LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
Python ai-agentsdata-engineeringdocument-aigenerative-aiidpjson-extractionllmmcp-serverocrpdf-extractionprompt-engineeringstructured-output
fix: Classifier fix to remove stop and max_tokens kwargs in llm call (#501)
Chandrasekharan M37d1da518ac7
diff --git a/tools/classifier/src/config/properties.json b/tools/classifier/src/config/properties.jsonindex b29db5c35..b095314dd 100644--- a/tools/classifier/src/config/properties.json+++ b/tools/classifier/src/config/properties.json@@ -2,7 +2,7 @@"schemaVersion": "0.0.1","displayName": "File Classifier","functionName": "classify",- "toolVersion": "0.0.25",+ "toolVersion": "0.0.26","description": "Classifies a file into a bin based on its contents","input": {"description": "File to be classified"diff --git a/tools/classifier/src/helper.py b/tools/classifier/src/helper.pyindex dd7e4ac65..488b67b8f 100644--- a/tools/classifier/src/helper.py+++ b/tools/classifier/src/helper.py@@ -132,7 +132,7 @@ class ClassifierHelper:str: Classification"""try:- completion = llm.complete(prompt, max_tokens=50, stop=["\n"])[LLM.RESPONSE]+ completion = llm.complete(prompt)[LLM.RESPONSE]classification: str = completion.text.strip()self.tool.stream_log(f"LLM response: {completion}")return classification