[Deprecation][2/N] Replace --task with --runner and --convert (#21470)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

tests/v1/core/test_kv_cache_utils.py

@@ -695,11 +695,7 @@ def test_estimate_max_model_len(model_id, max_model_len,
     # Create a VllmConfig
     model_config = ModelConfig(
         model_id,
-        task="generate",
-        tokenizer=model_id,
-        tokenizer_mode="auto",
-        trust_remote_code=False,
-        seed=0,
+        runner="generate",
         dtype="float16",
         max_model_len=max_model_len,
     )
@@ -733,11 +729,7 @@ def test_get_max_concurrency_for_kv_cache_config():
     max_model_len = 16384
     model_config = ModelConfig(
         model_id,
-        task="generate",
-        tokenizer=model_id,
-        tokenizer_mode="auto",
-        trust_remote_code=False,
-        seed=0,
+        runner="generate",
         dtype="float16",
         max_model_len=max_model_len,
     )

tests/v1/core/test_scheduler.py

@@ -1248,9 +1248,6 @@ def create_scheduler_with_priority(
     )
     model_config = ModelConfig(
         model=model,
-        task="auto",
-        tokenizer=model,
-        tokenizer_mode="auto",
         trust_remote_code=True,
         dtype="float16",
         seed=42,

tests/v1/core/utils.py

@@ -59,9 +59,6 @@ def create_scheduler(
     )
     model_config = ModelConfig(
         model=model,
-        task="auto",
-        tokenizer=model,
-        tokenizer_mode="auto",
         trust_remote_code=True,
         dtype="float16",
         seed=42,

tests/v1/kv_connector/unit/utils.py

@@ -68,9 +68,6 @@ def create_vllm_config(
     )
     model_config = ModelConfig(
         model=model,
-        task="auto",
-        tokenizer=model,
-        tokenizer_mode="auto",
         trust_remote_code=True,
         dtype="float16",
         seed=42,

tests/v1/spec_decode/test_eagle.py

@@ -24,13 +24,8 @@ eagle3_dir = "yuhuili/EAGLE3-LLaMA3.1-Instruct-8B"
 
 def _create_proposer(method: str, k: int) -> EagleProposer:
     model_config = ModelConfig(model=model_dir,
-                               task="generate",
-                               max_model_len=100,
-                               tokenizer=model_dir,
-                               tokenizer_mode="auto",
-                               dtype="auto",
-                               seed=None,
-                               trust_remote_code=False)
+                               runner="generate",
+                               max_model_len=100)
 
     # Choose model directory based on method
     draft_model_dir = eagle_dir if method == "eagle" else eagle3_dir

tests/v1/spec_decode/test_ngram.py

@@ -44,14 +44,7 @@ def test_ngram_proposer():
 
     def ngram_proposer(min_n: int, max_n: int, k: int) -> NgramProposer:
         # Dummy model config. Just to set max_model_len.
-        model_config = ModelConfig(model="facebook/opt-125m",
-                                   task="generate",
-                                   max_model_len=100,
-                                   tokenizer="facebook/opt-125m",
-                                   tokenizer_mode="auto",
-                                   dtype="auto",
-                                   seed=None,
-                                   trust_remote_code=False)
+        model_config = ModelConfig(model="facebook/opt-125m")
         return NgramProposer(
             vllm_config=VllmConfig(model_config=model_config,
                                    speculative_config=SpeculativeConfig.

tests/v1/tpu/worker/test_tpu_model_runner.py

@@ -26,10 +26,6 @@ def get_vllm_config():
     )
     model_config = ModelConfig(
         model="facebook/opt-125m",
-        task="generate",
-        tokenizer="facebook/opt-125m",
-        tokenizer_mode="auto",
-        trust_remote_code=True,
         dtype="bfloat16",  # TPUs typically use bfloat16
         seed=42,
     )

tests/v1/worker/test_gpu_model_runner.py

@@ -76,10 +76,6 @@ def get_vllm_config():
     )
     model_config = ModelConfig(
         model="facebook/opt-125m",
-        task="generate",
-        tokenizer="facebook/opt-125m",
-        tokenizer_mode="auto",
-        trust_remote_code=True,
         dtype="float16",
         seed=42,
     )