Update app.py

app.py

@@ -94,8 +94,8 @@ def run_interpretation(global_state, raw_interpretation_prompt, max_new_tokens,
     interpretation_prompt = InterpretationPrompt(tokenizer, interpretation_prompt)
 
     # generate the interpretations
-    generate = generate_interpretation_gpu if use_gpu else lambda lambda interpretation_prompt, **kwargs: interpretation_prompt.generate(**kwargs)
-    generated = generate(model, {0: interpreted_vectors}, k=3, **generation_kwargs)
+    generate = generate_interpretation_gpu if use_gpu else lambda interpretation_prompt, **kwargs: interpretation_prompt.generate(**kwargs)
+    generated = generate(interpretation_prompt, model, {0: interpreted_vectors}, k=3, **generation_kwargs)
     generation_texts = tokenizer.batch_decode(generated)
     progress_dummy_output = ''
     return ([progress_dummy_output] +