ONNX generation
#9
by davesoma - opened
Hello everyone,
Could someone share an example of Python code to run the ONNX version for text generation?
Thank you!
I tried to build a genai_config given the tensor config but dont work
{
"model": {
"bos_token_id": 0,
"context_length": 2072,
"decoder": {
"session_options": {
"log_id": "onnxruntime-genai",
"provider_options": []
},
"filename": "model.onnx",
"head_size": 64,
"hidden_size": 2560,
"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": 9,
"num_hidden_layers": 30,
"num_key_value_heads": 3
},
"eos_token_id": 0,
"pad_token_id": 0,
"type": "llama",
"vocab_size": 49152
},
"search": {
"diversity_penalty": 0.0,
"do_sample": false,
"early_stopping": true,
"length_penalty": 1.0,
"max_length": 2072,
"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": 50,
"top_p": 1.0
}