| { | |
| "activation_function": "silu", | |
| "architectures": [ | |
| "MidmLMHeadModel" | |
| ], | |
| "attn_pdrop": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_midm.MidmBitextConfig", | |
| "AutoModelForCausalLM": "modeling_midm.MidmLMHeadModel" | |
| }, | |
| "bos_token_id": 2, | |
| "embd_pdrop": 0.0, | |
| "eos_token_id": 3, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "model_type": "midm-bitext-S", | |
| "n_embd": 4096, | |
| "n_head": 32, | |
| "n_inner": 10880, | |
| "n_layer": 32, | |
| "n_positions": 8192, | |
| "normalization_type": "layernorm1p", | |
| "pad_token_id": 1, | |
| "reorder_and_upcast_attn": false, | |
| "resid_pdrop": 0.0, | |
| "rotary_percentage": 0.5, | |
| "scale_attn_by_inverse_layer_idx": false, | |
| "scale_attn_weights": true, | |
| "scale_qk_by_inverse_layer_idx": true, | |
| "summary_activation": null, | |
| "summary_first_dropout": 0.1, | |
| "summary_proj_to_labels": true, | |
| "summary_type": "cls_index", | |
| "summary_use_proj": true, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.20.0", | |
| "use_absolute_position_embedding": false, | |
| "use_cache": true, | |
| "use_rotary_position_embedding": true, | |
| "vocab_size": 72192 | |
| } | |