Auto Daily Leaderboard udpate Fri Aug 30 12:00:24 PM EDT 2024

arena_elo/results/20240830/elo_results_t2i_generation.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b07bd1eb73ba2441e1a2826c961f95e0340ed7ed04ad558150f03309cd5687a
+size 88248

arena_elo/results/20240830/t2i_generation_leaderboard.csv

@@ -0,0 +1,18 @@
+key,Model,Arena Elo rating (anony),Arena Elo rating (full),License,Organization,Link
+FLUX.1-dev,FLUX.1-dev,1128.1076601793302,1145.4292259194874,flux-1-dev-non-commercial-license (other),Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
+PlayGround V2.5,PlayGround V2.5,1126.2242175792326,1127.0304762360083,Playground v2.5 Community License,Playground,https://huggingface.co/playgroundai/playground-v2.5-1024px-aesthetic
+FLUX.1-schnell,FLUX.1-schnell,1086.3891359749045,1082.431753147752,Apache-2.0,Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
+PlayGround V2,PlayGround V2,1072.3338566106129,1070.3959486055644,Playground v2 Community License,Playground,https://huggingface.co/playgroundai/playground-v2-1024px-aesthetic
+Kolors,Kolors,1055.6435171321828,1054.0770126985808,Apache-2.0,Kwai Kolors,https://huggingface.co/Kwai-Kolors/Kolors
+HunyuanDiT,HunyuanDiT,1038.8528018338375,1026.9064910406098,tencent-hunyuan-community,Tencent,https://github.com/Tencent/HunyuanDiT
+StableCascade,StableCascade,1037.482462728124,1039.871994721617,stable-cascade-nc-community (other),Stability AI,https://fal.ai/models/stable-cascade/api
+SDXL-Lightning,SDXL-Lightning,1023.5139956954258,1026.8635094413457,openrail++,ByteDance,https://huggingface.co/ByteDance/SDXL-Lightning
+PixArtAlpha,PixArtAlpha,1021.8191124795862,1011.2525795471853,openrail++,PixArt-alpha,https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS
+PixArtSigma,PixArtSigma,1018.5453640621988,1017.5527842992569,openrail++,PixArt-alpha,https://github.com/PixArt-alpha/PixArt-sigma
+AuraFlow,AuraFlow,1003.1773899329969,999.230314834677,Apache-2.0,Fal.AI,https://huggingface.co/fal/AuraFlow
+SD3,SD3,990.1711320470132,985.5201090260039,stabilityai-nc-research-community,Stability AI,https://huggingface.co/blog/sd3
+SDXL,SDXL,964.1291628891177,963.834050856154,openrail++,Stability AI,https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0
+SDXLTurbo,SDXLTurbo,912.3456630523833,908.9499820955415,sai-nc-community (other),Stability AI,https://huggingface.co/stabilityai/sdxl-turbo
+LCM(v1.5/XL),LCM(v1.5/XL),904.8799595749739,897.333770100635,openrail++,Latent Consistency,https://fal.ai/models/fast-lcm-diffusion-turbo
+OpenJourney,OpenJourney,827.3303069916949,821.2404802457792,creativeml-openrail-m,PromptHero,https://huggingface.co/prompthero/openjourney
+LCM,LCM,789.0542612363914,802.0876918790069,MIT License,Tsinghua University,https://huggingface.co/SimianLuo/LCM_Dreamshaper_v7

arena_elo/results/latest/clean_battle_t2i_generation.json

@@ -95852,5 +95852,103 @@
     "judge": "arena_user_10.16.34.60",
     "anony": true,
     "tstamp": 1724957244.8733
+  },
+  {
+    "model_a_conv_id": "fa377430c44e4740b8e19a6a52331bf8",
+    "model_b_conv_id": "7425e0586f214a5a9aba4d4a21b96ad6",
+    "inputs": {
+      "prompt": "una filiale di banca popolare di sondrio sul pianeta  marte con i clienti che fanno la coda per entrare"
+    },
+    "model_a": "SDXL",
+    "model_b": "OpenJourney",
+    "vote_type": "bothbad_vote",
+    "winner": "tie (bothbad)",
+    "judge": "arena_user_10.16.18.162",
+    "anony": true,
+    "tstamp": 1724962761.8085
+  },
+  {
+    "model_a_conv_id": "c918ef993350454f82b4833e789d4cf4",
+    "model_b_conv_id": "08c316715a5e466494b93996c57a8e65",
+    "inputs": {
+      "prompt": "pope in mountain bike with guitar"
+    },
+    "model_a": "AuraFlow",
+    "model_b": "PixArtSigma",
+    "vote_type": "bothbad_vote",
+    "winner": "tie (bothbad)",
+    "judge": "arena_user_10.16.10.147",
+    "anony": true,
+    "tstamp": 1724963166.0098
+  },
+  {
+    "model_a_conv_id": "c8e2c90f233049c7807f177240dad575",
+    "model_b_conv_id": "34f0d757308f4b7dbf99030849030ae1",
+    "inputs": {
+      "prompt": "Figure: Fine-Tuning Process\nPre-trained Model:\nIllustration: Show a large neural network with multiple layers. Label this as \"Pre-trained Model\" with an arrow indicating it has been trained on a \"Large General Dataset\" (e.g., ImageNet for images, large text corpus for NLP).\nLayers: Highlight the layers that are pre-trained, using a color or shading to indicate they contain learned features.\nTask-Specific Dataset:\nIllustration: Next to the pre-trained model, show a smaller dataset labeled \"Task-Specific Dataset\" (e.g., labeled images for a specific classification task).\nModel Adaptation:\nIllustration: Show the pre-trained model with an additional layer on top, labeled \"Task-Specific Layer\" or \"Output Layer.\" This layer is specific to the new task.\nAnnotations: Use arrows to indicate that the new layer is added, and some layers of the pre-trained model might be frozen initially.\nFine-Tuning:\nIllustration: Show the entire model (pre-trained layers plus the new task-specific layer) with arrows indicating the flow of data during training.\nAnnotations: Indicate that backpropagation is used to update weights, with some layers gradually unfrozen.\nOutput:\nIllustration: Show the final model with an arrow pointing to \"Task-Specific Predictions\" or \"Task-Specific Output.\"\nFlowchart: Fine-Tuning Steps\nStart:\nBox: \"Start with Pre-trained Model\"\nArrow: Pointing to the next step.\nLoad Task-Specific Dataset:\nBox: \"Load Task-Specific Dataset\"\nArrow: Pointing to the next step.\nModify Model Architecture:\nBox: \"Add Task-Specific Layer\"\nArrow: Pointing to the next step.\nFreeze Layers:\nBox: \"Freeze Some Pre-trained Layers\"\nArrow: Pointing to the next step.\nTrain Model:\nBox: \"Fine-Tune Model on Task-Specific Data\"\nArrow: Pointing to the next step.\nEvaluate Model:\nBox: \"Evaluate on Validation Set\"\nArrow: Pointing to the next step.\nAdjust Hyperparameters:\nBox: \"Adjust Hyperparameters if Necessary\"\nArrow: Pointing back to \"Train Model\" for iteration or forward to the next step.\nDeploy Model:\nBox: \"Deploy Fine-Tuned Model\"\nArrow: Pointing to \"End.\"\nEnd:\nBox: \"End\"\nExplanation\nPre-trained Model: Represents the starting point with general knowledge.\nTask-Specific Dataset: Shows the new data that will guide the fine-tuning.\nModel Adaptation: Illustrates the addition of new layers and potential freezing of existing ones.\nFine-Tuning: Highlights the training process where the model learns task-specific features.\nOutput: Demonstrates the model's ability to make predictions for the specific task.\nThis visualization helps convey the concept of fine-tuning by showing how a pre-trained model is adapted and trained to perform a specific task using a smaller, task-specific dataset."
+    },
+    "model_a": "FLUX.1-schnell",
+    "model_b": "HunyuanDiT",
+    "vote_type": "leftvote",
+    "winner": "model_a",
+    "judge": "arena_user_10.16.18.162",
+    "anony": true,
+    "tstamp": 1724969910.1761
+  },
+  {
+    "model_a_conv_id": "601acf65eec34c349ebf1e11f5ef38e7",
+    "model_b_conv_id": "e211933bb74745b988d325f77f96e3c3",
+    "inputs": {
+      "prompt": "portait of mystical witch, hyper detailed, flowing background, intricate and detailed, trippy, 8 k "
+    },
+    "model_a": "PlayGround V2.5",
+    "model_b": "HunyuanDiT",
+    "vote_type": "leftvote",
+    "winner": "model_a",
+    "judge": "arena_user_10.16.34.60",
+    "anony": true,
+    "tstamp": 1725013008.3565
+  },
+  {
+    "model_a_conv_id": "9a98c6652dea4da9a6c34e2af29dca9e",
+    "model_b_conv_id": "04ed72a7dcab447fa2dda9bfeac8d028",
+    "inputs": {
+      "prompt": "Young woman wearing red dress and black sweater gives approval while holding cell phone ."
+    },
+    "model_a": "SDXL",
+    "model_b": "Kolors",
+    "vote_type": "rightvote",
+    "winner": "model_b",
+    "judge": "arena_user_10.16.10.147",
+    "anony": true,
+    "tstamp": 1725013023.6849
+  },
+  {
+    "model_a_conv_id": "78b783666c2944db91d1d65249e3afcc",
+    "model_b_conv_id": "5dd887c139cc4d7f93e954687e681fac",
+    "inputs": {
+      "prompt": "a cute dog is playing a ball"
+    },
+    "model_a": "PlayGround V2.5",
+    "model_b": "AuraFlow",
+    "vote_type": "leftvote",
+    "winner": "model_a",
+    "judge": "arena_user_10.16.1.165",
+    "anony": true,
+    "tstamp": 1725013066.5854
+  },
+  {
+    "model_a_conv_id": "ab920d339c9040d588bd699e8737ece0",
+    "model_b_conv_id": "b70c63a515da46d2bd8c5c7ce4bf2579",
+    "inputs": {
+      "prompt": "A cat riding a bamboo raft"
+    },
+    "model_a": "FLUX.1-dev",
+    "model_b": "SD3",
+    "vote_type": "leftvote",
+    "winner": "model_a",
+    "judge": "arena_user_10.16.34.60",
+    "anony": true,
+    "tstamp": 1725019838.6535
   }
 ]

arena_elo/results/latest/elo_results_t2i_generation.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ef671f77ad5445c22d89e41197fb29a719cd60ec5371370b7d9d3b8369b2c03
-size 88249
+oid sha256:7b07bd1eb73ba2441e1a2826c961f95e0340ed7ed04ad558150f03309cd5687a
+size 88248

arena_elo/results/latest/t2i_generation_leaderboard.csv

@@ -1,18 +1,18 @@
 key,Model,Arena Elo rating (anony),Arena Elo rating (full),License,Organization,Link
-PlayGround V2.5,PlayGround V2.5,1125.9647113219344,1126.8179449346965,Playground v2.5 Community License,Playground,https://huggingface.co/playgroundai/playground-v2.5-1024px-aesthetic
-FLUX.1-dev,FLUX.1-dev,1122.977921149468,1141.8553966548204,flux-1-dev-non-commercial-license (other),Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
-FLUX.1-schnell,FLUX.1-schnell,1083.197039124354,1078.930126686188,Apache-2.0,Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
-PlayGround V2,PlayGround V2,1072.626941667019,1070.726547878426,Playground v2 Community License,Playground,https://huggingface.co/playgroundai/playground-v2-1024px-aesthetic
-Kolors,Kolors,1052.908730084688,1050.9850967380553,Apache-2.0,Kwai Kolors,https://huggingface.co/Kwai-Kolors/Kolors
-HunyuanDiT,HunyuanDiT,1041.9179059135952,1029.5490717475357,tencent-hunyuan-community,Tencent,https://github.com/Tencent/HunyuanDiT
-StableCascade,StableCascade,1037.9168175776274,1040.2341802192325,stable-cascade-nc-community (other),Stability AI,https://fal.ai/models/stable-cascade/api
-SDXL-Lightning,SDXL-Lightning,1023.8917986436493,1027.1783912113383,openrail++,ByteDance,https://huggingface.co/ByteDance/SDXL-Lightning
-PixArtAlpha,PixArtAlpha,1022.2807006610543,1011.6090234692596,openrail++,PixArt-alpha,https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS
-PixArtSigma,PixArtSigma,1019.027621124307,1017.9548990662921,openrail++,PixArt-alpha,https://github.com/PixArt-alpha/PixArt-sigma
-AuraFlow,AuraFlow,1005.6783096127664,1001.6148483504423,Apache-2.0,Fal.AI,https://huggingface.co/fal/AuraFlow
-SD3,SD3,991.5967272859457,987.1118479592413,stabilityai-nc-research-community,Stability AI,https://huggingface.co/blog/sd3
-SDXL,SDXL,964.6926541747703,964.4225091224126,openrail++,Stability AI,https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0
-SDXLTurbo,SDXLTurbo,912.792093681317,909.3024088309655,sai-nc-community (other),Stability AI,https://huggingface.co/stabilityai/sdxl-turbo
-LCM(v1.5/XL),LCM(v1.5/XL),905.2873322742623,897.684769464938,openrail++,Latent Consistency,https://fal.ai/models/fast-lcm-diffusion-turbo
-OpenJourney,OpenJourney,827.6690928608571,821.5243787424323,creativeml-openrail-m,PromptHero,https://huggingface.co/prompthero/openjourney
-LCM,LCM,789.5736028423817,802.4526781048648,MIT License,Tsinghua University,https://huggingface.co/SimianLuo/LCM_Dreamshaper_v7
+FLUX.1-dev,FLUX.1-dev,1128.1076601793302,1145.4292259194874,flux-1-dev-non-commercial-license (other),Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
+PlayGround V2.5,PlayGround V2.5,1126.2242175792326,1127.0304762360083,Playground v2.5 Community License,Playground,https://huggingface.co/playgroundai/playground-v2.5-1024px-aesthetic
+FLUX.1-schnell,FLUX.1-schnell,1086.3891359749045,1082.431753147752,Apache-2.0,Black Forest Labs,https://huggingface.co/docs/diffusers/main/en/api/pipelines/flux
+PlayGround V2,PlayGround V2,1072.3338566106129,1070.3959486055644,Playground v2 Community License,Playground,https://huggingface.co/playgroundai/playground-v2-1024px-aesthetic
+Kolors,Kolors,1055.6435171321828,1054.0770126985808,Apache-2.0,Kwai Kolors,https://huggingface.co/Kwai-Kolors/Kolors
+HunyuanDiT,HunyuanDiT,1038.8528018338375,1026.9064910406098,tencent-hunyuan-community,Tencent,https://github.com/Tencent/HunyuanDiT
+StableCascade,StableCascade,1037.482462728124,1039.871994721617,stable-cascade-nc-community (other),Stability AI,https://fal.ai/models/stable-cascade/api
+SDXL-Lightning,SDXL-Lightning,1023.5139956954258,1026.8635094413457,openrail++,ByteDance,https://huggingface.co/ByteDance/SDXL-Lightning
+PixArtAlpha,PixArtAlpha,1021.8191124795862,1011.2525795471853,openrail++,PixArt-alpha,https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS
+PixArtSigma,PixArtSigma,1018.5453640621988,1017.5527842992569,openrail++,PixArt-alpha,https://github.com/PixArt-alpha/PixArt-sigma
+AuraFlow,AuraFlow,1003.1773899329969,999.230314834677,Apache-2.0,Fal.AI,https://huggingface.co/fal/AuraFlow
+SD3,SD3,990.1711320470132,985.5201090260039,stabilityai-nc-research-community,Stability AI,https://huggingface.co/blog/sd3
+SDXL,SDXL,964.1291628891177,963.834050856154,openrail++,Stability AI,https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0
+SDXLTurbo,SDXLTurbo,912.3456630523833,908.9499820955415,sai-nc-community (other),Stability AI,https://huggingface.co/stabilityai/sdxl-turbo
+LCM(v1.5/XL),LCM(v1.5/XL),904.8799595749739,897.333770100635,openrail++,Latent Consistency,https://fal.ai/models/fast-lcm-diffusion-turbo
+OpenJourney,OpenJourney,827.3303069916949,821.2404802457792,creativeml-openrail-m,PromptHero,https://huggingface.co/prompthero/openjourney
+LCM,LCM,789.0542612363914,802.0876918790069,MIT License,Tsinghua University,https://huggingface.co/SimianLuo/LCM_Dreamshaper_v7