| { | |
| "block_size": 256, | |
| "vocab_size": 50257, | |
| "n_layer": 36, | |
| "n_head": 16, | |
| "n_embd": 1024, | |
| "dropout": 0.1, | |
| "bias": true, | |
| "use_checkpointing": true, | |
| "model_type": "custom_gpt_slm", | |
| "tokenizer": "tiktoken_gpt2", | |
| "dataset": "PleIAs/common_corpus", | |
| "dataset_filter": "language == English" | |
| } |