Upload demo.sh with huggingface_hub

demo.sh

@@ -0,0 +1,30 @@
+clear
+
+scripts/evaluate.py \
+    --task AvatarGPTEvaluator \
+    --config configs/llm_t5/config_t5_large.yaml \
+    --eval_folder demo_outputs \
+    --eval_name demo \
+    --repeat_times 1 \
+    --eval_mode se1 \
+    --eval_task t2m,m2m,m2t,ct2t,cs2s,ct2s,cs2t,t2c,s2c,t2s,s2t \
+    --eval_pipeline p0 \
+    --topk 1 \
+    --use_semantic_sampling True \
+    --temperature 1.0 \
+    --demo_list demo_inputs/test.txt \
+    --demo_data demo_inputs/samples
+
+render/convert_hml_to_mesh.py \
+    --filedir demo_outputs/demo/output/se1_p0/ \
+    --mesh_dir demo_outputs/demo/meshes/se1_p0/ \
+    --video_dir demo_outputs/demo/videos/se1_p0/ \
+    --visualize False \
+    --convert_gt False
+
+render/mesh_renderer.py \
+    --mesh_dir demo_outputs/demo/meshes/se1_p0/ \
+    --video_dir demo_outputs/demo/videos/se1_p0/ \
+    --run_mode dynamic \
+    --sample_rate 4 \
+    --num_frame 20