lambda-160m / model_config.json
MK0727's picture
Upload lambda-160m pretrained model
134df9b verified
{"max_len": 1024, "d_model": 768, "num_layers": 16, "num_heads": 12, "d_ff": 3072, "learning_rate": 0.0002, "lr_schedule": "warmup_cosine", "lr_warmup_steps": 2000, "min_learning_rate": 2e-05, "min_learning_rate_ratio": 0.1, "loss_chunk_size": 32, "pad_token_id": 0, "bos_token_id": 2, "eos_token_id": 3, "corpus_signature": "551ac72eceb57f5f", "dataset_cases": [{"name": "fineweb2-edu-ja", "genre": "web", "language": "ja", "dataset_path": "hotchpotch/fineweb-2-edu-japanese", "config_name": "default", "split": "train", "text_column": "text", "token_percentage": 30.0, "is_ramped": false, "repeat_on_end": true, "excluded_url_domains": ["wikipedia.org"]}, {"name": "cleanedwiki-jp", "genre": "wiki", "language": "ja", "dataset_path": "MK0727/CleanedWiki-jp", "config_name": "all", "split": "train", "text_column": "text", "token_percentage": 70.0, "is_ramped": true, "repeat_on_end": true, "excluded_url_domains": []}], "mix_cycle_tokens": 100000, "ramp_start_progress": 0.5, "val_split_modulo": 100, "val_split_index": 0, "validation_cache_path": "models/lambda-160m/validation-cache-551ac72eceb57f5f-bos-eos-text-hash-len1024-samples6144-split100-0.pt", "validation_sample_count": 6144, "trained_steps": 40960}