From bb089c09d0e3311eb027872c35617cd98c861f62 Mon Sep 17 00:00:00 2001 From: GamerBoss101 Date: Sun, 30 Mar 2025 11:45:44 -0400 Subject: [PATCH] Demo Update 30 --- Backend/server.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/Backend/server.py b/Backend/server.py index f266ee1..1e87780 100644 --- a/Backend/server.py +++ b/Backend/server.py @@ -540,11 +540,13 @@ def generate_ai_response(user_text, session_id): # Generate response inputs = llm_tokenizer(prompt, return_tensors="pt").to(device) output = llm_model.generate( - inputs.input_ids, + inputs.input_ids, + attention_mask=inputs.attention_mask, # Add attention mask max_new_tokens=100, # Keep responses shorter for voice temperature=0.7, top_p=0.9, - do_sample=True + do_sample=True, + pad_token_id=llm_tokenizer.eos_token_id # Explicitly set pad_token_id ) response = llm_tokenizer.decode(output[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)