|
{ |
|
"module": "keras_hub.src.models.llama.llama_causal_lm", |
|
"class_name": "LlamaCausalLM", |
|
"config": { |
|
"backbone": { |
|
"module": "keras_hub.src.models.llama.llama_backbone", |
|
"class_name": "LlamaBackbone", |
|
"config": { |
|
"name": "llama_backbone_1", |
|
"trainable": true, |
|
"vocabulary_size": 32000, |
|
"num_layers": 32, |
|
"num_query_heads": 32, |
|
"hidden_dim": 4096, |
|
"intermediate_dim": 11008, |
|
"rope_max_wavelength": 10000.0, |
|
"rope_position_scaling_factor": 1.0, |
|
"rope_frequency_adjustment_factor": null, |
|
"rope_low_freq_factor": null, |
|
"rope_high_freq_factor": null, |
|
"rope_pretraining_sequence_length": null, |
|
"num_key_value_heads": 32, |
|
"layer_norm_epsilon": 1e-05, |
|
"dropout": 0 |
|
}, |
|
"registered_name": "keras_hub>LlamaBackbone" |
|
}, |
|
"preprocessor": { |
|
"module": "keras_hub.src.models.llama.llama_causal_lm_preprocessor", |
|
"class_name": "LlamaCausalLMPreprocessor", |
|
"config": { |
|
"name": "llama_causal_lm_preprocessor", |
|
"trainable": true, |
|
"dtype": { |
|
"module": "keras", |
|
"class_name": "DTypePolicy", |
|
"config": { |
|
"name": "bfloat16" |
|
}, |
|
"registered_name": null |
|
}, |
|
"tokenizer": { |
|
"module": "keras_hub.src.models.llama.llama_tokenizer", |
|
"class_name": "LlamaTokenizer", |
|
"config": { |
|
"name": "llama_tokenizer", |
|
"trainable": true, |
|
"dtype": { |
|
"module": "keras", |
|
"class_name": "DTypePolicy", |
|
"config": { |
|
"name": "int32" |
|
}, |
|
"registered_name": null |
|
}, |
|
"config_file": "tokenizer.json", |
|
"proto": null, |
|
"sequence_length": null, |
|
"add_bos": false, |
|
"add_eos": false |
|
}, |
|
"registered_name": "keras_hub>LlamaTokenizer" |
|
}, |
|
"config_file": "preprocessor.json", |
|
"sequence_length": 1024, |
|
"add_start_token": true, |
|
"add_end_token": true |
|
}, |
|
"registered_name": "keras_hub>LlamaCausalLMPreprocessor" |
|
}, |
|
"name": "llama_causal_lm" |
|
}, |
|
"registered_name": "keras_hub>LlamaCausalLM" |
|
} |