Buckets:
| run_id,engine,model,hardware,config,max_model_len,weights_reported_bytes,kv_pool_reported_bytes,recurrent_state_reported_bytes,other_device_memory_residual_bytes,device_ready_bytes,device_text_bytes,device_probe_bytes,estimate_error,estimation_assumptions,hf_mem_version,hf_mem_model_revision,hf_mem_weights_bytes,hf_mem_offloaded_bytes,hf_mem_requested_kv_bytes,hf_mem_overhead_bytes,hf_mem_requested_total_bytes,hf_mem_requested_batch_size,hf_mem_overhead_activations_bytes,hf_mem_overhead_encoder_bytes,hf_mem_overhead_graphs_bytes,hf_mem_overhead_runtime_bytes | |
| mm/sglang-a100-dsv2-lite/default,sglang,deepseek-ai/DeepSeek-V2-Lite-Chat,A100 80GB,default,32768,31632434135,37430639985,,4741996216,73805070336,73805070336,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,85864749cd611b4353ce1decdb286193298f64c7,31412968448,0,1019215872,5480658737,37912843057,1,2520723947,0,941300162,2018634628 | |
| mm/sglang-a100-gemma3-12b/default,sglang,google/gemma-3-12b-it,A100 80GB,default,32768,24835648389,45333379810,,5357803929,75526832128,75526832128,75526832128,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/sglang-a100-gemma3-1b/default,sglang,google/gemma-3-1b-it,A100 80GB,default,32768,2147483648,68010807132,,4322062500,74480353280,74482450432,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/sglang-a100-gemma3-4b/default,sglang,google/gemma-3-4b-it,A100 80GB,default,32768,8922794557,61246233640,,4732852635,74901880832,74901880832,74901880832,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/sglang-a100-gemma4-e4b/default,sglang,google/gemma-4-E4B-it,A100 80GB,default,32768,16449724744,35970351104,,21577932472,73998008320,73998008320,74115448832,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/sglang-a100-gpt-oss-20b/default,sglang,openai/gpt-oss-20b,A100 80GB,default,32768,14012330803,24953759990,,35105317847,74071408640,74071408640,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6cee5e81ee83917806bbde320786a8fb61efebee,13761264768,0,1211498496,6349153636,21321916900,1,3864226744,0,466292264,2018634628 | |
| mm/sglang-a100-internvl35-8b/default,sglang,OpenGVLab/InternVL3_5-8B,A100 80GB,default,32768,17126182093,49907519980,,3640320327,70674022400,70678216704,71332528128,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/sglang-a100-llama31-8b/default,sglang,meta-llama/Llama-3.1-8B-Instruct,A100 80GB,default,32768,16127602196,54052163420,,3835019920,74014785536,74014785536,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/sglang-a100-llama32-1b/default,sglang,meta-llama/Llama-3.2-1B-Instruct,A100 80GB,default,32768,2523293286,67645734912,,3153697178,73322725376,73322725376,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9213176726f574b556790deb65791e0c5aa438b6,2471628800,0,1073741824,5432137214,8977507838,1,2849146859,0,564355727,2018634628 | |
| mm/sglang-a100-llava-ov-7b/default,sglang,llava-hf/llava-onevision-qwen2-7b-ov-hf,A100 80GB,default,32768,16213501542,48339856916,,3100765062,67654123520,67654123520,68211965952,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/sglang-a100-mistral-7b/default,sglang,mistralai/Mistral-7B-Instruct-v0.3,A100 80GB,default,32768,14516989460,55662776156,,3734356624,73914122240,73914122240,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c170c708c41dac9275d15a8fff4eca08d52bab71,14496047104,0,4294967296,4661020260,23452034660,1,1857717534,0,784668098,2018634628 | |
| mm/sglang-a100-mistral-small31/default,sglang,mistralai/Mistral-Small-3.1-24B-Instruct-2503,A100 80GB,default,32768,48092896297,18940805776,,2797265223,69830967296,69830967296,71257030656,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/sglang-a100-nemotron-9b/default,sglang,nvidia/NVIDIA-Nemotron-Nano-9B-v2,A100 80GB,default,32768,17802639442,27681064222,,26622673552,72106377216,72108474368,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/sglang-a100-phi4mm/default,sglang,microsoft/Phi-4-multimodal-instruct,A100 80GB,default,32768,9524089979,55018531062,,2822095503,67364716544,67364716544,67618471936,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,93f923e1a7727d1c4f446756212d9d3e8fcc5d81,11148920768,0,4294967296,6495759205,21939647269,1,3871981701,0,605142876,2018634628 | |
| mm/sglang-a100-qwen25vl-7b/default,sglang,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,default,32768,16857746637,46407121634,,3114186833,66379055104,66383249408,70413975552,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/sglang-a100-qwen2audio-7b-ml8k/default,sglang,Qwen/Qwen2-Audio-7B-Instruct,A100 80GB,default,8192,16943645983,53644141528,,2061752073,72649539584,72693579776,72785854464,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/sglang-a100-qwen3-06b/default,sglang,Qwen/Qwen3-0.6B,A100 80GB,default,32768,1267015352,68891275428,,3126685860,73284976640,73284976640,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c1899de289a04d12100db370d81485cdf75e47ca,1503264768,0,3758096384,5474281111,10735642263,1,3039200593,0,416445890,2018634628 | |
| mm/sglang-a100-qwen3-30b/default,sglang,Qwen/Qwen3-30B-A3B-Instruct-2507,A100 80GB,default,32768,61224758804,8912057140,,3693420216,73830236160,73832333312,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/sglang-a100-qwen3-32b/default,sglang,Qwen/Qwen3-32B,A100 80GB,default,32768,65541200937,4617089844,,4930236579,75088527360,75088527360,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9216db5781bf21249d130ec9da846c4624c16137,65524246528,0,8589934592,6628119396,80742300516,1,3482119096,0,1127365672,2018634628 | |
| mm/sglang-a100-qwen3-8b/default,sglang,Qwen/Qwen3-8B,A100 80GB,default,32768,16406775071,53772990546,,3965043343,74144808960,74144808960,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,b968826d9c46dd6066d109eabc6255188de91218,16381470720,0,4831838208,6129804951,27343113879,1,3371389470,0,739780853,2018634628 | |
| mm/sglang-a100-qwen35-4b/default,sglang,Qwen/Qwen3.5-4B,A100 80GB,default,32768,9255654523,30193620090,,30480258827,69929533440,69929533440,72146223104,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/sglang-a100-qwen3vl-30b/default,sglang,Qwen/Qwen3-VL-30B-A3B-Instruct,A100 80GB,default,32768,62341450301,2899102924,,3023841527,68264394752,68264394752,71722598400,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/sglang-a100-qwen3vl-8b/default,sglang,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,default,32768,17738214932,47523813130,,3174333154,68436361216,68436361216,71894564864,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/sglang-h200-dsv2-lite/default,sglang,deepseek-ai/DeepSeek-V2-Lite-Chat,H200 141GB,default,32768,31643171553,102198746808,,3467011687,137308930048,137308930048,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,85864749cd611b4353ce1decdb286193298f64c7,31412968448,0,1019215872,5480658737,37912843057,1,2520723947,0,941300162,2018634628 | |
| mm/sglang-h200-gemma3-12b/default,sglang,google/gemma-3-12b-it,H200 141GB,default,32768,24835648389,110058536960,,4029551739,138923737088,138923737088,138923737088,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/sglang-h200-gemma3-1b/default,sglang,google/gemma-3-1b-it,H200 141GB,default,32768,2147483648,132735964282,,3008490374,137891938304,137891938304,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/sglang-h200-gemma3-4b/default,sglang,google/gemma-3-4b-it,H200 141GB,default,32768,8922794557,125971390792,,3282965627,138177150976,138177150976,138177150976,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/sglang-h200-gemma4-31b/default,sglang,google/gemma-4-31B-it,H200 141GB,default,32768,64005750129,52570399704,,10007944887,126584094720,126584094720,126806392832,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,842da3794eaa0b77d5f08bae87a17459d91ff475,62546177752,0,16958095360,8029191882,87533464994,1,4909663518,66608640,1034285096,2018634628 | |
| mm/sglang-h200-gemma4-e4b/default,sglang,google/gemma-4-E4B-it,H200 141GB,default,32768,16449724744,79070347920,,43183463464,138703536128,138703536128,138925834240,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/sglang-h200-gpt-oss-20b/default,sglang,openai/gpt-oss-20b,H200 141GB,default,32768,18865643848,51561082388,,67144347812,137571074048,137571074048,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6cee5e81ee83917806bbde320786a8fb61efebee,13761264768,0,1211498496,6349153636,21321916900,1,3864226744,0,466292264,2018634628 | |
| mm/sglang-h200-internvl35-8b/default,sglang,OpenGVLab/InternVL3_5-8B,H200 141GB,default,32768,17126182093,111475876168,,3141030379,131743088640,131745185792,132399497216,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/sglang-h200-llama31-8b/default,sglang,meta-llama/Llama-3.1-8B-Instruct,H200 141GB,default,32768,16127602196,118755845734,,3165776774,138049224704,138049224704,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/sglang-h200-llama32-1b/default,sglang,meta-llama/Llama-3.2-1B-Instruct,H200 141GB,default,32768,2523293286,132370892062,,2637042812,137531228160,137531228160,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9213176726f574b556790deb65791e0c5aa438b6,2471628800,0,1073741824,5432137214,8977507838,1,2849146859,0,564355727,2018634628 | |
| mm/sglang-h200-llava-ov-7b/default,sglang,llava-hf/llava-onevision-qwen2-7b-ov-hf,H200 141GB,default,32768,16213501542,107610405602,,2466586296,126290493440,126290493440,127060148224,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/sglang-h200-mistral-7b/default,sglang,mistralai/Mistral-7B-Instruct-v0.3,H200 141GB,default,32768,14516989460,120366458470,,3065113478,137948561408,137948561408,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c170c708c41dac9275d15a8fff4eca08d52bab71,14496047104,0,4294967296,4661020260,23452034660,1,1857717534,0,784668098,2018634628 | |
| mm/sglang-h200-mistral-small31/default,sglang,mistralai/Mistral-Small-3.1-24B-Instruct-2503,H200 141GB,default,32768,48092896297,80509161964,,2597868011,131199926272,131202023424,132584046592,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/sglang-h200-nemotron-9b/default,sglang,nvidia/NVIDIA-Nemotron-Nano-9B-v2,H200 141GB,default,32768,17802639442,61675730370,,57239163372,136717533184,136719630336,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/sglang-h200-phi4mm/default,sglang,microsoft/Phi-4-multimodal-instruct,H200 141GB,default,32768,9524089979,114289079746,,2265511363,126078681088,126078681088,126338727936,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,93f923e1a7727d1c4f446756212d9d3e8fcc5d81,11148920768,0,4294967296,6495759205,21939647269,1,3871981701,0,605142876,2018634628 | |
| mm/sglang-h200-qwen25vl-7b/default,sglang,Qwen/Qwen2.5-VL-7B-Instruct,H200 141GB,default,32768,16857746637,104496554312,,2614645227,123968946176,123971043328,128108724224,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/sglang-h200-qwen2audio-7b-ml8k/default,sglang,Qwen/Qwen2-Audio-7B-Instruct,H200 141GB,default,8192,16943645983,118455198024,,1434032537,136832876544,136897888256,136944025600,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/sglang-h200-qwen3-06b/default,sglang,Qwen/Qwen3-0.6B,H200 141GB,default,32768,1267015352,133616432578,,2297555846,137181003776,137181003776,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c1899de289a04d12100db370d81485cdf75e47ca,1503264768,0,3758096384,5474281111,10735642263,1,3039200593,0,416445890,2018634628 | |
| mm/sglang-h200-qwen3-30b/default,sglang,Qwen/Qwen3-30B-A3B-Instruct-2507,H200 141GB,default,32768,61224758804,73637214290,,3115948442,137977921536,137977921536,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/sglang-h200-qwen3-32b/default,sglang,Qwen/Qwen3-32B,H200 141GB,default,32768,65541200937,69342246994,,4264684421,139148132352,139148132352,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9216db5781bf21249d130ec9da846c4624c16137,65524246528,0,8589934592,6628119396,80742300516,1,3482119096,0,1127365672,2018634628 | |
| mm/sglang-h200-qwen3-8b/default,sglang,Qwen/Qwen3-8B,H200 141GB,default,32768,16406775071,118476672860,,3161582469,138045030400,138045030400,,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,b968826d9c46dd6066d109eabc6255188de91218,16381470720,0,4831838208,6129804951,27343113879,1,3371389470,0,739780853,2018634628 | |
| mm/sglang-h200-qwen35-4b/default,sglang,Qwen/Qwen3.5-4B,H200 141GB,default,32768,9255654523,62534723830,,59654914703,131445293056,131445293056,133779423232,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/sglang-h200-qwen36-35b/default,sglang,Qwen/Qwen3.6-35B-A3B,H200 141GB,default,32768,70319352054,28583007354,,29791470224,128693829632,128693829632,131879403520,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,345d467b117d1fd5eda9b879a5988de9ae5f47a7,71903645408,0,821821440,9020433815,81745900663,1,4374199787,1732247552,895351848,2018634628 | |
| mm/sglang-h200-qwen3vl-30b/default,sglang,Qwen/Qwen3-VL-30B-A3B-Instruct,H200 141GB,default,32768,62341450301,60408715018,,4712636089,127462801408,127462801408,131529179136,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/sglang-h200-qwen3vl-8b/default,sglang,Qwen/Qwen3-VL-8B-Instruct,H200 141GB,default,32768,17738214932,105033425224,,4636635300,127408275456,127408275456,131474653184,,SGLang chunked prefill is supplied as hf-mem's token budget; engine scheduling and graph profiles differ; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-dsv2-lite/default,vllm,deepseek-ai/DeepSeek-V2-Lite-Chat,A100 80GB,default,32768,31482110280,44860933407,,2267650457,78610694144,78610694144,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,85864749cd611b4353ce1decdb286193298f64c7,31412968448,0,1019215872,5480658737,37912843057,1,2520723947,0,941300162,2018634628 | |
| mm/vllm-a100-gemma3-12b/default,vllm,google/gemma-3-12b-it,A100 80GB,default,32768,25136296100,51174535332,,1714757304,78025588736,78025588736,78168195072,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-a100-gemma3-12b/text-only,vllm,google/gemma-3-12b-it,A100 80GB,text-only,32768,24288040059,51002736640,,1692527493,76983304192,76983304192,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-a100-gemma3-1b-swap/default,vllm,google/gemma-3-1b-it,A100 80GB,default,32768,2104533975,70824010711,,2412689490,75341234176,75345428480,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/vllm-a100-gemma3-1b/default,vllm,google/gemma-3-1b-it,A100 80GB,default,32768,2104533975,74571369677,,1379045212,78054948864,78057046016,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/vllm-a100-gemma3-4b/default,vllm,google/gemma-3-4b-it,A100 80GB,default,32768,9244917105,67226975601,,1553696030,78025588736,78027685888,78170292224,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/vllm-a100-gemma3-4b/text-only,vllm,google/gemma-3-4b-it,A100 80GB,text-only,32768,8396661064,67763846513,,1454039367,77614546944,77614546944,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/vllm-a100-gemma4-e4b/default,vllm,google/gemma-4-E4B-it,A100 80GB,default,32768,16181289288,60193966653,,1809716347,78184972288,78189166592,78281441280,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/vllm-a100-gemma4-e4b/text-only,vllm,google/gemma-4-E4B-it,A100 80GB,text-only,32768,15214921646,56446607688,,1804850954,73466380288,73470574592,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/vllm-a100-gpt-oss-20b/default,vllm,openai/gpt-oss-20b,A100 80GB,default,32768,14817637171,61299920732,,1798978929,77916536832,77916536832,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6cee5e81ee83917806bbde320786a8fb61efebee,13761264768,0,1211498496,6349153636,21321916900,1,3864226744,0,466292264,2018634628 | |
| mm/vllm-a100-internvl35-8b/default,vllm,OpenGVLab/InternVL3_5-8B,A100 80GB,default,32768,17072495002,57101590200,,1385252782,75559337984,75559337984,76022808576,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/vllm-a100-internvl35-8b/text-only,vllm,OpenGVLab/InternVL3_5-8B,A100 80GB,text-only,32768,16396037652,59947006034,,1344903578,77687947264,77687947264,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/vllm-a100-llama31-8b-swap/default,vllm,meta-llama/Llama-3.1-8B-Instruct,A100 80GB,default,32768,16106127360,59130962248,,1834294968,77071384576,77073481728,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/vllm-a100-llama31-8b/default,vllm,meta-llama/Llama-3.1-8B-Instruct,A100 80GB,default,32768,16106127360,60451664691,,1297927373,77855719424,77855719424,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/vllm-a100-llama32-1b/default,vllm,meta-llama/Llama-3.2-1B-Instruct,A100 80GB,default,32768,2491081032,74786118042,,1044088094,78321287168,78321287168,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9213176726f574b556790deb65791e0c5aa438b6,2471628800,0,1073741824,5432137214,8977507838,1,2849146859,0,564355727,2018634628 | |
| mm/vllm-a100-llava-ov-7b/default,vllm,llava-hf/llava-onevision-qwen2-7b-ov-hf,A100 80GB,default,32768,16181289288,60805999493,,1315124019,78302412800,78302412800,79506178048,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/vllm-a100-llava-ov-7b/text-only,vllm,llava-hf/llava-onevision-qwen2-7b-ov-hf,A100 80GB,text-only,32768,15343770665,61181809132,,1273516523,77799096320,77799096320,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/vllm-a100-mistral-7b/default,vllm,mistralai/Mistral-7B-Instruct-v0.3,A100 80GB,default,32768,14506252042,62942745723,,1188959355,78637957120,78637957120,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c170c708c41dac9275d15a8fff4eca08d52bab71,14496047104,0,4294967296,4661020260,23452034660,1,1857717534,0,784668098,2018634628 | |
| mm/vllm-a100-mistral-small31/default,vllm,mistralai/Mistral-Small-3.1-24B-Instruct-2503,A100 80GB,default,32768,48060684042,28808493138,,1510830244,78380007424,78380007424,79365668864,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/vllm-a100-mistral-small31/text-only,vllm,mistralai/Mistral-Small-3.1-24B-Instruct-2503,A100 80GB,text-only,32768,47180215747,28915867320,,1478617989,77574701056,77574701056,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/vllm-a100-nemotron-9b-swap/default,vllm,nvidia/NVIDIA-Nemotron-Nano-9B-v2,A100 80GB,default,32768,17802639442,57348550820,,1756616458,76907806720,76907806720,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/vllm-a100-nemotron-9b/default,vllm,nvidia/NVIDIA-Nemotron-Nano-9B-v2,A100 80GB,default,32768,17802639442,58604828754,,1267896156,77675364352,77696335872,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/vllm-a100-qwen25omni-7b/default,vllm,Qwen/Qwen2.5-Omni-7B,A100 80GB,default,32768,18017387807,58841051955,,1481721774,78340161536,78340161536,78522613760,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,ae9e1690543ffd5c0221dc27f79834d0294cba00,22362553472,0,1879048192,8227124631,32468726295,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25omni-7b/text-only,vllm,Qwen/Qwen2.5-Omni-7B,A100 80GB,text-only,32768,15375982920,61160334295,,1285847777,77822164992,77822164992,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,ae9e1690543ffd5c0221dc27f79834d0294cba00,22362553472,0,1879048192,8227124631,32468726295,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b-ml64k/default,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,default,65536,16825534382,51604032061,,1425518101,69855084544,69857181696,73283928064,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,3758096384,8227124631,28569554327,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b-ml64k/text-only,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,text-only,65536,15461882266,61074434949,,1283750625,77820067840,77820067840,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,3758096384,8227124631,28569554327,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b-ml8k/default,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,default,8192,16825534382,58443767480,,1380555162,76649857024,76649857024,76649857024,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,469762048,8227124631,25281219991,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b-ml8k/text-only,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,text-only,8192,15461882266,61074434949,,1264876257,77801193472,77801193472,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,469762048,8227124631,25281219991,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b/default,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,default,32768,16825534382,56167434813,,1381477909,74374447104,74376544256,77803290624,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen25vl-7b/text-only,vllm,Qwen/Qwen2.5-VL-7B-Instruct,A100 80GB,text-only,32768,15461882266,61074434949,,1283750625,77820067840,77820067840,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-a100-qwen2audio-7b-ml8k/default,vllm,Qwen/Qwen2-Audio-7B-Instruct,A100 80GB,default,8192,16804059546,60118804726,,1408908656,78331772928,78333870080,78430339072,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/vllm-a100-qwen2audio-7b-ml8k/text-only,vllm,Qwen/Qwen2-Audio-7B-Instruct,A100 80GB,text-only,8192,15515569357,60805999493,,1290880942,77612449792,77614546944,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/vllm-a100-qwen3-06b/default,vllm,Qwen/Qwen3-0.6B,A100 80GB,default,32768,1202590843,75924284375,,1272006574,78398881792,78400978944,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c1899de289a04d12100db370d81485cdf75e47ca,1503264768,0,3758096384,5474281111,10735642263,1,3039200593,0,416445890,2018634628 | |
| mm/vllm-a100-qwen3-30b-swap/default,vllm,Qwen/Qwen3-30B-A3B-Instruct-2507,A100 80GB,default,32768,61128122040,12970801234,,2926323958,77025247232,77027344384,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/vllm-a100-qwen3-30b/default,vllm,Qwen/Qwen3-30B-A3B-Instruct-2507,A100 80GB,default,32768,61128122040,15558519030,,1840166994,78526808064,78528905216,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/vllm-a100-qwen3-32b/default,vllm,Qwen/Qwen3-32B,A100 80GB,default,32768,65530463519,10114647982,,1770206003,77415317504,77417414656,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9216db5781bf21249d130ec9da846c4624c16137,65524246528,0,8589934592,6628119396,80742300516,1,3482119096,0,1127365672,2018634628 | |
| mm/vllm-a100-qwen3-8b/default,vllm,Qwen/Qwen3-8B,A100 80GB,default,32768,16396037652,59947006034,,1344903578,77687947264,77690044416,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,b968826d9c46dd6066d109eabc6255188de91218,16381470720,0,4831838208,6129804951,27343113879,1,3371389470,0,739780853,2018634628 | |
| mm/vllm-a100-qwen35-4b/default,vllm,Qwen/Qwen3.5-4B,A100 80GB,default,32768,9191230013,64156073984,,3849909699,77197213696,77199310848,80051437568,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/vllm-a100-qwen35-4b/text-only,vllm,Qwen/Qwen3.5-4B,A100 80GB,text-only,32768,8525510083,65283502899,,3822311178,77631324160,77631324160,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/vllm-a100-qwen3vl-30b/default,vllm,Qwen/Qwen3-VL-30B-A3B-Instruct,A100 80GB,default,32768,62255550956,12981538652,,1842683576,77079773184,77081870336,80173072384,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/vllm-a100-qwen3vl-30b/text-only,vllm,Qwen/Qwen3-VL-30B-A3B-Instruct,A100 80GB,text-only,32768,61128122040,15558519030,,1840166994,78526808064,78528905216,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/vllm-a100-qwen3vl-8b-ml128k/default,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,default,131072,17706002678,57498874675,,1386259415,76591136768,76593233920,79684435968,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,19327352832,7862052503,44723652727,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b-ml128k/text-only,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,text-only,131072,16481936998,59861106688,,1344903578,77687947264,77690044416,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,19327352832,7862052503,44723652727,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b-ml8k/default,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,default,8192,17706002678,57498874675,,1365287895,76570165248,76572262400,76572262400,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,1207959552,7862052503,26604259447,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b-ml8k/text-only,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,text-only,8192,16481936998,59861106688,,1323932058,77666975744,77666975744,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,1207959552,7862052503,26604259447,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b-swap/default,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,default,32768,17856326533,56317758669,,2052147118,76226232320,76228329472,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b/default,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,default,32768,17706002678,57498874675,,1384162263,76589039616,76591136768,79682338816,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-qwen3vl-8b/text-only,vllm,Qwen/Qwen3-VL-8B-Instruct,A100 80GB,text-only,32768,16481936998,59861106688,,1342806426,77685850112,77687947264,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-a100-smolvlm2-ml8k/text-only,vllm,HuggingFaceTB/SmolVLM2-2.2B-Instruct,A100 80GB,text-only,8192,3650722202,73927124582,,1062207488,78640054272,78640054272,,,Native batched token budget absent; hf-mem default used; hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,482adb537c021c86670beed01cd58990d01e72e4,8987139520,0,,,8987139520,1,,,, | |
| mm/vllm-a100-whisper-v3/default,vllm,openai/whisper-large-v3,A100 80GB,default,448,3092376453,74496207749,,1118578934,78707163136,78707163136,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,06f233fe06e710322aca913c1bc4249a0d71fce1,3086981120,0,,,3086981120,1,,,, | |
| mm/vllm-h200-dsv2-lite/default,vllm,deepseek-ai/DeepSeek-V2-Lite-Chat,H200 141GB,default,32768,31482110280,102402757755,,3171355197,137056223232,137068806144,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,85864749cd611b4353ce1decdb286193298f64c7,31412968448,0,1019215872,5480658737,37912843057,1,2520723947,0,941300162,2018634628 | |
| mm/vllm-h200-gemma3-12b-ml8k/default,vllm,google/gemma-3-12b-it,H200 141GB,default,8192,25136296100,106955423089,,2657636843,134749356032,134749356032,134891962368,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,3221225472,7687559652,35283435204,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-h200-gemma3-12b-ml8k/text-only,vllm,google/gemma-3-12b-it,H200 141GB,text-only,8192,24288040059,107352707564,,2940836249,134581583872,134581583872,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,3221225472,7687559652,35283435204,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-h200-gemma3-12b/default,vllm,google/gemma-3-12b-it,H200 141GB,default,32768,25136296100,106955423089,,2846380523,134938099712,134938099712,135038763008,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-h200-gemma3-12b/text-only,vllm,google/gemma-3-12b-it,H200 141GB,text-only,32768,24288040059,107352707564,,3024722329,134665469952,134686441472,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,96b6f1eccf38110c56df3a15bffe176da04bfd80,24374650080,0,7856979968,7687559652,39919189700,1,4744382008,108265472,816277544,2018634628 | |
| mm/vllm-h200-gemma3-1b-swap/default,vllm,google/gemma-3-1b-it,H200 141GB,default,32768,2104533975,134260677673,,1578106880,137943318528,137945415680,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/vllm-h200-gemma3-1b/default,vllm,google/gemma-3-1b-it,H200 141GB,default,32768,2104533975,130470369034,,2610660639,135185563648,135189757952,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,dcc83ea841ab6100d6b47a070329e1ba4cf78752,1999771904,0,515211264,6983169739,9498152907,1,4452903813,0,511631298,2018634628 | |
| mm/vllm-h200-gemma3-4b/default,vllm,google/gemma-3-4b-it,H200 141GB,default,32768,9244917105,123297773650,,2718370365,135261061120,135263158272,135365918720,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/vllm-h200-gemma3-4b/text-only,vllm,google/gemma-3-4b-it,H200 141GB,text-only,32768,8396661064,123823907144,,2738503024,134959071232,134959071232,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,093f9f388b31de276ce2de164bdc2081324b9767,8600158944,0,2740781056,7371933156,18712873156,1,4605969976,108265472,639063080,2018634628 | |
| mm/vllm-h200-gemma4-31b/default,vllm,google/gemma-4-31B-it,H200 141GB,default,32768,63340030198,68601365135,,3128824955,135070220288,135074414592,135204438016,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,842da3794eaa0b77d5f08bae87a17459d91ff475,62546177752,0,16958095360,8029191882,87533464994,1,4909663518,66608640,1034285096,2018634628 | |
| mm/vllm-h200-gemma4-31b/text-only,vllm,google/gemma-4-31B-it,H200 141GB,text-only,32768,62180389028,70007966925,,2911224463,135099580416,135101677568,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,842da3794eaa0b77d5f08bae87a17459d91ff475,62546177752,0,16958095360,8029191882,87533464994,1,4909663518,66608640,1034285096,2018634628 | |
| mm/vllm-h200-gemma4-e4b/default,vllm,google/gemma-4-E4B-it,H200 141GB,default,32768,16320875725,116050016338,,2711911137,135082803200,135086997504,135162494976,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/vllm-h200-gemma4-e4b/text-only,vllm,google/gemma-4-E4B-it,H200 141GB,text-only,32768,15343770665,115105123533,,2681460490,133130354688,133134548992,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,4ecc2ba63bcac23ba4a767fc87025b1450f84c9c,15992314836,0,2151776256,7317102180,25461193272,1,4605157048,49152000,644158504,2018634628 | |
| mm/vllm-h200-gpt-oss-20b/default,vllm,openai/gpt-oss-20b,H200 141GB,default,32768,13990855967,119249766973,,3159191716,136399814656,136437563392,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6cee5e81ee83917806bbde320786a8fb61efebee,13761264768,0,1211498496,6349153636,21321916900,1,3864226744,0,466292264,2018634628 | |
| mm/vllm-h200-internvl35-8b/default,vllm,OpenGVLab/InternVL3_5-8B,H200 141GB,default,32768,17072495002,114707839058,,2383916564,134164250624,134164250624,134659178496,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/vllm-h200-internvl35-8b/text-only,vllm,OpenGVLab/InternVL3_5-8B,H200 141GB,text-only,32768,16396037652,117682103910,,2283924358,136362065920,136383037440,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9bb6a56ad9cc69db95e2d4eeb15a52bbcac4ef79,17056636928,0,4831838208,6456960663,28345435799,1,3371389470,327155712,739780853,2018634628 | |
| mm/vllm-h200-llama31-8b-swap/default,vllm,meta-llama/Llama-3.1-8B-Instruct,H200 141GB,default,32768,16106127360,119861799813,,1555960955,137523888128,137544859648,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/vllm-h200-llama31-8b/default,vllm,meta-llama/Llama-3.1-8B-Instruct,H200 141GB,default,32768,16106127360,118422985769,,2164302807,136693415936,136716484608,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0e9e39f249a16976918f6564b8830bc894c89659,16060522496,0,4294967296,5873908836,26229398628,1,3070606110,0,784668098,2018634628 | |
| mm/vllm-h200-llama32-1b/default,vllm,meta-llama/Llama-3.2-1B-Instruct,H200 141GB,default,32768,2491081032,132510478500,,1844948500,136846508032,136869576704,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9213176726f574b556790deb65791e0c5aa438b6,2471628800,0,1073741824,5432137214,8977507838,1,2849146859,0,564355727,2018634628 | |
| mm/vllm-h200-llava-ov-7b/default,vllm,llava-hf/llava-onevision-qwen2-7b-ov-hf,H200 141GB,default,32768,16181289288,117649891656,,2281323888,136112504832,136135573504,138134159360,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/vllm-h200-llava-ov-7b/text-only,vllm,llava-hf/llava-onevision-qwen2-7b-ov-hf,H200 141GB,text-only,32768,15343770665,118831007662,,2210356265,136385134592,136385134592,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d50680527681998e456c7b78950205bedd8a068,16061615168,0,1879048192,6404739479,24345402839,1,3318463441,192689280,874952130,2018634628 | |
| mm/vllm-h200-mistral-7b/default,vllm,mistralai/Mistral-7B-Instruct-v0.3,H200 141GB,default,32768,14506252042,121461675131,,1761481851,137729409024,137729409024,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c170c708c41dac9275d15a8fff4eca08d52bab71,14496047104,0,4294967296,4661020260,23452034660,1,1857717534,0,784668098,2018634628 | |
| mm/vllm-h200-mistral-small31/default,vllm,mistralai/Mistral-Small-3.1-24B-Instruct-2503,H200 141GB,default,32768,48060684042,86114094285,,2438945833,136613724160,136613724160,137955901440,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/vllm-h200-mistral-small31/text-only,vllm,mistralai/Mistral-Small-3.1-24B-Instruct-2503,H200 141GB,text-only,32768,47180215747,87155623854,,2319827599,136655667200,136655667200,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,68faf511d618ef198fef186659617cfd2eb8e33a,48022722560,0,5368709120,6900397668,60291829348,1,3217104568,396492800,1268165672,2018634628 | |
| mm/vllm-h200-nemotron-9b-swap/default,vllm,nvidia/NVIDIA-Nemotron-Nano-9B-v2,H200 141GB,default,32768,17802639442,118025701294,,1592786944,137421127680,137421127680,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/vllm-h200-nemotron-9b/default,vllm,nvidia/NVIDIA-Nemotron-Nano-9B-v2,H200 141GB,default,32768,17802639442,116458038231,,2210440151,136471117824,136492089344,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,6533e8de2c68e4536bf7c411d7a3ce5734111476,17776454656,0,988807168,6468596478,25233858302,1,3147898552,0,1302063298,2018634628 | |
| mm/vllm-h200-qwen25omni-7b/default,vllm,Qwen/Qwen2.5-Omni-7B,H200 141GB,default,32768,18017387807,115674206700,,2508990709,136200585216,136200585216,136347385856,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,ae9e1690543ffd5c0221dc27f79834d0294cba00,22362553472,0,1879048192,8227124631,32468726295,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-h200-qwen25omni-7b/text-only,vllm,Qwen/Qwen2.5-Omni-7B,H200 141GB,text-only,32768,15375982920,118820270244,,2199367188,136395620352,136416591872,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,ae9e1690543ffd5c0221dc27f79834d0294cba00,22362553472,0,1879048192,8227124631,32468726295,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-h200-qwen25vl-7b/default,vllm,Qwen/Qwen2.5-VL-7B-Instruct,H200 141GB,default,32768,16825534382,114697101640,,2364790538,133887426560,133912592384,137339338752,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-h200-qwen25vl-7b/text-only,vllm,Qwen/Qwen2.5-VL-7B-Instruct,H200 141GB,text-only,32768,15461882266,118723633480,,2208007454,136393523200,136414494720,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,cc594898137f460bfe9f0759e9844b3ce807cfb5,16584333312,0,1879048192,8227124631,26690506135,1,3317650513,2015887360,874952130,2018634628 | |
| mm/vllm-h200-qwen2audio-7b-ml8k/default,vllm,Qwen/Qwen2-Audio-7B-Instruct,H200 141GB,default,8192,16804059546,116715736269,,2424936857,135944732672,135959412736,136060076032,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/vllm-h200-qwen2audio-7b-ml8k/text-only,vllm,Qwen/Qwen2-Audio-7B-Instruct,H200 141GB,text-only,8192,15515569357,118508885115,,2247433912,136271888384,136273985536,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0a095220c30b7b31434169c3086508ef3ea5bf0a,16794189824,0,4294967296,6193233764,27282390884,1,3423416862,46080000,705102274,2018634628 | |
| mm/vllm-h200-qwen3-06b/default,vllm,Qwen/Qwen3-0.6B,H200 141GB,default,32768,1202590843,133262097777,,2071440916,136536129536,136538226688,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,c1899de289a04d12100db370d81485cdf75e47ca,1503264768,0,3758096384,5474281111,10735642263,1,3039200593,0,416445890,2018634628 | |
| mm/vllm-h200-qwen3-30b-swap/default,vllm,Qwen/Qwen3-30B-A3B-Instruct-2507,H200 141GB,default,32768,61128122040,75226352189,,1857279755,138211753984,138234822656,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/vllm-h200-qwen3-30b/default,vllm,Qwen/Qwen3-30B-A3B-Instruct-2507,H200 141GB,default,32768,61128122040,72123238318,,2966002074,136217362432,136240431104,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0d7cf23991f47feeb3a57ecb4c9cee8ea4a17bfe,61064245248,0,3221225472,6349540606,70635011326,1,3149930219,0,1180975759,2018634628 | |
| mm/vllm-h200-qwen3-32b/default,vllm,Qwen/Qwen3-32B,H200 141GB,default,32768,65530463519,67162551091,,2922465198,135615479808,135630159872,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9216db5781bf21249d130ec9da846c4624c16137,65524246528,0,8589934592,6628119396,80742300516,1,3482119096,0,1127365672,2018634628 | |
| mm/vllm-h200-qwen3-8b/default,vllm,Qwen/Qwen3-8B,H200 141GB,default,32768,16396037652,117682103910,,2283924358,136362065920,136385134592,,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,b968826d9c46dd6066d109eabc6255188de91218,16381470720,0,4831838208,6129804951,27343113879,1,3371389470,0,739780853,2018634628 | |
| mm/vllm-h200-qwen35-4b/default,vllm,Qwen/Qwen3.5-4B,H200 141GB,default,32768,9191230013,122921964012,,2753602519,134866796544,134887768064,137741991936,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/vllm-h200-qwen35-4b/text-only,vllm,Qwen/Qwen3.5-4B,H200 141GB,text-only,32768,8525510083,123705795543,,2719376998,134950682624,134950682624,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a,9319737856,0,1193803776,8927133080,19440674712,1,4429564600,1610612736,868321116,2018634628 | |
| mm/vllm-h200-qwen36-35b/default,vllm,Qwen/Qwen3.6-35B-A3B,H200 141GB,default,32768,70276402381,61052960113,,3784897986,135114260480,135116357632,138207559680,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,345d467b117d1fd5eda9b879a5988de9ae5f47a7,71903645408,0,821821440,9020433815,81745900663,1,4374199787,1732247552,895351848,2018634628 | |
| mm/vllm-h200-qwen36-35b/text-only,vllm,Qwen/Qwen3.6-35B-A3B,H200 141GB,text-only,32768,69374459249,62073014845,,3733895250,135181369344,135185563648,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,345d467b117d1fd5eda9b879a5988de9ae5f47a7,71903645408,0,821821440,9020433815,81745900663,1,4374199787,1732247552,895351848,2018634628 | |
| mm/vllm-h200-qwen3vl-30b/default,vllm,Qwen/Qwen3-VL-30B-A3B-Instruct,H200 141GB,default,32768,62330712883,70609262346,,3008951747,135948926976,135971995648,139063197696,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/vllm-h200-qwen3vl-30b/text-only,vllm,Qwen/Qwen3-VL-30B-A3B-Instruct,H200 141GB,text-only,32768,61128122040,72123238318,,2966002074,136217362432,136240431104,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,9c4b90e1e4ba969fd3b5378b57d966d725f1b86c,62141508064,0,3221225472,7432509899,72795243435,1,3149930219,1732247552,531697500,2018634628 | |
| mm/vllm-h200-qwen3vl-8b-swap/default,vllm,Qwen/Qwen3-VL-8B-Instruct,H200 141GB,default,32768,17706002678,116909009797,,1623321477,136238333952,136261402624,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-h200-qwen3vl-8b/default,vllm,Qwen/Qwen3-VL-8B-Instruct,H200 141GB,default,32768,17856326533,115362821571,,2373263032,135592411136,135615479808,138706681856,,Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-h200-qwen3vl-8b/text-only,vllm,Qwen/Qwen3-VL-8B-Instruct,H200 141GB,text-only,32768,16481936998,117585467146,,2296758928,136364163072,136387231744,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,0c351dd01ed87e9c1b53cbc748cba10e6187ff3b,17534247392,0,4831838208,7862052503,30228138103,1,3371389470,1732247552,739780853,2018634628 | |
| mm/vllm-h200-smolvlm2-ml8k/text-only,vllm,HuggingFaceTB/SmolVLM2-2.2B-Instruct,H200 141GB,text-only,8192,3650722202,132725226865,,1596729589,137972678656,137972678656,,,hf-mem includes model encoders; native launch disabled multimodal components; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,482adb537c021c86670beed01cd58990d01e72e4,8987139520,0,,,8987139520,1,,,, | |
| mm/vllm-h200-whisper-v3/default,vllm,openai/whisper-large-v3,H200 141GB,default,448,3092376453,133305047450,,1652849377,138050273280,138050273280,,,Native batched token budget absent; hf-mem default used; Historical runtime model SHA not recorded; estimation pins the current main SHA recorded in model_revisions.json,hf-mem@1.0.0,06f233fe06e710322aca913c1bc4249a0d71fce1,3086981120,0,,,3086981120,1,,,, | |
| startup/llamacpp-v-gemma3-12b-q4km,llama.cpp,unsloth/gemma-3-12b-it-GGUF,A100 80GB,startup,32768,7294293115,2650800128,0,1646914437,11592007680,,,,Native GPU weights exclude host-mapped tensors; hf-mem offloaded bytes are recorded separately,hf-mem@1.0.0,d15e4c7dc21dc55d56bf8549db57a71ad8a2a35d,8148401472,825955200,2650800128,792650054,11591851654,1,161061273,126930124,0,504658657 | |
| startup/llamacpp-v-lfm3-17b-q4km,llama.cpp,LiquidAI/LFM3-1.7B-GGUF,A100 80GB,startup,32768,1091378872,410517504,639631,601956025,2104492032,,,,Native GPU weights exclude host-mapped tensors; hf-mem offloaded bytes are recorded separately,hf-mem@1.0.0,cd0c74c575b5c416f3011b5b4c298077f26ca8c2,1091380480,215040000,411156480,665719930,2168256890,1,161061273,0,0,504658657 | |
| startup/llamacpp-v-qwen3-8b-q4km,llama.cpp,unsloth/Qwen3-8B-GGUF,A100 80GB,startup,32768,4671762596,4831838208,0,645566300,10149167104,,,,Native GPU weights exclude host-mapped tensors; hf-mem offloaded bytes are recorded separately,hf-mem@1.0.0,a6adef130ffb23ddaf1a62fec9dced968c9bc482,4671766528,350060544,4831838208,665719930,10169324666,1,161061273,0,0,504658657 | |
| startup/llamacpp-v-qwen38-27b-q4km,llama.cpp,unsloth/Qwen3.8-27B-GGUF,A100 80GB,startup,32768,15387276083,2147483648,156887941,1880023368,19571671040,,,,Native GPU weights exclude host-mapped tensors; hf-mem offloaded bytes are recorded separately,hf-mem@1.0.0,4ca720788d1e01f1bff70c033e0d0028fd02e502,16665869248,715161600,2304376832,1173440429,20143686509,1,161061273,507720499,0,504658657 | |
Xet Storage Details
- Size:
- 73.3 kB
- Xet hash:
- b8b3eb542fefc0819d64d8cc60835702c2e23aab0ff6d376141b19a57c16d0be
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.