Tensor-23m / training_config.json
VGribAI's picture
Upload 6 files
8d1bab6 verified
Raw
History Blame Contribute Delete
358 Bytes
{
"d_model": 512,
"n_heads": 8,
"n_kv_heads": 4,
"n_layers": 6,
"max_seq_len": 512,
"device": "cuda",
"dataset_files": [
"deepseek_text_20260707_487aef.txt",
"deepseek_text_20260707_4f6f18.txt",
"deepseek_text_20260707_62ba67.txt",
"deepseek_text_20260707_94a4ea.txt",
"deepseek_text_20260707_da3020.txt"
]
}