| { | |
| "_name_or_path": "tiny_models/flaubert/FlaubertWithLMHeadModel", | |
| "architectures": [ | |
| "FlaubertWithLMHeadModel" | |
| ], | |
| "asm": false, | |
| "attention_dropout": 0.1, | |
| "bos_index": 0, | |
| "bos_token_id": 0, | |
| "causal": false, | |
| "dropout": 0.1, | |
| "emb_dim": 32, | |
| "embed_init_std": 0.02209708691207961, | |
| "end_n_top": 5, | |
| "eos_index": 1, | |
| "gelu_activation": true, | |
| "init_std": 0.02, | |
| "initializer_range": 0.02, | |
| "is_decoder": true, | |
| "is_encoder": true, | |
| "lang_id": 0, | |
| "layer_norm_eps": 1e-12, | |
| "layerdrop": 0.0, | |
| "mask_index": 5, | |
| "mask_token_id": 5, | |
| "max_position_embeddings": 512, | |
| "model_type": "flaubert", | |
| "n_heads": 4, | |
| "n_langs": 2, | |
| "n_layers": 5, | |
| "n_special": 0, | |
| "pad_index": 2, | |
| "pad_token_id": 2, | |
| "pre_norm": false, | |
| "sinusoidal_embeddings": false, | |
| "start_n_top": 5, | |
| "summary_activation": null, | |
| "summary_first_dropout": 0.1, | |
| "summary_proj_to_labels": true, | |
| "summary_type": "last", | |
| "summary_use_proj": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.28.0.dev0", | |
| "unk_index": 3, | |
| "use_lang_emb": true, | |
| "use_proj": null, | |
| "vocab_size": 67542 | |
| } | |