{ "module": "keras_nlp.src.models.gemma.gemma_causal_lm", "class_name": "GemmaCausalLM", "config": { "backbone": { "module": "keras_nlp.src.models.gemma.gemma_backbone", "class_name": "GemmaBackbone", "config": { "name": "gemma_backbone", "trainable": true, "dtype": { "module": "keras.dtype_policies", "class_name": "DTypePolicyMap", "config": { "default_policy": null, "policy_map": { "token_embedding": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_0/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_1/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_2/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_3/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_4/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_5/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_6/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_7/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_8/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_9/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_10/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_11/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_12/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_13/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_14/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_15/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_16/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_17/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_18/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_19/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_20/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_21/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_22/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_23/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_24/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/ffw_linear": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/ffw_gating_2": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/ffw_gating": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/attention/attention_output": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/attention/value": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/attention/key": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null }, "decoder_block_25/attention/query": { "module": "keras.dtype_policies", "class_name": "QuantizedDTypePolicy", "config": { "mode": "int8", "source_name": null }, "registered_name": null } } }, "registered_name": null }, "vocabulary_size": 256000, "num_layers": 26, "num_query_heads": 8, "num_key_value_heads": 4, "hidden_dim": 2304, "intermediate_dim": 18432, "head_dim": 256, "layer_norm_epsilon": 1e-06, "dropout": 0, "query_head_dim_normalize": true, "use_post_ffw_norm": true, "use_post_attention_norm": true, "final_logit_soft_cap": 30.0, "attention_logit_soft_cap": 50.0, "sliding_window_size": 4096, "use_sliding_window_attention": true }, "registered_name": "keras_nlp>GemmaBackbone" }, "preprocessor": { "module": "keras_nlp.src.models.gemma.gemma_causal_lm_preprocessor", "class_name": "GemmaCausalLMPreprocessor", "config": { "name": "gemma_causal_lm_preprocessor", "trainable": true, "dtype": { "module": "keras", "class_name": "DTypePolicy", "config": { "name": "mixed_bfloat16" }, "registered_name": null }, "tokenizer": { "module": "keras_nlp.src.models.gemma.gemma_tokenizer", "class_name": "GemmaTokenizer", "config": { "name": "gemma_tokenizer", "trainable": true, "dtype": { "module": "keras", "class_name": "DTypePolicy", "config": { "name": "int32" }, "registered_name": null }, "proto": null, "sequence_length": null }, "registered_name": "keras_nlp>GemmaTokenizer" }, "sequence_length": 2048, "add_start_token": true, "add_end_token": true }, "registered_name": "keras_nlp>GemmaCausalLMPreprocessor" }, "name": "gemma_causal_lm" }, "registered_name": "keras_nlp>GemmaCausalLM" }