Upload ms-swift/examples/export/quantize/omni/gptq.sh with huggingface_hub

ms-swift/examples/export/quantize/omni/gptq.sh

@@ -0,0 +1,18 @@
+# OMP_NUM_THREADS=14 please Check issue: https://github.com/AutoGPTQ/AutoGPTQ/issues/439
+OMP_NUM_THREADS=14 \
+CUDA_VISIBLE_DEVICES=0 \
+MAX_PIXELS=1003520 \
+VIDEO_MAX_PIXELS=50176 \
+FPS_MAX_FRAMES=12 \
+swift export \
+    --model Qwen/Qwen2.5-Omni-7B \
+    --dataset 'AI-ModelScope/alpaca-gpt4-data-zh#500' \
+              'AI-ModelScope/alpaca-gpt4-data-en#500' \
+              'modelscope/coco_2014_caption:validation#500' \
+              'swift/VideoChatGPT:Generic#500' \
+    --quant_n_samples 256 \
+    --quant_batch_size 1 \
+    --max_length 2048 \
+    --quant_method gptq \
+    --quant_bits 4 \
+    --output_dir Qwen2.5-Omni-7B-GPTQ-Int4