Add files using upload-large-folder tool

.gitattributes

@@ -47,3 +47,5 @@ SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#int
 SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#inter_0.4/0filter_ambiguous/largest_intra_inter.jsonl filter=lfs diff=lfs merge=lfs -text
 SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#inter_0.4/filter_ambiguous/largest_intra_inter.jsonl filter=lfs diff=lfs merge=lfs -text
 SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#inter_0.4/0filter_ambiguous/smallest_intra_inter.jsonl filter=lfs diff=lfs merge=lfs -text
+SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#inter_0.4/different_text_inter.jsonl filter=lfs diff=lfs merge=lfs -text
+SFR-Embedding-Code-2B_R\#0.8\#0.6\#dinov2-large\#0.0\#0.2\#rouge_0.2\#top_8\#inter_0.4/0filter_ambiguous/intra_fixed.jsonl filter=lfs diff=lfs merge=lfs -text

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/0filter_ambiguous/intra_fixed.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af3768b624539180f9f3f96bced76935478a013682b7e60514a31fc956398fa8
+size 1670544155

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_image_closest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb7bd3d8472707c27dd3f5bc4244ef688a24f686b1712ae6ae55bcff9a353e6f
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_image_farest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f50c62b1be871d2b358015a319f9092c70b18ebb03e76de2e514102feca377b
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_image_uniform/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d8cc209e213f198adfdf3e6d15537489975651102fa5f34e0bfefe0a6fc1b1c
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_random/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d8e8f114c1eabd176a00d35a285561e88e30d27421f4388dbc6f335f4f4a5ae
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_text_closest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7d428a17bb5367923de68261f218c7bdb6d491f89944f31773e2803eceed47b
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_text_closest_image_closest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5b12cdf9e2909e257e5c914c893518ed2433c3ae34e4a14286dae2582ed8f36
+size 1512951586

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_text_farest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eea37af0b6e97b8efe426718b0dc9c64dd4b330a7811f53e46f91a30b8b6e4fd
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_text_farest_image_farest/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62067d0f84aa2205f6ca5ed384827e08e801e5bb08e0a6f604e005fc98abae38
+size 1512951586

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/keep_text_uniform/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9acf98a396f6f9f7f7776c77d26a33005f516a7a1ce067f613a21ffac15a2613
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune-lock-text/low_inter_only/checkpoints/epoch_1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:774f38dcf66e7a359d55b7c6a93322e54a0002c54177eec4670866fc59015cc5
+size 1512951522

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_caltech101_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "caltech101", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.8611111111111112, "acc5": 0.9646762904636921, "mean_per_class_recall": 0.9117574143020897}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_cars_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cars", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.8604651162790697, "acc5": 0.9906728018903121, "mean_per_class_recall": 0.8603007472001762}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_cifar100_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cifar100", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.7329, "acc5": 0.9282, "mean_per_class_recall": 0.7332000000000002}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_cifar10_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cifar10", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.9239, "acc5": 0.9975, "mean_per_class_recall": 0.9239}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_country211_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "country211", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.1642654028436019, "acc5": 0.37867298578199055, "mean_per_class_recall": 0.1641706161137441}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_dtd_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "dtd", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.5569148936170213, "acc5": 0.851063829787234, "mean_per_class_recall": 0.5558510638297871}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_eurosat_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "eurosat", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.5003333333333333, "acc5": 0.9582962962962963, "mean_per_class_recall": 0.5124266666666667}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_fgvc_aircraft_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "fgvc_aircraft", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.23492349234923493, "acc5": 0.5643564356435643, "mean_per_class_recall": 0.23520499108734402}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_flickr30k_epoch_1.pt_ViT-B-32_en_zeroshot_retrieval.json

@@ -0,0 +1 @@
+{"dataset": "flickr30k", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_retrieval", "metrics": {"image_retrieval_recall@1": 0.6736000180244446, "text_retrieval_recall@1": 0.8360000252723694, "image_retrieval_recall@5": 0.8812000155448914, "text_retrieval_recall@5": 0.9599999785423279, "image_retrieval_recall@10": 0.932200014591217, "text_retrieval_recall@10": 0.984000027179718}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_flowers_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "flowers", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.7049926817368678, "acc5": 0.8697349162465442, "mean_per_class_recall": 0.7014735954769374}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_food101_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "food101", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.819009900990099, "acc5": 0.9664950495049505, "mean_per_class_recall": 0.8192475247524755}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_gtsrb_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "gtsrb", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.5064133016627078, "acc5": 0.7681710213776722, "mean_per_class_recall": 0.4390340643422039}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_imagenet1k_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "imagenet1k", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.65884, "acc5": 0.89428, "mean_per_class_recall": 0.6587000000000001}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_mscoco_captions_epoch_1.pt_ViT-B-32_en_zeroshot_retrieval.json

@@ -0,0 +1 @@
+{"dataset": "mscoco_captions", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_retrieval", "metrics": {"image_retrieval_recall@1": 0.39972010254859924, "text_retrieval_recall@1": 0.5735999941825867, "image_retrieval_recall@5": 0.6586565375328064, "text_retrieval_recall@5": 0.8055999875068665, "image_retrieval_recall@10": 0.759576141834259, "text_retrieval_recall@10": 0.878000020980835}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_pets_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "pets", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.9067865903515945, "acc5": 0.9942763695829926, "mean_per_class_recall": 0.9051993440098055}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_stl10_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "stl10", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.96425, "acc5": 0.99975, "mean_per_class_recall": 0.96425}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_sun397_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "sun397", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.6866689960829027, "acc5": 0.9371057616271585, "mean_per_class_recall": 0.6837170321383007}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/benchmark_vtab_resisc45_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "vtab/resisc45", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.6, "acc5": 0.9061904761904762, "mean_per_class_recall": 0.6079051385663736}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/out.log

@@ -0,0 +1,159 @@
+2025-04-24,00:12:07 | INFO | Running with a single process. Device cuda.
+2025-04-24,00:12:07 | INFO | Loaded ViT-B-32 model config.
+2025-04-24,00:12:09 | INFO | Loading pretrained ViT-B-32 weights (laion2b_s34b_b79k).
+2025-04-24,00:12:09 | INFO | Model:
+2025-04-24,00:12:09 | INFO | CLIP(
+  (visual): VisionTransformer(
+    (conv1): Conv2d(3, 768, kernel_size=(32, 32), stride=(32, 32), bias=False)
+    (patch_dropout): Identity()
+    (ln_pre): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+    (transformer): Transformer(
+      (resblocks): ModuleList(
+        (0-11): 12 x ResidualAttentionBlock(
+          (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+          (attn): MultiheadAttention(
+            (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
+          )
+          (ls_1): Identity()
+          (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+          (mlp): Sequential(
+            (c_fc): Linear(in_features=768, out_features=3072, bias=True)
+            (gelu): GELU(approximate='none')
+            (c_proj): Linear(in_features=3072, out_features=768, bias=True)
+          )
+          (ls_2): Identity()
+        )
+      )
+    )
+    (ln_post): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
+  )
+  (transformer): Transformer(
+    (resblocks): ModuleList(
+      (0-11): 12 x ResidualAttentionBlock(
+        (ln_1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
+        (attn): MultiheadAttention(
+          (out_proj): NonDynamicallyQuantizableLinear(in_features=512, out_features=512, bias=True)
+        )
+        (ls_1): Identity()
+        (ln_2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
+        (mlp): Sequential(
+          (c_fc): Linear(in_features=512, out_features=2048, bias=True)
+          (gelu): GELU(approximate='none')
+          (c_proj): Linear(in_features=2048, out_features=512, bias=True)
+        )
+        (ls_2): Identity()
+      )
+    )
+  )
+  (token_embedding): Embedding(49408, 512)
+  (ln_final): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
+)
+2025-04-24,00:12:09 | INFO | Params:
+2025-04-24,00:12:09 | INFO |   accum_freq: 4
+2025-04-24,00:12:09 | INFO |   aug_cfg: {}
+2025-04-24,00:12:09 | INFO |   batch_size: 2048
+2025-04-24,00:12:09 | INFO |   beta1: 0.9
+2025-04-24,00:12:09 | INFO |   beta2: 0.98
+2025-04-24,00:12:09 | INFO |   cache_dir: None
+2025-04-24,00:12:09 | INFO |   caption_ratio: 0.1
+2025-04-24,00:12:09 | INFO |   checkpoint_path: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints
+2025-04-24,00:12:09 | INFO |   coca_caption_loss_weight: 2.0
+2025-04-24,00:12:09 | INFO |   coca_contrastive_loss_weight: 1.0
+2025-04-24,00:12:09 | INFO |   copy_codebase: False
+2025-04-24,00:12:09 | INFO |   csv_caption_key: title
+2025-04-24,00:12:09 | INFO |   csv_img_key: filepath
+2025-04-24,00:12:09 | INFO |   csv_separator: 	
+2025-04-24,00:12:09 | INFO |   dataset_resampled: False
+2025-04-24,00:12:09 | INFO |   dataset_type: synthetic
+2025-04-24,00:12:09 | INFO |   ddp_static_graph: False
+2025-04-24,00:12:09 | INFO |   debug: False
+2025-04-24,00:12:09 | INFO |   delete_previous_checkpoint: False
+2025-04-24,00:12:09 | INFO |   device: cuda
+2025-04-24,00:12:09 | INFO |   dist_backend: None
+2025-04-24,00:12:09 | INFO |   dist_url: None
+2025-04-24,00:12:09 | INFO |   distill: False
+2025-04-24,00:12:09 | INFO |   distill_model: None
+2025-04-24,00:12:09 | INFO |   distill_pretrained: None
+2025-04-24,00:12:09 | INFO |   distributed: False
+2025-04-24,00:12:09 | INFO |   epochs: 1
+2025-04-24,00:12:09 | INFO |   epochs_cooldown: None
+2025-04-24,00:12:09 | INFO |   eps: 1e-08
+2025-04-24,00:12:09 | INFO |   force_custom_text: False
+2025-04-24,00:12:09 | INFO |   force_image_size: None
+2025-04-24,00:12:09 | INFO |   force_patch_dropout: None
+2025-04-24,00:12:09 | INFO |   force_quick_gelu: False
+2025-04-24,00:12:09 | INFO |   gather_with_grad: True
+2025-04-24,00:12:09 | INFO |   grad_checkpointing: True
+2025-04-24,00:12:09 | INFO |   grad_clip_norm: None
+2025-04-24,00:12:09 | INFO |   horovod: False
+2025-04-24,00:12:09 | INFO |   image_interpolation: None
+2025-04-24,00:12:09 | INFO |   image_mean: None
+2025-04-24,00:12:09 | INFO |   image_resize_mode: None
+2025-04-24,00:12:09 | INFO |   image_std: None
+2025-04-24,00:12:09 | INFO |   imagenet_v2: None
+2025-04-24,00:12:09 | INFO |   imagenet_val: None
+2025-04-24,00:12:09 | INFO |   keep_func_name: keep_image_farest
+2025-04-24,00:12:09 | INFO |   local_loss: False
+2025-04-24,00:12:09 | INFO |   local_rank: 0
+2025-04-24,00:12:09 | INFO |   lock_image: False
+2025-04-24,00:12:09 | INFO |   lock_image_freeze_bn_stats: False
+2025-04-24,00:12:09 | INFO |   lock_image_unlocked_groups: 0
+2025-04-24,00:12:09 | INFO |   lock_text: False
+2025-04-24,00:12:09 | INFO |   lock_text_freeze_layer_norm: False
+2025-04-24,00:12:09 | INFO |   lock_text_unlocked_layers: 0
+2025-04-24,00:12:09 | INFO |   log_every_n_steps: 100
+2025-04-24,00:12:09 | INFO |   log_level: 20
+2025-04-24,00:12:09 | INFO |   log_local: False
+2025-04-24,00:12:09 | INFO |   log_path: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/out.log
+2025-04-24,00:12:09 | INFO |   logs: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs
+2025-04-24,00:12:09 | INFO |   loss_dist_impl: None
+2025-04-24,00:12:09 | INFO |   lr: 1e-05
+2025-04-24,00:12:09 | INFO |   lr_cooldown_end: 0.0
+2025-04-24,00:12:09 | INFO |   lr_cooldown_power: 1.0
+2025-04-24,00:12:09 | INFO |   lr_scheduler: cosine
+2025-04-24,00:12:09 | INFO |   map_func_name: use_all
+2025-04-24,00:12:09 | INFO |   model: ViT-B-32
+2025-04-24,00:12:09 | INFO |   momentum: None
+2025-04-24,00:12:09 | INFO |   name: ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest
+2025-04-24,00:12:09 | INFO |   no_set_device_rank: False
+2025-04-24,00:12:09 | INFO |   opt: adamw
+2025-04-24,00:12:09 | INFO |   precision: amp
+2025-04-24,00:12:09 | INFO |   pretrained: laion2b_s34b_b79k
+2025-04-24,00:12:09 | INFO |   pretrained_image: False
+2025-04-24,00:12:09 | INFO |   rank: 0
+2025-04-24,00:12:09 | INFO |   remote_sync: None
+2025-04-24,00:12:09 | INFO |   remote_sync_frequency: 300
+2025-04-24,00:12:09 | INFO |   remote_sync_protocol: s3
+2025-04-24,00:12:09 | INFO |   report_to: wandb
+2025-04-24,00:12:09 | INFO |   resume: None
+2025-04-24,00:12:09 | INFO |   save_frequency: 1
+2025-04-24,00:12:09 | INFO |   save_most_recent: False
+2025-04-24,00:12:09 | INFO |   seed: 0
+2025-04-24,00:12:09 | INFO |   siglip: False
+2025-04-24,00:12:09 | INFO |   skip_scheduler: False
+2025-04-24,00:12:09 | INFO |   tensorboard: False
+2025-04-24,00:12:09 | INFO |   tensorboard_path: 
+2025-04-24,00:12:09 | INFO |   torchcompile: False
+2025-04-24,00:12:09 | INFO |   torchscript: False
+2025-04-24,00:12:09 | INFO |   trace: False
+2025-04-24,00:12:09 | INFO |   train_data: /mnt/personal/zhudongy/CapsFusion120M-wds/{00000..01426}.tar
+2025-04-24,00:12:09 | INFO |   train_data_upsampling_factors: None
+2025-04-24,00:12:09 | INFO |   train_num_samples: 10006295
+2025-04-24,00:12:09 | INFO |   use_bn_sync: False
+2025-04-24,00:12:09 | INFO |   use_bnb_linear: None
+2025-04-24,00:12:09 | INFO |   val_data: None
+2025-04-24,00:12:09 | INFO |   val_frequency: 1
+2025-04-24,00:12:09 | INFO |   val_num_samples: None
+2025-04-24,00:12:09 | INFO |   wandb: True
+2025-04-24,00:12:09 | INFO |   wandb_notes: 
+2025-04-24,00:12:09 | INFO |   wandb_project_name: open-clip
+2025-04-24,00:12:09 | INFO |   warmup: 122
+2025-04-24,00:12:09 | INFO |   wd: 0.5
+2025-04-24,00:12:09 | INFO |   workers: 16
+2025-04-24,00:12:09 | INFO |   world_size: 1
+2025-04-24,00:12:09 | INFO |   zeroshot_frequency: 2
+2025-04-24,00:12:09 | INFO | Created AdamW (adamw) optimizer: lr: 1e-05, betas: (0.9, 0.98), eps: 1e-08, weight_decay: 0.5, amsgrad: False, foreach: None, maximize: False, capturable: False, differentiable: False, fused: None
+2025-04-24,00:15:17 | INFO | Start epoch 0
+2025-04-24,00:15:49 | INFO | Train Epoch: 0 [   8192/1015808 (1%)] Data (t): 23.724 Batch (t): 31.607, 259.181/s, 259.181/s/gpu LR: 0.000000 Logit Scale: 100.000 Imm_image: 32.957 (32.957) Imm_text: 32.957 (32.957) Isd_image: 4.5893 (4.5893) Isd_text: 4.5893 (4.5893) Contrastive_loss: 0.61915 (0.61915) Loss: 0.61915 (0.61915)
+2025-04-24,00:28:52 | INFO | Train Epoch: 0 [ 827392/1015808 (81%)] Data (t): 1.129 Batch (t): 7.828, 1053.99/s, 1053.99/s/gpu LR: 0.000008 Logit Scale: 99.989 Imm_image: 33.944 (33.451) Imm_text: 33.944 (33.451) Isd_image: 5.0255 (4.8074) Isd_text: 5.0255 (4.8074) Contrastive_loss: 0.37668 (0.49792) Loss: 0.37668 (0.49792)
+2025-04-24,00:31:57 | INFO | Train Epoch: 0 [1015808/1015808 (100%)] Data (t): 1.280 Batch (t): 8.069, 1047.99/s, 1047.99/s/gpu LR: 0.000005 Logit Scale: 99.987 Imm_image: 34.089 (33.663) Imm_text: 34.089 (33.663) Isd_image: 4.6706 (4.7618) Isd_text: 4.6706 (4.7618) Contrastive_loss: 0.36459 (0.45347) Loss: 0.36459 (0.45347)

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_farest/params.txt

@@ -0,0 +1,103 @@
+accum_freq: 4
+aug_cfg: {}
+batch_size: 2048
+beta1: 0.9
+beta2: 0.98
+cache_dir: None
+caption_ratio: 0.1
+checkpoint_path: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/checkpoints
+coca_caption_loss_weight: 2.0
+coca_contrastive_loss_weight: 1.0
+copy_codebase: False
+csv_caption_key: title
+csv_img_key: filepath
+csv_separator: 	
+dataset_resampled: False
+dataset_type: synthetic
+ddp_static_graph: False
+debug: False
+delete_previous_checkpoint: False
+device: cuda
+dist_backend: None
+dist_url: None
+distill: False
+distill_model: None
+distill_pretrained: None
+distributed: False
+epochs: 1
+epochs_cooldown: None
+eps: 1e-08
+force_custom_text: False
+force_image_size: None
+force_patch_dropout: None
+force_quick_gelu: False
+gather_with_grad: True
+grad_checkpointing: True
+grad_clip_norm: None
+horovod: False
+image_interpolation: None
+image_mean: None
+image_resize_mode: None
+image_std: None
+imagenet_v2: None
+imagenet_val: None
+keep_func_name: keep_image_farest
+local_loss: False
+local_rank: 0
+lock_image: False
+lock_image_freeze_bn_stats: False
+lock_image_unlocked_groups: 0
+lock_text: False
+lock_text_freeze_layer_norm: False
+lock_text_unlocked_layers: 0
+log_every_n_steps: 100
+log_level: 20
+log_local: False
+log_path: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest/out.log
+logs: /mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs
+loss_dist_impl: None
+lr: 1e-05
+lr_cooldown_end: 0.0
+lr_cooldown_power: 1.0
+lr_scheduler: cosine
+map_func_name: use_all
+model: ViT-B-32
+momentum: None
+name: ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_farest
+no_set_device_rank: False
+opt: adamw
+precision: amp
+pretrained: laion2b_s34b_b79k
+pretrained_image: False
+rank: 0
+remote_sync: None
+remote_sync_frequency: 300
+remote_sync_protocol: s3
+report_to: wandb
+resume: None
+save_frequency: 1
+save_most_recent: False
+seed: 0
+siglip: False
+skip_scheduler: False
+tensorboard: False
+tensorboard_path: 
+torchcompile: False
+torchscript: False
+trace: False
+train_data: /mnt/personal/zhudongy/CapsFusion120M-wds/{00000..01426}.tar
+train_data_upsampling_factors: None
+train_num_samples: 10006295
+use_bn_sync: False
+use_bnb_linear: None
+val_data: None
+val_frequency: 1
+val_num_samples: None
+wandb: True
+wandb_notes: 
+wandb_project_name: open-clip
+warmup: 122
+wd: 0.5
+workers: 16
+world_size: 1
+zeroshot_frequency: 2

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_caltech101_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "caltech101", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.8632983377077865, "acc5": 0.9641294838145232, "mean_per_class_recall": 0.9132225597261372}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_cars_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cars", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.8614600174107698, "acc5": 0.9898022634000746, "mean_per_class_recall": 0.862527756274282}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_cifar100_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cifar100", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.7413, "acc5": 0.9325, "mean_per_class_recall": 0.7419}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_cifar10_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "cifar10", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.928, "acc5": 0.9973, "mean_per_class_recall": 0.9279}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_country211_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "country211", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.1638388625592417, "acc5": 0.37668246445497633, "mean_per_class_recall": 0.16421800947867302}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_dtd_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "dtd", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.576063829787234, "acc5": 0.8563829787234043, "mean_per_class_recall": 0.5771276595744682}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_eurosat_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "eurosat", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.4793703703703704, "acc5": 0.9591481481481482, "mean_per_class_recall": 0.49027666666666664}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_fgvc_aircraft_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "fgvc_aircraft", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.24002400240024002, "acc5": 0.5553555355535553, "mean_per_class_recall": 0.23985739750445628}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_flickr30k_epoch_1.pt_ViT-B-32_en_zeroshot_retrieval.json

@@ -0,0 +1 @@
+{"dataset": "flickr30k", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_retrieval", "metrics": {"image_retrieval_recall@1": 0.6711999773979187, "text_retrieval_recall@1": 0.8299999833106995, "image_retrieval_recall@5": 0.8858000040054321, "text_retrieval_recall@5": 0.9570000171661377, "image_retrieval_recall@10": 0.9312000274658203, "text_retrieval_recall@10": 0.9810000061988831}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_flowers_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "flowers", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.7088957554073834, "acc5": 0.8736379899170597, "mean_per_class_recall": 0.6985781923439345}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_food101_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "food101", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.8213069306930693, "acc5": 0.9667326732673267, "mean_per_class_recall": 0.821069306930693}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_gtsrb_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "gtsrb", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.498812351543943, "acc5": 0.7555027711797307, "mean_per_class_recall": 0.42383596334759127}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_imagenet1k_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "imagenet1k", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.66026, "acc5": 0.89522, "mean_per_class_recall": 0.6604000000000001}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_mscoco_captions_epoch_1.pt_ViT-B-32_en_zeroshot_retrieval.json

@@ -0,0 +1 @@
+{"dataset": "mscoco_captions", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_retrieval", "metrics": {"image_retrieval_recall@1": 0.400479793548584, "text_retrieval_recall@1": 0.5802000164985657, "image_retrieval_recall@5": 0.6611355543136597, "text_retrieval_recall@5": 0.8051999807357788, "image_retrieval_recall@10": 0.7623750567436218, "text_retrieval_recall@10": 0.8772000074386597}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_pets_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "pets", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.9051512673753066, "acc5": 0.9953665849005179, "mean_per_class_recall": 0.9041328844971016}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_stl10_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "stl10", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.96075, "acc5": 0.999625, "mean_per_class_recall": 0.960625}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_sun397_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "sun397", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.6863011935193188, "acc5": 0.9371057616271585, "mean_per_class_recall": 0.6836321109384585}, "language": "en"}

SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/ViT-B-32-laion2b_s34b_b79k-010-1e5-recaption-finetune/keep_image_uniform/benchmark_vtab_resisc45_epoch_1.pt_ViT-B-32_en_zeroshot_classification.json

@@ -0,0 +1 @@
+{"dataset": "vtab/resisc45", "model": "ViT-B-32", "pretrained": "/mnt/personal/zhudongy/capsfusion10m_results/SFR-Embedding-Code-2B_R#0.8#0.6#dinov2-large#0.0#0.2#rouge_0.2#top_8#inter_0.4/logs/ViT-B-32-laion2b_s34b_b79k-finetune-dc10m-010-1e5-recaption-keep_image_uniform/checkpoints/epoch_1.pt", "task": "zeroshot_classification", "metrics": {"acc1": 0.5992063492063492, "acc5": 0.9063492063492063, "mean_per_class_recall": 0.60694300093902}, "language": "en"}