feat: bump transformers to allow Qwen3-VL (#225)

remove unsupported cuda versions (12.1-12.3) from hub.json and tests.json
to fix compatibility issues with worker deployment

- hub.json: remove 12.1, 12.2, 12.3 from allowedCudaVersions
- tests.json: remove 12.1, 12.2, 12.3, 12.4 from allowedCudaVersions

refs: AE-1452

.runpod/README.md

@@ -24,6 +24,7 @@ All behaviour is controlled through environment variables:
 | `CUSTOM_CHAT_TEMPLATE`              | Custom chat template override                     |                     | Jinja2 template string                                             |
 | `ENABLE_AUTO_TOOL_CHOICE`           | Enable automatic tool selection                   | false               | boolean (true or false)                                            |
 | `TOOL_CALL_PARSER`                  | Parser for tool calls                             |                     | "mistral", "hermes", "llama3_json", "granite", "deepseek_v3", etc. |
+| `REASONING_PARSER`                  | Parser for reasoning-capable models               |                     | "deepseek_r1", "qwen3", "granite", "hunyuan_a13b"                  |
 | `OPENAI_SERVED_MODEL_NAME_OVERRIDE` | Override served model name in API                 |                     | String                                                             |
 | `MAX_CONCURRENCY`                   | Maximum concurrent requests                       | 300                 | Integer                                                            |
 

.runpod/hub.json

@@ -1,25 +1,15 @@
 {
   "title": "vLLM",
-  "description": "Deploy OpenAI-Compatible Blazing-Fast LLM Endpoints powered by the vLLM Inference Engine on RunPod Serverless",
+  "description": "Deploy OpenAI-Compatible Blazing-Fast LLM Endpoints powered by vLLM",
   "type": "serverless",
   "category": "language",
   "iconUrl": "https://registry.npmmirror.com/@lobehub/icons-static-png/latest/files/dark/vllm-color.png",
   "config": {
     "runsOn": "GPU",
-    "containerDiskInGb": 200,
-    "gpuIds": "ADA_80_PRO, AMPERE_80",
+    "containerDiskInGb": 150,
+    "gpuIds": "ADA_80_PRO,AMPERE_80",
     "gpuCount": 1,
-    "allowedCudaVersions": [
-      "12.9",
-      "12.8",
-      "12.7",
-      "12.6",
-      "12.5",
-      "12.4",
-      "12.3",
-      "12.2",
-      "12.1"
-    ],
+    "allowedCudaVersions": ["12.9", "12.8", "12.7", "12.6", "12.5", "12.4"],
     "presets": [
       {
         "name": "deepseek-ai/deepseek-r1-distill-llama-8b",
@@ -38,16 +28,6 @@
           "required": true
         }
       },
-      {
-        "key": "HF_TOKEN",
-        "input": {
-          "name": "Access Token",
-          "type": "string",
-          "description": "Hugging Face access token for gated & private models",
-          "default": "",
-          "required": false
-        }
-      },
       {
         "key": "TOKENIZER",
         "input": {
@@ -945,7 +925,7 @@
           "name": "Max Concurrency",
           "type": "number",
           "description": "Max concurrent requests per worker. vLLM has an internal queue, so you don't have to worry about limiting by VRAM, this is for improving scaling/load balancing efficiency",
-          "default": 300,
+          "default": 30,
           "advanced": true
         }
       },
@@ -1023,6 +1003,23 @@
           "default": "",
           "advanced": true
         }
+      },
+      {
+        "key": "REASONING_PARSER",
+        "input": {
+          "name": "Reasoning Parser",
+          "type": "string",
+          "description": "Parser for reasoning-capable models (enables reasoning mode)",
+          "options": [
+            { "label": "None", "value": "" },
+            { "label": "DeepSeek R1", "value": "deepseek_r1" },
+            { "label": "Qwen3", "value": "qwen3" },
+            { "label": "Granite", "value": "granite" },
+            { "label": "Hunyuan A13B", "value": "hunyuan_a13b" }
+          ],
+          "default": "",
+          "advanced": true
+        }
       }
     ]
   }

.runpod/tests.json

@@ -38,14 +38,6 @@
         "value": "HuggingFaceTB/SmolLM2-135M-Instruct"
       }
     ],
-    "allowedCudaVersions": [
-      "12.7",
-      "12.6",
-      "12.5",
-      "12.4",
-      "12.3",
-      "12.2",
-      "12.1"
-    ]
+    "allowedCudaVersions": ["12.9", "12.8", "12.7", "12.6", "12.5"]
   }
 }

Dockerfile

@@ -12,7 +12,7 @@ RUN --mount=type=cache,target=/root/.cache/pip \
     python3 -m pip install --upgrade -r /requirements.txt
 
 # Install vLLM (switching back to pip installs since issues that required building fork are fixed and space optimization is not as important since caching) and FlashInfer 
-RUN python3 -m pip install vllm==0.10.0 && \
+RUN python3 -m pip install vllm==0.11.0 && \
     python3 -m pip install flashinfer -i https://flashinfer.ai/whl/cu121/torch2.3
 
 # Setup for Option 2: Building the Image with the Model included

README.md

@@ -57,7 +57,7 @@ Configure worker-vllm using environment variables:
 | `ENABLE_AUTO_TOOL_CHOICE`           | Enable automatic tool selection                   | false               | boolean (true or false)                                            |
 | `TOOL_CALL_PARSER`                  | Parser for tool calls                             |                     | "mistral", "hermes", "llama3_json", "granite", "deepseek_v3", etc. |
 | `OPENAI_SERVED_MODEL_NAME_OVERRIDE` | Override served model name in API                 |                     | String                                                             |
-| `MAX_CONCURRENCY`                   | Maximum concurrent requests                       | 300                 | Integer                                                            |
+| `MAX_CONCURRENCY`                   | Maximum concurrent requests                       | 30                  | Integer                                                            |
 
 For the complete list of all available environment variables, examples, and detailed descriptions: **[Configuration](docs/configuration.md)**
 

builder/requirements.txt

@@ -8,7 +8,7 @@ typing-extensions>=4.8.0
 pydantic
 pydantic-settings
 hf-transfer
-transformers>=4.55.0
+transformers>=4.57.0
 bitsandbytes>=0.45.0
 kernels
 torch==2.6.0

docs/configuration.md

@@ -113,12 +113,13 @@ The way this works is that the first request will have a batch size of `DEFAULT_
 | `OPENAI_RESPONSE_ROLE`              | `assistant` | `str`            | Role of the LLM's Response in OpenAI Chat Completions.                                                                                                                                                            |
 | `ENABLE_AUTO_TOOL_CHOICE`           | `false`     | `bool`           | Enables automatic tool selection for supported models. Set to `true` to activate.                                                                                                                                 |
 | `TOOL_CALL_PARSER`                  | `None`      | `str`            | Specifies the parser for tool calls. Options: `mistral`, `hermes`, `llama3_json`, `llama4_json`, `llama4_pythonic`, `granite`, `granite-20b-fc`, `deepseek_v3`, `internlm`, `jamba`, `phi4_mini_json`, `pythonic` |
+| `REASONING_PARSER`                  | `None`      | `str`            | Parser for reasoning-capable models (enables reasoning mode). Examples: `deepseek_r1`, `qwen3`, `granite`, `hunyuan_a13b`. Leave unset to disable.                                                                |
 
 ## Serverless & Concurrency Settings
 
 | Variable               | Default | Type/Choices | Description                                                                                                                                                                |
 | ---------------------- | ------- | ------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `MAX_CONCURRENCY`      | `300`   | `int`        | Max concurrent requests per worker. vLLM has an internal queue, so you don't have to worry about limiting by VRAM, this is for improving scaling/load balancing efficiency |
+| `MAX_CONCURRENCY`      | `30`    | `int`        | Max concurrent requests per worker. vLLM has an internal queue, so you don't have to worry about limiting by VRAM, this is for improving scaling/load balancing efficiency |
 | `DISABLE_LOG_STATS`    | False   | `bool`       | Enables or disables vLLM stats logging.                                                                                                                                    |
 | `DISABLE_LOG_REQUESTS` | False   | `bool`       | Enables or disables vLLM request logging.                                                                                                                                  |
 

docs/conventions.md

@@ -51,7 +51,8 @@ RunPod Request → handler.py → JobInput → Engine Selection → vLLM Generat
 
 - `src/engine_args.py`: Centralized configuration management
 - `src/constants.py`: Default values for core settings
-- `worker-config.json`: UI form generation for RunPod console
+- `.runpod/hub.json`: Hub UI configuration (CRITICAL: always update when changing defaults)
+- `worker-config.json`: UI form generation for RunPod console (if exists)
 
 ## Core Development Concepts
 
@@ -222,7 +223,7 @@ src/
 
 ### 2. **Concurrency Patterns**
 
-- **Max Concurrency**: 300 concurrent requests by default
+- **Max Concurrency**: 30 concurrent requests by default
 - **vLLM Queuing**: Internal request batching and scheduling
 - **RunPod Integration**: Concurrency modifier for auto-scaling
 

src/constants.py

@@ -1,4 +1,4 @@
 DEFAULT_BATCH_SIZE = 50
-DEFAULT_MAX_CONCURRENCY = 300
+DEFAULT_MAX_CONCURRENCY = 30
 DEFAULT_BATCH_SIZE_GROWTH_FACTOR = 3
 DEFAULT_MIN_BATCH_SIZE = 1