{ "model": "Novi-Nano-Base", "version": 2, "vocab_size": 8192, "block_size": 256, "n_embd": 96, "n_layer": 4, "n_head": 4, "n_inner": 384, "parameters": 1258560, "training_tokens": 300000000, "tokens_seen": 300023808, "optimizer_steps": 2289, "batch_size": 256, "grad_accum_steps": 2, "learning_rate": 0.0003, "min_learning_rate": 3e-05, "warmup_steps": 500, "weight_decay": 0.1, "grad_clip": 1.0, "validation_tokens": 5000000, "final_validation_loss": 5.418698577137737, "final_validation_perplexity": 225.5853494508474, "best_validation_loss": 5.418698577137737 }