Update app.py

app.py

@@ -6,36 +6,38 @@ from gradio.mix import Parallel, Series
 #import torch.nn.functional as F
 from aitextgen import aitextgen
 
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+#is fine tuning worth is?????????????????????????????????????????????
 
-from datasets import load_dataset
-dataset = load_dataset("bananabot/engMollywoodSummaries")
-tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neo-1.3B")
+#device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 
-tokenizer.pad_token = tokenizer.eos_token
+#from datasets import load_dataset
+#dataset = load_dataset("bananabot/engMollywoodSummaries")
+#tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neo-1.3B")
 
-def tokenize_function(examples):
-    return tokenizer(examples["text"], padding="max_length", truncation=True)
-tokenized_datasets = dataset.map(tokenize_function, batched=True)
+#tokenizer.pad_token = tokenizer.eos_token
 
-model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-1.3B").to(device)
-training_args = TrainingArguments(output_dir="test_trainer")
-small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))
-small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))
-def compute_metrics(eval_pred):
-    logits, labels = eval_pred
-    predictions = np.argmax(logits, axis=-1)
-    return metric.compute(predictions=predictions, references=labels)
+#def tokenize_function(examples):
+#    return tokenizer(examples["text"], padding="max_length", truncation=True)
+#tokenized_datasets = dataset.map(tokenize_function, batched=True)
 
-trainer = Trainer(
-    model=model,
-    args=training_args,
-    train_dataset=small_train_dataset,
-    eval_dataset=small_eval_dataset,
-    compute_metrics=compute_metrics,
-)
+#model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-1.3B").to(device)
+#training_args = TrainingArguments(output_dir="test_trainer")
+#small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))
+#small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))
+#def compute_metrics(eval_pred):
+#    logits, labels = eval_pred
+#    predictions = np.argmax(logits, axis=-1)
+#    return metric.compute(predictions=predictions, references=labels)
 
-trainer.train()
+#trainer = Trainer(
+#    model=model,
+#    args=training_args,
+#    train_dataset=small_train_dataset,
+#    eval_dataset=small_eval_dataset,
+#    compute_metrics=compute_metrics,
+#)
+
+#trainer.train()
 
 
 
@@ -59,7 +61,7 @@ ai = aitextgen(model="EleutherAI/gpt-neo-1.3B")
 #    print (output)
 
 def ai_text(inp):
-  generated_text = ai.generate_one(max_length=123, prompt = inp, no_repeat_ngram_size=3, num_beams=5, do_sample=True, temperature=1.37, top_k=69, top_p=0.96)
+  generated_text = ai.generate_one(max_length=333, prompt = inp, no_repeat_ngram_size=3, num_beams=7, do_sample=True, temperature=1.37, top_k=69, top_p=0.96)
   print(type(generated_text))
   return generated_text