43 lines
1.3 KiB
JSON
43 lines
1.3 KiB
JSON
{
|
|
"_name_or_path": "inceptionai/jais-family-590m",
|
|
"activation_function": "swiglu",
|
|
"alibi_scaling": null,
|
|
"architectures": [
|
|
"JAISLMHeadModel"
|
|
],
|
|
"attn_pdrop": 0.0,
|
|
"auto_map": {
|
|
"AutoConfig": "configuration_jais.JAISConfig",
|
|
"AutoModel": "modeling_jais.JAISModel",
|
|
"AutoModelForCausalLM": "modeling_jais.JAISLMHeadModel",
|
|
"AutoModelForQuestionAnswering": "modeling_jais.JAISForQuestionAnswering",
|
|
"AutoModelForSequenceClassification": "modeling_jais.JAISForSequenceClassification",
|
|
"AutoModelForTokenClassification": "modeling_jais.JAISForTokenClassification"
|
|
},
|
|
"bos_token_id": 0,
|
|
"embd_pdrop": 0.0,
|
|
"eos_token_id": 0,
|
|
"initializer_range": 0.02,
|
|
"layer_norm_epsilon": 1e-05,
|
|
"model_type": "jais",
|
|
"mup_embeddings_scale": 9.1705785388303,
|
|
"mup_output_alpha": 1.09518349815769,
|
|
"mup_scale_qk_dot_by_d": true,
|
|
"mup_width_scale": 0.16666666666666666,
|
|
"n_embd": 1536,
|
|
"n_head": 12,
|
|
"n_inner": 4096,
|
|
"n_layer": 18,
|
|
"n_positions": 2048,
|
|
"pad_token_id": 0,
|
|
"position_embedding_type": "alibi",
|
|
"reorder_and_upcast_attn": false,
|
|
"resid_pdrop": 0.0,
|
|
"scale_attn_by_inverse_layer_idx": false,
|
|
"scale_attn_weights": true,
|
|
"torch_dtype": "float32",
|
|
"transformers_version": "4.40.1",
|
|
"use_cache": true,
|
|
"vocab_size": 84992
|
|
}
|