Update README.md

README.md

@@ -29,26 +29,28 @@ So while the original LIAR dataset supplies factual claims from political discou
 
 ### Task Format
 
-The model reframes classification as a text-to-text task, generating a numeric label (as a string) for each claim:
+This model treats classification as a **text-to-text generation task**. Each input is a short claim or quote, and the model responds with one of six factuality labels, generated directly as a lowercase string:
 
-- `0`: pants-fire  
-- `1`: false  
-- `2`: barely-true  
-- `3`: half-true  
-- `4`: mostly-true  
-- `5`: true
+- `pants-fire`  
+- `false`  
+- `barely-true`  
+- `half-true`  
+- `mostly-true`  
+- `true`
 
-**Example Input**:  
+The input format uses a summarization-style prefix to frame the task:
+
+**Example Input**:
 ```
-veracity: Open-source AI systems cannot hallucinate because they're transparent.
+summarize: Python is the fastest programming language available.
 ```
 
 **Example Output**:  
 ```
-0
+half-true
 ```
 
-You can map the numeric labels back to human-readable categories using a simple dictionary.
+This response reflects the model’s ability to evaluate short-form claims with nuance, producing a graded label based on its understanding of truthfulness.
 
 ### Training Details
 
@@ -75,8 +77,11 @@ It is **not** intended for production-grade fact-checking or regulatory enforcem
 ### Example Usage
 
 ```python
+### Example Usage
+
 from transformers import T5ForConditionalGeneration, T5Tokenizer
 
+# Load the fine-tuned model and tokenizer
 model = T5ForConditionalGeneration.from_pretrained(
     "gcuomo/open-source-ai-t5-liar-lens"
 )
@@ -84,20 +89,18 @@ tokenizer = T5Tokenizer.from_pretrained(
     "gcuomo/open-source-ai-t5-liar-lens"
 )
 
-label_map = {
-    "0": "pants-fire",
-    "1": "false",
-    "2": "barely-true",
-    "3": "half-true",
-    "4": "mostly-true",
-    "5": "true"
-}
-
-input_text = "veracity: Blockchain guarantees ethical outcomes in all AI systems."
-inputs = tokenizer(input_text, return_tensors="pt")
-output = model.generate(**inputs)
-prediction = tokenizer.decode(output[0], skip_special_tokens=True).strip()
-print("Predicted label:", label_map.get(prediction, prediction))
+# Prepare input
+statement = "Blockchain guarantees ethical outcomes in all AI systems."
+prompt = f"summarize: {statement}"
+inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True, max_length=128)
+
+# Generate prediction
+output = model.generate(**inputs, max_new_tokens=8)
+prediction = tokenizer.decode(output[0], skip_special_tokens=True).strip().lower()
+
+# Print result
+print("Predicted label:", prediction)
+
 ```
 
 ### Citation