Training in progress, step 9000, checkpoint

checkpoint-9000/config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "openai/clip-vit-large-patch14-336",
+  "architectures": [
+    "CLIPModel"
+  ],
+  "initializer_factor": 1.0,
+  "logit_scale_init_value": 2.6592,
+  "model_type": "clip",
+  "projection_dim": 768,
+  "text_config": {
+    "dropout": 0.0,
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "model_type": "clip_text_model",
+    "num_attention_heads": 12,
+    "projection_dim": 768
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.0.dev0",
+  "vision_config": {
+    "dropout": 0.0,
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768
+  }
+}

checkpoint-9000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adb9d6ca55e0d44f526ca359db1f2df7973e7e89f9a1be68d478034b85336772
+size 1711848436

checkpoint-9000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1005daa8a00f477e721fda1d9644171715374b913eb900eb550b65762d738475
+size 3424043887

checkpoint-9000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75e7a66179ff03d27874bf0d0676cd79fe7137c908da603458cd4b183abc4a42
+size 14503

checkpoint-9000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a72fba01b28c5359d8cfb82e817d7b9af9c8e725475a079a4b45150d9b0ed1e7
+size 623

checkpoint-9000/trainer_state.json

@@ -0,0 +1,303 @@
+{
+  "best_metric": 0.6901102066040039,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l37-l/checkpoint-9000",
+  "epoch": 1.0131712259371835,
+  "eval_steps": 500,
+  "global_step": 9000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.056287290329843524,
+      "grad_norm": 96.92656707763672,
+      "learning_rate": 2.9437127096701565e-07,
+      "loss": 0.3748,
+      "step": 500
+    },
+    {
+      "epoch": 0.056287290329843524,
+      "eval_loss": 1.2226382493972778,
+      "eval_runtime": 131.2779,
+      "eval_samples_per_second": 15.037,
+      "eval_steps_per_second": 1.882,
+      "step": 500
+    },
+    {
+      "epoch": 0.11257458065968705,
+      "grad_norm": 402.31158447265625,
+      "learning_rate": 2.887425419340313e-07,
+      "loss": 0.3057,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11257458065968705,
+      "eval_loss": 1.0702259540557861,
+      "eval_runtime": 131.0109,
+      "eval_samples_per_second": 15.067,
+      "eval_steps_per_second": 1.885,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16886187098953057,
+      "grad_norm": 515.0911865234375,
+      "learning_rate": 2.831138129010469e-07,
+      "loss": 0.2239,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16886187098953057,
+      "eval_loss": 0.9957238435745239,
+      "eval_runtime": 134.8377,
+      "eval_samples_per_second": 14.64,
+      "eval_steps_per_second": 1.832,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2251491613193741,
+      "grad_norm": 3.1967105865478516,
+      "learning_rate": 2.774850838680626e-07,
+      "loss": 0.2229,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2251491613193741,
+      "eval_loss": 0.9504629373550415,
+      "eval_runtime": 129.0169,
+      "eval_samples_per_second": 15.3,
+      "eval_steps_per_second": 1.914,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2814364516492176,
+      "grad_norm": 276.22998046875,
+      "learning_rate": 2.718563548350782e-07,
+      "loss": 0.2098,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2814364516492176,
+      "eval_loss": 0.9006705284118652,
+      "eval_runtime": 129.1816,
+      "eval_samples_per_second": 15.281,
+      "eval_steps_per_second": 1.912,
+      "step": 2500
+    },
+    {
+      "epoch": 0.33772374197906113,
+      "grad_norm": 1105.317138671875,
+      "learning_rate": 2.6622762580209386e-07,
+      "loss": 0.1938,
+      "step": 3000
+    },
+    {
+      "epoch": 0.33772374197906113,
+      "eval_loss": 0.8782906532287598,
+      "eval_runtime": 128.6387,
+      "eval_samples_per_second": 15.345,
+      "eval_steps_per_second": 1.92,
+      "step": 3000
+    },
+    {
+      "epoch": 0.39401103230890466,
+      "grad_norm": 0.00043465400813147426,
+      "learning_rate": 2.605988967691095e-07,
+      "loss": 0.1688,
+      "step": 3500
+    },
+    {
+      "epoch": 0.39401103230890466,
+      "eval_loss": 0.8405746221542358,
+      "eval_runtime": 128.6989,
+      "eval_samples_per_second": 15.338,
+      "eval_steps_per_second": 1.919,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4502983226387482,
+      "grad_norm": 0.027426382526755333,
+      "learning_rate": 2.549701677361252e-07,
+      "loss": 0.1457,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4502983226387482,
+      "eval_loss": 0.813768208026886,
+      "eval_runtime": 128.7209,
+      "eval_samples_per_second": 15.336,
+      "eval_steps_per_second": 1.919,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5065856129685917,
+      "grad_norm": 0.0012935090344399214,
+      "learning_rate": 2.4934143870314085e-07,
+      "loss": 0.179,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5065856129685917,
+      "eval_loss": 0.7965527176856995,
+      "eval_runtime": 128.3969,
+      "eval_samples_per_second": 15.374,
+      "eval_steps_per_second": 1.924,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5628729032984352,
+      "grad_norm": 0.0017458726651966572,
+      "learning_rate": 2.4371270967015646e-07,
+      "loss": 0.1224,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5628729032984352,
+      "eval_loss": 0.7788484692573547,
+      "eval_runtime": 128.4939,
+      "eval_samples_per_second": 15.363,
+      "eval_steps_per_second": 1.922,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6191601936282788,
+      "grad_norm": 65.76844787597656,
+      "learning_rate": 2.380839806371721e-07,
+      "loss": 0.1551,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6191601936282788,
+      "eval_loss": 0.7626588344573975,
+      "eval_runtime": 128.7302,
+      "eval_samples_per_second": 15.334,
+      "eval_steps_per_second": 1.919,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6754474839581223,
+      "grad_norm": 5.014094829559326,
+      "learning_rate": 2.3245525160418776e-07,
+      "loss": 0.1721,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6754474839581223,
+      "eval_loss": 0.7464810609817505,
+      "eval_runtime": 128.749,
+      "eval_samples_per_second": 15.332,
+      "eval_steps_per_second": 1.918,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7317347742879657,
+      "grad_norm": 0.0022696161177009344,
+      "learning_rate": 2.268265225712034e-07,
+      "loss": 0.1532,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7317347742879657,
+      "eval_loss": 0.7335842251777649,
+      "eval_runtime": 128.8165,
+      "eval_samples_per_second": 15.324,
+      "eval_steps_per_second": 1.917,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7880220646178093,
+      "grad_norm": 5.508715730684344e-06,
+      "learning_rate": 2.2119779353821906e-07,
+      "loss": 0.1991,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7880220646178093,
+      "eval_loss": 0.7244272232055664,
+      "eval_runtime": 128.9211,
+      "eval_samples_per_second": 15.312,
+      "eval_steps_per_second": 1.916,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8443093549476528,
+      "grad_norm": 0.30320531129837036,
+      "learning_rate": 2.155690645052347e-07,
+      "loss": 0.1551,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8443093549476528,
+      "eval_loss": 0.718368411064148,
+      "eval_runtime": 128.8149,
+      "eval_samples_per_second": 15.324,
+      "eval_steps_per_second": 1.917,
+      "step": 7500
+    },
+    {
+      "epoch": 0.9005966452774964,
+      "grad_norm": 1.9788849385804497e-05,
+      "learning_rate": 2.0994033547225037e-07,
+      "loss": 0.1439,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9005966452774964,
+      "eval_loss": 0.7040167450904846,
+      "eval_runtime": 128.4921,
+      "eval_samples_per_second": 15.363,
+      "eval_steps_per_second": 1.922,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9568839356073399,
+      "grad_norm": 1.0750063665909693e-06,
+      "learning_rate": 2.04311606439266e-07,
+      "loss": 0.1361,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9568839356073399,
+      "eval_loss": 0.6983720660209656,
+      "eval_runtime": 128.5929,
+      "eval_samples_per_second": 15.351,
+      "eval_steps_per_second": 1.921,
+      "step": 8500
+    },
+    {
+      "epoch": 1.0131712259371835,
+      "grad_norm": 187.1415557861328,
+      "learning_rate": 1.9868287740628167e-07,
+      "loss": 0.1144,
+      "step": 9000
+    },
+    {
+      "epoch": 1.0131712259371835,
+      "eval_loss": 0.6901102066040039,
+      "eval_runtime": 128.5322,
+      "eval_samples_per_second": 15.358,
+      "eval_steps_per_second": 1.922,
+      "step": 9000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 26649,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3238364798866890.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-9000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:074eb2b3aae1f3310fb0b27d83d259baa7ad74a32c8433c81af819e28558f6b0
+size 4847