babylm_baseline_modernbert / final_model_metadata.json
eligran12's picture
Upload folder using huggingface_hub
2e5a419 verified
Raw
History Blame Contribute Delete
1.66 kB
{
"timestamp": "2026-07-06T11:13:55.659948",
"model_architecture": "ModernBERT",
"objective": "masked_language_modeling",
"tokenizer_strategy": "default_modernbert_tokenizer",
"tokenizer_name": "answerdotai/ModernBERT-base",
"vocab_size": 50368,
"training_config": {
"model_name": "modernbert_default_tokenizer",
"base_model_for_tokenizer": "answerdotai/ModernBERT-base",
"hidden_size": 768,
"num_hidden_layers": 22,
"num_attention_heads": 12,
"intermediate_size": 1152,
"training": {
"batch_size": 16,
"gradient_accumulation_steps": 8,
"learning_rate": 0.0005,
"num_epochs": 10,
"warmup_steps": 1000,
"weight_decay": 0.01,
"logging_steps": 100,
"seed": 42,
"max_grad_norm": 1.0,
"mlm_probability": 0.15
},
"data": {
"max_seq_length": 256,
"adjusted_budget_per_lang": 33333333,
"tokenize_batch_size": 1000,
"chunk_batch_size": 1000
},
"output_dir": "./model_modernbert_default",
"babylm_checkpoint_dir": "./babylm_checkpoints_modernbert_default",
"detailed_checkpoint_dir": "./checkpoints_detailed_modernbert_default",
"checkpoint_intervals": [
1000000,
2000000,
3000000,
4000000,
5000000,
6000000,
7000000,
8000000,
9000000,
10000000,
20000000,
30000000,
40000000,
50000000,
60000000,
70000000,
80000000,
90000000,
100000000,
200000000,
300000000,
400000000,
500000000,
600000000,
700000000,
800000000,
900000000,
1000000000
]
}
}