From 325b352f8fddae20d3273b528a11ff5db0e17b10 Mon Sep 17 00:00:00 2001
From: kevin-us <kevin.us@kakaocorp.com>
Date: Sat, 26 Oct 2024 01:57:46 +0900
Subject: [PATCH] remove unnecessary print and codes

---
 tests/models/decoder_only/language/test_cls_models.py | 4 +---
 vllm/model_executor/models/qwen2_cls.py               | 5 -----
 2 files changed, 1 insertion(+), 8 deletions(-)

diff --git a/tests/models/decoder_only/language/test_cls_models.py b/tests/models/decoder_only/language/test_cls_models.py
index f2dc3c3c..352f8a4e 100644
--- a/tests/models/decoder_only/language/test_cls_models.py
+++ b/tests/models/decoder_only/language/test_cls_models.py
@@ -8,15 +8,13 @@
 import pytest
 import torch
 
-from ...utils import check_logprobs_close, check_outputs_equal
-
 CLASSIFICATION_MODELS = [
     "jason9693/Qwen2.5-1.5B-apeach"
 ]
 
 
 @pytest.mark.parametrize("model", CLASSIFICATION_MODELS)
-@pytest.mark.parametrize("dtype", ["bfloat16"])
+@pytest.mark.parametrize("dtype", ["float"])
 def test_classification_models(
     hf_runner,
     vllm_runner,
diff --git a/vllm/model_executor/models/qwen2_cls.py b/vllm/model_executor/models/qwen2_cls.py
index 5286b70b..84b5e72e 100644
--- a/vllm/model_executor/models/qwen2_cls.py
+++ b/vllm/model_executor/models/qwen2_cls.py
@@ -85,7 +85,6 @@ def __init__(
         self.lora_config = lora_config
 
         self.quant_config = quant_config
-        print(f"config: {config}\ncache_config: {cache_config}\nquant_config: {quant_config}")
         self.model = Qwen2Model(config, cache_config, quant_config)
 
         self.score = RowParallelLinear(config.hidden_size,
@@ -101,11 +100,9 @@ def forward(
         attn_metadata: AttentionMetadata,
         intermediate_tensors: Optional[IntermediateTensors] = None,
     ) -> torch.Tensor:
-        print(f"{input_ids}\n{positions}\n{kv_caches}\n{attn_metadata}\n{intermediate_tensors}")
         hidden_states = self.model(input_ids, positions, kv_caches,
                                    attn_metadata, intermediate_tensors)
         logits, _ = self.score(hidden_states)
-        print(logits)
         return logits
 
     def pooler(
@@ -137,7 +134,6 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
                 name = name.replace(weight_name, param_name)
                 # Skip loading extra bias for GPTQ models.
                 if name.endswith(".bias") and name not in params_dict:
-                    print(f"bias is ignored: {name}")
                     continue
                 if is_pp_missing_parameter(name, self):
                     continue
@@ -148,7 +144,6 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
             else:
                 # Skip loading extra bias for GPTQ models.
                 if name.endswith(".bias") and name not in params_dict:
-                    print(f"bias is ignored: {name}")
                     continue
                 # Remapping the name of FP8 kv-scale.
                 name = maybe_remap_kv_scale_name(name, params_dict)