| { |
| "_name_or_path": "Spravil/caption-via-translation-0_4B-ft", |
| "model_type": "florence2", |
| "architectures": [ |
| "Florence2ForConditionalGeneration" |
| ], |
| "auto_map": { |
| "AutoConfig": "Spravil/caption-via-translation-0_4B-ft--configuration_florence2.Florence2Config", |
| "AutoModelForCausalLM": "Spravil/caption-via-translation-0_4B-ft--modeling_florence2.Florence2ForConditionalGeneration" |
| }, |
| "is_encoder_decoder": true, |
| "projection_dim": 768, |
| "text_config": { |
| "activation_dropout": 0.1, |
| "attention_dropout": 0.1, |
| "bos_token_id": 2, |
| "d_model": 768, |
| "decoder_attention_heads": 12, |
| "decoder_ffn_dim": 3072, |
| "decoder_layers": 6, |
| "decoder_start_token_id": 1, |
| "early_stopping": true, |
| "encoder_attention_heads": 12, |
| "encoder_ffn_dim": 3072, |
| "encoder_layers": 6, |
| "encoder_vocab_size": 51328, |
| "eos_token_id": 1, |
| "forced_eos_token_id": 1, |
| "no_repeat_ngram_size": 3, |
| "num_beams": 3, |
| "num_hidden_layers": 6, |
| "pad_token_id": 0, |
| "use_encoder_tokenizer": true, |
| "vocab_size": 257024 |
| }, |
| "vision_config": { |
| "model_type": "davit", |
| "drop_path_rate": 0.1, |
| "patch_size": [ |
| 7, |
| 3, |
| 3, |
| 3 |
| ], |
| "patch_stride": [ |
| 4, |
| 2, |
| 2, |
| 2 |
| ], |
| "patch_padding": [ |
| 3, |
| 1, |
| 1, |
| 1 |
| ], |
| "patch_prenorm": [ |
| false, |
| true, |
| true, |
| true |
| ], |
| "enable_checkpoint": false, |
| "dim_embed": [ |
| 128, |
| 256, |
| 512, |
| 1024 |
| ], |
| "num_heads": [ |
| 4, |
| 8, |
| 16, |
| 32 |
| ], |
| "num_groups": [ |
| 4, |
| 8, |
| 16, |
| 32 |
| ], |
| "depths": [ |
| 1, |
| 1, |
| 9, |
| 1 |
| ], |
| "window_size": 12, |
| "projection_dim": 768, |
| "visual_temporal_embedding": { |
| "type": "COSINE", |
| "max_temporal_embeddings": 100 |
| }, |
| "image_pos_embed": { |
| "type": "learned_abs_2d", |
| "max_pos_embeddings": 50 |
| }, |
| "image_feature_source": [ |
| "spatial_avg_pool", |
| "temporal_avg_pool" |
| ] |
| }, |
| "vocab_size": 257024 |
| } |
