[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6GHTS0M2DYP6W407ATZN9VR":3,"skill-same:01M6GHTS0M2DYP6W407ATZN9VR":106},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":26,"publishedAt":47,"linkOnly":48,"markdownContent":49,"markdownBody":50,"frontMatter":51,"refName":66,"commitSha":67,"skillDirectory":68,"forksCount":69,"repositoryHtmlUrl":70,"repositoryHomepageUrl":6,"ownerHtmlUrl":71,"ownerBlogUrl":6,"ownerDescription":72,"updatedAt":73,"install":74,"resources":87},"01M6GHTS0M2DYP6W407ATZN9VR","inference-sh-cli",null,"Run 150+ AI apps via inference.sh CLI (infsh) — image generation, video creation, LLMs, search, 3D, social automation. Uses the terminal tool. Triggers: inference.sh, infsh, ai apps, flux, veo, image generation, video generation, seedream, seedance, tavily","zapabob\u002FSkills","zapabob","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F130830376?v=4","01M6DBMYZAGRTFQDSQYW3W764P","峯岸 亮","01M6DBMZ3GKDSJ01NZ9XTZD7JD",2,"MIT","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[18,23],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46],"cli","inference sh","inference","inference sh cli","seedance","image","apps","triggers","creation","social","image-generation","tool","3D","Video","search","windows","automation","linux","llms","tavily","2026-11-10 10:12:30",false,"---\nname: inference-sh-cli\ndescription: \"Run 150+ AI apps via inference.sh CLI (infsh) — image generation, video creation, LLMs, search, 3D, social automation. Uses the terminal tool. Triggers: inference.sh, infsh, ai apps, flux, veo, image generation, video generation, seedream, seedance, tavily\"\nversion: 1.0.0\nauthor: okaris\nlicense: MIT\nplatforms: [linux, macos, windows]\nmetadata:\n  hermes:\n    tags: [AI, image-generation, video, LLM, search, inference, FLUX, Veo, Claude]\n    related_skills: []\n---\n\n# inference.sh CLI\n\nRun 150+ AI apps in the cloud with a simple CLI. No GPU required.\n\nAll commands use the **terminal tool** to run `infsh` commands.\n\n## When to Use\n\n- User asks to generate images (FLUX, Reve, Seedream, Grok, Gemini image)\n- User asks to generate video (Veo, Wan, Seedance, OmniHuman)\n- User asks about inference.sh or infsh\n- User wants to run AI apps without managing individual provider APIs\n- User asks for AI-powered search (Tavily, Exa)\n- User needs avatar\u002Flipsync generation\n\n## Prerequisites\n\nThe `infsh` CLI must be installed and authenticated. Check with:\n\n```bash\ninfsh me\n```\n\nIf not installed:\n\n```bash\ncurl -fsSL https:\u002F\u002Fcli.inference.sh | sh\ninfsh login\n```\n\nSee `references\u002Fauthentication.md` for full setup details.\n\n## Workflow\n\n### 1. Always Search First\n\nNever guess app names — always search to find the correct app ID:\n\n```bash\ninfsh app list --search flux\ninfsh app list --search video\ninfsh app list --search image\n```\n\n### 2. Run an App\n\nUse the exact app ID from the search results. Always use `--json` for machine-readable output:\n\n```bash\ninfsh app run \u003Capp-id> --input '{\"prompt\": \"your prompt here\"}' --json\n```\n\n### 3. Parse the Output\n\nThe JSON output contains URLs to generated media. Present these to the user with `MEDIA:\u003Curl>` for inline display.\n\n## Common Commands\n\n### Image Generation\n\n```bash\n# Search for image apps\ninfsh app list --search image\n\n# FLUX Dev with LoRA\ninfsh app run falai\u002Fflux-dev-lora --input '{\"prompt\": \"sunset over mountains\", \"num_images\": 1}' --json\n\n# Gemini image generation\ninfsh app run google\u002Fgemini-2-5-flash-image --input '{\"prompt\": \"futuristic city\", \"num_images\": 1}' --json\n\n# Seedream (ByteDance)\ninfsh app run bytedance\u002Fseedream-5-lite --input '{\"prompt\": \"nature scene\"}' --json\n\n# Grok Imagine (xAI)\ninfsh app run xai\u002Fgrok-imagine-image --input '{\"prompt\": \"abstract art\"}' --json\n```\n\n### Video Generation\n\n```bash\n# Search for video apps\ninfsh app list --search video\n\n# Veo 3.1 (Google)\ninfsh app run google\u002Fveo-3-1-fast --input '{\"prompt\": \"drone shot of coastline\"}' --json\n\n# Seedance (ByteDance)\ninfsh app run bytedance\u002Fseedance-1-5-pro --input '{\"prompt\": \"dancing figure\", \"resolution\": \"1080p\"}' --json\n\n# Wan 2.5\ninfsh app run falai\u002Fwan-2-5 --input '{\"prompt\": \"person walking through city\"}' --json\n```\n\n### Local File Uploads\n\nThe CLI automatically uploads local files when you provide a path:\n\n```bash\n# Upscale a local image\ninfsh app run falai\u002Ftopaz-image-upscaler --input '{\"image\": \"\u002Fpath\u002Fto\u002Fphoto.jpg\", \"upscale_factor\": 2}' --json\n\n# Image-to-video from local file\ninfsh app run falai\u002Fwan-2-5-i2v --input '{\"image\": \"\u002Fpath\u002Fto\u002Fimage.png\", \"prompt\": \"make it move\"}' --json\n\n# Avatar with audio\ninfsh app run bytedance\u002Fomnihuman-1-5 --input '{\"audio\": \"\u002Fpath\u002Fto\u002Faudio.mp3\", \"image\": \"\u002Fpath\u002Fto\u002Fface.jpg\"}' --json\n```\n\n### Search & Research\n\n```bash\ninfsh app list --search search\ninfsh app run tavily\u002Ftavily-search --input '{\"query\": \"latest AI news\"}' --json\ninfsh app run exa\u002Fexa-search --input '{\"query\": \"machine learning papers\"}' --json\n```\n\n### Other Categories\n\n```bash\n# 3D generation\ninfsh app list --search 3d\n\n# Audio \u002F TTS\ninfsh app list --search tts\n\n# Twitter\u002FX automation\ninfsh app list --search twitter\n```\n\n## Pitfalls\n\n1. **Never guess app IDs** — always run `infsh app list --search \u003Cterm>` first. App IDs change and new apps are added frequently.\n2. **Always use `--json`** — raw output is hard to parse. The `--json` flag gives structured output with URLs.\n3. **Check authentication** — if commands fail with auth errors, run `infsh login` or verify `INFSH_API_KEY` is set.\n4. **Long-running apps** — video generation can take 30-120 seconds. The terminal tool timeout should be sufficient, but warn the user it may take a moment.\n5. **Input format** — the `--input` flag takes a JSON string. Make sure to properly escape quotes.\n\n## Reference Docs\n\n- `references\u002Fauthentication.md` — Setup, login, API keys\n- `references\u002Fapp-discovery.md` — Searching and browsing the app catalog\n- `references\u002Frunning-apps.md` — Running apps, input formats, output handling\n- `references\u002Fcli-reference.md` — Complete CLI command reference\n","# inference.sh CLI\n\nRun 150+ AI apps in the cloud with a simple CLI. No GPU required.\n\nAll commands use the **terminal tool** to run `infsh` commands.\n\n## When to Use\n\n- User asks to generate images (FLUX, Reve, Seedream, Grok, Gemini image)\n- User asks to generate video (Veo, Wan, Seedance, OmniHuman)\n- User asks about inference.sh or infsh\n- User wants to run AI apps without managing individual provider APIs\n- User asks for AI-powered search (Tavily, Exa)\n- User needs avatar\u002Flipsync generation\n\n## Prerequisites\n\nThe `infsh` CLI must be installed and authenticated. Check with:\n\n```bash\ninfsh me\n```\n\nIf not installed:\n\n```bash\ncurl -fsSL https:\u002F\u002Fcli.inference.sh | sh\ninfsh login\n```\n\nSee `references\u002Fauthentication.md` for full setup details.\n\n## Workflow\n\n### 1. Always Search First\n\nNever guess app names — always search to find the correct app ID:\n\n```bash\ninfsh app list --search flux\ninfsh app list --search video\ninfsh app list --search image\n```\n\n### 2. Run an App\n\nUse the exact app ID from the search results. Always use `--json` for machine-readable output:\n\n```bash\ninfsh app run \u003Capp-id> --input '{\"prompt\": \"your prompt here\"}' --json\n```\n\n### 3. Parse the Output\n\nThe JSON output contains URLs to generated media. Present these to the user with `MEDIA:\u003Curl>` for inline display.\n\n## Common Commands\n\n### Image Generation\n\n```bash\n# Search for image apps\ninfsh app list --search image\n\n# FLUX Dev with LoRA\ninfsh app run falai\u002Fflux-dev-lora --input '{\"prompt\": \"sunset over mountains\", \"num_images\": 1}' --json\n\n# Gemini image generation\ninfsh app run google\u002Fgemini-2-5-flash-image --input '{\"prompt\": \"futuristic city\", \"num_images\": 1}' --json\n\n# Seedream (ByteDance)\ninfsh app run bytedance\u002Fseedream-5-lite --input '{\"prompt\": \"nature scene\"}' --json\n\n# Grok Imagine (xAI)\ninfsh app run xai\u002Fgrok-imagine-image --input '{\"prompt\": \"abstract art\"}' --json\n```\n\n### Video Generation\n\n```bash\n# Search for video apps\ninfsh app list --search video\n\n# Veo 3.1 (Google)\ninfsh app run google\u002Fveo-3-1-fast --input '{\"prompt\": \"drone shot of coastline\"}' --json\n\n# Seedance (ByteDance)\ninfsh app run bytedance\u002Fseedance-1-5-pro --input '{\"prompt\": \"dancing figure\", \"resolution\": \"1080p\"}' --json\n\n# Wan 2.5\ninfsh app run falai\u002Fwan-2-5 --input '{\"prompt\": \"person walking through city\"}' --json\n```\n\n### Local File Uploads\n\nThe CLI automatically uploads local files when you provide a path:\n\n```bash\n# Upscale a local image\ninfsh app run falai\u002Ftopaz-image-upscaler --input '{\"image\": \"\u002Fpath\u002Fto\u002Fphoto.jpg\", \"upscale_factor\": 2}' --json\n\n# Image-to-video from local file\ninfsh app run falai\u002Fwan-2-5-i2v --input '{\"image\": \"\u002Fpath\u002Fto\u002Fimage.png\", \"prompt\": \"make it move\"}' --json\n\n# Avatar with audio\ninfsh app run bytedance\u002Fomnihuman-1-5 --input '{\"audio\": \"\u002Fpath\u002Fto\u002Faudio.mp3\", \"image\": \"\u002Fpath\u002Fto\u002Fface.jpg\"}' --json\n```\n\n### Search & Research\n\n```bash\ninfsh app list --search search\ninfsh app run tavily\u002Ftavily-search --input '{\"query\": \"latest AI news\"}' --json\ninfsh app run exa\u002Fexa-search --input '{\"query\": \"machine learning papers\"}' --json\n```\n\n### Other Categories\n\n```bash\n# 3D generation\ninfsh app list --search 3d\n\n# Audio \u002F TTS\ninfsh app list --search tts\n\n# Twitter\u002FX automation\ninfsh app list --search twitter\n```\n\n## Pitfalls\n\n1. **Never guess app IDs** — always run `infsh app list --search \u003Cterm>` first. App IDs change and new apps are added frequently.\n2. **Always use `--json`** — raw output is hard to parse. The `--json` flag gives structured output with URLs.\n3. **Check authentication** — if commands fail with auth errors, run `infsh login` or verify `INFSH_API_KEY` is set.\n4. **Long-running apps** — video generation can take 30-120 seconds. The terminal tool timeout should be sufficient, but warn the user it may take a moment.\n5. **Input format** — the `--input` flag takes a JSON string. Make sure to properly escape quotes.\n\n## Reference Docs\n\n- `references\u002Fauthentication.md` — Setup, login, API keys\n- `references\u002Fapp-discovery.md` — Searching and browsing the app catalog\n- `references\u002Frunning-apps.md` — Running apps, input formats, output handling\n- `references\u002Fcli-reference.md` — Complete CLI command reference\n",{"name":5,"author":52,"license":15,"version":53,"metadata":54,"platforms":64,"description":7},"okaris","1.0.0",{"hermes":55},{"tags":56,"related_skills":63},[57,37,58,59,41,29,60,61,62],"AI","video","LLM","FLUX","Veo","Claude",[],[44,65,42],"macos","main","0d1434f89f4bc6b07e9d272accd8c1013768820f","registry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional",0,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills","https:\u002F\u002Fgithub.com\u002Fzapabob","The Open University of JAPAN　2019-now\r\n\r\nTokyo University of Science　2007-2010","2026-10-26 16:29:01",{"source":8,"skillName":5,"commands":75,"prompt":85,"zipAvailable":86},[76,79,82],{"tool":77,"command":78},"npx","npx skills add zapabob\u002FSkills --skill inference-sh-cli",{"tool":80,"command":81},"bunx","bunx skills add zapabob\u002FSkills --skill inference-sh-cli",{"tool":83,"command":84},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill inference-sh-cli","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",true,[88,94,98,102],{"path":89,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":92,"sourceHtmlUrl":93},"references\u002Fapp-discovery.md","REFERENCE","参考文档",2120,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002Freferences\u002Fapp-discovery.md",{"path":95,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":96,"sourceHtmlUrl":97},"references\u002Fauthentication.md",1084,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002Freferences\u002Fauthentication.md",{"path":99,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":100,"sourceHtmlUrl":101},"references\u002Fcli-reference.md",3181,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002Freferences\u002Fcli-reference.md",{"path":103,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":104,"sourceHtmlUrl":105},"references\u002Frunning-apps.md",3705,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fcli\u002Fvariants\u002Fhermes-agent-optional\u002Freferences\u002Frunning-apps.md",[107,135,170,195,222,238],{"skillNo":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":122,"publishedAt":47},"01M6GJTE1GESGGF9K028Y7753E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fxurl\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[113,116,117],{"categoryNo":114,"categoryName":115,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000010","后端与接口",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":118,"categoryName":119,"taxonomyType":120,"taxonomyTypeDesc":121},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[109,123,65,27,124,44,42,125,41,126,127,128,129,130,131,132,66,36,133,134],"openclaw","github","official","https","social-media","api","Media","hermes","hermes agent","twitter","via","post",{"skillNo":136,"name":137,"nameZh":6,"description":138,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":139,"categories":140,"keywords":152,"publishedAt":47},"01M6GJT8S1PXQ8WPXT1HSNBWH2","writing-plans","Write implementation plans: bite-sized tasks, paths, code.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwriting-plans\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[141,144,147,148,149],{"categoryNo":142,"categoryName":143,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":145,"categoryName":146,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":118,"categoryName":119,"taxonomyType":120,"taxonomyTypeDesc":121},{"categoryNo":150,"categoryName":151,"taxonomyType":120,"taxonomyTypeDesc":121},"01JBSEEDCAT000000000000036","内容创作者",[153,154,155,156,65,157,44,158,42,159,160,161,162,163,164,165,166,167,168,169],"writing","writing plans","plans","design","review","workflow","documentation","tasks","development","code","planning","code-review","subagent","write","superpowers","paths","test-driven-development",{"skillNo":171,"name":172,"nameZh":6,"description":173,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":174,"categories":175,"keywords":177,"publishedAt":47},"01M6GJST0H5X7WHBJ3BZK0DVKH","whisper","OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwhisper\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[176],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[172,65,44,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194],"openai","speech","multimodal","Transformers","ASR","Audio","text","model","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":196,"name":197,"nameZh":6,"description":198,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":199,"categories":200,"keywords":204,"publishedAt":47},"01M6GJSMRQ8EETZTYMG72P70EA","weights-and-biases","W&B: log ML experiments, sweeps, model registry, dashboards.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fweights-and-biases\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[201],{"categoryNo":202,"categoryName":203,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000013","分析与可视化",[205,65,206,207,42,208,209,210,211,212,213,185,214,215,216,217,218,219,220,221],"weights and biases","tensorflow","experiments","pytorch","visualization","registry","real-time","tuning","dashboards","experiment","log","time","Experiment Tracking","Model Registry","Hyperparameter Tuning","WandB","Real-Time Visualization",{"skillNo":223,"name":224,"nameZh":6,"description":225,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":226,"categories":227,"keywords":229,"publishedAt":47},"01M6GJS19KDPRY112JTR326S06","watchers","Poll RSS, JSON APIs, and GitHub with watermark dedup.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwatchers\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[228],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[224,65,230,124,44,43,231,232,233,234,235,130,131,236,237],"terminal","http","monitoring","cron","json","rss","devops","with",{"skillNo":239,"name":240,"nameZh":6,"description":241,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":242,"categories":243,"keywords":246,"publishedAt":47},"01M6GJRHJD2A6Q830SWF9W48BC","serving-llms-vllm","vLLM: high-throughput LLM serving, OpenAI API, quantization.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fvllm\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[244,245],{"categoryNo":114,"categoryName":115,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":118,"categoryName":119,"taxonomyType":120,"taxonomyTypeDesc":121},[45,247,248,44,249,178,250,251,252,181,253,128,254,255,256,257,258,259,260,261],"vLLM","serving llms vllm","research","openai api","tensor","Inference Serving","Production","llm serving","High Throughput","PagedAttention","Continuous Batching","Quantization","Tensor Parallelism","high-throughput","throughput"]