chore: import upstream snapshot with attribution

examples/infer/vllm/dp_tp.sh

@@ -0,0 +1,11 @@
+NPROC_PER_NODE=4 \
+CUDA_VISIBLE_DEVICES=0,1,2,3 \
+swift infer \
+    --model Qwen/Qwen2.5-7B-Instruct \
+    --infer_backend vllm \
+    --val_dataset AI-ModelScope/alpaca-gpt4-data-zh#2000 \
+    --vllm_gpu_memory_utilization 0.9 \
+    --vllm_max_model_len 8192 \
+    --vllm_tensor_parallel_size 2 \
+    --max_new_tokens 2048 \
+    --write_batch_size 1000

examples/infer/vllm/mllm_ddp.sh

@@ -0,0 +1,11 @@
+# You need to use flash-attn (manual installation) instead of xformers.
+NPROC_PER_NODE=2 \
+CUDA_VISIBLE_DEVICES=0,1 \
+swift infer \
+    --model Qwen/Qwen2.5-Omni-7B \
+    --infer_backend vllm \
+    --val_dataset speech_asr/speech_asr_aishell1_trainsets:validation#1000 \
+    --vllm_gpu_memory_utilization 0.9 \
+    --vllm_max_model_len 8192 \
+    --max_new_tokens 2048 \
+    --vllm_limit_mm_per_prompt '{"audio": 5}'

examples/infer/vllm/mllm_tp.sh

@@ -0,0 +1,11 @@
+CUDA_VISIBLE_DEVICES=0,1 \
+MAX_PIXELS=1003520 \
+swift infer \
+    --model Qwen/Qwen2.5-VL-3B-Instruct \
+    --infer_backend vllm \
+    --val_dataset AI-ModelScope/LaTeX_OCR#1000 \
+    --vllm_gpu_memory_utilization 0.9 \
+    --vllm_tensor_parallel_size 2 \
+    --vllm_max_model_len 32768 \
+    --max_new_tokens 2048 \
+    --vllm_limit_mm_per_prompt '{"image": 5, "video": 2}'

examples/infer/vllm/mtp.sh

@@ -0,0 +1,10 @@
+CUDA_VISIBLE_DEVICES=0,1,2,3 \
+swift infer \
+    --model Qwen/Qwen3-Next-80B-A3B-Instruct \
+    --vllm_tensor_parallel_size 4 \
+    --infer_backend vllm \
+    --vllm_max_model_len 8192 \
+    --val_dataset AI-ModelScope/alpaca-gpt4-data-zh#100 \
+    --vllm_speculative_config '{"method":"qwen3_next_mtp","num_speculative_tokens":2}' \
+    --vllm_gpu_memory_utilization 0.9 \
+    --max_new_tokens 2048