|
{ |
|
"model": { |
|
"bos_token_id": 1, |
|
"context_length": 32768, |
|
"decoder": { |
|
"session_options": { |
|
"log_id": "onnxruntime-genai", |
|
"custom_ops_library": "onnx_custom_ops.dll", |
|
"custom_allocator": "shared_d3d_xrt", |
|
"external_data_file": "Mistral-7B-Instruct-v0.3_jit.pb.bin", |
|
"provider_options": [ |
|
] |
|
}, |
|
"filename": "Mistral-7B-Instruct-v0.3_jit.onnx", |
|
"head_size": 128, |
|
"hidden_size": 4096, |
|
"inputs": { |
|
"input_ids": "input_ids", |
|
"attention_mask": "attention_mask", |
|
"position_ids": "position_ids", |
|
"past_key_names": "past_key_values.%d.key", |
|
"past_value_names": "past_key_values.%d.value" |
|
}, |
|
"outputs": { |
|
"logits": "logits", |
|
"present_key_names": "present.%d.key", |
|
"present_value_names": "present.%d.value" |
|
}, |
|
"num_attention_heads": 32, |
|
"num_hidden_layers": 32, |
|
"num_key_value_heads": 8 |
|
}, |
|
"eos_token_id": 2, |
|
"pad_token_id": 2, |
|
"type": "mistral", |
|
"vocab_size": 32768 |
|
}, |
|
"search": { |
|
"diversity_penalty": 0.0, |
|
"do_sample": false, |
|
"early_stopping": true, |
|
"length_penalty": 1.0, |
|
"max_length": 32768, |
|
"min_length": 0, |
|
"no_repeat_ngram_size": 0, |
|
"num_beams": 1, |
|
"num_return_sequences": 1, |
|
"past_present_share_buffer": true, |
|
"repetition_penalty": 1.0, |
|
"temperature": 1.0, |
|
"top_k": 1, |
|
"top_p": 1.0 |
|
} |
|
} |