🐦 Twitter Post Details

Viewing enriched Twitter post

@lqiao

πŸš€ Eval Protocol is Open Sourced! Reinforcement fine-tuning is complicated, because there are hundreds of environments and tens of trainers you can pick and choose and integrate with. Even worse, in production, agents don’t live in clean β€œgyms.” They operate in messy, async environments - flaky APIs, partial observability, conflicting objectives, long feedback loops. We solve that problem by open sourcing Eval Protocol. The goal is for you to build your production RFT flow without reinventing the wheel of managing such complex integration. πŸ‘‰ Day 0 support for trainers and environments like TRL (@huggingface), @rllm_project , OpenEnv (@PyTorch),Β  as well as support for proprietary trainers like @OpenAI RFT and @thinkymachines Tinker. More to come. πŸ‘‰ Instrument agents in production instead of toy or simulated environments πŸ‘‰ Move from offline benchmarks to live, continuous improvement

Media 1

πŸ“Š Media Metadata

{
  "media": [
    {
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1992309156859228413/media_0.jpg?",
      "media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1992309156859228413/media_0.jpg?",
      "type": "photo",
      "filename": "media_0.jpg"
    }
  ],
  "processed_at": "2025-11-27T21:03:21.528702",
  "pipeline_version": "2.0"
}

πŸ”§ Raw API Response

{
  "type": "tweet",
  "id": "1992309156859228413",
  "url": "https://x.com/lqiao/status/1992309156859228413",
  "twitterUrl": "https://twitter.com/lqiao/status/1992309156859228413",
  "text": "πŸš€ Eval Protocol is Open Sourced!\n\nReinforcement fine-tuning is complicated, because there are hundreds of environments and tens of trainers you can pick and choose and integrate with. Even worse, in production, agents don’t live in clean β€œgyms.” They operate in messy, async environments - flaky APIs, partial observability, conflicting objectives, long feedback loops. \n\nWe solve that problem by open sourcing Eval Protocol. The goal is for you to build your production RFT flow without reinventing the wheel of managing such complex integration. \nπŸ‘‰ Day 0 support for trainers and environments like TRL (@huggingface), @rllm_project , OpenEnv (@PyTorch),Β  as well as support for proprietary trainers like @OpenAI  RFT and @thinkymachines Tinker. More to come.\nπŸ‘‰ Instrument agents in production instead of toy or simulated environments\nπŸ‘‰ Move from offline benchmarks to live, continuous improvement",
  "source": "Twitter for iPhone",
  "retweetCount": 39,
  "replyCount": 9,
  "likeCount": 207,
  "quoteCount": 3,
  "viewCount": 23187,
  "createdAt": "Sat Nov 22 19:08:12 +0000 2025",
  "lang": "en",
  "bookmarkCount": 199,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "1992309156859228413",
  "displayTextRange": [
    0,
    275
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "lqiao",
    "url": "https://x.com/lqiao",
    "twitterUrl": "https://twitter.com/lqiao",
    "id": "51670467",
    "name": "Lin Qiao",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1594101249254731776/-cNhmmTk_normal.jpg",
    "coverPicture": "",
    "description": "",
    "location": "Redwood City, CA",
    "followers": 5384,
    "following": 241,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sun Jun 28 06:41:23 +0000 2009",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 308,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 54,
    "statusesCount": 378,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [],
    "profile_bio": {
      "description": "Cofounder and CEO of @FireworksAI_HQ",
      "entities": {
        "description": {
          "user_mentions": [
            {
              "id_str": "0",
              "indices": [
                21,
                36
              ],
              "name": "",
              "screen_name": "FireworksAI_HQ"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "fireworks.ai",
              "expanded_url": "http://fireworks.ai",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/yq19EbpJyP"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/9GXFEWuTjz",
        "expanded_url": "https://twitter.com/lqiao/status/1992309156859228413/photo/1",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {},
          "orig": {}
        },
        "id_str": "1992307163314978816",
        "indices": [
          276,
          299
        ],
        "media_key": "3_1992307163314978816",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARumGM8w2wAACgACG6Yan1laMP0AAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABG6YYzzDbAAAKAAIbphqfWVow/QAA",
            "media_key": "3_1992307163314978816"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/G6YYzzDbAAAw-bc.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 613,
              "w": 1094,
              "x": 0,
              "y": 71
            },
            {
              "h": 684,
              "w": 684,
              "x": 0,
              "y": 0
            },
            {
              "h": 684,
              "w": 600,
              "x": 0,
              "y": 0
            },
            {
              "h": 684,
              "w": 342,
              "x": 0,
              "y": 0
            },
            {
              "h": 684,
              "w": 1094,
              "x": 0,
              "y": 0
            }
          ],
          "height": 684,
          "width": 1094
        },
        "sizes": {
          "large": {
            "h": 684,
            "w": 1094
          }
        },
        "type": "photo",
        "url": "https://t.co/9GXFEWuTjz"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "user_mentions": [
      {
        "id_str": "778764142412984320",
        "indices": [
          605,
          617
        ],
        "name": "Hugging Face",
        "screen_name": "huggingface"
      },
      {
        "id_str": "1975088327431999488",
        "indices": [
          620,
          633
        ],
        "name": "rLLM",
        "screen_name": "rllm_project"
      },
      {
        "id_str": "776585502606721024",
        "indices": [
          645,
          653
        ],
        "name": "PyTorch",
        "screen_name": "PyTorch"
      },
      {
        "id_str": "4398626122",
        "indices": [
          706,
          713
        ],
        "name": "OpenAI",
        "screen_name": "OpenAI"
      },
      {
        "id_str": "1891523745929732096",
        "indices": [
          723,
          738
        ],
        "name": "Thinking Machines",
        "screen_name": "thinkymachines"
      }
    ]
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}