|
{ |
|
"_name_or_path": "checkpoints/jat_small_v100/checkpoint-250000", |
|
"action_loss_coef": 0.995, |
|
"activation_function": "gelu_new", |
|
"architectures": [ |
|
"JatModel" |
|
], |
|
"attention_dropout": 0.0, |
|
"attention_layers": [ |
|
"global", |
|
"local", |
|
"global", |
|
"local", |
|
"global", |
|
"local", |
|
"global", |
|
"local", |
|
"global", |
|
"local", |
|
"global", |
|
"local" |
|
], |
|
"attention_types": [ |
|
[ |
|
[ |
|
"global", |
|
"local" |
|
], |
|
6 |
|
] |
|
], |
|
"auto_map": { |
|
"AutoConfig": "configuration_jat.JatConfig", |
|
"AutoModelForCausalLM": "modeling_jat.JatModel" |
|
}, |
|
"bos_token_id": 50256, |
|
"classifier_dropout": 0.1, |
|
"embed_dropout": 0.0, |
|
"eos_token_id": 50256, |
|
"hidden_size": 768, |
|
"image_size": 224, |
|
"initializer_range": 0.02, |
|
"intermediate_size": null, |
|
"layer_norm_epsilon": 1e-05, |
|
"max_continuous_size": 377, |
|
"max_discrete_value": 212, |
|
"max_position_embeddings": 512, |
|
"model_type": "jat", |
|
"num_channels": 3, |
|
"num_heads": 12, |
|
"num_layers": 12, |
|
"observation_loss_coef": 0.005, |
|
"patch_size": 16, |
|
"resid_dropout": 0.0, |
|
"tokenizer_class": "GPT2TokenizerFast", |
|
"torch_dtype": "float32", |
|
"transformers_version": "4.36.1", |
|
"use_cache": true, |
|
"vocab_size": 50257, |
|
"window_size": 256 |
|
} |
|
|