{ "model_config": { "model_name": "OpenLLM-Small-10k", "model_size": "small", "vocab_size": 32000, "n_layer": 6, "n_head": 8, "n_embd": 512, "block_size": 1024, "dropout": 0.1, "bias": true, "training_info": { "step": 10000, "best_loss": 5.22, "model_type": "gpt-small" } }, "tokenizer_config": { "type": "sentencepiece", "vocab_size": 32000, "model_file": "data/tokenizer/tokenizer.model" }, "training_config": { "learning_rate": 0.0001, "batch_size": 4, "gradient_accumulation_steps": 4, "max_steps": 10000, "warmup_steps": 100, "weight_decay": 0.01 } }