v. 1.1

* updated library versions
* cache directory defined as subdirectory of home dir
* cache dir for Transformers set as environment variable, instead of
in config file

CHANGES.md

@@ -0,0 +1,10 @@
+# ChangeLog
+
+## v. 1.1
+* updated library versions
+* cache directory defined as subdirectory of home dir
+* cache dir for Transformers set as environment variable, instead of
+  in config file 
+
+## v. 1.0
+* initial version

app.py

@@ -1,8 +1,22 @@
-import gradio as gr
 import os
-from pii_transform.api.e2e import PiiTextProcessor
+from pathlib import Path
+
+# Set the HF cache directory for the Transformers plugin
+cache_dir = Path.home() / "app"/ "cache"
+cache_dir.mkdir(parents=True, exist_ok=True)
+os.environ["HUGGINGFACE_HUB_CACHE"] = str(cache_dir)
+
+from pii_process.api import PiiTextProcessor
 from pii_extract.defs import FMT_CONFIG_PLUGIN
+import gradio as gr
 
+# Test if the cache directory exists
+if cache_dir.is_dir():
+    gr.Info("Cache directory created at "+str(cache_dir))
+else:
+    gr.Warning("Cache directory creation error")
+
+# Read examples
 examples = []
 with open("examples.txt", "r") as f:
     examples = f.readlines()
@@ -15,19 +29,14 @@ language_choices = {
     "German": "de",
     "French": "fr",
 }
+
 language_code = "en"
-cache_dir = "/home/user/app/cache"
-os.makedirs(cache_dir, exist_ok=True)
-if os.path.isdir(cache_dir):
-    gr.Info("Cache directory created at "+cache_dir)
-else:
-    gr.Warning("Cache directory creation error")
 
 policy_help_string = """
 Policies are defined as follows:
 
-1. **Annotate** - replace the PII instance by a \<TYPE:VALUE\> string, i.e. include both the PII type and its value
-2. **Redact** - all PII instances are replaced by a \<PII\> generic string
+1. **Annotate** - replace the PII instance by a <TYPE:VALUE> string, i.e. include both the PII type and its value
+2. **Redact** - all PII instances are replaced by a <PII> generic string
 3. **Placeholder** - replace with a prototypical value
 4. **Synthetic** - substitute with synthetic data
 
@@ -170,4 +179,3 @@ with gr.Blocks() as demo:
     with gr.Accordion(label="Help Panel", open=False):
         gr.Markdown(value=policy_help_string)
 demo.queue().launch()
-

config.json

@@ -4,7 +4,6 @@
         {
             "format": "piisa:config:pii-extract-plg-transformers:main:v1",
             "task_config": {
-                "cachedir": "/home/user/app/cache",
                 "reuse_engine": true,
                 "aggregation": "max",
                 "models": [
@@ -74,4 +73,4 @@
             }
         }
     ]
-}
+}

requirements.txt

@@ -1,9 +1,5 @@
-pii-data==0.4.0
-pii-extract-base==0.5.0
-pii-extract-plg-presidio==0.3.2
-pii-extract-plg-regex==0.4.1
-pii-extract-plg-transformers==0.1.1
-pii-preprocess==0.0.4
-pii-transform==0.5.1
-torch==2.1.0
-transformers==4.34.0
+wheel
+pii-process[transformers] >= 0.1.1
+
+torch>=2.2
+transformers>=4.34.0