|
{ |
|
"_name_or_path": "mistralai/Mistral-7B-v0.1", |
|
"architectures": [ |
|
"ZettHypernet" |
|
], |
|
"attention_dropout": 0.0, |
|
"auto_map": { |
|
"AutoConfig": "configuration_hypernet.ZettHypernetConfig", |
|
"AutoModel": "modeling_hypernet.ZettHypernet" |
|
}, |
|
"bos_token_id": 1, |
|
"eos_token_id": 2, |
|
"hidden_act": "silu", |
|
"hidden_size": 4096, |
|
"hn_add_inter_token_attention": false, |
|
"hn_concat_last_hidden_state": false, |
|
"hn_embed_lang_id": true, |
|
"hn_embed_target_priors": false, |
|
"hn_embed_using_source_embeddings": true, |
|
"hn_hidden_size": 4096, |
|
"hn_inter_token_attention_bias_by_priors": true, |
|
"hn_inter_token_attention_bias_scaler": 1.0, |
|
"hn_intermediate_size": 8192, |
|
"hn_language_adapter_bottleneck_dim": 0, |
|
"hn_model_name_or_path": "roberta-base", |
|
"hn_model_type": "roberta", |
|
"hn_n_extra_tokens": 522, |
|
"hn_n_inter_token_blocks": 16, |
|
"hn_n_layers": 3, |
|
"hn_num_attention_heads": 32, |
|
"hn_predict_bias": true, |
|
"hn_rescale_embeddings": true, |
|
"hn_single_head": false, |
|
"hn_surface_maxlen": 15, |
|
"initializer_range": 0.02, |
|
"intermediate_size": 14336, |
|
"langs": [ |
|
"en", |
|
"ru", |
|
"de", |
|
"es", |
|
"fr", |
|
"it", |
|
"pt", |
|
"el", |
|
"ko", |
|
"fi", |
|
"id", |
|
"tr", |
|
"ar", |
|
"vi", |
|
"bg", |
|
"ca", |
|
"hi", |
|
"et", |
|
"bn", |
|
"ta", |
|
"ur", |
|
"sw", |
|
"te", |
|
"eu", |
|
"ht", |
|
"qu" |
|
], |
|
"max_position_embeddings": 32768, |
|
"n_embd": 4096, |
|
"n_langs": 26, |
|
"name": "v7:mistral7b_multilingual:lw=0.5_long_length=15_alpha=0.1_26l_resume", |
|
"num_attention_heads": 32, |
|
"num_hidden_layers": 32, |
|
"num_key_value_heads": 8, |
|
"original_vocab_size": 32000, |
|
"pad_token_id": 2, |
|
"rms_norm_eps": 1e-05, |
|
"rope_theta": 10000.0, |
|
"separate_out_embeddings": true, |
|
"sliding_window": 4096, |
|
"tie_word_embeddings": false, |
|
"torch_dtype": "float32", |
|
"transformers_version": "4.39.0.dev0", |
|
"use_cache": true, |
|
"use_unigram_bias": true, |
|
"vocab_size": 32896 |
|
} |
|
|