Anujgr8/indic-bert-large-cs

1{2  "_name_or_path": "ai4bharat/indic-bert",3  "activation_dropout": 0.1,4  "adapter_attn_dim": null,5  "adapter_kernel_size": 3,6  "adapter_stride": 2,7  "add_adapter": false,8  "apply_spec_augment": true,9  "architectures": [10    "Wav2Vec2ForCTC"11  ],12  "attention_dropout": 0.0,13  "attention_probs_dropout_prob": 0,14  "bos_token_id": 1,15  "classifier_proj_size": 256,16  "codevector_dim": 256,17  "contrastive_logits_temperature": 0.1,18  "conv_bias": false,19  "conv_dim": [20    512,21    512,22    512,23    512,24    512,25    512,26    51227  ],28  "conv_kernel": [29    10,30    3,31    3,32    3,33    3,34    2,35    236  ],37  "conv_stride": [38    5,39    2,40    2,41    2,42    2,43    2,44    245  ],46  "ctc_loss_reduction": "mean",47  "ctc_zero_infinity": false,48  "diversity_loss_weight": 0.1,49  "do_stable_layer_norm": false,50  "down_scale_factor": 1,51  "embedding_size": 128,52  "eos_token_id": 2,53  "feat_extract_activation": "gelu",54  "feat_extract_norm": "group",55  "feat_proj_dropout": 0.0,56  "feat_quantizer_dropout": 0.0,57  "final_dropout": 0.1,58  "gap_size": 0,59  "hidden_act": "gelu",60  "hidden_dropout": 0.0,61  "hidden_dropout_prob": 0,62  "hidden_size": 768,63  "initializer_range": 0.02,64  "inner_group_num": 1,65  "intermediate_size": 3072,66  "layer_norm_eps": 1e-05,67  "layerdrop": 0.0,68  "mask_feature_length": 10,69  "mask_feature_min_masks": 0,70  "mask_feature_prob": 0.0,71  "mask_time_length": 10,72  "mask_time_min_masks": 2,73  "mask_time_prob": 0.05,74  "max_position_embeddings": 512,75  "model_type": "wav2vec2",76  "net_structure_type": 0,77  "num_adapter_layers": 3,78  "num_attention_heads": 12,79  "num_codevector_groups": 2,80  "num_codevectors_per_group": 320,81  "num_conv_pos_embedding_groups": 16,82  "num_conv_pos_embeddings": 128,83  "num_feat_extract_layers": 7,84  "num_hidden_groups": 1,85  "num_hidden_layers": 12,86  "num_memory_blocks": 0,87  "num_negatives": 100,88  "output_hidden_size": 768,89  "pad_token_id": 94,90  "proj_codevector_dim": 256,91  "tdnn_dilation": [92    1,93    2,94    3,95    1,96    197  ],98  "tdnn_dim": [99    512,100    512,101    512,102    512,103    1500104  ],105  "tdnn_kernel": [106    5,107    3,108    3,109    1,110    1111  ],112  "torch_dtype": "float32",113  "transformers_version": "4.35.2",114  "type_vocab_size": 2,115  "use_weighted_layer_sum": false,116  "vocab_size": 97,117  "xvector_output_dim": 512118}119