Upload metadata.json
Browse files- metadata.json +35 -0
metadata.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"actual_num_records": 100,
|
| 3 |
+
"buffer_size": 100,
|
| 4 |
+
"column_statistics": [
|
| 5 |
+
{
|
| 6 |
+
"column_name": "llm_structured_1",
|
| 7 |
+
"column_type": "llm-structured",
|
| 8 |
+
"input_tokens_mean": 793.96,
|
| 9 |
+
"input_tokens_median": 823.0,
|
| 10 |
+
"input_tokens_stddev": 135.64,
|
| 11 |
+
"num_null": 0,
|
| 12 |
+
"num_records": 100,
|
| 13 |
+
"num_unique": 94,
|
| 14 |
+
"output_tokens_mean": 89.48,
|
| 15 |
+
"output_tokens_median": 73.5,
|
| 16 |
+
"output_tokens_stddev": 41.28,
|
| 17 |
+
"pyarrow_dtype": "struct<answer: string, evidence_quote: string, question: string>",
|
| 18 |
+
"simple_dtype": "dict"
|
| 19 |
+
}
|
| 20 |
+
],
|
| 21 |
+
"dataset_name": "recipe_economic-700",
|
| 22 |
+
"file_paths": {
|
| 23 |
+
"data": [
|
| 24 |
+
"data/batch_00000.parquet"
|
| 25 |
+
]
|
| 26 |
+
},
|
| 27 |
+
"num_completed_batches": 1,
|
| 28 |
+
"schema": {
|
| 29 |
+
"chunk_text": "string",
|
| 30 |
+
"llm_structured_1": "struct<answer: string, evidence_quote: string, question: string>",
|
| 31 |
+
"source_file": "string"
|
| 32 |
+
},
|
| 33 |
+
"target_num_records": 100,
|
| 34 |
+
"total_num_batches": 1
|
| 35 |
+
}
|