How to convert model to onnx format and do inference?

[kishoreb4]

Hi, Im' trying to convert the model to onnx format with the help of below code.
pipeline = ChronosPipeline.from_pretrained(
"amazon/chronos-t5-small",
device_map = "cuda",
torch_dtype = torch.bfloat16)

df = pd.read_csv("AirPassengers.csv")

dummy_input = torch.tensor(df['#Passengers']).to(device)

torch.onnx.export(
pipeline.model,
dummy_input,
"chronos_model.onnx",
verbose=True,
input_names=["input"],
output_names=["output"],
)

Facing and error as below:

File /libraries/gpu_py_3_11/lib/python3.11/site-packages/transformers/models/t5/modeling_t5.py:1023, in T5Stack.forward(self, input_ids, attention_mask, encoder_hidden_states, encoder_attention_mask, inputs_embeds, head_mask, cross_attn_head_mask, past_key_values, use_cache, output_attentions, output_hidden_states, return_dict)
1020 raise ValueError("You have to initialize the model with valid token embeddings")
1021 inputs_embeds = self.embed_tokens(input_ids)
-> 1023 batch_size, seq_length = input_shape
1025 # required mask seq length can be calculated via length of past
1026 mask_seq_length = past_key_values[0][0].shape[2] + seq_length if past_key_values is not None else seq_length

ValueError: not enough values to unpack (expected 2, got 1)

Could anyone please help?

#ONNX