diff --git a/tests/quantization/ggml/test_ggml.py b/tests/quantization/ggml/test_ggml.py index 42b05f18449d..1171e82e5285 100644 --- a/tests/quantization/ggml/test_ggml.py +++ b/tests/quantization/ggml/test_ggml.py @@ -623,8 +623,8 @@ def test_falcon7b_q2_k(self): torch_dtype=torch.float16, ) - text = tokenizer(self.example_text, return_tensors="pt").to(torch_device) - out = model.generate(**text, max_new_tokens=10) + text = tokenizer(self.example_text, return_tensors="pt")["input_ids"].to(torch_device) + out = model.generate(text, max_new_tokens=10) EXPECTED_TEXT = "Hello All,\nI am new to this forum." self.assertEqual(tokenizer.decode(out[0], skip_special_tokens=True), EXPECTED_TEXT)