UltraFinBERT-150M / experiment.json
beethogedeon's picture
Publish validated UltraFinBERT-150M FP16 weights, model card and evaluation
42dd9d2 verified
Raw History Blame Contribute Delete
2.04 kB
{
"base_model": "dmis-lab/UltraBERT-150M",
"hub_repo_id": "beethogedeon/UltraFinBERT-150M",
"model_revisions": {
"dmis-lab/UltraBERT-150M": "4cc8ca415eb1958184c491297fe5aaa94c4a8ad2",
"ProsusAI/finbert": "0bc2ac10dda6929bcb1f77ccc43c7ec5c424f031",
"beethogedeon/Modern-FinBERT-large": "a5dd04dae984a46e729698ca39f398181c9aaeea"
},
"seed": 42,
"label2id": {
"negative": 0,
"neutral": 1,
"positive": 2
},
"max_length": 256,
"split_strategy": "Stratified 70/15/15 split generated from TXT with seed 42 (fallback mode)",
"learning_rates": [
8e-06
],
"max_epochs": 10,
"early_stopping_patience": 3,
"train_batch_size": 8,
"gradient_accumulation_steps": 4,
"effective_batch_size": 32,
"optimizer": "AdamW",
"betas": [
0.9,
0.98
],
"epsilon": 1e-06,
"weight_decay": 0.1,
"scheduler": "linear",
"warmup_fraction": 0.05,
"training_compute_dtype": "bfloat16",
"evaluation_compute_dtype": "torch.float16",
"export_dtype": "float16",
"attention": "sdpa",
"data_audit": {
"raw_rows": 2264,
"encoding": "cp1252",
"conflict_groups": 0,
"conflict_rows_removed": 0,
"duplicate_rows_removed": 5,
"retained_rows": 2259,
"raw_label_counts": {
"neutral": 1391,
"positive": 570,
"negative": 303
},
"retained_label_counts": {
"neutral": 1386,
"positive": 570,
"negative": 303
},
"source_sha256": "5fb370b3005b9a2759f708849a691ad2e4d2593d4c995c1afde229398a516594",
"source": "provided TXT file",
"split_strategy_note": "Stratified 70/15/15 split generated from TXT with seed 42 (fallback mode)",
"split_sizes": {
"train": 1581,
"validation": 339,
"test": 339
}
},
"environment": {
"torch": "2.11.0+cu128",
"transformers": "5.16.1",
"accelerate": "1.14.0",
"datasets": "4.8.5",
"huggingface_hub": "1.29.0",
"safetensors": "0.8.0",
"numpy": "2.1.3",
"pandas": "2.3.3",
"scikit-learn": "1.6.1"
},
"device": "Tesla T4"
}