components oss_distillation_generate_data_batch_postprocess - Azure/azureml-assets GitHub Wiki

OSS Distillation Generate Data Postprocess Batch Scoring

oss_distillation_generate_data_batch_postprocess

Overview

Component to prepare data returned from teacher model enpoint in batch

Version: 0.0.1

View in Studio: https://ml.azure.com/registries/azureml/components/oss_distillation_generate_data_batch_postprocess/version/0.0.1

Inputs

Inputs

Name Description Type Default Optional Enum
train_file_path Path to the registered training data asset. The supported data formats are jsonl, json, csv, tsv and parquet. uri_file
validation_file_path Path to the registered validation data asset. The supported data formats are jsonl, json, csv, tsv and parquet. uri_file True
hash_train_data jsonl file containing the hash for each payload. uri_file False
hash_validation_data jsonl file containing the hash for each payload. uri_file True
batch_score_train_result Path to the directory containing jsonl file(s) that have the result for each payload. uri_folder
batch_score_validation_result Path to the directory containing jsonl file(s) that have the result for each payload. uri_folder True
min_endpoint_success_ratio The minimum value of (successful_requests / total_requests) required for classifying inference as successful. If (successful_requests / total_requests) < min_endpoint_success_ratio, the experiment will be marked as failed. By default it is 0.7 (0 means all requests are allowed to fail while 1 means no request should fail.) number 0.7
enable_chain_of_thought Enable Chain of thought for data generation string false True
enable_chain_of_density Enable Chain of density for text summarization string false True
data_generation_task_type Data generation task type. Supported values are: 1. NLI: Generate Natural Language Inference data 2. CONVERSATION: Generate conversational data (multi/single turn) 3. NLU_QA: Generate Natural Language Understanding data for Question Answering data 4. MATH: Generate Math data for numerical responses 5. SUMMARIZATION: Generate Key Summary for an Article string ['NLI', 'CONVERSATION', 'NLU_QA', 'MATH', 'SUMMARIZATION']
connection_config_file Connection config file for batch scoring uri_file

Outputs

Name Description Type
generated_batch_train_file_path Generated train data uri_file
generated_batch_validation_file_path Generated validation data uri_file

Environment

azureml://registries/azureml/environments/acft-hf-nlp-gpu/versions/76

⚠️ **GitHub.com Fallback** ⚠️