|
{ |
|
"architectures": [ |
|
"MUDDFormer" |
|
], |
|
"auto_map": { |
|
"AutoConfig": "configuration_muddformer.MUDDFormerConfig", |
|
"AutoModelForCausalLM": "modeling_muddformer.MUDDFormer" |
|
}, |
|
"block_size": 2048, |
|
"bos_token_id": 1, |
|
"dense": true, |
|
"dense_type": "qkvr", |
|
"dim": 2560, |
|
"dynamic_dense": true, |
|
"eos_token_id": 2, |
|
"expand_last": true, |
|
"head_dim": 80, |
|
"intermediate_size": 6912, |
|
"is_training": false, |
|
"model_type": "muddformer", |
|
"n_head": 32, |
|
"n_layer": 32, |
|
"n_local_heads": 32, |
|
"norm_eps": 1e-06, |
|
"rope_base": 10000, |
|
"round64": true, |
|
"sepln": true, |
|
"stack_hidden": false, |
|
"tie_word_embeddings": false, |
|
"torch_dtype": "bfloat16", |
|
"transformers_version": "4.35.0", |
|
"use_gradient_checkpointing": false, |
|
"use_layer_cache": true, |
|
"use_qk_norm": true, |
|
"vocab_size": 50432 |
|
} |
|
|