok_s": 21.39, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4738428, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915429888, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r6_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.2, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.09, + "tok_s": 13.71, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5618153, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926366208, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r6_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.9, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.53, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.75, + "tok_s": 14.25, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.61, + "tok_s": 21.39, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 1.09, + "tok_s": 13.71, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.47, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.77397500000001, + "max_expert_requests": 9311 + }, + { + "level_mb": 8192, + "repeat": 7, + "status": "OK", + "wall_s": 50.96, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.561, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.7, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.51, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8846, + "expert_hits": 6713, + "expert_misses": 2133, + "hit_rate": 75.8874, + "ssb_demand_bytes": 3040358400, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3040358400, + "ssd_bytes_per_token": 18315400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5183656, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3042476032, + "rss_bytes": 7926935552, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 341, + "hit": 254, + "miss": 87, + "hit_rate": 74.4868, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r7_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7556.9, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.57, + "tok_s": 15.88, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5568726, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7924023296, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r7_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.517, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.6, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.3, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4801452, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915241472, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r7_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7558.9, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.9, + "tok_s": 16.75, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5644775, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926087680, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r7_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.7, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.67, + "tok_s": 16.51, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.57, + "tok_s": 15.88, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.3, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.9, + "tok_s": 16.75, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.61, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.80375000000001, + "max_expert_requests": 9311 + }, + { + "level_mb": 8192, + "repeat": 8, + "status": "OK", + "wall_s": 50.3, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.559, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.4, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.54, + "tok_s": 20.5, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8786, + "expert_hits": 6657, + "expert_misses": 2129, + "hit_rate": 75.7683, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5027522, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7926644736, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 281, + "hit": 198, + "miss": 83, + "hit_rate": 70.4626, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r8_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7556.9, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.91, + "tok_s": 13.08, + "ssd_mb": 243.0, + "ssd_mb_per_token": 9.7 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9316, + "expert_hits": 7028, + "expert_misses": 2288, + "hit_rate": 75.4401, + "ssb_demand_bytes": 3259643904, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3259643904, + "ssd_bytes_per_token": 20895200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5491540, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3261898752, + "rss_bytes": 7923986432, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 395, + "hit": 301, + "miss": 94, + "hit_rate": 76.2025, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r8_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.517, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.9, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.62, + "tok_s": 20.94, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5232524, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915589632, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r8_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.4, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.49, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5735514, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926603776, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r8_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.4, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.54, + "tok_s": 20.5, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.91, + "tok_s": 13.08, + "ssd_mb": 243.0, + "ssd_mb_per_token": 9.7 + }, + { + "tokens": 13, + "seconds": 0.62, + "tok_s": 20.94, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.49, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.5025, + "mean_ssd_mb_per_token": 7.125, + "mean_hot_rate_pct": 74.774575, + "max_expert_requests": 9316 + }, + { + "level_mb": 8192, + "repeat": 9, + "status": "OK", + "wall_s": 46.37, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.547, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7548.3, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.51, + "tok_s": 21.72, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8755, + "expert_hits": 6631, + "expert_misses": 2124, + "hit_rate": 75.7396, + "ssb_demand_bytes": 3026589696, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3026589696, + "ssd_bytes_per_token": 18232500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4730982, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3028688896, + "rss_bytes": 7914926080, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 250, + "hit": 172, + "miss": 78, + "hit_rate": 68.8, + "demand_bytes": 119328768, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r9_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7557.4, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.43, + "tok_s": 17.47, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9310, + "expert_hits": 7023, + "expert_misses": 2287, + "hit_rate": 75.435, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4627149, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7924482048, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 389, + "hit": 296, + "miss": 93, + "hit_rate": 76.0925, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r9_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.545, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.4, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.86, + "tok_s": 15.15, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4687292, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915094016, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r9_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.3, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.96, + "tok_s": 15.59, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4946151, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926468608, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r9_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.3, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.51, + "tok_s": 21.72, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.43, + "tok_s": 17.47, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.86, + "tok_s": 15.15, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.96, + "tok_s": 15.59, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.482499999999998, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.766125, + "max_expert_requests": 9310 + }, + { + "level_mb": 8192, + "repeat": 10, + "status": "OK", + "wall_s": 45.95, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:30", + "interval": 900, + "temperature_2m": 20.7, + "relative_humidity_2m": 72, + "precipitation": 0.0 + }, + "tool_seconds": 0.552, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.63, + "tok_s": 17.33, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8783, + "expert_hits": 6654, + "expert_misses": 2129, + "hit_rate": 75.76, + "ssb_demand_bytes": 3034238976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3034238976, + "ssd_bytes_per_token": 18278500.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5211997, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3036348416, + "rss_bytes": 7926669312, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 278, + "hit": 195, + "miss": 83, + "hit_rate": 70.1439, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r10_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently has a temperature of 20.7°C, 72% humidity, and no precipitation.", + "peak_rss_mb": 7556.5, + "peak_anon_mb": 7074.1, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 25, + "seconds": 1.37, + "tok_s": 18.27, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + "pager": { + "phase": "decode", + "tokens": 156, + "expert_requests": 9311, + "expert_hits": 7024, + "expert_misses": 2287, + "hit_rate": 75.4377, + "ssb_demand_bytes": 3258114048, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3258114048, + "ssd_bytes_per_token": 20885300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5085189, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3260366848, + "rss_bytes": 7923609600, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 479, + "hit": 362, + "miss": 117, + "hit_rate": 75.5741, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 421, + "hit": 310, + "miss": 111, + "hit_rate": 73.6342, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 426, + "hit": 323, + "miss": 103, + "hit_rate": 75.8216, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 414, + "hit": 311, + "miss": 103, + "hit_rate": 75.1208, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 397, + "hit": 298, + "miss": 99, + "hit_rate": 75.063, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 397, + "hit": 292, + "miss": 105, + "hit_rate": 73.5516, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 416, + "hit": 316, + "miss": 100, + "hit_rate": 75.9615, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 416, + "hit": 308, + "miss": 108, + "hit_rate": 74.0385, + "demand_bytes": 165224448, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 400, + "hit": 300, + "miss": 100, + "hit_rate": 75, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 378, + "hit": 290, + "miss": 88, + "hit_rate": 76.7196, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 390, + "hit": 299, + "miss": 91, + "hit_rate": 76.6667, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 392, + "hit": 289, + "miss": 103, + "hit_rate": 73.7245, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 389, + "hit": 292, + "miss": 97, + "hit_rate": 75.0643, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 394, + "hit": 299, + "miss": 95, + "hit_rate": 75.8883, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 406, + "hit": 309, + "miss": 97, + "hit_rate": 76.1084, + "demand_bytes": 128729088, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 416, + "hit": 312, + "miss": 104, + "hit_rate": 75, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 389, + "hit": 294, + "miss": 95, + "hit_rate": 75.5784, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 298, + "miss": 93, + "hit_rate": 76.2148, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 293, + "miss": 94, + "hit_rate": 75.7106, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 408, + "hit": 313, + "miss": 95, + "hit_rate": 76.7157, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 401, + "hit": 304, + "miss": 97, + "hit_rate": 75.8105, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 414, + "hit": 315, + "miss": 99, + "hit_rate": 76.087, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r10_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.532, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 7548.6, + "peak_anon_mb": 7064.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.69, + "tok_s": 18.73, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9118, + "expert_hits": 7038, + "expert_misses": 2080, + "hit_rate": 77.188, + "ssb_demand_bytes": 2963533824, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2963533824, + "ssd_bytes_per_token": 17031800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4810864, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 2965585920, + "rss_bytes": 7915311104, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 273, + "hit": 192, + "miss": 81, + "hit_rate": 70.3297, + "demand_bytes": 123918336, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r10_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 7559.1, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.45, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 4255933, + "prefetch_wait_us": 0, + "arena_bytes": 7275913216, + "arena_used_bytes": 3114770432, + "rss_bytes": 7926300672, + "n_slots": 5279, + "n_small_slots": 3960, + "n_big_slots": 1319, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_8192_r10_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 7559.5, + "peak_anon_mb": 7075.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.63, + "tok_s": 17.33, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 25, + "seconds": 1.37, + "tok_s": 18.27, + "ssd_mb": 244.4, + "ssd_mb_per_token": 9.8 + }, + { + "tokens": 13, + "seconds": 0.69, + "tok_s": 18.73, + "ssd_mb": 29.6, + "ssd_mb_per_token": 2.3 + }, + { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.45, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.445, + "mean_ssd_mb_per_token": 7.15, + "mean_hot_rate_pct": 74.7719, + "max_expert_requests": 9311 + }, + { + "level_mb": 16384, + "repeat": 1, + "status": "OK", + "wall_s": 73.94, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.56, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15096.1, + "peak_anon_mb": 14611.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.64, + "tok_s": 17.12, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8838, + "expert_hits": 6707, + "expert_misses": 2131, + "hit_rate": 75.8882, + "ssb_demand_bytes": 3037298688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3037298688, + "ssd_bytes_per_token": 18297000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5726073, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3039412224, + "rss_bytes": 15829430272, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 333, + "hit": 248, + "miss": 85, + "hit_rate": 74.4745, + "demand_bytes": 130037760, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r1_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.81, + "tok_s": 17.68, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6336199, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827816448, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r1_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.524, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15088.6, + "peak_anon_mb": 14604.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.67, + "tok_s": 19.47, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 7088, + "expert_misses": 2082, + "hit_rate": 77.2955, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6707878, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2968649728, + "rss_bytes": 15821578240, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 242, + "miss": 83, + "hit_rate": 74.4615, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r1_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15095.3, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.39, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7469, + "expert_hits": 5283, + "expert_misses": 2186, + "hit_rate": 70.7324, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5890696, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15828598784, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 298, + "hit": 211, + "miss": 87, + "hit_rate": 70.8054, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r1_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15096.1, + "peak_anon_mb": 14611.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.64, + "tok_s": 17.12, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.81, + "tok_s": 17.68, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.67, + "tok_s": 19.47, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 0.97, + "tok_s": 15.39, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.415, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.56597500000001, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 2, + "status": "OK", + "wall_s": 77.93, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.557, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15095.0, + "peak_anon_mb": 14610.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.65, + "tok_s": 16.95, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 6706, + "expert_misses": 2130, + "hit_rate": 75.8941, + "ssb_demand_bytes": 3035768832, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3035768832, + "ssd_bytes_per_token": 18287800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5794905, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3037880320, + "rss_bytes": 15828283392, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 247, + "miss": 84, + "hit_rate": 74.6224, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r2_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.2, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.81, + "tok_s": 17.65, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6178972, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827513344, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r2_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.512, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15088.1, + "peak_anon_mb": 14604.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.73, + "tok_s": 17.79, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 7088, + "expert_misses": 2082, + "hit_rate": 77.2955, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5951084, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2968649728, + "rss_bytes": 15821041664, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 242, + "miss": 83, + "hit_rate": 74.4615, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r2_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15097.1, + "peak_anon_mb": 14612.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.01, + "tok_s": 14.88, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6745795, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3114770432, + "rss_bytes": 15830405120, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r2_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15097.1, + "peak_anon_mb": 14612.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.65, + "tok_s": 16.95, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.81, + "tok_s": 17.65, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.73, + "tok_s": 17.79, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 1.01, + "tok_s": 14.88, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 16.8175, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.559825, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 3, + "status": "OK", + "wall_s": 76.92, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.551, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15090.6, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.66, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8779, + "expert_hits": 6653, + "expert_misses": 2126, + "hit_rate": 75.7831, + "ssb_demand_bytes": 3029649408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3029649408, + "ssd_bytes_per_token": 18250900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6353012, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3031752704, + "rss_bytes": 15823671296, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 194, + "miss": 80, + "hit_rate": 70.8029, + "demand_bytes": 122388480, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r3_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.2, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.95, + "tok_s": 16.41, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7254888, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827443712, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r3_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.57, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15088.9, + "peak_anon_mb": 14604.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.6, + "tok_s": 21.72, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 7088, + "expert_misses": 2082, + "hit_rate": 77.2955, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5893681, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2968649728, + "rss_bytes": 15821824000, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 242, + "miss": 83, + "hit_rate": 74.4615, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r3_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15095.1, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.92, + "tok_s": 16.3, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7458, + "expert_hits": 5272, + "expert_misses": 2186, + "hit_rate": 70.6892, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6606064, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15828402176, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 287, + "hit": 200, + "miss": 87, + "hit_rate": 69.6864, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r3_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15095.1, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.66, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.95, + "tok_s": 16.41, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.6, + "tok_s": 21.72, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 0.92, + "tok_s": 16.3, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 18.2725, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.52890000000001, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 4, + "status": "OK", + "wall_s": 81.4, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.558, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15090.8, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.61, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 6691, + "expert_misses": 2130, + "hit_rate": 75.8531, + "ssb_demand_bytes": 3035768832, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3035768832, + "ssd_bytes_per_token": 18287800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7341633, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3037880320, + "rss_bytes": 15823835136, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 232, + "miss": 84, + "hit_rate": 73.4177, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r4_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15093.4, + "peak_anon_mb": 14610.5, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.69, + "tok_s": 18.9, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10586, + "expert_hits": 8291, + "expert_misses": 2295, + "hit_rate": 78.3204, + "ssb_demand_bytes": 3270150144, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3270150144, + "ssd_bytes_per_token": 19236200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5840233, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 3272417280, + "rss_bytes": 15826563072, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 433, + "hit": 338, + "miss": 95, + "hit_rate": 78.06, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r4_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.516, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15094.8, + "peak_anon_mb": 14610.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.23, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9168, + "expert_hits": 7086, + "expert_misses": 2082, + "hit_rate": 77.2906, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5882820, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 2968649728, + "rss_bytes": 15828070400, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 323, + "hit": 240, + "miss": 83, + "hit_rate": 74.3034, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r4_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15094.8, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.02, + "tok_s": 14.76, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7469, + "expert_hits": 5283, + "expert_misses": 2186, + "hit_rate": 70.7324, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7189708, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15827935232, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 298, + "hit": 211, + "miss": 87, + "hit_rate": 70.8054, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r4_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15094.8, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.59, + "tok_s": 18.61, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.69, + "tok_s": 18.9, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.75, + "tok_s": 17.23, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 1.02, + "tok_s": 14.76, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.375, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.549125, + "max_expert_requests": 10586 + }, + { + "level_mb": 16384, + "repeat": 5, + "status": "OK", + "wall_s": 81.28, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.916, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15090.9, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.53, + "tok_s": 20.79, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8779, + "expert_hits": 6653, + "expert_misses": 2126, + "hit_rate": 75.7831, + "ssb_demand_bytes": 3029649408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3029649408, + "ssd_bytes_per_token": 18250900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5892548, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3031752704, + "rss_bytes": 15823925248, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 194, + "miss": 80, + "hit_rate": 70.8029, + "demand_bytes": 122388480, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r5_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.1, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.9, + "tok_s": 16.81, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6657761, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827374080, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r5_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.529, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15091.1, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.65, + "tok_s": 20.15, + "ssd_mb": 25.2, + "ssd_mb_per_token": 1.9 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9180, + "expert_hits": 7097, + "expert_misses": 2083, + "hit_rate": 77.3094, + "ssb_demand_bytes": 2968123392, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2968123392, + "ssd_bytes_per_token": 17058200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6545242, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2970181632, + "rss_bytes": 15824179200, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 335, + "hit": 251, + "miss": 84, + "hit_rate": 74.9254, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r5_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.04, + "tok_s": 14.47, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7469, + "expert_hits": 5283, + "expert_misses": 2186, + "hit_rate": 70.7324, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7110120, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15828869120, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 298, + "hit": 211, + "miss": 87, + "hit_rate": 70.8054, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r5_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.53, + "tok_s": 20.79, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.9, + "tok_s": 16.81, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.65, + "tok_s": 20.15, + "ssd_mb": 25.2, + "ssd_mb_per_token": 1.9 + }, + { + "tokens": 15, + "seconds": 1.04, + "tok_s": 14.47, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 18.055, + "mean_ssd_mb_per_token": 6.5, + "mean_hot_rate_pct": 75.543175, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 6, + "status": "OK", + "wall_s": 79.78, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.937, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15093.1, + "peak_anon_mb": 14610.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.61, + "tok_s": 18.05, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8836, + "expert_hits": 6706, + "expert_misses": 2130, + "hit_rate": 75.8941, + "ssb_demand_bytes": 3035768832, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3035768832, + "ssd_bytes_per_token": 18287800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6511636, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3037880320, + "rss_bytes": 15825883136, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 331, + "hit": 247, + "miss": 84, + "hit_rate": 74.6224, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r6_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.7, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.75, + "tok_s": 18.33, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10599, + "expert_hits": 8303, + "expert_misses": 2296, + "hit_rate": 78.3376, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6924978, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827914752, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 350, + "miss": 96, + "hit_rate": 78.4753, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r6_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.527, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15094.8, + "peak_anon_mb": 14610.6, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.76, + "tok_s": 17.19, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9168, + "expert_hits": 7086, + "expert_misses": 2082, + "hit_rate": 77.2906, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7882430, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 2968649728, + "rss_bytes": 15828000768, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 323, + "hit": 240, + "miss": 83, + "hit_rate": 74.3034, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r6_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15093.9, + "peak_anon_mb": 14610.4, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 1.0, + "tok_s": 15.07, + "ssd_mb": 208.7, + "ssd_mb_per_token": 13.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7445, + "expert_hits": 5260, + "expert_misses": 2185, + "hit_rate": 70.6514, + "ssb_demand_bytes": 3115653120, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3115653120, + "ssd_bytes_per_token": 23783600.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5589144, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 3117834240, + "rss_bytes": 15827128320, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 274, + "hit": 188, + "miss": 86, + "hit_rate": 68.6131, + "demand_bytes": 131567616, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r6_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15094.8, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.61, + "tok_s": 18.05, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.75, + "tok_s": 18.33, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.76, + "tok_s": 17.19, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 1.0, + "tok_s": 15.07, + "ssd_mb": 208.7, + "ssd_mb_per_token": 13.9 + } + ], + "mean_decode_tok_s": 17.16, + "mean_ssd_mb_per_token": 6.575, + "mean_hot_rate_pct": 75.543425, + "max_expert_requests": 10599 + }, + { + "level_mb": 16384, + "repeat": 7, + "status": "OK", + "wall_s": 77.9, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.943, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15087.1, + "peak_anon_mb": 14603.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.58, + "tok_s": 19.12, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8776, + "expert_hits": 6651, + "expert_misses": 2125, + "hit_rate": 75.7862, + "ssb_demand_bytes": 3028119552, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3028119552, + "ssd_bytes_per_token": 18241700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6270594, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 3030220800, + "rss_bytes": 15819976704, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 271, + "hit": 192, + "miss": 79, + "hit_rate": 70.8487, + "demand_bytes": 120858624, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r7_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.2, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.67, + "tok_s": 19.15, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10599, + "expert_hits": 8303, + "expert_misses": 2296, + "hit_rate": 78.3376, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6764145, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827406848, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 446, + "hit": 350, + "miss": 96, + "hit_rate": 78.4753, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r7_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.505, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15088.5, + "peak_anon_mb": 14604.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.81, + "tok_s": 16.12, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 7088, + "expert_misses": 2082, + "hit_rate": 77.2955, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5636142, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2968649728, + "rss_bytes": 15821451264, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 242, + "miss": 83, + "hit_rate": 74.4615, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r7_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.98, + "tok_s": 15.36, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7469, + "expert_hits": 5283, + "expert_misses": 2186, + "hit_rate": 70.7324, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5820820, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15828860928, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 298, + "hit": 211, + "miss": 87, + "hit_rate": 70.8054, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r7_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.58, + "tok_s": 19.12, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.67, + "tok_s": 19.15, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.81, + "tok_s": 16.12, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 0.98, + "tok_s": 15.36, + "ssd_mb": 207.3, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.4375, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.537925, + "max_expert_requests": 10599 + }, + { + "level_mb": 16384, + "repeat": 8, + "status": "OK", + "wall_s": 76.46, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.916, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15096.6, + "peak_anon_mb": 14611.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.18, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8828, + "expert_hits": 6697, + "expert_misses": 2131, + "hit_rate": 75.8609, + "ssb_demand_bytes": 3037298688, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3037298688, + "ssd_bytes_per_token": 18297000.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6344462, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3039412224, + "rss_bytes": 15829970944, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 323, + "hit": 238, + "miss": 85, + "hit_rate": 73.6842, + "demand_bytes": 130037760, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r8_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.59, + "tok_s": 20.09, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 7186356, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827828736, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r8_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.55, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15090.1, + "peak_anon_mb": 14605.5, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.67, + "tok_s": 19.5, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9172, + "expert_hits": 7089, + "expert_misses": 2083, + "hit_rate": 77.2896, + "ssb_demand_bytes": 2968123392, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2968123392, + "ssd_bytes_per_token": 17058200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6319629, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2970181632, + "rss_bytes": 15823110144, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 327, + "hit": 243, + "miss": 84, + "hit_rate": 74.3119, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r8_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15088.5, + "peak_anon_mb": 14605.3, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.92, + "tok_s": 16.3, + "ssd_mb": 208.7, + "ssd_mb_per_token": 13.9 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7440, + "expert_hits": 5256, + "expert_misses": 2184, + "hit_rate": 70.6452, + "ssb_demand_bytes": 3114123264, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3114123264, + "ssd_bytes_per_token": 23771900.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5635511, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 3116302336, + "rss_bytes": 15821422592, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 269, + "hit": 184, + "miss": 85, + "hit_rate": 68.4015, + "demand_bytes": 130037760, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r8_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15096.6, + "peak_anon_mb": 14611.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.72, + "tok_s": 15.18, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.59, + "tok_s": 20.09, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.67, + "tok_s": 19.5, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 0.92, + "tok_s": 16.3, + "ssd_mb": 208.7, + "ssd_mb_per_token": 13.9 + } + ], + "mean_decode_tok_s": 17.7675, + "mean_ssd_mb_per_token": 6.575, + "mean_hot_rate_pct": 75.535875, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 9, + "status": "OK", + "wall_s": 76.95, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 0.931, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15090.6, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.57, + "tok_s": 19.16, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8821, + "expert_hits": 6691, + "expert_misses": 2130, + "hit_rate": 75.8531, + "ssb_demand_bytes": 3035768832, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3035768832, + "ssd_bytes_per_token": 18287800.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6782883, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3037880320, + "rss_bytes": 15823589376, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 316, + "hit": 232, + "miss": 84, + "hit_rate": 73.4177, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r9_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.9, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.78, + "tok_s": 17.99, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6658725, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15828156416, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r9_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.52, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15091.0, + "peak_anon_mb": 14606.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.79, + "tok_s": 16.44, + "ssd_mb": 25.2, + "ssd_mb_per_token": 1.9 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9180, + "expert_hits": 7097, + "expert_misses": 2083, + "hit_rate": 77.3094, + "ssb_demand_bytes": 2968123392, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2968123392, + "ssd_bytes_per_token": 17058200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5313521, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 2970181632, + "rss_bytes": 15824039936, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 335, + "hit": 251, + "miss": 84, + "hit_rate": 74.9254, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r9_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.88, + "tok_s": 16.99, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7469, + "expert_hits": 5283, + "expert_misses": 2186, + "hit_rate": 70.7324, + "ssb_demand_bytes": 3117182976, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3117182976, + "ssd_bytes_per_token": 23795300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5714652, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3119366144, + "rss_bytes": 15828926464, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 298, + "hit": 211, + "miss": 87, + "hit_rate": 70.8054, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r9_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15095.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.57, + "tok_s": 19.16, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.78, + "tok_s": 17.99, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.79, + "tok_s": 16.44, + "ssd_mb": 25.2, + "ssd_mb_per_token": 1.9 + }, + { + "tokens": 15, + "seconds": 0.88, + "tok_s": 16.99, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 17.645, + "mean_ssd_mb_per_token": 6.5, + "mean_hot_rate_pct": 75.560675, + "max_expert_requests": 10604 + }, + { + "level_mb": 16384, + "repeat": 10, + "status": "OK", + "wall_s": 79.63, + "tool_calls": [ + { + "tool": "get_weather", + "call_output": "get_weathercity\nTokyo", + "result": { + "city": "Tokyo", + "country": "Japan", + "time": "2026-10-13T10:45", + "interval": 900, + "temperature_2m": 20.9, + "relative_humidity_2m": 71, + "precipitation": 0.0 + }, + "tool_seconds": 1.129, + "model_call": { + "exit_code": 0, + "text": "get_weathercity\nTokyo", + "peak_rss_mb": 15086.9, + "peak_anon_mb": 14603.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 11, + "seconds": 0.62, + "tok_s": 17.67, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + "pager": { + "phase": "decode", + "tokens": 166, + "expert_requests": 8776, + "expert_hits": 6651, + "expert_misses": 2125, + "hit_rate": 75.7862, + "ssb_demand_bytes": 3028119552, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3028119552, + "ssd_bytes_per_token": 18241700.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 5493391, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 3030220800, + "rss_bytes": 15819747328, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 451, + "hit": 341, + "miss": 110, + "hit_rate": 75.6098, + "demand_bytes": 168284160, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 444, + "hit": 337, + "miss": 107, + "hit_rate": 75.9009, + "demand_bytes": 163694592, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 431, + "hit": 325, + "miss": 106, + "hit_rate": 75.406, + "demand_bytes": 140673024, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 410, + "hit": 309, + "miss": 101, + "hit_rate": 75.3659, + "demand_bytes": 134037504, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 416, + "hit": 319, + "miss": 97, + "hit_rate": 76.6827, + "demand_bytes": 148396032, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 386, + "hit": 290, + "miss": 96, + "hit_rate": 75.1295, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 406, + "hit": 306, + "miss": 100, + "hit_rate": 75.3695, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 411, + "hit": 306, + "miss": 105, + "hit_rate": 74.4526, + "demand_bytes": 160634880, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 362, + "hit": 276, + "miss": 86, + "hit_rate": 76.2431, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 337, + "hit": 253, + "miss": 84, + "hit_rate": 75.0742, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 315, + "hit": 241, + "miss": 74, + "hit_rate": 76.5079, + "demand_bytes": 113209344, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 352, + "hit": 269, + "miss": 83, + "hit_rate": 76.4205, + "demand_bytes": 110149632, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 354, + "hit": 267, + "miss": 87, + "hit_rate": 75.4237, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 372, + "hit": 287, + "miss": 85, + "hit_rate": 77.1505, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 390, + "hit": 297, + "miss": 93, + "hit_rate": 76.1538, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 386, + "hit": 294, + "miss": 92, + "hit_rate": 76.1658, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 391, + "hit": 301, + "miss": 90, + "hit_rate": 76.9821, + "demand_bytes": 119439360, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 357, + "hit": 269, + "miss": 88, + "hit_rate": 75.3501, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 401, + "hit": 312, + "miss": 89, + "hit_rate": 77.8055, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 389, + "hit": 298, + "miss": 91, + "hit_rate": 76.6067, + "demand_bytes": 139216896, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 390, + "hit": 295, + "miss": 95, + "hit_rate": 75.641, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 271, + "hit": 192, + "miss": 79, + "hit_rate": 70.8487, + "demand_bytes": 120858624, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r10_get_weather.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": ", Japan, currently 10:45, with a temperature of 20.9°C, 71% humidity, and no precipitation.", + "peak_rss_mb": 15094.6, + "peak_anon_mb": 14611.7, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 32, + "seconds": 1.74, + "tok_s": 18.35, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + "pager": { + "phase": "decode", + "tokens": 170, + "expert_requests": 10604, + "expert_hits": 8308, + "expert_misses": 2296, + "hit_rate": 78.3478, + "ssb_demand_bytes": 3271680000, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3271680000, + "ssd_bytes_per_token": 19245200.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6993422, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3273949184, + "rss_bytes": 15827857408, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 534, + "hit": 417, + "miss": 117, + "hit_rate": 78.0899, + "demand_bytes": 178993152, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 479, + "hit": 368, + "miss": 111, + "hit_rate": 76.8267, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 483, + "hit": 380, + "miss": 103, + "hit_rate": 78.6749, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 468, + "hit": 364, + "miss": 104, + "hit_rate": 77.7778, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 453, + "hit": 355, + "miss": 98, + "hit_rate": 78.3664, + "demand_bytes": 149925888, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 451, + "hit": 348, + "miss": 103, + "hit_rate": 77.1619, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 472, + "hit": 372, + "miss": 100, + "hit_rate": 78.8136, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 473, + "hit": 364, + "miss": 109, + "hit_rate": 76.9556, + "demand_bytes": 166754304, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 455, + "hit": 355, + "miss": 100, + "hit_rate": 78.022, + "demand_bytes": 132710400, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 431, + "hit": 344, + "miss": 87, + "hit_rate": 79.8144, + "demand_bytes": 115458048, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 444, + "hit": 349, + "miss": 95, + "hit_rate": 78.6036, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 447, + "hit": 340, + "miss": 107, + "hit_rate": 76.0626, + "demand_bytes": 142000128, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 446, + "hit": 347, + "miss": 99, + "hit_rate": 77.8027, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 450, + "hit": 355, + "miss": 95, + "hit_rate": 78.8889, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 467, + "hit": 373, + "miss": 94, + "hit_rate": 79.8715, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 472, + "hit": 368, + "miss": 104, + "hit_rate": 77.9661, + "demand_bytes": 138018816, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 449, + "hit": 353, + "miss": 96, + "hit_rate": 78.6192, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 447, + "hit": 353, + "miss": 94, + "hit_rate": 78.9709, + "demand_bytes": 124747776, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 442, + "hit": 349, + "miss": 93, + "hit_rate": 78.9593, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 458, + "hit": 362, + "miss": 96, + "hit_rate": 79.0393, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 460, + "hit": 364, + "miss": 96, + "hit_rate": 79.1304, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 472, + "hit": 373, + "miss": 99, + "hit_rate": 79.0254, + "demand_bytes": 151455744, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 451, + "hit": 355, + "miss": 96, + "hit_rate": 78.714, + "demand_bytes": 146866176, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r10_get_weather_return.log" + } + }, + { + "tool": "compile_project", + "call_output": "compile_projectproject\ntiny_fixture\n", + "result": { + "returncode": 0, + "tool": "cmake --build", + "tail": "[ 50%] Building CXX object CMakeFiles/tiny_fixture.dir/main.cpp.o\n[100%] Linking CXX executable tiny_fixture\n[100%] Built target tiny_fixture\n\n" + }, + "tool_seconds": 0.554, + "model_call": { + "exit_code": 0, + "text": "compile_projectproject\ntiny_fixture\n", + "peak_rss_mb": 15087.2, + "peak_anon_mb": 14603.0, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 13, + "seconds": 0.68, + "tok_s": 19.17, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + "pager": { + "phase": "decode", + "tokens": 174, + "expert_requests": 9170, + "expert_hits": 7088, + "expert_misses": 2082, + "hit_rate": 77.2955, + "ssb_demand_bytes": 2966593536, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 2966593536, + "ssd_bytes_per_token": 17049400.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6072346, + "prefetch_wait_us": 0, + "arena_bytes": 15177883648, + "arena_used_bytes": 2968649728, + "rss_bytes": 15820062720, + "n_slots": 11012, + "n_small_slots": 8259, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 476, + "hit": 365, + "miss": 111, + "hit_rate": 76.6807, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 462, + "hit": 356, + "miss": 106, + "hit_rate": 77.0563, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 451, + "hit": 346, + "miss": 105, + "hit_rate": 76.7184, + "demand_bytes": 139345920, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 430, + "hit": 331, + "miss": 99, + "hit_rate": 76.9767, + "demand_bytes": 131383296, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 428, + "hit": 333, + "miss": 95, + "hit_rate": 77.8037, + "demand_bytes": 145336320, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 400, + "hit": 307, + "miss": 93, + "hit_rate": 76.75, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 414, + "hit": 319, + "miss": 95, + "hit_rate": 77.0531, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 430, + "hit": 329, + "miss": 101, + "hit_rate": 76.5116, + "demand_bytes": 154515456, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 365, + "hit": 279, + "miss": 86, + "hit_rate": 76.4384, + "demand_bytes": 114130944, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 348, + "hit": 266, + "miss": 82, + "hit_rate": 76.4368, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 332, + "hit": 261, + "miss": 71, + "hit_rate": 78.6145, + "demand_bytes": 108619776, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 362, + "hit": 282, + "miss": 80, + "hit_rate": 77.9006, + "demand_bytes": 106168320, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 371, + "hit": 287, + "miss": 84, + "hit_rate": 77.3585, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 353, + "hit": 270, + "miss": 83, + "hit_rate": 76.4873, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 380, + "hit": 298, + "miss": 82, + "hit_rate": 78.4211, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 415, + "hit": 322, + "miss": 93, + "hit_rate": 77.5904, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 411, + "hit": 321, + "miss": 90, + "hit_rate": 78.1022, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 407, + "hit": 316, + "miss": 91, + "hit_rate": 77.6413, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 387, + "hit": 299, + "miss": 88, + "hit_rate": 77.261, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 406, + "hit": 319, + "miss": 87, + "hit_rate": 78.5714, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 412, + "hit": 323, + "miss": 89, + "hit_rate": 78.3981, + "demand_bytes": 136157184, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 405, + "hit": 317, + "miss": 88, + "hit_rate": 78.2716, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 325, + "hit": 242, + "miss": 83, + "hit_rate": 74.4615, + "demand_bytes": 126978048, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r10_compile_project.log" + }, + "tool_result_summary": { + "exit_code": 0, + "text": "build completed successfully, building and linking the `tiny_fixture` executable.", + "peak_rss_mb": 15097.1, + "peak_anon_mb": 14612.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode": { + "tokens": 15, + "seconds": 0.88, + "tok_s": 17.05, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + }, + "pager": { + "phase": "decode", + "tokens": 131, + "expert_requests": 7451, + "expert_hits": 5268, + "expert_misses": 2183, + "hit_rate": 70.7019, + "ssb_demand_bytes": 3112593408, + "ssb_prefetch_bytes": 0, + "ssb_total_bytes": 3112593408, + "ssd_bytes_per_token": 23760300.0, + "prefetch_issued": 0, + "prefetch_used": 0, + "prefetch_used_rate": 0, + "prefetch_bytes_used": 0, + "prefetch_bytes_wasted": 0, + "prefetch_waste_rate": 0, + "prefetch_duplicate": 0, + "evictions": 0, + "demand_wait_us": 6669350, + "prefetch_wait_us": 0, + "arena_bytes": 15179210752, + "arena_used_bytes": 3114770432, + "rss_bytes": 15830413312, + "n_slots": 11013, + "n_small_slots": 8260, + "n_big_slots": 2753, + "small_slot_bytes": 1327104, + "big_slot_bytes": 1531904, + "slot_bytes": 1531904, + "n_moe_layers": 23, + "expert_bytes_total": 4192468992, + "arch": "bailingmoe3", + "name": "Ling-3.0-tiny", + "per_layer": [ + { + "layer": 0, + "req": 0, + "hit": 0, + "miss": 0, + "hit_rate": 0, + "demand_bytes": 0, + "prefetch_bytes": 0 + }, + { + "layer": 1, + "req": 407, + "hit": 293, + "miss": 114, + "hit_rate": 71.9902, + "demand_bytes": 174403584, + "prefetch_bytes": 0 + }, + { + "layer": 2, + "req": 367, + "hit": 256, + "miss": 111, + "hit_rate": 69.7548, + "demand_bytes": 169814016, + "prefetch_bytes": 0 + }, + { + "layer": 3, + "req": 344, + "hit": 241, + "miss": 103, + "hit_rate": 70.0581, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 4, + "req": 353, + "hit": 250, + "miss": 103, + "hit_rate": 70.8215, + "demand_bytes": 136691712, + "prefetch_bytes": 0 + }, + { + "layer": 5, + "req": 345, + "hit": 241, + "miss": 104, + "hit_rate": 69.8551, + "demand_bytes": 159105024, + "prefetch_bytes": 0 + }, + { + "layer": 6, + "req": 324, + "hit": 228, + "miss": 96, + "hit_rate": 70.3704, + "demand_bytes": 127401984, + "prefetch_bytes": 0 + }, + { + "layer": 7, + "req": 341, + "hit": 243, + "miss": 98, + "hit_rate": 71.261, + "demand_bytes": 130056192, + "prefetch_bytes": 0 + }, + { + "layer": 8, + "req": 348, + "hit": 242, + "miss": 106, + "hit_rate": 69.5402, + "demand_bytes": 162164736, + "prefetch_bytes": 0 + }, + { + "layer": 9, + "req": 297, + "hit": 212, + "miss": 85, + "hit_rate": 71.3805, + "demand_bytes": 112803840, + "prefetch_bytes": 0 + }, + { + "layer": 10, + "req": 284, + "hit": 202, + "miss": 82, + "hit_rate": 71.1268, + "demand_bytes": 108822528, + "prefetch_bytes": 0 + }, + { + "layer": 11, + "req": 300, + "hit": 212, + "miss": 88, + "hit_rate": 70.6667, + "demand_bytes": 134627328, + "prefetch_bytes": 0 + }, + { + "layer": 12, + "req": 315, + "hit": 220, + "miss": 95, + "hit_rate": 69.8413, + "demand_bytes": 126074880, + "prefetch_bytes": 0 + }, + { + "layer": 13, + "req": 327, + "hit": 234, + "miss": 93, + "hit_rate": 71.5596, + "demand_bytes": 123420672, + "prefetch_bytes": 0 + }, + { + "layer": 14, + "req": 318, + "hit": 228, + "miss": 90, + "hit_rate": 71.6981, + "demand_bytes": 137687040, + "prefetch_bytes": 0 + }, + { + "layer": 15, + "req": 314, + "hit": 223, + "miss": 91, + "hit_rate": 71.0191, + "demand_bytes": 120766464, + "prefetch_bytes": 0 + }, + { + "layer": 16, + "req": 323, + "hit": 221, + "miss": 102, + "hit_rate": 68.4211, + "demand_bytes": 135364608, + "prefetch_bytes": 0 + }, + { + "layer": 17, + "req": 324, + "hit": 231, + "miss": 93, + "hit_rate": 71.2963, + "demand_bytes": 142276608, + "prefetch_bytes": 0 + }, + { + "layer": 18, + "req": 293, + "hit": 209, + "miss": 84, + "hit_rate": 71.3311, + "demand_bytes": 111476736, + "prefetch_bytes": 0 + }, + { + "layer": 19, + "req": 298, + "hit": 210, + "miss": 88, + "hit_rate": 70.4698, + "demand_bytes": 116785152, + "prefetch_bytes": 0 + }, + { + "layer": 20, + "req": 315, + "hit": 228, + "miss": 87, + "hit_rate": 72.381, + "demand_bytes": 133097472, + "prefetch_bytes": 0 + }, + { + "layer": 21, + "req": 323, + "hit": 229, + "miss": 94, + "hit_rate": 70.8978, + "demand_bytes": 143806464, + "prefetch_bytes": 0 + }, + { + "layer": 22, + "req": 311, + "hit": 219, + "miss": 92, + "hit_rate": 70.418, + "demand_bytes": 140746752, + "prefetch_bytes": 0 + }, + { + "layer": 23, + "req": 280, + "hit": 196, + "miss": 84, + "hit_rate": 70, + "demand_bytes": 128507904, + "prefetch_bytes": 0 + } + ] + }, + "stderr_path": "results/tool_use_runs/level_16384_r10_compile_project_return.log" + } + } + ], + "tools_passed": 2, + "peak_rss_mb": 15097.1, + "peak_anon_mb": 14612.9, + "peak_swap_mb": 0.0, + "anon_within_limit": true, + "swap_free": true, + "decode_segments": [ + { + "tokens": 11, + "seconds": 0.62, + "tok_s": 17.67, + "ssd_mb": 30.0, + "ssd_mb_per_token": 2.7 + }, + { + "tokens": 32, + "seconds": 1.74, + "tok_s": 18.35, + "ssd_mb": 244.6, + "ssd_mb_per_token": 7.6 + }, + { + "tokens": 13, + "seconds": 0.68, + "tok_s": 19.17, + "ssd_mb": 26.7, + "ssd_mb_per_token": 2.1 + }, + { + "tokens": 15, + "seconds": 0.88, + "tok_s": 17.05, + "ssd_mb": 207.2, + "ssd_mb_per_token": 13.8 + } + ], + "mean_decode_tok_s": 18.060000000000002, + "mean_ssd_mb_per_token": 6.55, + "mean_hot_rate_pct": 75.53285, + "max_expert_requests": 10604 + } + ] +} \ No newline at end of file