Training in progress, step 8500, checkpoint

checkpoint-8500/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-8500/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1f0568f00824ff249c59f2cd181602247653153a95085f77798a3b975bc23c0
+size 1711848436

checkpoint-8500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a842e73a8d58d76cd50f68284525c2c8eb0851a972d8cd307cd1cfffc8fc4df
+size 3424043887

checkpoint-8500/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26c99df967694e66808fb8d4e4814367449c627fb6d4991b78dafe1ced9ade3e
+size 14503

checkpoint-8500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80504ed63f7820876db765aa8e0a9197638387e505e28d185e8975f843408573
+size 623

checkpoint-8500/trainer_state.json

@@ -0,0 +1,288 @@
+{
+  "best_metric": 0.980699360370636,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l58-l/checkpoint-8500",
+  "epoch": 1.565954310980103,
+  "eval_steps": 500,
+  "global_step": 8500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.09211495946941783,
+      "grad_norm": 19.808820724487305,
+      "learning_rate": 4.84647506755097e-08,
+      "loss": 0.3758,
+      "step": 500
+    },
+    {
+      "epoch": 0.09211495946941783,
+      "eval_loss": 1.418845295906067,
+      "eval_runtime": 75.2377,
+      "eval_samples_per_second": 16.042,
+      "eval_steps_per_second": 2.007,
+      "step": 500
+    },
+    {
+      "epoch": 0.18422991893883567,
+      "grad_norm": 76.37991333007812,
+      "learning_rate": 4.69295013510194e-08,
+      "loss": 0.4108,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18422991893883567,
+      "eval_loss": 1.3513683080673218,
+      "eval_runtime": 74.9686,
+      "eval_samples_per_second": 16.1,
+      "eval_steps_per_second": 2.014,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2763448784082535,
+      "grad_norm": 0.5795987844467163,
+      "learning_rate": 4.539425202652911e-08,
+      "loss": 0.4335,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2763448784082535,
+      "eval_loss": 1.2895005941390991,
+      "eval_runtime": 76.713,
+      "eval_samples_per_second": 15.734,
+      "eval_steps_per_second": 1.968,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36845983787767134,
+      "grad_norm": 0.1463615745306015,
+      "learning_rate": 4.385900270203881e-08,
+      "loss": 0.3436,
+      "step": 2000
+    },
+    {
+      "epoch": 0.36845983787767134,
+      "eval_loss": 1.2414860725402832,
+      "eval_runtime": 76.9749,
+      "eval_samples_per_second": 15.68,
+      "eval_steps_per_second": 1.962,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46057479734708917,
+      "grad_norm": 323.8576354980469,
+      "learning_rate": 4.232375337754851e-08,
+      "loss": 0.366,
+      "step": 2500
+    },
+    {
+      "epoch": 0.46057479734708917,
+      "eval_loss": 1.1949334144592285,
+      "eval_runtime": 77.0381,
+      "eval_samples_per_second": 15.668,
+      "eval_steps_per_second": 1.96,
+      "step": 2500
+    },
+    {
+      "epoch": 0.552689756816507,
+      "grad_norm": 339.28533935546875,
+      "learning_rate": 4.078850405305821e-08,
+      "loss": 0.3191,
+      "step": 3000
+    },
+    {
+      "epoch": 0.552689756816507,
+      "eval_loss": 1.1560280323028564,
+      "eval_runtime": 77.0026,
+      "eval_samples_per_second": 15.675,
+      "eval_steps_per_second": 1.961,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6448047162859248,
+      "grad_norm": 60.27859878540039,
+      "learning_rate": 3.925325472856792e-08,
+      "loss": 0.2779,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6448047162859248,
+      "eval_loss": 1.1233855485916138,
+      "eval_runtime": 76.9006,
+      "eval_samples_per_second": 15.696,
+      "eval_steps_per_second": 1.964,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7369196757553427,
+      "grad_norm": 18.268455505371094,
+      "learning_rate": 3.7718005404077616e-08,
+      "loss": 0.2942,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7369196757553427,
+      "eval_loss": 1.0993207693099976,
+      "eval_runtime": 77.2834,
+      "eval_samples_per_second": 15.618,
+      "eval_steps_per_second": 1.954,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8290346352247605,
+      "grad_norm": 505.22723388671875,
+      "learning_rate": 3.6182756079587324e-08,
+      "loss": 0.2615,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8290346352247605,
+      "eval_loss": 1.080470085144043,
+      "eval_runtime": 77.1214,
+      "eval_samples_per_second": 15.651,
+      "eval_steps_per_second": 1.958,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9211495946941783,
+      "grad_norm": 1.2971173524856567,
+      "learning_rate": 3.4647506755097025e-08,
+      "loss": 0.2715,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9211495946941783,
+      "eval_loss": 1.0614728927612305,
+      "eval_runtime": 77.2982,
+      "eval_samples_per_second": 15.615,
+      "eval_steps_per_second": 1.953,
+      "step": 5000
+    },
+    {
+      "epoch": 1.013264554163596,
+      "grad_norm": 0.17611829936504364,
+      "learning_rate": 3.3112257430606727e-08,
+      "loss": 0.2509,
+      "step": 5500
+    },
+    {
+      "epoch": 1.013264554163596,
+      "eval_loss": 1.0461839437484741,
+      "eval_runtime": 75.3494,
+      "eval_samples_per_second": 16.019,
+      "eval_steps_per_second": 2.004,
+      "step": 5500
+    },
+    {
+      "epoch": 1.105379513633014,
+      "grad_norm": 228.2562713623047,
+      "learning_rate": 3.1577008106116435e-08,
+      "loss": 0.2653,
+      "step": 6000
+    },
+    {
+      "epoch": 1.105379513633014,
+      "eval_loss": 1.0305675268173218,
+      "eval_runtime": 77.0542,
+      "eval_samples_per_second": 15.664,
+      "eval_steps_per_second": 1.96,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1974944731024317,
+      "grad_norm": 432.4784240722656,
+      "learning_rate": 3.0041758781626136e-08,
+      "loss": 0.2199,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1974944731024317,
+      "eval_loss": 1.0182703733444214,
+      "eval_runtime": 76.296,
+      "eval_samples_per_second": 15.82,
+      "eval_steps_per_second": 1.979,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2896094325718497,
+      "grad_norm": 0.6385967135429382,
+      "learning_rate": 2.8506509457135834e-08,
+      "loss": 0.1844,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2896094325718497,
+      "eval_loss": 1.0078601837158203,
+      "eval_runtime": 76.5441,
+      "eval_samples_per_second": 15.769,
+      "eval_steps_per_second": 1.973,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3817243920412676,
+      "grad_norm": 57.47743606567383,
+      "learning_rate": 2.697126013264554e-08,
+      "loss": 0.2143,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3817243920412676,
+      "eval_loss": 0.9969784617424011,
+      "eval_runtime": 75.6186,
+      "eval_samples_per_second": 15.962,
+      "eval_steps_per_second": 1.997,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4738393515106853,
+      "grad_norm": 0.11704438179731369,
+      "learning_rate": 2.5436010808155244e-08,
+      "loss": 0.2011,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4738393515106853,
+      "eval_loss": 0.9886976480484009,
+      "eval_runtime": 75.238,
+      "eval_samples_per_second": 16.042,
+      "eval_steps_per_second": 2.007,
+      "step": 8000
+    },
+    {
+      "epoch": 1.565954310980103,
+      "grad_norm": 39.71356964111328,
+      "learning_rate": 2.3900761483664948e-08,
+      "loss": 0.1995,
+      "step": 8500
+    },
+    {
+      "epoch": 1.565954310980103,
+      "eval_loss": 0.980699360370636,
+      "eval_runtime": 75.7533,
+      "eval_samples_per_second": 15.933,
+      "eval_steps_per_second": 1.993,
+      "step": 8500
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 16284,
+  "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": 3058665502412520.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-8500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbcece3d46fd39b0ba80eb3e495fdf46c9f891af5b4bddce9de77ad5a467b625
+size 4847