Training in progress, step 100

all_results.json

@@ -0,0 +1,20 @@
+{
+    "epoch": 20.0,
+    "eval_average_metrics": 89.30256169796215,
+    "eval_loss": 0.8872480988502502,
+    "eval_pearson": 89.49792307719518,
+    "eval_runtime": 2.7224,
+    "eval_samples_per_second": 275.492,
+    "eval_spearmanr": 89.10720031872911,
+    "test_average_metrics": 90.79137097057375,
+    "test_loss": 0.8804214000701904,
+    "test_pearson": 90.90568048271486,
+    "test_runtime": 2.8444,
+    "test_samples_per_second": 263.679,
+    "test_spearmanr": 90.67706145843265,
+    "train_loss": 0.8372542282630657,
+    "train_runtime": 586.0447,
+    "train_samples": 5749,
+    "train_samples_per_second": 196.197,
+    "train_steps_per_second": 1.979
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "eval_average_metrics": 89.30256169796215,
+    "eval_loss": 0.8872480988502502,
+    "eval_pearson": 89.49792307719518,
+    "eval_runtime": 2.7224,
+    "eval_samples_per_second": 275.492,
+    "eval_spearmanr": 89.10720031872911
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:57cd12837eb39b6fc4acbd3be4c26886d60426d3fa48b7270d3d26744e8d0407
-size 1084131
+oid sha256:0eca3ac1b8c9afb3b644aee2fb42980f649a93eff5c912ef9e18ff24e234fb59
+size 7551621

runs/Feb01_01-54-14_node2/1643651750.748784/events.out.tfevents.1643651750.node2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6754ad87577fb4e26d7b2954320a597e3aab3cde8a9c3dd1104a66253200dd8e
+size 5011

runs/Feb01_01-54-14_node2/events.out.tfevents.1643651750.node2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5bd80d3e949b4b9fab40d66191cda117fd9b53d4bf24023c1c65b5f838d3862
+size 4300

runs/Jan31_20-46-39_node1/events.out.tfevents.1643633306.node1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:52c3ba3443bdfc85fdcd834aafdb110aedde7fb7282436445478a07c4440db50
-size 6114
+oid sha256:3f344d6a7f8597fa84e48a683460e2ff75d7810d92110b331b7876a6bcb28a40
+size 6842

runs/Jan31_20-46-39_node1/events.out.tfevents.1643633894.node1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcb9494d3587cbde4e393254fcbfa34e00810a9e0b5510dce0d149602f2cdf7a
+size 788

test_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 20.0,
+    "test_average_metrics": 90.79137097057375,
+    "test_loss": 0.8804214000701904,
+    "test_pearson": 90.90568048271486,
+    "test_runtime": 2.8444,
+    "test_samples_per_second": 263.679,
+    "test_spearmanr": 90.67706145843265
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "train_loss": 0.8372542282630657,
+    "train_runtime": 586.0447,
+    "train_samples": 5749,
+    "train_samples_per_second": 196.197,
+    "train_steps_per_second": 1.979
+}

trainer_state.json

@@ -0,0 +1,87 @@
+{
+  "best_metric": 89.30256169796215,
+  "best_model_checkpoint": "outputs/bitfit/t5-base/stsb/checkpoint-600",
+  "epoch": 20.0,
+  "global_step": 1160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 3.45,
+      "eval_average_metrics": 88.9566249550781,
+      "eval_loss": 0.8938751816749573,
+      "eval_pearson": 88.95684320357462,
+      "eval_runtime": 2.4661,
+      "eval_samples_per_second": 304.124,
+      "eval_spearmanr": 88.95640670658156,
+      "step": 200
+    },
+    {
+      "epoch": 6.9,
+      "eval_average_metrics": 88.95231484069625,
+      "eval_loss": 0.8857536911964417,
+      "eval_pearson": 89.08767005696347,
+      "eval_runtime": 2.4501,
+      "eval_samples_per_second": 306.105,
+      "eval_spearmanr": 88.81695962442902,
+      "step": 400
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 0.0003,
+      "loss": 0.8558,
+      "step": 500
+    },
+    {
+      "epoch": 10.34,
+      "eval_average_metrics": 89.30256169796215,
+      "eval_loss": 0.8872480988502502,
+      "eval_pearson": 89.49792307719518,
+      "eval_runtime": 2.7534,
+      "eval_samples_per_second": 272.387,
+      "eval_spearmanr": 89.10720031872911,
+      "step": 600
+    },
+    {
+      "epoch": 13.79,
+      "eval_average_metrics": 89.00274406096713,
+      "eval_loss": 0.8818585872650146,
+      "eval_pearson": 89.11339528387046,
+      "eval_runtime": 2.7298,
+      "eval_samples_per_second": 274.743,
+      "eval_spearmanr": 88.8920928380638,
+      "step": 800
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 0.8251,
+      "step": 1000
+    },
+    {
+      "epoch": 17.24,
+      "eval_average_metrics": 89.29097960305327,
+      "eval_loss": 0.8865894079208374,
+      "eval_pearson": 89.46184016190372,
+      "eval_runtime": 2.3462,
+      "eval_samples_per_second": 319.662,
+      "eval_spearmanr": 89.12011904420282,
+      "step": 1000
+    },
+    {
+      "epoch": 20.0,
+      "step": 1160,
+      "total_flos": 1.3780641944564136e+16,
+      "train_loss": 0.8372542282630657,
+      "train_runtime": 586.0447,
+      "train_samples_per_second": 196.197,
+      "train_steps_per_second": 1.979
+    }
+  ],
+  "max_steps": 1160,
+  "num_train_epochs": 20,
+  "total_flos": 1.3780641944564136e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d6d3c60071912ff06d193cf80a0e4b597f88ab9f45a9d5e53699f1d6e7c1ecd2
+oid sha256:4532ae8df18f200354792e9d75508579d85eda3d88a52d81a7dd97a8a6facb9c
 size 3183

training_config.json

@@ -1 +1 @@
-{"dataset_config_name": ["en"], "delta_type": "bitfit", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "stsb", "eval_steps": 200, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "max_source_length": 128, "metric_for_best_model": "average_metrics", "model_name_or_path": "../../../../plm_cache/t5-base", "num_train_epochs": 20, "output_dir": "outputs/bitfit/t5-base/stsb", "overwrite_output_dir": true, "per_device_eval_batch_size": 100, "per_device_train_batch_size": 100, "predict_with_generate": true, "push_to_hub": true, "save_steps": 200, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "split_validation_test": true, "task_name": "stsb", "test_dataset_config_name": ["en"], "test_dataset_name": "stsb", "tokenizer_name": "../../../../plm_cache/t5-base", "warmup_steps": 500}
+{"bottleneck_dim": 24, "dataset_config_name": ["en"], "delta_type": "adapter", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "stsb", "eval_steps": 100, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "max_source_length": 128, "metric_for_best_model": "average_metrics", "model_name_or_path": "../../../../plm_cache/t5-base", "num_train_epochs": 20, "output_dir": "outputs/bitfit/t5-base/stsb", "overwrite_output_dir": true, "per_device_eval_batch_size": 32, "per_device_train_batch_size": 32, "predict_with_generate": true, "push_to_hub": true, "save_steps": 100, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "split_validation_test": true, "task_name": "stsb", "test_dataset_config_name": ["en"], "test_dataset_name": "stsb", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "warmup_steps": 0}