| { | |
| "architectures": [ | |
| "Cheers" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_umm.UMMConfig", | |
| "AutoModel": "modeling_umm.UMMModel", | |
| "AutoModelForCausalLM": "modeling_umm.Cheers" | |
| }, | |
| "vae_encoder_config": { | |
| "resolution": 512 | |
| }, | |
| "vae_decoder_config": { | |
| "resolution": 512 | |
| }, | |
| "vision_representation_config": { | |
| "attention_dropout": 0.0, | |
| "hidden_act": "gelu_pytorch_tanh", | |
| "hidden_size": 1152, | |
| "image_size": 512, | |
| "intermediate_size": 4304, | |
| "layer_norm_eps": 1e-06, | |
| "model_type": "umm", | |
| "num_attention_heads": 16, | |
| "num_channels": 3, | |
| "num_hidden_layers": 27, | |
| "num_patches": 1024, | |
| "patch_size": 16 | |
| }, | |
| "text_config":{ | |
| "hidden_size": 1536, | |
| "intermediate_size": 8960, | |
| "max_window_layers": 21, | |
| "num_attention_heads": 12, | |
| "num_key_value_heads": 2, | |
| "sliding_window": 32768, | |
| "tie_word_embeddings": true, | |
| "vocab_size": 151936, | |
| "max_position_embeddings": 32768 | |
| }, | |
| "model_type": "umm", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.51.3" | |
| } |