| { | |
| "version": "QNN_SYSTEM_CONTEXT_BINARY_INFO_VERSION_3", | |
| "info": { | |
| "backendId": 6, | |
| "buildId": "v2.28.0.241029232508_102474", | |
| "coreApiVersion": "2.21.0", | |
| "backendApiVersion": "5.28.0", | |
| "socVersion": "", | |
| "contextBlobVersion": "3.2.0", | |
| "contextBlobSize": 467141712, | |
| "numContextTensors": 0, | |
| "contextTensors": [], | |
| "numGraphs": 2, | |
| "graphs": [ | |
| { | |
| "version": "QNN_SYSTEM_CONTEXT_GRAPH_INFO_VERSION_3", | |
| "info": { | |
| "graphName": "ar128_cl4096_3_of_4", | |
| "numGraphInputs": 20, | |
| "graphInputs": [ | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 1, | |
| "name": "past_key_16_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15907233953475953, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 6, | |
| "name": "past_key_21_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14844174683094026, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 9, | |
| "name": "past_value_21_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07443267107009888, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 16, | |
| "name": "past_value_16_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07209276407957077, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 125, | |
| "name": "past_key_22_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1573055237531662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 128, | |
| "name": "past_value_22_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06790729612112045, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 267, | |
| "name": "past_key_23_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1642313301563263, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 270, | |
| "name": "past_value_23_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0965195819735527, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 441, | |
| "name": "past_key_17_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.17103533446788789, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 444, | |
| "name": "past_value_17_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06805716454982758, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 583, | |
| "name": "past_key_18_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14796161651611329, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 586, | |
| "name": "past_value_18_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06999378651380539, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 731, | |
| "name": "past_key_19_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15470433235168458, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 734, | |
| "name": "past_value_19_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06925888359546662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 875, | |
| "name": "past_key_20_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 3968 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.16168494522571565, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 878, | |
| "name": "past_value_20_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 3968, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.062198661267757419, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 1041, | |
| "name": "_model_layers_15_Add_1_Add_output_0", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 3, | |
| "dimensions": [ | |
| 1, | |
| 128, | |
| 3072 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1906648725271225, | |
| "offset": -41533 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 1946, | |
| "name": "position_ids_cos", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 128, | |
| 48 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.00003632373773143627, | |
| "offset": -32768 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 1948, | |
| "name": "position_ids_sin", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 128, | |
| 48 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.00003632373773143627, | |
| "offset": -32768 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 2653, | |
| "name": "attention_mask", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 128, | |
| 4096 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0007629510946571827, | |
| "offset": -65535 | |
| } | |
| } | |
| } | |
| } | |
| ], | |
| "numGraphOutputs": 17, | |
| "graphOutputs": [ | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 1913, | |
| "name": "past_value_16_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07209276407957077, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 2556, | |
| "name": "past_key_16_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15907233953475953, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 3676, | |
| "name": "past_value_17_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06805716454982758, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 4317, | |
| "name": "past_key_17_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.17103533446788789, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 5436, | |
| "name": "past_value_18_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06999378651380539, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 6077, | |
| "name": "past_key_18_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14796161651611329, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 7196, | |
| "name": "past_value_19_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06925888359546662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 7837, | |
| "name": "past_key_19_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15470433235168458, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 8956, | |
| "name": "past_value_20_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.062198661267757419, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 9597, | |
| "name": "past_key_20_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.16168494522571565, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 10716, | |
| "name": "past_value_21_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07443267107009888, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 11357, | |
| "name": "past_key_21_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14844174683094026, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 12476, | |
| "name": "past_value_22_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06790729612112045, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 13117, | |
| "name": "past_key_22_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1573055237531662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 14236, | |
| "name": "past_value_23_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 128, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0965195819735527, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 14877, | |
| "name": "past_key_23_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 128 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1642313301563263, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15124, | |
| "name": "_model_layers_23_Add_1_Add_output_0", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 3, | |
| "dimensions": [ | |
| 1, | |
| 128, | |
| 3072 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.191017746925354, | |
| "offset": -41581 | |
| } | |
| } | |
| } | |
| } | |
| ], | |
| "numUpdateableTensors": 0, | |
| "updateableTensors": [], | |
| "graphBlobInfoSize": 40, | |
| "graphBlobInfo": [ | |
| { | |
| "version": "QNN_SYSTEM_CONTEXT_HTP_GRAPH_INFO_BLOB_VERSION_V1", | |
| "info": { | |
| "spillFillBufferSize": 0, | |
| "optimizationLevel": 3, | |
| "vtcmSize": 8, | |
| "htpDlbc": 0, | |
| "numHvxThreads": 0 | |
| } | |
| } | |
| ] | |
| } | |
| }, | |
| { | |
| "version": "QNN_SYSTEM_CONTEXT_GRAPH_INFO_VERSION_3", | |
| "info": { | |
| "graphName": "ar1_cl4096_3_of_4", | |
| "numGraphInputs": 20, | |
| "graphInputs": [ | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15125, | |
| "name": "past_key_16_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15907233953475953, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15130, | |
| "name": "past_key_21_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14844174683094026, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15133, | |
| "name": "past_value_21_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07443267107009888, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15140, | |
| "name": "past_value_16_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07209276407957077, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15249, | |
| "name": "past_key_22_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1573055237531662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15252, | |
| "name": "past_value_22_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06790729612112045, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15391, | |
| "name": "past_key_23_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1642313301563263, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15394, | |
| "name": "past_value_23_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0965195819735527, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15565, | |
| "name": "past_key_17_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.17103533446788789, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15568, | |
| "name": "past_value_17_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06805716454982758, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15707, | |
| "name": "past_key_18_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14796161651611329, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15710, | |
| "name": "past_value_18_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06999378651380539, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15855, | |
| "name": "past_key_19_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15470433235168458, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15858, | |
| "name": "past_value_19_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06925888359546662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 15999, | |
| "name": "past_key_20_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 4095 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.16168494522571565, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 16002, | |
| "name": "past_value_20_in", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 4095, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.062198661267757419, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 16165, | |
| "name": "_model_layers_15_Add_1_Add_output_0", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 3, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 3072 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1906648725271225, | |
| "offset": -41533 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 17070, | |
| "name": "position_ids_cos", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 1, | |
| 48 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.00003632373773143627, | |
| "offset": -32768 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 17072, | |
| "name": "position_ids_sin", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 1, | |
| 48 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.00003632373773143627, | |
| "offset": -32768 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 17777, | |
| "name": "attention_mask", | |
| "type": "QNN_TENSOR_TYPE_APP_WRITE", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 4, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 1, | |
| 4096 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0007629510946571827, | |
| "offset": -65535 | |
| } | |
| } | |
| } | |
| } | |
| ], | |
| "numGraphOutputs": 17, | |
| "graphOutputs": [ | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 17037, | |
| "name": "past_value_16_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07209276407957077, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 17680, | |
| "name": "past_key_16_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15907233953475953, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 18800, | |
| "name": "past_value_17_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06805716454982758, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 19441, | |
| "name": "past_key_17_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.17103533446788789, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 20560, | |
| "name": "past_value_18_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06999378651380539, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 21201, | |
| "name": "past_key_18_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14796161651611329, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 22320, | |
| "name": "past_value_19_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06925888359546662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 22961, | |
| "name": "past_key_19_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.15470433235168458, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 24080, | |
| "name": "past_value_20_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.062198661267757419, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 24721, | |
| "name": "past_key_20_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.16168494522571565, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 25840, | |
| "name": "past_value_21_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.07443267107009888, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 26481, | |
| "name": "past_key_21_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.14844174683094026, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 27600, | |
| "name": "past_value_22_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.06790729612112045, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 28241, | |
| "name": "past_key_22_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1573055237531662, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 29360, | |
| "name": "past_value_23_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 1, | |
| 96 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.0965195819735527, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 30001, | |
| "name": "past_key_23_out", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_8", | |
| "rank": 4, | |
| "dimensions": [ | |
| 32, | |
| 1, | |
| 96, | |
| 1 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.1642313301563263, | |
| "offset": -128 | |
| } | |
| } | |
| } | |
| }, | |
| { | |
| "version": "QNN_TENSOR_VERSION_1", | |
| "info": { | |
| "id": 30248, | |
| "name": "_model_layers_23_Add_1_Add_output_0", | |
| "type": "QNN_TENSOR_TYPE_APP_READ", | |
| "dataFormat": "QNN_TENSOR_DATA_FORMAT_FLAT_BUFFER", | |
| "dataType": "QNN_DATATYPE_UFIXED_POINT_16", | |
| "rank": 3, | |
| "dimensions": [ | |
| 1, | |
| 1, | |
| 3072 | |
| ], | |
| "memType": "QNN_TENSORMEMTYPE_MEMHANDLE", | |
| "quantizeParams": { | |
| "definition": "QNN_DEFINITION_DEFINED", | |
| "quantizationEncoding": "QNN_QUANTIZATION_ENCODING_SCALE_OFFSET", | |
| "scaleOffset": { | |
| "scale": 0.191017746925354, | |
| "offset": -41581 | |
| } | |
| } | |
| } | |
| } | |
| ], | |
| "numUpdateableTensors": 0, | |
| "updateableTensors": [], | |
| "graphBlobInfoSize": 40, | |
| "graphBlobInfo": [ | |
| { | |
| "version": "QNN_SYSTEM_CONTEXT_HTP_GRAPH_INFO_BLOB_VERSION_V1", | |
| "info": { | |
| "spillFillBufferSize": 0, | |
| "optimizationLevel": 3, | |
| "vtcmSize": 8, | |
| "htpDlbc": 0, | |
| "numHvxThreads": 0 | |
| } | |
| } | |
| ] | |
| } | |
| } | |
| ], | |
| "contextMetadataSize": 8, | |
| "contextMetadata": { | |
| "version": "QNN_SYSTEM_CONTEXT_HTP_CONTEXT_INFO_BLOB_VERSION_V1", | |
| "info": { | |
| "dsp arch": 73 | |
| } | |
| }, | |
| "soc model": 43 | |
| } | |
| } | |