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:e9792e813c62ecb94df1dc098fa74bd42d77ad7259d6268bd3506050baf27a5b
+size 1711848436

checkpoint-8500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18c063aac0e84ccd980d6c703e5658969e6551dfa516f6aa3504cfd59387a07b
+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 sha256:6d0daf5959a332ed79de6edb9cd8968043cb16cc97faaaaeb5a80dbdb2b397a0
+size 623

checkpoint-8500/trainer_state.json

@@ -0,0 +1,288 @@
+{
+  "best_metric": 0.943097710609436,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l56-l/checkpoint-6000",
+  "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": 18.028329849243164,
+      "learning_rate": 4.846475067550971e-06,
+      "loss": 0.2253,
+      "step": 500
+    },
+    {
+      "epoch": 0.09211495946941783,
+      "eval_loss": 1.4693936109542847,
+      "eval_runtime": 74.2802,
+      "eval_samples_per_second": 16.249,
+      "eval_steps_per_second": 2.033,
+      "step": 500
+    },
+    {
+      "epoch": 0.18422991893883567,
+      "grad_norm": 0.4609026312828064,
+      "learning_rate": 4.692950135101941e-06,
+      "loss": 0.2544,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18422991893883567,
+      "eval_loss": 1.5420819520950317,
+      "eval_runtime": 74.0766,
+      "eval_samples_per_second": 16.294,
+      "eval_steps_per_second": 2.038,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2763448784082535,
+      "grad_norm": 71.96737670898438,
+      "learning_rate": 4.5394252026529115e-06,
+      "loss": 0.3236,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2763448784082535,
+      "eval_loss": 1.4642689228057861,
+      "eval_runtime": 74.2336,
+      "eval_samples_per_second": 16.259,
+      "eval_steps_per_second": 2.034,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36845983787767134,
+      "grad_norm": 6.366127069945549e-10,
+      "learning_rate": 4.385900270203882e-06,
+      "loss": 0.1888,
+      "step": 2000
+    },
+    {
+      "epoch": 0.36845983787767134,
+      "eval_loss": 1.3193777799606323,
+      "eval_runtime": 76.1214,
+      "eval_samples_per_second": 15.856,
+      "eval_steps_per_second": 1.984,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46057479734708917,
+      "grad_norm": 452.8532409667969,
+      "learning_rate": 4.232375337754851e-06,
+      "loss": 0.2563,
+      "step": 2500
+    },
+    {
+      "epoch": 0.46057479734708917,
+      "eval_loss": 1.37644362449646,
+      "eval_runtime": 76.3265,
+      "eval_samples_per_second": 15.814,
+      "eval_steps_per_second": 1.978,
+      "step": 2500
+    },
+    {
+      "epoch": 0.552689756816507,
+      "grad_norm": 0.0002858435327652842,
+      "learning_rate": 4.078850405305822e-06,
+      "loss": 0.2794,
+      "step": 3000
+    },
+    {
+      "epoch": 0.552689756816507,
+      "eval_loss": 1.3007307052612305,
+      "eval_runtime": 76.3512,
+      "eval_samples_per_second": 15.809,
+      "eval_steps_per_second": 1.978,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6448047162859248,
+      "grad_norm": 0.324462890625,
+      "learning_rate": 3.925325472856792e-06,
+      "loss": 0.1749,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6448047162859248,
+      "eval_loss": 1.3360542058944702,
+      "eval_runtime": 74.5379,
+      "eval_samples_per_second": 16.193,
+      "eval_steps_per_second": 2.026,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7369196757553427,
+      "grad_norm": 76.63452911376953,
+      "learning_rate": 3.7718005404077624e-06,
+      "loss": 0.2672,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7369196757553427,
+      "eval_loss": 1.268362283706665,
+      "eval_runtime": 76.6253,
+      "eval_samples_per_second": 15.752,
+      "eval_steps_per_second": 1.971,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8290346352247605,
+      "grad_norm": 1.9390699863433838,
+      "learning_rate": 3.6182756079587327e-06,
+      "loss": 0.218,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8290346352247605,
+      "eval_loss": 1.1064503192901611,
+      "eval_runtime": 75.0503,
+      "eval_samples_per_second": 16.083,
+      "eval_steps_per_second": 2.012,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9211495946941783,
+      "grad_norm": 3.37814665130054e-10,
+      "learning_rate": 3.464750675509703e-06,
+      "loss": 0.1665,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9211495946941783,
+      "eval_loss": 1.0620027780532837,
+      "eval_runtime": 76.9535,
+      "eval_samples_per_second": 15.685,
+      "eval_steps_per_second": 1.962,
+      "step": 5000
+    },
+    {
+      "epoch": 1.013264554163596,
+      "grad_norm": 8.782347293845305e-09,
+      "learning_rate": 3.311225743060673e-06,
+      "loss": 0.1842,
+      "step": 5500
+    },
+    {
+      "epoch": 1.013264554163596,
+      "eval_loss": 0.9443314075469971,
+      "eval_runtime": 75.2095,
+      "eval_samples_per_second": 16.049,
+      "eval_steps_per_second": 2.008,
+      "step": 5500
+    },
+    {
+      "epoch": 1.105379513633014,
+      "grad_norm": 0.002741970354691148,
+      "learning_rate": 3.1577008106116434e-06,
+      "loss": 0.1183,
+      "step": 6000
+    },
+    {
+      "epoch": 1.105379513633014,
+      "eval_loss": 0.943097710609436,
+      "eval_runtime": 74.832,
+      "eval_samples_per_second": 16.129,
+      "eval_steps_per_second": 2.018,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1974944731024317,
+      "grad_norm": 9.241516818292439e-05,
+      "learning_rate": 3.004175878162614e-06,
+      "loss": 0.1066,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1974944731024317,
+      "eval_loss": 1.005082607269287,
+      "eval_runtime": 74.7497,
+      "eval_samples_per_second": 16.147,
+      "eval_steps_per_second": 2.02,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2896094325718497,
+      "grad_norm": 3.972076889624532e-09,
+      "learning_rate": 2.850650945713584e-06,
+      "loss": 0.0901,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2896094325718497,
+      "eval_loss": 1.001724362373352,
+      "eval_runtime": 74.9393,
+      "eval_samples_per_second": 16.106,
+      "eval_steps_per_second": 2.015,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3817243920412676,
+      "grad_norm": 0.6726647615432739,
+      "learning_rate": 2.6971260132645544e-06,
+      "loss": 0.112,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3817243920412676,
+      "eval_loss": 1.0029873847961426,
+      "eval_runtime": 76.2263,
+      "eval_samples_per_second": 15.834,
+      "eval_steps_per_second": 1.981,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4738393515106853,
+      "grad_norm": 0.0007436455925926566,
+      "learning_rate": 2.543601080815525e-06,
+      "loss": 0.1265,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4738393515106853,
+      "eval_loss": 0.9462646842002869,
+      "eval_runtime": 74.6539,
+      "eval_samples_per_second": 16.168,
+      "eval_steps_per_second": 2.023,
+      "step": 8000
+    },
+    {
+      "epoch": 1.565954310980103,
+      "grad_norm": 3.9008838825793646e-08,
+      "learning_rate": 2.390076148366495e-06,
+      "loss": 0.1193,
+      "step": 8500
+    },
+    {
+      "epoch": 1.565954310980103,
+      "eval_loss": 1.0377594232559204,
+      "eval_runtime": 74.635,
+      "eval_samples_per_second": 16.172,
+      "eval_steps_per_second": 2.023,
+      "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 sha256:46089b82f4ed355b4512be05866138fa1fe20f53cc2bb3deace890abfba0e5f2
+size 4847