Request V2 files from graphql

[horheynm]

Using this branch on deepsparse:

(.venv) george@quad-mle-2:~/nm/deepsparse$ deepsparse.benchmark zoo:nlg/text_generation/opt-1.3b/pytorch/huggingface/opt_pretrain/pruned50_quantW8A8-none
2023-09-05 14:56:20 deepsparse.benchmark.helpers INFO     Thread pinning to cores enabled
Downloading (…)A8/model.onnx.tar.gz: 100%|████████████████████████████████████████████████████| 1.01G/1.01G [01:33<00:00, 11.6MB/s]
2023-09-05 14:58:29 deepsparse.benchmark.benchmark_model INFO     Found model with KV cache support. Benchmarking the autoregressive model with input_ids_length: 1 and sequence length: 512.
2023-09-05 14:58:30 deepsparse.transformers.utils.helpers INFO     Overwriting in-place the input shapes of the transformer model at /home/george/.cache/sparsezoo/neuralmagic/opt-1.3b-opt_pretrain-pruned50_quantW8A8/model.onnx
DeepSparse, Copyright 2021-present / Neuralmagic, Inc. version: 1.6.0.20230815 COMMUNITY | (134dba40) (release) (optimized) (system=avx2, binary=avx2)
[7f1214ded700 >WARN<  operator() ./src/include/wand/utility/warnings.hpp:14] Generating emulated code for quantized (INT8) operations since no VNNI instructions were detected. Set NM_FAST_VNNI_EMULATION=1 to increase performance at the expense of accuracy.
2023-09-05 15:00:33 deepsparse.benchmark.benchmark_model INFO     deepsparse.engine.Engine:
...
2023-09-05 15:00:33 deepsparse.utils.onnx INFO     Generating input 'causal_mask', type = int64, shape = [1, 1, 1, 512]
2023-09-05 15:00:33 deepsparse.benchmark.benchmark_model INFO     Starting 'singlestream' performance measurements for 10 seconds
Original Model Path: zoo:nlg/text_generation/opt-1.3b/pytorch/huggingface/opt_pretrain/pruned50_quantW8A8-none
Batch Size: 1
Sequence Length: 512
Input IDs Length: 1
Scenario: sync
Throughput (items/sec): 17.4194
Latency Mean (ms/batch): 57.3149
Latency Median (ms/batch): 57.0588
Latency Std (ms/batch): 1.6795
Iterations: 175