Update app.py

app.py

@@ -1,15 +1,43 @@
 import gradio as gr
 from ctransformers import AutoModelForCausalLM
 from transformers import AutoTokenizer, pipeline
+import datasets
 import torch
 import re
 import random
 
+
+
+
+from pathlib import Path
+from huggingface_hub import CommitScheduler
+
+JSON_DATASET_DIR = Path("json_dataset")
+JSON_DATASET_DIR.mkdir(parents=True, exist_ok=True)
+
+JSON_DATASET_PATH = JSON_DATASET_DIR / f"feedbacks.json"
+
+scheduler = CommitScheduler(
+    repo_id="Detsutut/feedbacks_test",
+    repo_type="dataset",
+    folder_path=JSON_DATASET_DIR,
+    path_in_repo="data",
+)
+
+def save_json(last_state: dict, pos_or_neg: str) -> None:
+    last_state["feedback"]=pos_or_neg
+    with scheduler.lock:
+        with JSON_DATASET_PATH.open("a") as f:
+            json.dump(last_state, f)
+            f.write("\n")
+
+
+
+
 # Initialize the model
 model = AutoModelForCausalLM.from_pretrained("Detsutut/Igea-1B-instruct-v0.3-test4epochs-GGUF", model_file="unsloth.Q4_K_M.gguf", model_type="mistral", hf=True)
 tokenizer = AutoTokenizer.from_pretrained( "Detsutut/Igea-1B-instruct-v0.3-test4epochs")
 
-
 gen_pipeline = pipeline(
     "text-generation",
     model=model,
@@ -67,10 +95,12 @@ def generate_text(input_text, max_new_tokens=512, temperature=1, system_prompt="
 def positive_feedback(last_generated_text):
     print("positive")
     print(last_generated_text)
+    save_json(last_generated_text,"positive")
 
 def negative_feedback(last_generated_text):
     print("negative")
     print(last_generated_text)
+    save_json(last_generated_text,"negative")
     
 
 # Create the Gradio interface