Update weekly accuracy reference

.github/ci_expected_accuracy/check_expected.py

@@ -3,6 +3,7 @@
 import pandas as pd
 import pathlib
 
+# Reference last updated is https://github.com/intel/torch-xpu-ops/pull/1223
 
 parser = argparse.ArgumentParser(description="Accuracy Check", formatter_class=argparse.ArgumentDefaultsHelpFormatter)
 parser.add_argument("--category", type=str, default="inductor", help="inductor")
@@ -67,7 +68,8 @@
     else:
         if refer_accuracy == "N/A":
             new_models.append([model_name, test_accuracy])
-            real_failed_models.append([model_name, test_accuracy])
+            # Not failed for new models
+            expected_failed_models.append([model_name, test_accuracy])
             refer_data.loc[len(refer_data),:] = "N/A"
             refer_data.at[len(refer_data) - 1, "name"] = model_name
             refer_data.at[len(refer_data) - 1, args.dtype] = test_accuracy

.github/ci_expected_accuracy/inductor_timm_models_inference.csv

@@ -59,4 +59,4 @@ twins_pcpvt_base,pass,pass,pass,pass,pass
 visformer_small,pass,pass,pass,pass,pass
 vit_base_patch16_224,pass,pass,pass,pass,pass
 volo_d1_224,pass,pass,pass,pass,pass
-xcit_large_24_p8_224,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+xcit_large_24_p8_224,pass,pass,pass,pass,pass

.github/ci_expected_accuracy/inductor_timm_models_training.csv

@@ -6,27 +6,30 @@ cait_m36_384,pass,pass,pass,pass,pass
 coat_lite_mini,pass,pass,pass,pass,pass
 convit_base,pass,pass,pass,pass,pass
 convmixer_768_32,pass,pass,pass,pass,pass
-convnext_base,pass,pass,fail_accuracy,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1274
+convnext_base,pass,fail_accuracy,fail_accuracy,pass,pass
 crossvit_9_240,pass,pass,pass,pass,pass
 cspdarknet53,pass,pass,pass,pass,pass
 deit_base_distilled_patch16_224,pass,pass,pass,pass,pass
 dla102,pass,pass,pass,pass,pass
 dm_nfnet_f0,pass,pass,pass,pass,pass
 dpn107,pass,pass,pass,pass,pass
 eca_botnext26ts_256,pass,pass,pass,pass,pass
-eca_halonext26ts,pass,pass,fail_to_run,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1275
+eca_halonext26ts,pass,pass,pass,fail_accuracy,pass
 ese_vovnet19b_dw,pass,pass,pass,pass,pass
 fbnetc_100,pass,pass,pass,pass,pass
-fbnetv3_b,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1275
+fbnetv3_b,pass,pass,pass,fail_accuracy,pass
 gernet_l,pass,pass,pass,pass,pass
 ghostnet_100,pass,pass,pass,pass,pass
 gluon_inception_v3,pass,pass,pass,pass,pass
 gmixer_24_224,pass,pass,pass,pass,pass
 gmlp_s16_224,pass,pass,pass,pass,pass
 hrnet_w18,pass,pass,pass,pass,pass
 inception_v3,pass,pass,pass,pass,pass
-jx_nest_base,pass,pass,fail_accuracy,pass,pass
-lcnet_050,pass,pass,fail_accuracy,pass,pass
+jx_nest_base,pass,pass,pass,pass,pass
+lcnet_050,pass,pass,pass,pass,pass
 levit_128,pass,pass,pass,pass,pass
 mixer_b16_224,pass,pass,pass,pass,pass
 mixnet_l,pass,pass,pass,pass,pass
@@ -59,4 +62,4 @@ twins_pcpvt_base,pass,pass,pass,pass,pass
 visformer_small,pass,pass,pass,pass,pass
 vit_base_patch16_224,pass,pass,pass,pass,pass
 volo_d1_224,pass,pass,pass,pass,pass
-xcit_large_24_p8_224,pass_due_to_skip,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+xcit_large_24_p8_224,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip

.github/ci_expected_accuracy/inductor_torchbench_inference.csv

@@ -1,9 +1,11 @@
 name,float32,bfloat16,float16,amp_bf16,amp_fp16
-torchrec_dlrm,pass,eager_fail_to_run,eager_fail_to_run,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1221
+torchrec_dlrm,pass,eager_fail_to_run,eager_fail_to_run,fail_to_run,fail_to_run
 BERT_pytorch,pass,pass,pass,pass,pass
 Background_Matting,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 DALLE2_pytorch,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
-LearningToPaint,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1263
+LearningToPaint,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 Super_SloMo,pass,pass,pass,pass,pass
 alexnet,pass,pass,pass,pass,pass
 basic_gnn_edgecnn,pass,pass,pass,pass,pass
@@ -14,6 +16,7 @@ cm3leon_generate,pass,pass,pass,pass,pass
 dcgan,pass,pass,pass,pass,pass
 demucs,pass,pass,pass,pass,pass
 densenet121,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1278
 detectron2_fasterrcnn_r_101_c4,pass,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
 detectron2_fasterrcnn_r_101_dc5,pass,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
 detectron2_fasterrcnn_r_101_fpn,pass,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
@@ -25,9 +28,9 @@ detectron2_maskrcnn,fail_to_run,eager_fail_to_run,fail_to_run,eager_fail_to_run,
 detectron2_maskrcnn_r_101_c4,fail_accuracy,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
 detectron2_maskrcnn_r_101_fpn,fail_accuracy,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
 detectron2_maskrcnn_r_50_c4,pass,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
-detectron2_maskrcnn_r_50_fpn,fail_accuracy,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
+detectron2_maskrcnn_r_50_fpn,pass,eager_fail_to_run,fail_accuracy,eager_fail_to_run,fail_accuracy
 dlrm,pass,pass,pass,pass,pass
-doctr_det_predictor,pass,pass,pass,eager_fail_to_run,pass
+doctr_det_predictor,pass,pass,pass,pass,pass
 doctr_reco_predictor,pass,pass,pass,pass,pass
 drq,pass,pass,pass,pass,pass
 fastNLP_Bert,pass,pass,pass,pass,pass
@@ -42,9 +45,11 @@ hf_DistilBert,pass,pass,pass,pass,pass
 hf_GPT2,pass,pass,pass,pass,pass
 hf_GPT2_large,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 hf_Longformer,pass,pass,pass,pass,pass
-hf_Reformer,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1262
+hf_Reformer,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ
 hf_T5,pass,pass,pass,pass,pass
-hf_T5_base,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1276
+hf_T5_base,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 hf_T5_generate,pass,pass,pass,pass,pass
 hf_T5_large,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 hf_Whisper,pass,pass,pass,pass,pass
@@ -53,22 +58,24 @@ hf_distil_whisper,pass,pass,pass,pass,pass
 lennard_jones,pass,pass,pass,pass,pass
 llama,pass,pass,pass,pass,pass
 llama_v2_7b_16h,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
-llava,eager_fail_to_run,pass,pass,eager_fail_to_run,eager_fail_to_run
+# https://github.com/intel/torch-xpu-ops/issues/1277
+llava,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 maml,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 maml_omniglot,pass,pass,pass,pass,pass
 microbench_unbacked_tolist_sum,pass,pass,pass,pass,pass
 mnasnet1_0,pass,pass,pass,pass,pass
 mobilenet_v2,pass,pass,pass,pass,pass
 mobilenet_v2_quantized_qat,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
 mobilenet_v3_large,pass,pass,pass,pass,pass
-moco,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
+moco,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 moondream,pass,pass,pass,pass,pass
 nanogpt,pass,pass,pass,pass,pass
-nvidia_deeprecommender,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1260
+nvidia_deeprecommender,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 opacus_cifar10,pass,pass,pass,pass,pass
 phlippe_densenet,pass,pass,pass,pass,pass
 phlippe_resnet,pass,pass,pass,pass,pass
-pyhpc_equation_of_state,pass,pass,fail_accuracy,pass,pass
+pyhpc_equation_of_state,pass,pass,pass,pass,pass
 pyhpc_isoneutral_mixing,pass,pass,pass,pass,pass
 pyhpc_turbulent_kinetic_energy,pass,pass,pass,pass,pass
 pytorch_CycleGAN_and_pix2pix,pass,pass,pass,pass,pass
@@ -84,11 +91,13 @@ sam_fast,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 shufflenet_v2_x1_0,pass,pass,pass,pass,pass
 simple_gpt,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
 simple_gpt_tp_manual,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
-soft_actor_critic,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1273
+soft_actor_critic,pass,fail_accuracy,pass,pass,pass
 speech_transformer,pass,pass,pass,pass,pass
-squeezenet1_1,pass,fail_accuracy,fail_accuracy,pass,pass
+squeezenet1_1,pass,fail_accuracy,pass,pass,pass
 stable_diffusion_text_encoder,pass,pass,pass,pass,pass
-stable_diffusion_unet,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
+# https://github.com/intel/torch-xpu-ops/issues/1261
+stable_diffusion_unet,eager_fail_to_run,pass_due_to_skip,pass_due_to_skip,eager_fail_to_run,eager_fail_to_run
 tacotron2,pass,pass,pass,fail_to_run,fail_to_run
 timm_efficientdet,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
 timm_efficientnet,pass,pass,pass,pass,pass
@@ -98,10 +107,9 @@ timm_resnest,pass,pass,pass,pass,pass
 timm_vision_transformer,pass,pass,pass,pass,pass
 timm_vision_transformer_large,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 timm_vovnet,pass,pass,pass,pass,pass
-torch_multimodal_clip,pass,pass,pass,eager_fail_to_run,eager_fail_to_run
-tts_angular,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+torch_multimodal_clip,pass,pass,pass,pass,pass
+tts_angular,pass,pass,pass,pass,pass
 vgg16,pass,pass,pass,pass,pass
 vision_maskrcnn,pass,pass,pass,eager_fail_to_run,eager_fail_to_run
-# Skip yolov3 for known torchbench issue: https://github.com/intel/torch-xpu-ops/issues/1229
-yolov3,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+yolov3,pass,pass,pass,pass,pass
 hf_Roberta_base,pass,pass,pass,pass,pass

.github/ci_expected_accuracy/inductor_torchbench_training.csv

@@ -1,18 +1,20 @@
 name,float32,bfloat16,float16,amp_bf16,amp_fp16
-torchrec_dlrm,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+torchrec_dlrm,pass,eager_fail_to_run,eager_fail_to_run,pass,pass
 BERT_pytorch,pass,pass,pass,pass,pass
 Background_Matting,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 DALLE2_pytorch,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
-LearningToPaint,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1263
+LearningToPaint,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+# https://github.com/intel/torch-xpu-ops/issues/1256
 Super_SloMo,eager_two_runs_differ,pass,pass,eager_two_runs_differ,pass
-alexnet,eager_two_runs_differ,pass,pass,pass,pass
+alexnet,pass,pass,pass,pass,pass
 basic_gnn_edgecnn,pass,pass,pass,pass,pass
 basic_gnn_gcn,pass,pass,pass,pass,pass
 basic_gnn_gin,pass,pass,pass,pass,pass
 basic_gnn_sage,pass,pass,pass,pass,pass
 cm3leon_generate,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 dcgan,pass,pass,pass,pass,pass
-demucs,fail_accuracy,eager_fail_to_run,eager_fail_to_run,fail_accuracy,fail_accuracy
+demucs,eager_two_runs_differ,eager_fail_to_run,eager_fail_to_run,eager_two_runs_differ,eager_two_runs_differ
 densenet121,pass,pass,pass,pass,pass
 detectron2_fasterrcnn_r_101_c4,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 detectron2_fasterrcnn_r_101_dc5,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
@@ -26,12 +28,14 @@ detectron2_maskrcnn_r_101_c4,eager_fail_to_run,eager_fail_to_run,eager_fail_to_r
 detectron2_maskrcnn_r_101_fpn,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 detectron2_maskrcnn_r_50_c4,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 detectron2_maskrcnn_r_50_fpn,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
-dlrm,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+dlrm,pass,pass,pass,pass,pass
 doctr_det_predictor,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 doctr_reco_predictor,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 drq,pass,pass,pass,pass,pass
 fastNLP_Bert,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/508
 functorch_dp_cifar10,fail_accuracy,fail_accuracy,fail_accuracy,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/510
 functorch_maml_omniglot,pass,pass,pass,fail_accuracy,pass
 hf_Albert,pass,pass,pass,pass,pass
 hf_Bart,pass,pass,pass,pass,pass
@@ -42,7 +46,8 @@ hf_DistilBert,pass,pass,pass,pass,pass
 hf_GPT2,pass,pass,pass,pass,pass
 hf_GPT2_large,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 hf_Longformer,pass,pass,pass,pass,pass
-hf_Reformer,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1262
+hf_Reformer,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ,eager_two_runs_differ
 hf_T5,pass,pass,pass,pass,pass
 hf_T5_base,pass,pass,pass,pass,pass
 hf_T5_generate,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
@@ -59,14 +64,16 @@ maml_omniglot,pass,pass,pass,pass,pass
 microbench_unbacked_tolist_sum,pass,pass,pass,pass,pass
 mnasnet1_0,pass,pass,pass,pass,pass
 mobilenet_v2,pass,pass,pass,pass,pass
-mobilenet_v2_quantized_qat,fail_accuracy,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+mobilenet_v2_quantized_qat,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 mobilenet_v3_large,pass,pass,pass,pass,pass
-moco,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
+moco,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 moondream,pass,pass,pass,pass,pass
 nanogpt,pass,pass,pass,pass,pass
-nvidia_deeprecommender,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/1260
+nvidia_deeprecommender,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 opacus_cifar10,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 phlippe_densenet,pass,pass,pass,pass,pass
+# https://github.com/intel/torch-xpu-ops/issues/509
 phlippe_resnet,pass,fail_accuracy,pass,fail_accuracy,pass
 pyhpc_equation_of_state,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 pyhpc_isoneutral_mixing,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
@@ -77,7 +84,7 @@ pytorch_unet,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 resnet152,pass,pass,pass,pass,pass
 resnet18,pass,pass,pass,pass,pass
 resnet50,pass,pass,pass,pass,pass
-resnet50_quantized_qat,eager_two_runs_differ,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+resnet50_quantized_qat,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 resnext50_32x4d,pass,pass,pass,pass,pass
 sam,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 sam_fast,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
@@ -86,22 +93,23 @@ simple_gpt,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_t
 simple_gpt_tp_manual,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
 soft_actor_critic,pass,pass,pass,pass,pass
 speech_transformer,pass,pass,pass,pass,pass
-squeezenet1_1,pass,fail_accuracy,fail_accuracy,fail_accuracy,pass
+squeezenet1_1,pass,pass,pass,pass,pass
 stable_diffusion_text_encoder,pass,pass,pass,pass,pass
-stable_diffusion_unet,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
+# https://github.com/intel/torch-xpu-ops/issues/1261
+stable_diffusion_unet,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 tacotron2,fail_to_run,fail_to_run,fail_to_run,fail_to_run,fail_to_run
 timm_efficientdet,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load,model_fail_to_load
 timm_efficientnet,pass,pass,pass,pass,pass
 timm_nfnet,pass,pass,pass,pass,pass
-timm_regnet,pass,fail_accuracy,pass,pass,pass
+timm_regnet,pass,pass,pass,pass,pass
 timm_resnest,pass,pass,pass,pass,pass
 timm_vision_transformer,pass,pass,pass,pass,pass
 timm_vision_transformer_large,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip,pass_due_to_skip
 timm_vovnet,pass,pass,pass,pass,pass
-torch_multimodal_clip,pass,pass,pass,eager_fail_to_run,eager_fail_to_run
-tts_angular,pass,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
+torch_multimodal_clip,pass,pass,pass,pass,pass
+tts_angular,pass,pass,pass,pass,pass
 vgg16,pass,pass,pass,pass,pass
-vision_maskrcnn,pass,pass,pass,eager_fail_to_run,eager_fail_to_run
-# Skip yolov3 for known torchbench issue: https://github.com/intel/torch-xpu-ops/issues/1229
+# https://github.com/intel/torch-xpu-ops/issues/1264
+vision_maskrcnn,eager_fail_to_run,pass,pass,eager_fail_to_run,eager_fail_to_run
 yolov3,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run,eager_fail_to_run
 hf_Roberta_base,pass,pass,pass,pass,pass