Support server

lightx2v/api_server.py

+import argparse
+from fastapi import FastAPI
+from pydantic import BaseModel
+import uvicorn
+import json
+
+from lightx2v.utils.profiler import ProfilingContext
+from lightx2v.utils.set_config import set_config
+from lightx2v.infer import init_runner
+
+
+class Message(BaseModel):
+    prompt: str
+    negative_prompt: str = ""
+    image_path: str = ""
+    save_video_path: str
+
+    def get(self, key, default=None):
+        return getattr(self, key, default)
+
+
+async def main(message):
+    runner.set_inputs(message)
+    runner.run_pipeline()
+    return {"response": "finished"}
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--model_cls", type=str, required=True, choices=["wan2.1", "hunyuan", "wan2.1_causal"], default="hunyuan")
+    parser.add_argument("--task", type=str, choices=["t2v", "i2v"], default="t2v")
+    parser.add_argument("--model_path", type=str, required=True)
+    parser.add_argument("--config_json", type=str, required=True)
+    parser.add_argument("--port", type=int, default=8000)
+    args = parser.parse_args()
+    print(f"args: {args}")
+
+    with ProfilingContext("Init Server Cost"):
+        config = set_config(args)
+        print(f"config:\n{json.dumps(config, ensure_ascii=False, indent=4)}")
+        runner = init_runner(config)
+
+    app = FastAPI()
+
+    @app.post("/v1/local/video/generate")
+    async def generate_video(message: Message):
+        response = await main(message)
+        return response
+
+    uvicorn.run(app, host="0.0.0.0", port=config.port)

lightx2v/__main__.py → lightx2v/infer.py

@@ -17,15 +17,30 @@ from lightx2v.models.runners.graph_runner import GraphRunner
 from lightx2v.common.ops import *
 
 
+def init_runner(config):
+    seed_all(config.seed)
+
+    if config.parallel_attn_type:
+        dist.init_process_group(backend="nccl")
+
+    if CHECK_ENABLE_GRAPH_MODE():
+        default_runner = RUNNER_REGISTER[config.model_cls](config)
+        runner = GraphRunner(default_runner)
+    else:
+        runner = RUNNER_REGISTER[config.model_cls](config)
+    return runner
+
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
     parser.add_argument("--model_cls", type=str, required=True, choices=["wan2.1", "hunyuan", "wan2.1_causal"], default="hunyuan")
     parser.add_argument("--task", type=str, choices=["t2v", "i2v"], default="t2v")
     parser.add_argument("--model_path", type=str, required=True)
+    parser.add_argument("--config_json", type=str, required=True)
+
     parser.add_argument("--prompt", type=str, required=True)
     parser.add_argument("--negative_prompt", type=str, default="")
-    parser.add_argument("--image_path", type=str, default=None, help="The path to input image file or path for image-to-video (i2v) task")
-    parser.add_argument("--config_json", type=str, required=True)
+    parser.add_argument("--image_path", type=str, default="", help="The path to input image file or path for image-to-video (i2v) task")
     parser.add_argument("--save_video_path", type=str, default="./output_lightx2v.mp4", help="The path to save video path/file")
     args = parser.parse_args()
     print(f"args: {args}")
@@ -33,15 +48,6 @@ if __name__ == "__main__":
     with ProfilingContext("Total Cost"):
         config = set_config(args)
         print(f"config:\n{json.dumps(config, ensure_ascii=False, indent=4)}")
+        runner = init_runner(config)
 
-        seed_all(config.seed)
-
-        if config.parallel_attn_type:
-            dist.init_process_group(backend="nccl")
-
-        if CHECK_ENABLE_GRAPH_MODE():
-            default_runner = RUNNER_REGISTER[config.model_cls](config)
-            runner = GraphRunner(default_runner)
-        else:
-            runner = RUNNER_REGISTER[config.model_cls](config)
         runner.run_pipeline()

lightx2v/models/runners/default_runner.py

@@ -11,6 +11,12 @@ class DefaultRunner:
         self.config = config
         self.model, self.text_encoders, self.vae_model, self.image_encoder = self.load_model()
 
+    def set_inputs(self, inputs):
+        self.config["prompt"] = inputs.get("prompt", "")
+        self.config["negative_prompt"] = inputs.get("negative_prompt", "")
+        self.config["image_path"] = inputs.get("image_path", "")
+        self.config["save_video_path"] = inputs.get("save_video_path", "")
+
     def run_input_encoder(self):
         image_encoder_output = None
         if self.config["task"] == "i2v":

lightx2v/utils/set_config.py

@@ -28,12 +28,12 @@ def set_config(args):
     config.update({k: v for k, v in vars(args).items()})
     config = EasyDict(config)
 
-    with open(args.config_json, "r") as f:
+    with open(config.config_json, "r") as f:
         config_json = json.load(f)
     config.update(config_json)
 
-    if os.path.exists(os.path.join(args.model_path, "config.json")):
-        with open(os.path.join(args.model_path, "config.json"), "r") as f:
+    if os.path.exists(os.path.join(config.model_path, "config.json")):
+        with open(os.path.join(config.model_path, "config.json"), "r") as f:
             model_config = json.load(f)
         config.update(model_config)
 

scripts/post.py

+import requests
+
+
+url = "http://localhost:8000/v1/local/video/generate"
+
+message = {
+    "prompt": "Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage.",
+    "negative_prompt": "色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    "image_path": "",
+    "save_video_path": "./output_lightx2v_wan_t2v_ap4.mp4",  # It is best to set it to an absolute path.
+}
+
+response = requests.post(url, json=message)
+
+print(f"response: {response.json()}")

scripts/run_hunyuan_i2v.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 export ENABLE_PROFILING_DEBUG=true
 export ENABLE_GRAPH_MODE=false
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls hunyuan \
 --task i2v \
 --model_path $model_path \

scripts/run_hunyuan_t2v.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 export ENABLE_PROFILING_DEBUG=true
 export ENABLE_GRAPH_MODE=false
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls hunyuan \
 --task t2v \
 --model_path $model_path \

scripts/run_hunyuan_t2v_dist.sh

@@ -27,7 +27,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 
 export ENABLE_PROFILING_DEBUG=true
 
-torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
+torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/infer.py \
 --model_cls hunyuan \
 --task t2v \
 --model_path $model_path \
@@ -35,7 +35,7 @@ torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
 --prompt "A cat walks on the grass, realistic style." \
 --save_video_path ${lightx2v_path}/save_results/output_lightx2v_hunyuan_t2v_dist_ulysses.mp4
 
-torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
+torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/infer.py \
 --model_cls hunyuan \
 --task t2v \
 --model_path $model_path \

scripts/run_hunyuan_t2v_taylorseer.sh

@@ -27,7 +27,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 
 export ENABLE_PROFILING_DEBUG=true
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls hunyuan \
 --task t2v \
 --model_path $model_path \

scripts/run_wan_i2v.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 export ENABLE_PROFILING_DEBUG=true
 export ENABLE_GRAPH_MODE=false
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls wan2.1 \
 --task i2v \
 --model_path $model_path \

scripts/run_wan_i2v_with_lora.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 
 export ENABLE_PROFILING_DEBUG=true
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls wan2.1 \
 --task i2v \
 --model_path $model_path \

scripts/run_wan_t2v.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 export ENABLE_PROFILING_DEBUG=true
 export ENABLE_GRAPH_MODE=false
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls wan2.1 \
 --task t2v \
 --model_path $model_path \

scripts/run_wan_t2v_causal.sh

@@ -28,7 +28,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 export ENABLE_PROFILING_DEBUG=true
 export ENABLE_GRAPH_MODE=false
 
-python -m lightx2v \
+python -m lightx2v.infer \
 --model_cls wan2.1_causal \
 --task t2v \
 --model_path $model_path \

scripts/run_wan_t2v_dist.sh

@@ -27,7 +27,7 @@ export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
 
 export ENABLE_PROFILING_DEBUG=true
 
-torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
+torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/infer.py \
 --model_cls wan2.1 \
 --task t2v \
 --model_path $model_path \
@@ -45,7 +45,7 @@ torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
 --parallel_vae \
 --save_video_path ${lightx2v_path}/save_results/output_lightx2v_wan_t2v_dist_ring.mp4
 
-torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/__main__.py \
+torchrun --nproc_per_node=4 ${lightx2v_path}/lightx2v/infer.py \
 --model_cls wan2.1 \
 --task t2v \
 --model_path $model_path \

scripts/start_server.sh

+#!/bin/bash
+
+# set path and first
+lightx2v_path=
+model_path=
+
+# check section
+if [ -z "${CUDA_VISIBLE_DEVICES}" ]; then
+    cuda_devices=0
+    echo "Warn: CUDA_VISIBLE_DEVICES is not set, using defalt value: ${cuda_devices}, change at shell script or set env variable."
+    export CUDA_VISIBLE_DEVICES=${cuda_devices}
+fi
+
+if [ -z "${lightx2v_path}" ]; then
+    echo "Error: lightx2v_path is not set. Please set this variable first."
+    exit 1
+fi
+
+if [ -z "${model_path}" ]; then
+    echo "Error: model_path is not set. Please set this variable first."
+    exit 1
+fi
+
+export TOKENIZERS_PARALLELISM=false
+
+export PYTHONPATH=${lightx2v_path}:$PYTHONPATH
+
+export ENABLE_PROFILING_DEBUG=true
+export ENABLE_GRAPH_MODE=false
+
+python -m lightx2v.api_server \
+--model_cls wan2.1 \
+--task t2v \
+--model_path $model_path \
+--config_json ${lightx2v_path}/configs/wan_t2v.json \
+--port 8000