[CI/Build] Further decouple HuggingFace implementation from ours during tests (#4166)

tests/conftest.py

@@ -1,19 +1,21 @@
 import contextlib
 import gc
 import os
-from typing import List, Optional, Tuple
+from typing import Any, Dict, List, Optional, Tuple
 
 import pytest
 import torch
 from PIL import Image
-from transformers import (AutoModelForCausalLM, AutoProcessor,
+from transformers import (AutoModelForCausalLM, AutoProcessor, AutoTokenizer,
-                          LlavaForConditionalGeneration)
+                          LlavaConfig, LlavaForConditionalGeneration)
 
 from vllm import LLM, SamplingParams
 from vllm.config import TokenizerPoolConfig, VisionLanguageConfig
 from vllm.distributed import destroy_model_parallel
+from vllm.logger import init_logger
 from vllm.sequence import MultiModalData
-from vllm.transformers_utils.tokenizer import get_tokenizer
+
+logger = init_logger(__name__)
 
 _TEST_DIR = os.path.dirname(__file__)
 _TEST_PROMPTS = [os.path.join(_TEST_DIR, "prompts", "example.txt")]
@@ -129,9 +131,7 @@ _STR_DTYPE_TO_TORCH_DTYPE = {
     "float": torch.float,
 }
 
-_VISION_LANGUAGE_MODELS = {
+AutoModelForCausalLM.register(LlavaConfig, LlavaForConditionalGeneration)
-    "llava-hf/llava-1.5-7b-hf": LlavaForConditionalGeneration,
-}
 
 _EMBEDDING_MODELS = [
     "intfloat/e5-mistral-7b-instruct",
@@ -143,23 +143,14 @@ class HfRunner:
     def __init__(
         self,
         model_name: str,
-        tokenizer_name: Optional[str] = None,
         dtype: str = "half",
     ) -> None:
         assert dtype in _STR_DTYPE_TO_TORCH_DTYPE
         torch_dtype = _STR_DTYPE_TO_TORCH_DTYPE[dtype]
+
         self.model_name = model_name
-        if model_name in _VISION_LANGUAGE_MODELS:
+
-            self.model = _VISION_LANGUAGE_MODELS[model_name].from_pretrained(
+        if model_name in _EMBEDDING_MODELS:
-                model_name,
-                torch_dtype=torch_dtype,
-                trust_remote_code=True,
-            ).cuda()
-            self.processor = AutoProcessor.from_pretrained(
-                model_name,
-                torch_dtype=torch_dtype,
-            )
-        elif model_name in _EMBEDDING_MODELS:
             # Lazy init required for AMD CI
             from sentence_transformers import SentenceTransformer
             self.model = SentenceTransformer(
@@ -172,10 +163,24 @@ class HfRunner:
                 torch_dtype=torch_dtype,
                 trust_remote_code=True,
             ).cuda()
-            self.processor = None
+
-        if tokenizer_name is None:
+        self.tokenizer = AutoTokenizer.from_pretrained(
-            tokenizer_name = model_name
+            model_name,
-        self.tokenizer = get_tokenizer(tokenizer_name, trust_remote_code=True)
+            torch_dtype=torch_dtype,
+            trust_remote_code=True,
+        )
+
+        try:
+            self.processor = AutoProcessor.from_pretrained(
+                model_name,
+                torch_dtype=torch_dtype,
+                trust_remote_code=True,
+            )
+        except Exception:
+            logger.warning(
+                "Unable to auto-load processor from HuggingFace for "
+                "model %s. Using tokenizer instead.", model_name)
+            self.processor = self.tokenizer
 
     def generate(
         self,
@@ -187,19 +192,19 @@ class HfRunner:
         if images:
             assert len(prompts) == len(images)
         for i, prompt in enumerate(prompts):
-            if self.model_name not in _VISION_LANGUAGE_MODELS:
+            processor_kwargs: Dict[str, Any] = {
-                input_ids = self.tokenizer(prompt,
+                "text": prompt,
-                                           return_tensors="pt").input_ids
+                "return_tensors": "pt",
-                inputs = {"input_ids": input_ids.cuda()}
+            }
-            else:
+            if images is not None and images[i] is not None:
-                image = images[i] if images else None
+                processor_kwargs["images"] = images[i]
-                inputs = self.processor(text=prompt,
+
-                                        images=image,
+            inputs = self.processor(**processor_kwargs)
-                                        return_tensors="pt")
+            inputs = {
-                inputs = {
+                key: value.cuda() if value is not None else None
-                    key: value.cuda() if value is not None else None
+                for key, value in inputs.items()
-                    for key, value in inputs.items()
+            }
-                }
+
             output_ids = self.model.generate(
                 **inputs,
                 use_cache=True,