{ "_name_or_path": "LLM360/CrystalCoder", "activation_function": "swiglu", "architectures": [ "CrystalCoderLMHeadModel" ], "attn_pdrop": 0.0, "auto_map": { "AutoConfig": "LLM360/CrystalCoder--configuration_crystalcoder.CrystalCoderConfig", "AutoModel": "LLM360/CrystalCoder--modeling_crystalcoder.CrystalCoderModel", "AutoModelForCausalLM": "LLM360/CrystalCoder--modeling_crystalcoder.CrystalCoderLMHeadModel" }, "bos_token_id": 1, "embd_pdrop": 0.0, "eos_token_id": 2, "initializer_range": 0.02, "layer_norm_epsilon": 1e-05, "model_type": "crystalcoder", "mup_embeddings_scale": 14.6, "mup_output_alpha": 2.22, "mup_scale_qk_dot_by_d": true, "mup_width_scale": 0.0625, "n_embd": 4096, "n_head": 32, "n_inner": 10922, "n_layer": 32, "n_positions": 2048, "position_embedding_type": "rotary", "reorder_and_upcast_attn": false, "resid_pdrop": 0.0, "rotary_dim": 32, "scale_attn_by_inverse_layer_idx": false, "scale_attn_weights": true, "torch_dtype": "bfloat16", "transformers_version": "4.35.2", "use_cache": true, "vocab_size": 32032 }