Training in progress, step 5000, checkpoint

checkpoint-5000/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-5000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6651e414cdf4f8da157a5653db8a5f35316a3de0d0c4769a201b6c9d3233370
+size 1711848436

checkpoint-5000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2be07ad9d1f47856adf648c539345fb8d07b0c3b17e6c622d2f6e1cef79fba16
+size 3424043887

checkpoint-5000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c75113044764a8910570c724906de201579e80685919d6e45c24460ff85166e4
+size 14503

checkpoint-5000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f04a200a6a0dc4d6f9fa1e8b0bebcbca3f5dd6af4d797c5ac279911c63e522b
+size 623

checkpoint-5000/trainer_state.json

@@ -0,0 +1,183 @@
+{
+  "best_metric": 1.3958221673965454,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l18-l/checkpoint-5000",
+  "epoch": 0.5332764505119454,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05332764505119454,
+      "grad_norm": 186.70851135253906,
+      "learning_rate": 9.866680887372013e-09,
+      "loss": 0.4804,
+      "step": 500
+    },
+    {
+      "epoch": 0.05332764505119454,
+      "eval_loss": 1.4967491626739502,
+      "eval_runtime": 60.9128,
+      "eval_samples_per_second": 16.204,
+      "eval_steps_per_second": 2.036,
+      "step": 500
+    },
+    {
+      "epoch": 0.10665529010238908,
+      "grad_norm": 86.50623321533203,
+      "learning_rate": 9.733361774744027e-09,
+      "loss": 0.4928,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10665529010238908,
+      "eval_loss": 1.482089877128601,
+      "eval_runtime": 62.7665,
+      "eval_samples_per_second": 15.725,
+      "eval_steps_per_second": 1.976,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1599829351535836,
+      "grad_norm": 603.98974609375,
+      "learning_rate": 9.600042662116041e-09,
+      "loss": 0.4325,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1599829351535836,
+      "eval_loss": 1.4686362743377686,
+      "eval_runtime": 61.6915,
+      "eval_samples_per_second": 15.999,
+      "eval_steps_per_second": 2.01,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21331058020477817,
+      "grad_norm": 13.387365341186523,
+      "learning_rate": 9.466723549488055e-09,
+      "loss": 0.3964,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21331058020477817,
+      "eval_loss": 1.4561632871627808,
+      "eval_runtime": 61.8331,
+      "eval_samples_per_second": 15.962,
+      "eval_steps_per_second": 2.005,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2666382252559727,
+      "grad_norm": 322.0577392578125,
+      "learning_rate": 9.333404436860068e-09,
+      "loss": 0.4294,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2666382252559727,
+      "eval_loss": 1.445016622543335,
+      "eval_runtime": 63.1527,
+      "eval_samples_per_second": 15.629,
+      "eval_steps_per_second": 1.963,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3199658703071672,
+      "grad_norm": 0.010774992406368256,
+      "learning_rate": 9.200085324232083e-09,
+      "loss": 0.3954,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3199658703071672,
+      "eval_loss": 1.4342260360717773,
+      "eval_runtime": 63.3562,
+      "eval_samples_per_second": 15.579,
+      "eval_steps_per_second": 1.957,
+      "step": 3000
+    },
+    {
+      "epoch": 0.37329351535836175,
+      "grad_norm": 42.48988723754883,
+      "learning_rate": 9.066766211604096e-09,
+      "loss": 0.4116,
+      "step": 3500
+    },
+    {
+      "epoch": 0.37329351535836175,
+      "eval_loss": 1.4254305362701416,
+      "eval_runtime": 63.1723,
+      "eval_samples_per_second": 15.624,
+      "eval_steps_per_second": 1.963,
+      "step": 3500
+    },
+    {
+      "epoch": 0.42662116040955633,
+      "grad_norm": 6.53631067276001,
+      "learning_rate": 8.933447098976109e-09,
+      "loss": 0.3366,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42662116040955633,
+      "eval_loss": 1.4154412746429443,
+      "eval_runtime": 63.1863,
+      "eval_samples_per_second": 15.62,
+      "eval_steps_per_second": 1.962,
+      "step": 4000
+    },
+    {
+      "epoch": 0.47994880546075086,
+      "grad_norm": 219.85145568847656,
+      "learning_rate": 8.800127986348124e-09,
+      "loss": 0.3941,
+      "step": 4500
+    },
+    {
+      "epoch": 0.47994880546075086,
+      "eval_loss": 1.406192421913147,
+      "eval_runtime": 63.3732,
+      "eval_samples_per_second": 15.574,
+      "eval_steps_per_second": 1.957,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5332764505119454,
+      "grad_norm": 0.5011767148971558,
+      "learning_rate": 8.666808873720137e-09,
+      "loss": 0.4351,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5332764505119454,
+      "eval_loss": 1.3958221673965454,
+      "eval_runtime": 61.9375,
+      "eval_samples_per_second": 15.935,
+      "eval_steps_per_second": 2.002,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 37504,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "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": 1799511273002520.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-5000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf62d80c5182e7768a9c6f2a2a23a186cdc24c25dbb6c6b5e997201f7ab22576
+size 4847