| { |
| "architectures": [ |
| "EMGMTSForCausalLM" |
| ], |
| "auto_map": { |
| "AutoConfig": "emg_mts_multilingual.EMGMTSConfig", |
| "AutoModel": "emg_mts_multilingual.EMGMTSModel", |
| "AutoModelForCausalLM": "emg_mts_multilingual.EMGMTSForCausalLM", |
| "AutoModelForSequenceClassification": "emg_mts_multilingual.EMGMTSForSequenceClassification", |
| "AutoModelForTokenClassification": "emg_mts_multilingual.EMGMTSForTokenClassification" |
| }, |
| "bos_token_id": 1, |
| "classifier_dropout": 0.1, |
| "conv_kernel": 4, |
| "d_k": 96, |
| "decay_fast": 0.85, |
| "decay_slow": 0.995, |
| "decays": null, |
| "dropout": 0.1, |
| "dtype": "float32", |
| "eos_token_id": 2, |
| "gumbel_temperature": 2.0, |
| "hidden_size": 704, |
| "k_params": 0, |
| "lambda_param_sup": 0.0, |
| "logit_cap": 6.0, |
| "max_position_embeddings": 1024, |
| "mlp_expansion": 3, |
| "model_type": "emg_mts", |
| "n_layer": 14, |
| "n_selective_states": 2, |
| "pad_token_id": 3, |
| "param_label_smooth": 0.05, |
| "param_reg": 0.01, |
| "qk_norm": true, |
| "tie_word_embeddings": true, |
| "transformers_version": "5.3.0", |
| "use_compile": false, |
| "val_dim_fast": 192, |
| "val_dim_slow": 96, |
| "vocab_size": 39953 |
| } |
|
|
