[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"model-mobilenetv3-small-100-lamb-in1k":3},{"id":4,"slug":5,"name":6,"hf_id":6,"organization":7,"architecture":8,"architecture_config":9,"parameters":10,"parameters_label":11,"context_length":8,"license":8,"downloads":12,"likes":13,"status":14,"description":8,"huggingface_url":15,"created_at":16,"seo":17,"category":20,"family":23,"score":25,"variants":26,"quantizations":27,"requirements":54,"gpu_recommendations":102,"fitting_gpus":155,"community":185},443,"mobilenetv3-small-100-lamb-in1k","timm\u002Fmobilenetv3_small_100.lamb_in1k","timm",null,[],2555002,"2.6M",18120763,101,"published","https:\u002F\u002Fhuggingface.co\u002Ftimm\u002Fmobilenetv3_small_100.lamb_in1k","2026-08-13T18:01:42+00:00",{"title":6,"description":18,"image":8,"robots":19,"canonical_url":8},"AI 模型部署决策引擎——开源 AI 的 GPU 需求、云端价格与成本分析。","index, follow",{"name":21,"slug":22},"OCR \u002F Vision","ocr-vision",{"name":24,"slug":7},"Timm",0.0296,[],[28,33,38,43,47,51],{"format":29,"bits":30,"size_factor":31,"quality_loss":32},"AWQ",4,0.263,0.03,{"format":34,"bits":35,"size_factor":36,"quality_loss":37},"FP16",16,1,0,{"format":39,"bits":40,"size_factor":41,"quality_loss":42},"FP8",8,0.5,0.01,{"format":44,"bits":30,"size_factor":45,"quality_loss":46},"GGUF",0.25,0.05,{"format":48,"bits":30,"size_factor":49,"quality_loss":50},"GPTQ",0.275,0.04,{"format":52,"bits":40,"size_factor":41,"quality_loss":53},"INT8",0.02,{"FP16":55,"FP8":76,"INT8":86,"AWQ":90,"GPTQ":94,"GGUF":98},{"minimum":56,"production":64,"recommended":71},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":61,"requirement_source":62,"confidence":63},"minimum",2048,3.58,32,39.01,"inferred",0.95,{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":68,"ram_gb":69,"disk_gb":70,"requirement_source":62,"confidence":63},"production",8192,2,21.61,32.41,104.01,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":74,"ram_gb":60,"disk_gb":75,"requirement_source":62,"confidence":63},"recommended",4096,5.76,65.01,{"minimum":77,"production":79,"recommended":83},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":78,"requirement_source":62,"confidence":63},39,{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":62,"confidence":63},21.6,32.4,104,{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":84,"ram_gb":60,"disk_gb":85,"requirement_source":62,"confidence":63},5.75,65,{"minimum":87,"production":88,"recommended":89},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":78,"requirement_source":62,"confidence":63},{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":62,"confidence":63},{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":84,"ram_gb":60,"disk_gb":85,"requirement_source":62,"confidence":63},{"minimum":91,"production":92,"recommended":93},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":78,"requirement_source":62,"confidence":63},{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":62,"confidence":63},{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":84,"ram_gb":60,"disk_gb":85,"requirement_source":62,"confidence":63},{"minimum":95,"production":96,"recommended":97},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":78,"requirement_source":62,"confidence":63},{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":62,"confidence":63},{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":84,"ram_gb":60,"disk_gb":85,"requirement_source":62,"confidence":63},{"minimum":99,"production":100,"recommended":101},{"scenario":57,"context_length":58,"batch_size":36,"weight_gb":37,"kv_cache_gb":45,"vram_gb":59,"ram_gb":60,"disk_gb":78,"requirement_source":62,"confidence":63},{"scenario":65,"context_length":66,"batch_size":67,"weight_gb":37,"kv_cache_gb":40,"vram_gb":80,"ram_gb":81,"disk_gb":82,"requirement_source":62,"confidence":63},{"scenario":72,"context_length":73,"batch_size":36,"weight_gb":37,"kv_cache_gb":36,"vram_gb":84,"ram_gb":60,"disk_gb":85,"requirement_source":62,"confidence":63},{"recommended":103,"cheapest":119,"performance":127,"min_complexity":140,"cluster_options":152,"no_match":153,"required_vram_gb":84,"candidate_count":121,"usage":154},{"gpu_model":104,"gpu_model_id":105,"gpu_count":36,"total_vram_gb":35,"hour_price":106,"monthly_cost":107,"providers":108,"bandwidth_gbps":110,"fp16_tflops":111,"interconnect_score":112,"parallel_efficiency":36,"score":113,"tier":114,"perf_metric":115,"perf_label":116,"perf_value":37,"workload":8,"cost_metric":117},"RTX 5080",23,0.1339,97.75,[109],"vast",960,113,40,0.5907,"cheapest","tok\u002Fs","0 tok\u002Fs",{"effective_value":37,"effective_unit":115,"capacity_factor":118},0.7,{"gpu_model":120,"gpu_model_id":121,"gpu_count":36,"total_vram_gb":35,"hour_price":122,"monthly_cost":123,"providers":124,"bandwidth_gbps":37,"fp16_tflops":37,"interconnect_score":112,"parallel_efficiency":36,"score":125,"tier":114,"perf_metric":115,"perf_label":116,"perf_value":37,"workload":8,"cost_metric":126},"Tesla V100",17,0.0289,21.1,[109],0.5854,{"effective_value":37,"effective_unit":115,"capacity_factor":118},{"gpu_model":128,"gpu_model_id":67,"gpu_count":40,"total_vram_gb":129,"hour_price":130,"monthly_cost":131,"providers":132,"bandwidth_gbps":134,"fp16_tflops":135,"interconnect_score":112,"parallel_efficiency":136,"score":137,"tier":138,"perf_metric":115,"perf_label":116,"perf_value":37,"workload":8,"cost_metric":139},"RTX 5090",256,0.3337,1948.81,[109,133],"runpod",1792,209,0.4,0.3604,"balanced",{"effective_value":37,"effective_unit":115,"capacity_factor":118},{"gpu_model":141,"gpu_model_id":142,"gpu_count":36,"total_vram_gb":143,"hour_price":144,"monthly_cost":145,"providers":146,"bandwidth_gbps":148,"fp16_tflops":149,"interconnect_score":112,"parallel_efficiency":36,"score":150,"tier":114,"perf_metric":115,"perf_label":116,"perf_value":37,"workload":8,"cost_metric":151},"RTX 3090",3,24,0.0622,45.41,[109,147],"tensordock",936,71,0.5596,{"effective_value":37,"effective_unit":115,"capacity_factor":118},[],false,"value",[156,160,164,171,178],{"name":120,"vram_gb":35,"cheapest_hour":122,"cost":157,"provider":109,"estimated_tps":37},{"hourly":122,"daily":158,"monthly":123,"yearly":159},0.69,253.16,{"name":141,"vram_gb":143,"cheapest_hour":144,"cost":161,"provider":109,"estimated_tps":37},{"hourly":144,"daily":162,"monthly":145,"yearly":163},1.49,544.87,{"name":165,"vram_gb":35,"cheapest_hour":166,"cost":167,"provider":109,"estimated_tps":37},"RTX 4070S Ti",0.0678,{"hourly":166,"daily":168,"monthly":169,"yearly":170},1.63,49.49,593.93,{"name":172,"vram_gb":35,"cheapest_hour":173,"cost":174,"provider":109,"estimated_tps":37},"RTX 4080S",0.0685,{"hourly":173,"daily":175,"monthly":176,"yearly":177},1.64,50.01,600.06,{"name":179,"vram_gb":40,"cheapest_hour":180,"cost":181,"provider":109,"estimated_tps":37},"RTX 5060 Ti",0.0719,{"hourly":180,"daily":182,"monthly":183,"yearly":184},1.73,52.49,629.84,{"posts_count":37,"benchmarks_count":37}]