| {"block_size": 1024, "vocab_size": 2048, "n_layer": 4, "n_head": 128, "d_head": 16, "d_model": 2048, "dropout": 0.0, "bias": true, "ln_bias": true, "rms_norm": true, "residual_activation_type": "identity", "activation_type": "gelu", "afrac": 0.125, "afrac_loctypes": "attn_in,attn_out,mlp_in,mlp_out,mlp_neuron,attn_k,attn_q,attn_v", "afrac_approx": false, "enable_sparse_kernels": false, "ignore_dw_grad": false, "debug_nans": false, "tied_unembed": false, "bigram_table_rank": null, "tokenizer_name": "tinypython_2k", "pfrac": 0.5, "expansion_factor": 8, "expansion_factor_mlp": 8, "debug_exact_topk": true, "grad_checkpointing": true, "sparse_matmul_impl": "cuda", "d_mlp": 8192, "enable_bigram_table": true, "learnable_bigram_table": false, "d_pos_emb": null, "dropout_cat_pos_emb": false, "sinusoidal_cat_pos_emb": false, "flash": true, "sink": false, "rtopk": false, "n_embd": null} |