Training in progress, step 100

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b7ab0e156d45c29135aef9f1e302ef29b711547cb1532cf35a03b2a5d46f4a58
-size 2631685
+oid sha256:a7e3a234abfaf95ec3e328fc4b32d4255980707d3a6be2169e74ad421aeb3788
+size 879301

runs/Feb02_23-11-48_node4/1643814814.5124707/events.out.tfevents.1643814814.node4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:214322adc4328b7f68666c727129eca77be34a0cbdef82b47ca6e8f9a5b2acbe
+size 5005

runs/Feb02_23-11-48_node4/events.out.tfevents.1643814814.node4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f833fee575e3c798efc382f815d82193b08810f226eb42e8da32ad1cffc94a37
+size 4294

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9fc389561a9087ed9ef38e78165fe8b7afe4b52382655fc678f176c464c9a458
+oid sha256:7968245360993fd5e701b5950bb95dda379b0ddaed0ae03b70cdccde5fa80c48
 size 3183

training_config.json

@@ -1 +1 @@
-{"dataset_config_name": ["en"], "delta_type": "lora", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "cola", "eval_steps": 100, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "lora_r": 8, "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/cola", "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": "cola", "test_dataset_config_name": ["en"], "test_dataset_name": "cola", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "warmup_steps": 0}
+{"dataset_config_name": ["en"], "delta_type": "compacter", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "cola", "eval_steps": 100, "evaluation_strategy": "steps", "factorized_phm": true, "factorized_phm_rule": false, "gradient_clip": false, "greater_is_better": true, "hypercomplex_adapters": true, "hypercomplex_division": 4, "hypercomplex_nonlinearity": "glorot-uniform", "learn_phm": 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", "non_linearity": "gelu_new", "normalize_phm_weight": false, "num_train_epochs": 20, "output_dir": "outputs/lora/t5-base/cola", "overwrite_output_dir": true, "per_device_eval_batch_size": 32, "per_device_train_batch_size": 32, "phm_c_init": "normal", "phm_clamp": false, "phm_init_range": 0.0001, "predict_with_generate": true, "push_to_hub": true, "save_steps": 100, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "shared_phm_rule": false, "split_validation_test": true, "task_name": "cola", "test_dataset_config_name": ["en"], "test_dataset_name": "cola", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "use_bias_down_sampler": true, "use_bias_up_sampler": true, "warmup_steps": 0}