Upload folder using huggingface_hub

.gitattributes

@@ -35,3 +35,10 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 Finetune-MULTI/wandb/run-20260324_040756-vgxlb14a/run-vgxlb14a.wandb filter=lfs diff=lfs merge=lfs -text
 Finetune-MULTI/wandb/run-20260324_041355-7nzklvmz/run-7nzklvmz.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_043205-qaue51mb/run-qaue51mb.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_043608-md2c6b0g/run-md2c6b0g.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_044256-f0nbbv1z/run-f0nbbv1z.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_045330-ljb653t8/run-ljb653t8.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_133444-pfwf0zeu/run-pfwf0zeu.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_134050-qxvbi54v/run-qxvbi54v.wandb filter=lfs diff=lfs merge=lfs -text
+Finetune-GenomicBenchmarks/wandb/run-20260324_153401-oww9zr78/run-oww9zr78.wandb filter=lfs diff=lfs merge=lfs -text

Finetune-GenomicBenchmarks/.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

Finetune-GenomicBenchmarks/data_download.py

@@ -0,0 +1,23 @@
+from datasets import load_dataset
+from pathlib import Path
+
+datasets = [
+    "demo_human_or_worm",
+    "dummy_mouse_enhancers_ensembl",
+    "human_enhancers_ensembl",
+    "human_nontata_promoters",
+    "demo_coding_vs_intergenomic_seqs",
+    "drosophila_enhancers_stark",
+    "human_enhancers_cohn",
+    "human_ensembl_regulatory",
+    "human_ocr_ensembl",
+]
+
+out_root = Path("hf_raw")
+out_root.mkdir(parents=True, exist_ok=True)
+
+for name in datasets:
+    hf_id = f"katarinagresova/Genomic_Benchmarks_{name}"
+    ds = load_dataset(hf_id)  # downloads to the HF cache
+    ds.save_to_disk(out_root / name)  # optional: persist locally for reuse
+    print(f"downloaded {hf_id}")

Finetune-GenomicBenchmarks/data_split.py

@@ -0,0 +1,28 @@
+from datasets import load_from_disk, concatenate_datasets
+from pathlib import Path
+import pandas as pd
+
+src_root = Path("hf_raw")
+dst_root = Path("ft_data")
+seed = 42
+
+for ds_dir in src_root.iterdir():
+    if not ds_dir.is_dir():
+        continue
+    ds = load_from_disk(ds_dir)
+
+    # Combine all available splits, shuffle, then 80/10/10
+    full = concatenate_datasets([ds[s] for s in ds.keys()])
+    full = full.shuffle(seed=seed)
+    split1 = full.train_test_split(test_size=0.2, seed=seed)
+    train = split1["train"]
+    split2 = split1["test"].train_test_split(test_size=0.5, seed=seed)
+    dev, test = split2["train"], split2["test"]
+
+    out = dst_root / ds_dir.name / "split"
+    out.mkdir(parents=True, exist_ok=True)
+    for name, subset in [("train", train), ("dev", dev), ("test", test)]:
+        subset.to_pandas()[["seq", "label"]].rename(
+            columns={"seq": "sequence", "label": "labels"}
+        ).to_csv(out / f"{name}.csv", sep="\t", index=False)
+        print(f"Wrote {out/f'{name}.csv'}")

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "/data/nanhuang/Nan/models/DNAbert2_Pretrained",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.35.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 4096
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:021ad14b2f38dbff7f8eeea18f7132f657a60426d38ab02c696c2744fa19ccf6
+size 356777880

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96fa9e420b35ab5abb870bfc4fbcf26b1e0681ccb0c1bd1cdd4dfdc78bb41f50
+size 713677451

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52f47a8b0e975cf59e02cf3ebc58d149a406e70a9b82323945324cc788745b08
+size 14645

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48c44eacc41015ff6a31c3be859e9d08482db72ae3540b67fdb6bec6cf82b0a8
+size 1465

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "cache_dir": null,
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "model_max_length": 100,
+  "pad_token": "[PAD]",
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "trust_remote_code": true,
+  "unk_token": "[UNK]",
+  "use_fast": true
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/trainer_state.json

@@ -0,0 +1,166 @@
+{
+  "best_metric": 0.9063933951179595,
+  "best_model_checkpoint": "genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875",
+  "epoch": 3.0,
+  "eval_steps": 100,
+  "global_step": 1875,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.6548672566371683e-05,
+      "loss": 0.4279,
+      "step": 100
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.851872871736663e-05,
+      "loss": 0.2941,
+      "step": 200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.68161180476731e-05,
+      "loss": 0.2896,
+      "step": 300
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.511350737797957e-05,
+      "loss": 0.2724,
+      "step": 400
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.341089670828604e-05,
+      "loss": 0.2618,
+      "step": 500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.170828603859251e-05,
+      "loss": 0.2538,
+      "step": 600
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8981,
+      "eval_f1": 0.8980013458817503,
+      "eval_loss": 0.24450437724590302,
+      "eval_matthews_correlation": 0.7970483436740018,
+      "eval_precision": 0.8991229435739876,
+      "eval_recall": 0.897926298388621,
+      "eval_runtime": 2.2999,
+      "eval_samples_per_second": 4347.921,
+      "eval_steps_per_second": 34.349,
+      "step": 625
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0005675368898978e-05,
+      "loss": 0.2381,
+      "step": 700
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.830306469920545e-05,
+      "loss": 0.2257,
+      "step": 800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6600454029511918e-05,
+      "loss": 0.2222,
+      "step": 900
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.4897843359818387e-05,
+      "loss": 0.2267,
+      "step": 1000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.3195232690124857e-05,
+      "loss": 0.2286,
+      "step": 1100
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.1492622020431328e-05,
+      "loss": 0.2205,
+      "step": 1200
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9036,
+      "eval_f1": 0.9035999035999036,
+      "eval_loss": 0.23238573968410492,
+      "eval_matthews_correlation": 0.8073154756092576,
+      "eval_precision": 0.9036641967999408,
+      "eval_recall": 0.9036512789126667,
+      "eval_runtime": 2.312,
+      "eval_samples_per_second": 4325.297,
+      "eval_steps_per_second": 34.17,
+      "step": 1250
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.790011350737798e-06,
+      "loss": 0.1951,
+      "step": 1300
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.087400681044268e-06,
+      "loss": 0.1857,
+      "step": 1400
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 6.384790011350738e-06,
+      "loss": 0.1822,
+      "step": 1500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 4.682179341657208e-06,
+      "loss": 0.175,
+      "step": 1600
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.9795686719636776e-06,
+      "loss": 0.1706,
+      "step": 1700
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.2769580022701475e-06,
+      "loss": 0.1733,
+      "step": 1800
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9064,
+      "eval_f1": 0.9063933951179595,
+      "eval_loss": 0.24301454424858093,
+      "eval_matthews_correlation": 0.8132168703204711,
+      "eval_precision": 0.906713678896274,
+      "eval_recall": 0.9065032186577142,
+      "eval_runtime": 2.322,
+      "eval_samples_per_second": 4306.719,
+      "eval_steps_per_second": 34.023,
+      "step": 1875
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 1875,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 6413331974400000.0,
+  "trial_name": null,
+  "trial_params": null
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/checkpoint-1875/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef267b331203305d86d178a26fc0947311a7d6fc6d91ae987e0fcc2751ebc0c3
+size 5457

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_coding_vs_intergenomic_seqs/DNAbert2_Pretrained/lr3e-5_wd0.0_wr0.06_ep3_seed42/results/base5120_demo_coding_vs_intergenomic_seqs_lr3e-5_wd0.0_wr0.06_ep3_seed42/eval_results.json

@@ -0,0 +1 @@
+{"eval_loss": 0.2542661130428314, "eval_accuracy": 0.9021, "eval_f1": 0.9020706907370306, "eval_matthews_correlation": 0.8049607381043299, "eval_precision": 0.9027773220988344, "eval_recall": 0.902183634938159, "eval_runtime": 2.1971, "eval_samples_per_second": 4551.358, "eval_steps_per_second": 35.956, "epoch": 3.0}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "/data/nanhuang/Nan/models/DNAbert2_Pretrained",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.35.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 4096
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92e25307685c51ad096f3aa92059c090823fb1cb5db99bdba9753c7a7e34c95f
+size 356777880

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c4705316037c1c05b2aabfbb133c8e6813d57377ad6979be8e9cc442d7e22a4
+size 713677451

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cad7b27eb5af31f3653dea61322524cbd98718a2c3f49a7f9e7fa7ba7942638c
+size 14645

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0820d74bc9c19dfa93c8aa4a6b85582725f6342ebcd7b58d2a0c7ac71726eee6
+size 1465

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "cache_dir": null,
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "trust_remote_code": true,
+  "unk_token": "[UNK]",
+  "use_fast": true
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/trainer_state.json

@@ -0,0 +1,368 @@
+{
+  "best_metric": 0.9556951153864713,
+  "best_model_checkpoint": "genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375",
+  "epoch": 7.0,
+  "eval_steps": 100,
+  "global_step": 4375,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 0.4352,
+      "step": 100
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.1944,
+      "step": 200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3e-05,
+      "loss": 0.1817,
+      "step": 300
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9361702127659574e-05,
+      "loss": 0.1653,
+      "step": 400
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.872340425531915e-05,
+      "loss": 0.15,
+      "step": 500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.8085106382978723e-05,
+      "loss": 0.1423,
+      "step": 600
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9492,
+      "eval_f1": 0.9491973663914737,
+      "eval_loss": 0.14096176624298096,
+      "eval_matthews_correlation": 0.8987776409510696,
+      "eval_precision": 0.9494807442016145,
+      "eval_recall": 0.949296915548862,
+      "eval_runtime": 9.3786,
+      "eval_samples_per_second": 1066.258,
+      "eval_steps_per_second": 8.423,
+      "step": 625
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.74468085106383e-05,
+      "loss": 0.1286,
+      "step": 700
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.6808510638297873e-05,
+      "loss": 0.1227,
+      "step": 800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.617021276595745e-05,
+      "loss": 0.1128,
+      "step": 900
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 0.1133,
+      "step": 1000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.4893617021276595e-05,
+      "loss": 0.1108,
+      "step": 1100
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.4255319148936168e-05,
+      "loss": 0.1174,
+      "step": 1200
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9448,
+      "eval_f1": 0.9447548116457338,
+      "eval_loss": 0.14714133739471436,
+      "eval_matthews_correlation": 0.8919111732730618,
+      "eval_precision": 0.9468665229355923,
+      "eval_recall": 0.9450465072788566,
+      "eval_runtime": 9.4606,
+      "eval_samples_per_second": 1057.01,
+      "eval_steps_per_second": 8.35,
+      "step": 1250
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.3617021276595744e-05,
+      "loss": 0.0896,
+      "step": 1300
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.297872340425532e-05,
+      "loss": 0.0771,
+      "step": 1400
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.2340425531914894e-05,
+      "loss": 0.0849,
+      "step": 1500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.170212765957447e-05,
+      "loss": 0.0786,
+      "step": 1600
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1063829787234043e-05,
+      "loss": 0.0872,
+      "step": 1700
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0425531914893616e-05,
+      "loss": 0.0827,
+      "step": 1800
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9533,
+      "eval_f1": 0.9532990543058497,
+      "eval_loss": 0.12726238369941711,
+      "eval_matthews_correlation": 0.9066100253951647,
+      "eval_precision": 0.9532948131792527,
+      "eval_recall": 0.9533152124454098,
+      "eval_runtime": 9.3211,
+      "eval_samples_per_second": 1072.834,
+      "eval_steps_per_second": 8.475,
+      "step": 1875
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.978723404255319e-05,
+      "loss": 0.0769,
+      "step": 1900
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 0.0553,
+      "step": 2000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.8510638297872342e-05,
+      "loss": 0.0561,
+      "step": 2100
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 0.0551,
+      "step": 2200
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.723404255319149e-05,
+      "loss": 0.0555,
+      "step": 2300
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 0.0574,
+      "step": 2400
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.5957446808510637e-05,
+      "loss": 0.0562,
+      "step": 2500
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9532,
+      "eval_f1": 0.9531823797204316,
+      "eval_loss": 0.16251493990421295,
+      "eval_matthews_correlation": 0.9066581426052277,
+      "eval_precision": 0.9535473452173776,
+      "eval_recall": 0.9531109024342193,
+      "eval_runtime": 9.3211,
+      "eval_samples_per_second": 1072.831,
+      "eval_steps_per_second": 8.475,
+      "step": 2500
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.531914893617021e-05,
+      "loss": 0.0343,
+      "step": 2600
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.4680851063829787e-05,
+      "loss": 0.0289,
+      "step": 2700
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.4042553191489362e-05,
+      "loss": 0.0365,
+      "step": 2800
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.3404255319148936e-05,
+      "loss": 0.0336,
+      "step": 2900
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 0.0362,
+      "step": 3000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.2127659574468084e-05,
+      "loss": 0.0355,
+      "step": 3100
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9553,
+      "eval_f1": 0.9552997206232539,
+      "eval_loss": 0.17249269783496857,
+      "eval_matthews_correlation": 0.9106381321410131,
+      "eval_precision": 0.9553088212352849,
+      "eval_recall": 0.9553293111362456,
+      "eval_runtime": 9.36,
+      "eval_samples_per_second": 1068.377,
+      "eval_steps_per_second": 8.44,
+      "step": 3125
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.148936170212766e-05,
+      "loss": 0.0224,
+      "step": 3200
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1.0851063829787235e-05,
+      "loss": 0.0207,
+      "step": 3300
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 0.0194,
+      "step": 3400
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 0.0188,
+      "step": 3500
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.0197,
+      "step": 3600
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 0.0183,
+      "step": 3700
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9552,
+      "eval_f1": 0.9551867424458852,
+      "eval_loss": 0.24090713262557983,
+      "eval_matthews_correlation": 0.91057237575741,
+      "eval_precision": 0.9554460306619799,
+      "eval_recall": 0.9551264011936584,
+      "eval_runtime": 9.3493,
+      "eval_samples_per_second": 1069.598,
+      "eval_steps_per_second": 8.45,
+      "step": 3750
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 7.659574468085105e-06,
+      "loss": 0.0129,
+      "step": 3800
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 7.021276595744681e-06,
+      "loss": 0.0136,
+      "step": 3900
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 0.0076,
+      "step": 4000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 5.74468085106383e-06,
+      "loss": 0.0079,
+      "step": 4100
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.0105,
+      "step": 4200
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.0083,
+      "step": 4300
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9557,
+      "eval_f1": 0.9556951153864713,
+      "eval_loss": 0.2684195339679718,
+      "eval_matthews_correlation": 0.9114136506343465,
+      "eval_precision": 0.9557403251037204,
+      "eval_recall": 0.9556733279930717,
+      "eval_runtime": 9.3532,
+      "eval_samples_per_second": 1069.157,
+      "eval_steps_per_second": 8.446,
+      "step": 4375
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 5000,
+  "num_train_epochs": 8,
+  "save_steps": 100,
+  "total_flos": 5.81310987936e+16,
+  "trial_name": null,
+  "trial_params": null
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/checkpoint-4375/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f20a373dc3f643d266ad74c7a0b0d13cb127512911e4b86a88d2e216be1b358c
+size 5393

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/demo_human_or_worm/DNAbert2_Pretrained/lr3e-5_wd0.03_wr0.06_ep8_seed42/results/base5120_demo_human_or_worm_lr3e-5_wd0.03_wr0.06_ep8_seed42/eval_results.json

@@ -0,0 +1 @@
+{"eval_loss": 0.24703867733478546, "eval_accuracy": 0.958, "eval_f1": 0.957999621996598, "eval_matthews_correlation": 0.9160011600280884, "eval_precision": 0.957997831991328, "eval_recall": 0.958003328053249, "eval_runtime": 9.2577, "eval_samples_per_second": 1080.186, "eval_steps_per_second": 8.533, "epoch": 8.0}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "/data/nanhuang/Nan/models/DNAbert2_Pretrained",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.35.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 4096
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:feeef0c39907da371ecf27c5b36c6e2e02993d3f49aa4c5cc63ef947ea3c0e8c
+size 356777880

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a67dec67c94f26952fbfce73585c99612150584c9e86209d83fc2ef1dab902c6
+size 713677451

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4098e4ab6e0c90fa8c2284ad8ab481b5d170da601ea7faab38e80d96bbe954b6
+size 14645

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:675091024a487fbecbd7ff59a446faf287a463a27c48ae3ddc55b3312f4aa745
+size 1465

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "cache_dir": null,
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "trust_remote_code": true,
+  "unk_token": "[UNK]",
+  "use_fast": true
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/trainer_state.json

@@ -0,0 +1,77 @@
+{
+  "best_metric": 0.6941549901360716,
+  "best_model_checkpoint": "genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172",
+  "epoch": 3.976878612716763,
+  "eval_steps": 100,
+  "global_step": 172,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.99,
+      "eval_accuracy": 0.5817655571635311,
+      "eval_f1": 0.5610827704893171,
+      "eval_loss": 0.6628125309944153,
+      "eval_matthews_correlation": 0.21314123722950265,
+      "eval_precision": 0.6214865572625698,
+      "eval_recall": 0.5934860367092647,
+      "eval_runtime": 2.2484,
+      "eval_samples_per_second": 307.332,
+      "eval_steps_per_second": 9.785,
+      "step": 43
+    },
+    {
+      "epoch": 1.99,
+      "eval_accuracy": 0.6743849493487699,
+      "eval_f1": 0.6528777821686682,
+      "eval_loss": 0.5978314876556396,
+      "eval_matthews_correlation": 0.3669122104687527,
+      "eval_precision": 0.7052053787162517,
+      "eval_recall": 0.6640119901257788,
+      "eval_runtime": 2.2606,
+      "eval_samples_per_second": 305.673,
+      "eval_steps_per_second": 9.732,
+      "step": 86
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.4794520547945205e-05,
+      "loss": 0.6252,
+      "step": 100
+    },
+    {
+      "epoch": 2.98,
+      "eval_accuracy": 0.6903039073806078,
+      "eval_f1": 0.687708020071297,
+      "eval_loss": 0.58121657371521,
+      "eval_matthews_correlation": 0.37804735134263795,
+      "eval_precision": 0.6905416012558869,
+      "eval_recall": 0.6875178424490755,
+      "eval_runtime": 2.2678,
+      "eval_samples_per_second": 304.702,
+      "eval_steps_per_second": 9.701,
+      "step": 129
+    },
+    {
+      "epoch": 3.98,
+      "eval_accuracy": 0.6960926193921853,
+      "eval_f1": 0.6941549901360716,
+      "eval_loss": 0.5836874842643738,
+      "eval_matthews_correlation": 0.38973274325358714,
+      "eval_precision": 0.6958638707926167,
+      "eval_recall": 0.6938739525432837,
+      "eval_runtime": 2.274,
+      "eval_samples_per_second": 303.868,
+      "eval_steps_per_second": 9.675,
+      "step": 172
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 172,
+  "num_train_epochs": 4,
+  "save_steps": 100,
+  "total_flos": 5788706328975360.0,
+  "trial_name": null,
+  "trial_params": null
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/checkpoint-172/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a68a7c1d06f3fae4b483d7ad97db3a2ebb2b62fc4919f1c9ca8106fbb000df3
+size 5457

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/drosophila_enhancers_stark/DNAbert2_Pretrained/lr3e-5_wd0.05_wr0.15_ep4_seed42/results/base5120_drosophila_enhancers_stark_lr3e-5_wd0.05_wr0.15_ep4_seed42/eval_results.json

@@ -0,0 +1 @@
+{"eval_loss": 0.6012589931488037, "eval_accuracy": 0.6907514450867052, "eval_f1": 0.685187074829932, "eval_matthews_correlation": 0.37509742426025894, "eval_precision": 0.6905166435506241, "eval_recall": 0.6846270161290322, "eval_runtime": 2.2655, "eval_samples_per_second": 305.451, "eval_steps_per_second": 9.711, "epoch": 3.98}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "/data/nanhuang/Nan/models/DNAbert2_Pretrained",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.35.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 4096
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbb5da133669d1b7e98a5f3c658941682f60b2082efc73aa9a6abb0247660563
+size 356777880

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:542ed2c97fd12d4b4b23702cbde98497b5cd3339e63658fbe28c1c18872936a0
+size 713677451

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab279adf128873bd2145c65a30f6ed93cbcac4b882bde6ac0f5067f8b73817e9
+size 14645

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99ee3b73fbe41c8fc4eee2f5d8cbed5b4be61fbf8b2cc81d5f85262c6ebf5b73
+size 1465

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "cache_dir": null,
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "trust_remote_code": true,
+  "unk_token": "[UNK]",
+  "use_fast": true
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/trainer_state.json

@@ -0,0 +1,71 @@
+{
+  "best_metric": 0.732448866777225,
+  "best_model_checkpoint": "genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31",
+  "epoch": 4.0,
+  "eval_steps": 100,
+  "global_step": 31,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.9,
+      "eval_accuracy": 0.5785123966942148,
+      "eval_f1": 0.5727933541017654,
+      "eval_loss": 0.6725832223892212,
+      "eval_matthews_correlation": 0.15919997829120064,
+      "eval_precision": 0.5816561242093157,
+      "eval_recall": 0.5775956284153005,
+      "eval_runtime": 0.4097,
+      "eval_samples_per_second": 295.354,
+      "eval_steps_per_second": 9.764,
+      "step": 7
+    },
+    {
+      "epoch": 1.94,
+      "eval_accuracy": 0.6859504132231405,
+      "eval_f1": 0.6857572443958447,
+      "eval_loss": 0.5833003520965576,
+      "eval_matthews_correlation": 0.3718896520690588,
+      "eval_precision": 0.6860974274767379,
+      "eval_recall": 0.6857923497267759,
+      "eval_runtime": 0.4058,
+      "eval_samples_per_second": 298.171,
+      "eval_steps_per_second": 9.857,
+      "step": 15
+    },
+    {
+      "epoch": 2.97,
+      "eval_accuracy": 0.6611570247933884,
+      "eval_f1": 0.6565593631014192,
+      "eval_loss": 0.7039547562599182,
+      "eval_matthews_correlation": 0.328770377703483,
+      "eval_precision": 0.6686313973548016,
+      "eval_recall": 0.6602459016393443,
+      "eval_runtime": 0.4059,
+      "eval_samples_per_second": 298.079,
+      "eval_steps_per_second": 9.854,
+      "step": 23
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7355371900826446,
+      "eval_f1": 0.732448866777225,
+      "eval_loss": 0.603266716003418,
+      "eval_matthews_correlation": 0.48516753947547236,
+      "eval_precision": 0.7488499137435307,
+      "eval_recall": 0.7364754098360655,
+      "eval_runtime": 0.41,
+      "eval_samples_per_second": 295.158,
+      "eval_steps_per_second": 9.757,
+      "step": 31
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 56,
+  "num_train_epochs": 8,
+  "save_steps": 100,
+  "total_flos": 1018766006353920.0,
+  "trial_name": null,
+  "trial_params": null
+}

Finetune-GenomicBenchmarks/genomic_bench_DNAbert2_output/dummy_mouse_enhancers_ensembl/DNAbert2_Pretrained/lr3e-5_wd0.01_wr0.10_ep8_seed42/checkpoint-31/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26f1d73872a5edc6087d5a86d94ab9dfb19e4292bbdef9e58372cec8c98799f6
+size 5457