| { |
| "_attn_implementation_autoset": true, |
| "architectures": [ |
| "LlamaForCausalLM" |
| ], |
| "auto_map":{ |
| "AutoConfig": "models/config.NextStepConfig", |
| "AutoModel": "models/nextstep_model.NextStep" |
| }, |
| "attention_bias": true, |
| "attention_dropout": 0.0, |
| "base_image_grid_size": 32, |
| "boi": 151667, |
| "bos_token_id": 151643, |
| "eoi": 151668, |
| "eos_token_id": 151643, |
| "fm_head_batch_mul": 4, |
| "fm_head_dim": 1536, |
| "fm_head_layers": 12, |
| "head_dim": 128, |
| "hidden_act": "silu", |
| "hidden_size": 5120, |
| "im_loss_weight": 1.0, |
| "image_placeholder_id": 151669, |
| "initializer_range": 0.02, |
| "intermediate_size": 13824, |
| "latent_channels": 16, |
| "latent_patch_size": 2, |
| "latent_size": 64, |
| "lm_loss_weight": 0.01, |
| "max_position_embeddings": 131072, |
| "max_window_layers": 48, |
| "mlp_bias": false, |
| "model_type": "nextstep", |
| "num_attention_heads": 40, |
| "num_hidden_layers": 48, |
| "num_key_value_heads": 8, |
| "o_attention_bias": false, |
| "pad_token_id_added": 151665, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 1000000.0, |
| "sliding_window": 131072, |
| "tie_word_embeddings": false, |
| "torch_dtype": "bfloat16", |
| "transformers_version": "4.55.0", |
| "use_cache": true, |
| "use_gen_pos_embed": true, |
| "use_sliding_window": false, |
| "vae_name_or_path": "./vae", |
| "vocab_size": 152064 |
| } |
|
|
