|
{ |
|
"activation_function": "silu", |
|
"architectures": [ |
|
"HawkForCausalLM" |
|
], |
|
"auto_map": { |
|
"AutoConfig": "configuration_hawk.HawkConfig", |
|
"AutoModelForCausalLM": "modelling_hawk.HawkForCausalLM" |
|
}, |
|
"attn_pdrop": 0.0, |
|
"bos_token_id": 65535, |
|
"embd_pdrop": 0.0, |
|
"eos_token_id": 65535, |
|
"initializer_range": 0.02, |
|
"layer_norm_epsilon": 1e-06, |
|
"model_type": "hawk", |
|
"n_embd": 1024, |
|
"n_head": 16, |
|
"n_inner": 2688, |
|
"n_layer": 20, |
|
"n_positions": 1024, |
|
"reorder_and_upcast_attn": false, |
|
"resid_pdrop": 0.0, |
|
"rotary_dim": 64, |
|
"scale_attn_by_inverse_layer_idx": false, |
|
"scale_attn_weights": true, |
|
"summary_activation": null, |
|
"summary_first_dropout": 0.0, |
|
"summary_proj_to_labels": true, |
|
"summary_type": "cls_index", |
|
"summary_use_proj": true, |
|
"tokenizer_class": "HawkTokenizer", |
|
"transformers_version": "4.31.0", |
|
"use_cache": true, |
|
"vocab_size": 65536 |
|
} |
|
|