Update app.py

app.py

@@ -1,47 +1,46 @@
-# app.py
-import os, random
-from collections import defaultdict
-import gradio as gr
-
-# ─── 1) Load & tokenize your corpus ─────────────────────────────────────────
-HERE         = os.path.dirname(__file__)
-DATA_DIR     = os.path.join(HERE, "training_corpus")
-CORPUS_FILE  = os.path.join(DATA_DIR, "corpus_earth_10k.txt")
-
-with open(CORPUS_FILE, "r", encoding="utf-8") as f:
-    tokens = f.read().split()
-
-# ─── 2) Build Markov chain (order = 2) ───────────────────────────────────────
-k = 2
-chain = defaultdict(list)
-for i in range(len(tokens) - k):
-    state = tuple(tokens[i : i + k])
-    chain[state].append(tokens[i + k])
-
-# ─── 3) Generation function ─────────────────────────────────────────────────
-def generate_text():
-    # pick a random starting state
-    state = random.choice(list(chain.keys()))
-    output = list(state)
-
-    # generate 200 words total (including the k seed words)
-    for _ in range(200 - k):
-        nxt_choices = chain.get(state)
-        if not nxt_choices:
-            # dead end → pick a fresh state
-            state = random.choice(list(chain.keys()))
-            output.extend(state)
-            continue
-        nxt = random.choice(nxt_choices)
-        output.append(nxt)
-        state = tuple(output[-k:])
-    return " ".join(output)
-
-# ─── 4) Gradio UI ────────────────────────────────────────────────────────────
-with gr.Blocks() as demo:
-    gr.Markdown("## Planet Earth Markov Generator")
-    output_box = gr.Textbox(label="Generated Text", lines=10)
-    gr.Button("Generate").click(fn=generate_text, outputs=output_box)
-
-if __name__ == "__main__":
+# app.py
+import os, random
+from collections import defaultdict
+import gradio as gr
+
+# 1) Load & tokenize the corpus
+HERE         = os.path.dirname(__file__)
+CORPUS_FILE = os.path.join(HERE, "corpus_earth_10k.txt")
+
+with open(CORPUS_FILE, "r", encoding="utf-8") as f:
+    tokens = f.read().split()
+
+# 2) Building Markov chain with order = 2
+k = 2
+chain = defaultdict(list)
+for i in range(len(tokens) - k):
+    state = tuple(tokens[i : i + k])
+    chain[state].append(tokens[i + k])
+
+# 3) Text-Generation function 
+def generate_text():
+    # pick a random starting state
+    state = random.choice(list(chain.keys()))
+    output = list(state)
+
+    # generate 200 words total (including the k seed words)
+    for _ in range(200 - k):
+        nxt_choices = chain.get(state)
+        if not nxt_choices:
+            # dead end → pick a fresh state
+            state = random.choice(list(chain.keys()))
+            output.extend(state)
+            continue
+        nxt = random.choice(nxt_choices)
+        output.append(nxt)
+        state = tuple(output[-k:])
+    return " ".join(output)
+
+# 4) Gradio UI 
+with gr.Blocks() as demo:
+    gr.Markdown("## Planet Earth Markov Generator")
+    output_box = gr.Textbox(label="Generated Text", lines=10)
+    gr.Button("Generate").click(fn=generate_text, outputs=output_box)
+
+if __name__ == "__main__":
     demo.launch(share=True)