Upload folder using huggingface_hub

.gitattributes

@@ -147,3 +147,10 @@ sft/1M3/Full_smoe_share/logs/0626_1043_llava_v1.5_gqa_llava_model_args_d02902/gq
 sft/1M3/Full_smoe_share/checkpoint-27572/trainer_state.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_share/checkpoint-34462/trainer_state.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_share/checkpoint-20679/trainer_state.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/mmbench_en_dev.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/mmerealworld_lite.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/pope.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/textvqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_sigmoidgating/logs/0626_1111_llava_v1.5_gqa_llava_model_args_0a2586/gqa.json filter=lfs diff=lfs merge=lfs -text

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/mmbench_en_dev.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:391bb5a6dfd251a405d86027d2dde4b5a2836744ba3159c3a5d90fbb6b153269
+size 14568198

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/mmerealworld_lite.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0abf054f869f32af15a13fa4b8df4c994079f0c26fa0f76372a9ae55559d51b8
+size 1994104231

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/pope.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c12e8dc6e4e1229e02dec7f36f200ee3335070e1d5a10fe941ea06ac00a62dc2
+size 17598180

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/rank1_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 1 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/rank2_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 2 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/rank3_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 3 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/results.json

@@ -0,0 +1,488 @@
+{
+    "results": {
+        "ai2d": {
+            "exact_match,flexible-extract": 0.6881476683937824,
+            "exact_match_stderr,flexible-extract": 0.008337714200821831,
+            "alias": "ai2d"
+        },
+        "mmbench_en_dev": {
+            "gpt_eval_score,none": 70.1890034364261,
+            "gpt_eval_score_stderr,none": "N/A",
+            "submission,none": null,
+            "submission_stderr,none": "N/A",
+            "alias": "mmbench_en_dev"
+        },
+        "mmerealworld_lite": {
+            "mme_realworld_score,none": 0.3418447107868682,
+            "mme_realworld_score_stderr,none": "N/A",
+            "alias": "mmerealworld_lite"
+        },
+        "mmmu_val": {
+            "mmmu_acc,none": 0.42778,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": "mmmu_val"
+        },
+        "mmstar": {
+            "coarse perception,none": 0.6973701963406836,
+            "coarse perception_stderr,none": "N/A",
+            "fine-grained perception,none": 0.34010071235568656,
+            "fine-grained perception_stderr,none": "N/A",
+            "instance reasoning,none": 0.536952831443589,
+            "instance reasoning_stderr,none": "N/A",
+            "logical reasoning,none": 0.36768523006146764,
+            "logical reasoning_stderr,none": "N/A",
+            "math,none": 0.3951929074222578,
+            "math_stderr,none": "N/A",
+            "science & technology,none": 0.2946540428415929,
+            "science & technology_stderr,none": "N/A",
+            "alias": "mmstar"
+        },
+        "pope": {
+            "pope_accuracy,none": 0.576,
+            "pope_accuracy_stderr,none": "N/A",
+            "pope_precision,none": 0.8904261206419479,
+            "pope_precision_stderr,none": "N/A",
+            "pope_recall,none": 0.8404283102637764,
+            "pope_recall_stderr,none": "N/A",
+            "pope_f1_score,none": 0.8647050920327825,
+            "pope_f1_score_stderr,none": "N/A",
+            "pope_yes_ratio,none": 0.5,
+            "pope_yes_ratio_stderr,none": "N/A",
+            "alias": "pope"
+        },
+        "textvqa_val": {
+            "exact_match,none": 0.4207800002217293,
+            "exact_match_stderr,none": 0.006767274000916751,
+            "submission,none": null,
+            "submission_stderr,none": "N/A",
+            "alias": "textvqa_val"
+        }
+    },
+    "configs": {
+        "ai2d": {
+            "task": "ai2d",
+            "dataset_path": "lmms-lab/ai2d",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function ai2d_doc_to_visual at 0x7f1c203e4280>",
+            "doc_to_text": "<function ai2d_doc_to_text at 0x7f1c203e4550>",
+            "doc_to_target": "<function ai2d_doc_to_target at 0x7f1c203e4af0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 512,
+                "temperature": 0.0,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "filter_list": [
+                {
+                    "name": "flexible-extract",
+                    "filter": [
+                        {
+                            "function": "<class 'utils.MultiChoiceRegexFilter'>",
+                            "group_select": 0,
+                            "ignore_case": true,
+                            "ignore_punctuation": true,
+                            "regex_pattern": "([A-Z])\\."
+                        }
+                    ]
+                }
+            ],
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "prompt_format": "mcq",
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer with the option's letter from the given choices directly."
+                },
+                "gpt4v": {
+                    "prompt_format": "mcq",
+                    "pre_prompt": "",
+                    "post_prompt": "\nAbove choices are given in {option}. {content} format.\nPlease answer with the option letter from the given choices directly."
+                },
+                "qwen_vl": {
+                    "prompt_format": "qa",
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                },
+                "xcomposer2_4khd": {
+                    "prompt_format": "mcq_xcomposer",
+                    "pre_prompt": "[UNUSED_TOKEN_146]user\nQuestion: ",
+                    "post_prompt": "[UNUSED_TOKEN_145]\n[UNUSED_TOKEN_146]assistant\nThe answer is"
+                }
+            },
+            "model_specific_target_kwargs": {
+                "default": "mcq",
+                "qwen_vl": "qa"
+            }
+        },
+        "mmbench_en_dev": {
+            "task": "mmbench_en_dev",
+            "dataset_path": "lmms-lab/MMBench",
+            "dataset_name": "en",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "dev",
+            "doc_to_visual": "<function mmbench_doc_to_visual at 0x7f1c7b5c2c10>",
+            "doc_to_text": "<function mmbench_doc_to_text at 0x7f1c7b5cd160>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmbench_process_results at 0x7f1c7b5cd670>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mmbench_aggregate_dev_results_eval at 0x7f1c7b5c24c0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "submission",
+                    "aggregation": "<function mmbench_aggregate_dev_results_submission at 0x7f1c7b5c29d0>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "until": [
+                    "ASSISTANT:"
+                ],
+                "max_new_tokens": 1024,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "image_aspect_ratio": "original"
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer with the option's letter from the given choices directly."
+                }
+            },
+            "model_specific_generation_kwargs": {
+                "llava": {
+                    "image_aspect_ratio": "original"
+                }
+            }
+        },
+        "mmerealworld_lite": {
+            "task": "mmerealworld_lite",
+            "dataset_path": "yifanzhang114/MME-RealWorld-lite-lmms-eval",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "train",
+            "doc_to_visual": "<function mme_realworld_doc_to_visual at 0x7f1c20b89430>",
+            "doc_to_text": "<function mme_realworld_doc_to_text at 0x7f1c20b89ee0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_realworld_process_results at 0x7f1c20b8eb80>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_realworld_score",
+                    "aggregation": "<function mme_realworld_aggregate_results at 0x7f1c20b93700>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nSelect the best answer to the above multiple-choice question based on the image. Respond with only the letter (A, B, C, D, or E) of the correct option."
+                },
+                "gpt4v": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nSelect the best answer to the above multiple-choice question based on the image. Respond with only the letter (A, B, C, D, or E) of the correct option."
+                },
+                "xcomposer2_4khd": {
+                    "pre_prompt": "[UNUSED_TOKEN_146]user\n",
+                    "post_prompt": " Answer this question with A, B, C, or D.[UNUSED_TOKEN_145]\n[UNUSED_TOKEN_146]assistant\n"
+                }
+            }
+        },
+        "mmmu_val": {
+            "task": "mmmu_val",
+            "dataset_path": "lmms-lab/MMMU",
+            "test_split": "validation",
+            "doc_to_visual": "<function mmmu_doc_to_visual at 0x7f1c32d90ee0>",
+            "doc_to_text": "<function mmmu_doc_to_text at 0x7f1c32c91c10>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmmu_process_results at 0x7f1c32c32af0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_aggregate_results at 0x7f1c32bb4a60>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 128,
+                "until": [
+                    "\n\n"
+                ],
+                "image_aspect_ratio": "original"
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_generation_kwargs": {
+                "llava": {
+                    "image_aspect_ratio": "original"
+                }
+            }
+        },
+        "mmstar": {
+            "task": "mmstar",
+            "dataset_path": "Lin-Chen/MMStar",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "val",
+            "doc_to_visual": "<function mmstar_doc_to_visual at 0x7f1c2033d8b0>",
+            "doc_to_text": "<function mmstar_doc_to_text at 0x7f1c2033dd30>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmstar_process_results at 0x7f1c20343280>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "coarse perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c20343700>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fine-grained perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c20343af0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "instance reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c20343ee0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "logical reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c2034a310>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "science & technology",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c2034a700>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "math",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f1c2034aaf0>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "until": [
+                    "\n\n"
+                ],
+                "do_sample": false
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer with the option's letter from the given choices directly"
+                }
+            }
+        },
+        "pope": {
+            "task": "pope",
+            "dataset_path": "lmms-lab/POPE",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function pope_doc_to_visual at 0x7f1c20c20ee0>",
+            "doc_to_text": "<function pope_doc_to_text at 0x7f1c20c254c0>",
+            "doc_to_target": "answer",
+            "process_results": "<function pope_process_results at 0x7f1c20c25a60>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "pope_accuracy",
+                    "aggregation": "<function pope_aggregate_accuracy at 0x7f1c20c29040>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_precision",
+                    "aggregation": "<function pope_aggregate_precision at 0x7f1c20c295e0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_recall",
+                    "aggregation": "<function pope_aggregate_recall at 0x7f1c20c29b80>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_f1_score",
+                    "aggregation": "<function pope_aggregate_f1_score at 0x7f1c20c2e160>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_yes_ratio",
+                    "aggregation": "<function pope_aggregate_yes_ratio at 0x7f1c20c2e700>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 128,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ]
+        },
+        "textvqa_val": {
+            "task": "textvqa_val",
+            "dataset_path": "lmms-lab/textvqa",
+            "test_split": "validation",
+            "doc_to_visual": "<function textvqa_doc_to_visual at 0x7f1c4cf59f70>",
+            "doc_to_text": "<function textvqa_doc_to_text at 0x7f1c4cf8a430>",
+            "doc_to_target": "answer",
+            "process_results": "<function textvqa_process_results at 0x7f1c4cf8a700>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                },
+                {
+                    "metric": "submission",
+                    "aggregation": "<function textvqa_aggregate_submissions at 0x7f1c4cf59d30>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "until": [
+                    "ASSISTANT:"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase.",
+                    "ocr": false
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                }
+            }
+        }
+    },
+    "versions": {
+        "ai2d": "Yaml",
+        "mmbench_en_dev": "Yaml",
+        "mmerealworld_lite": "Yaml",
+        "mmmu_val": "Yaml",
+        "mmstar": "Yaml",
+        "pope": "Yaml",
+        "textvqa_val": "Yaml"
+    },
+    "n-shot": {
+        "ai2d": 0,
+        "mmbench_en_dev": 0,
+        "mmerealworld_lite": 0,
+        "mmmu_val": 0,
+        "mmstar": 0,
+        "pope": 0,
+        "textvqa_val": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/submissions/mmbench_en_dev_results.json

@@ -0,0 +1 @@
+{"overall_acc": 0.7018900343642611, "category_acc": {"action_recognition": 0.8888888888888888, "attribute_comparison": 0.7045454545454546, "attribute_recognition": 0.9054054054054054, "celebrity_recognition": 0.7373737373737373, "function_reasoning": 0.8734177215189873, "future_prediction": 0.55, "identity_reasoning": 0.9777777777777777, "image_emotion": 0.84, "image_quality": 0.37735849056603776, "image_scene": 0.9423076923076923, "image_style": 0.9056603773584906, "image_topic": 0.8888888888888888, "nature_relation": 0.5833333333333334, "object_localization": 0.38271604938271603, "ocr": 0.6666666666666666, "physical_property_reasoning": 0.44, "physical_relation": 0.5833333333333334, "social_relation": 0.8604651162790697, "spatial_relationship": 0.2222222222222222, "structuralized_imagetext_understanding": 0.5641025641025641}, "l2_category_acc": {"attribute_reasoning": 0.7336683417085427, "coarse_perception": 0.8108108108108109, "finegrained_perception (cross-instance)": 0.6223776223776224, "finegrained_perception (instance-level)": 0.6723549488054608, "logic_reasoning": 0.559322033898305, "relation_reasoning": 0.6869565217391305}}

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/submissions/mmbench_en_dev_results.xlsx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:141930c19fa1b731ed0b56f1335ba844dbee7ff4320a980e63f8023c2a8b4f96
+size 865128

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2318_llava..._pope_llava_model_args_0a2586/textvqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f4c579bc06d043373aa46e4d11ef3bc70e54547e14ada9074d02165f9a76690
+size 13163093

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/rank1_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 1 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/rank2_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 2 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/rank3_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 3 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/results.json

@@ -0,0 +1,67 @@
+{
+    "results": {
+        "ocrbench": {
+            "ocrbench_accuracy,none": 0.402,
+            "ocrbench_accuracy_stderr,none": "N/A",
+            "alias": "ocrbench"
+        }
+    },
+    "configs": {
+        "ocrbench": {
+            "task": "ocrbench",
+            "dataset_path": "echo840/OCRBench",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function ocrbench_doc_to_visual at 0x7f2844d26310>",
+            "doc_to_text": "<function ocrbench_doc_to_text at 0x7f28449f18b0>",
+            "doc_to_target": "answer",
+            "process_results": "<function ocrbench_process_results at 0x7f28449f1c10>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "ocrbench_accuracy",
+                    "aggregation": "<function ocrbench_aggregate_accuracy at 0x7f28449f1f70>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 128,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ]
+        }
+    },
+    "versions": {
+        "ocrbench": "Yaml"
+    },
+    "n-shot": {
+        "ocrbench": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_sigmoidgating/logs/0616_2345_llava...bench_llava_model_args_0a2586/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 214
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 45
+Irregular Text Recognition(Total 50): 45
+Artistic Text Recognition(Total 50): 43
+Handwriting Recognition(Total 50): 18
+Digit String Recognition(Total 50): 36
+Non-Semantic Text Recognition(Total 50): 27
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 107
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 25
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 18
+Handwritten Mathematical Expression Recognition(Total 100): 38
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 402

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0312_llava...l_mme_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0312_llava...l_mme_llava_model_args_0a2586/rank1_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 1 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0312_llava...l_mme_llava_model_args_0a2586/rank2_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 2 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0312_llava...l_mme_llava_model_args_0a2586/rank3_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 3 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0312_llava...l_mme_llava_model_args_0a2586/results.json

@@ -0,0 +1,215 @@
+{
+    "results": {
+        "docvqa_val": {
+            "anls,none": 0.27911759207328474,
+            "anls_stderr,none": 0.006133807176340855,
+            "alias": "docvqa_val"
+        },
+        "mme": {
+            "mme_cognition_score,none": 365.0,
+            "mme_cognition_score_stderr,none": "N/A",
+            "mme_percetion_score,none": 1283.9111644657862,
+            "mme_percetion_score_stderr,none": "N/A",
+            "alias": "mme"
+        },
+        "scienceqa_img": {
+            "exact_match,none": 0.8284581060981656,
+            "exact_match_stderr,none": 0.008396051856378862,
+            "alias": "scienceqa_img"
+        }
+    },
+    "configs": {
+        "docvqa_val": {
+            "task": "docvqa_val",
+            "dataset_path": "lmms-lab/DocVQA",
+            "dataset_name": "DocVQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "validation",
+            "doc_to_visual": "<function docvqa_doc_to_visual at 0x7fab9dd96af0>",
+            "doc_to_text": "<function docvqa_doc_to_text at 0x7fab9dd96e50>",
+            "doc_to_target": "answers",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "anls",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 32,
+                "temperature": 0.0,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                }
+            }
+        },
+        "mme": {
+            "task": "mme",
+            "dataset_path": "lmms-lab/MME",
+            "dataset_kwargs": {
+                "token": false
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function mme_doc_to_visual at 0x7fabc86e0040>",
+            "doc_to_text": "<function mme_doc_to_text at 0x7fabc86e0790>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_process_results at 0x7fabc86e0d30>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_percetion_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7fabc6cd3280>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "mme_cognition_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7fabc6cd3700>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                },
+                "gpt4v": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question with Yes or No."
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                },
+                "otterhd": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                },
+                "xcomposer2_4khd": {
+                    "pre_prompt": "[UNUSED_TOKEN_146]user\n",
+                    "post_prompt": " Answer this question briefly[UNUSED_TOKEN_145]\n[UNUSED_TOKEN_146]assistant\n"
+                }
+            }
+        },
+        "scienceqa_img": {
+            "task": "scienceqa_img",
+            "dataset_path": "lmms-lab/ScienceQA",
+            "dataset_name": "ScienceQA-IMG",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function sqa_doc_to_visual at 0x7fab928ce280>",
+            "doc_to_text": "<function sqa_doc_to_text at 0x7fab928ce4c0>",
+            "doc_to_target": "<function sqa_doc_to_target at 0x7fab928ce8b0>",
+            "process_results": "<function sqa_process_results at 0x7fab928cec10>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ],
+                "image_aspect_ratio": "original"
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "format": "default",
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer with the option's letter from the given choices directly."
+                },
+                "qwen_vl": {
+                    "format": "qwen_vl"
+                },
+                "idefics2": {
+                    "format": "default",
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer:"
+                }
+            },
+            "model_specific_generation_kwargs": {
+                "llava": {
+                    "image_aspect_ratio": "original"
+                }
+            }
+        }
+    },
+    "versions": {
+        "docvqa_val": "Yaml",
+        "mme": "Yaml",
+        "scienceqa_img": "Yaml"
+    },
+    "n-shot": {
+        "docvqa_val": 0,
+        "mme": 0,
+        "scienceqa_img": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/infovqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79ae48adbc1028ff47ec51f89fda6585da1731315a941b15f7487e5f18877fa6
+size 576438848

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/rank1_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 1 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/rank2_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 2 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/rank3_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 3 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0617_0323_llava...a_val_llava_model_args_0a2586/results.json

@@ -0,0 +1,245 @@
+{
+    "results": {
+        "infovqa_val": {
+            "anls,none": 0.31453052481256694,
+            "anls_stderr,none": 0.008774986238770308,
+            "alias": "infovqa_val"
+        },
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.20259999999999997,
+            "mmmu_acc_stderr,none": 0.035410117438777894,
+            "alias": "mmmu_pro"
+        },
+        "mmmu_pro_standard": {
+            "mmmu_acc,none": 0.27341,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_standard"
+        },
+        "mmmu_pro_vision": {
+            "mmmu_acc,none": 0.13179,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_vision"
+        },
+        "realworldqa": {
+            "exact_match,flexible-extract": 0.4954248366013072,
+            "exact_match_stderr,flexible-extract": 0.018088607818930215,
+            "alias": "realworldqa"
+        }
+    },
+    "groups": {
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.20259999999999997,
+            "mmmu_acc_stderr,none": 0.035410117438777894,
+            "alias": "mmmu_pro"
+        }
+    },
+    "configs": {
+        "infovqa_val": {
+            "task": "infovqa_val",
+            "dataset_path": "lmms-lab/DocVQA",
+            "dataset_name": "InfographicVQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "validation",
+            "doc_to_visual": "<function infovqa_doc_to_visual at 0x7ff657b2d0d0>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7ff657b2d3a0>",
+            "doc_to_target": "answers",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "anls",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 32,
+                "temperature": 0.0,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                }
+            }
+        },
+        "mmmu_pro_standard": {
+            "task": "mmmu_pro_standard",
+            "dataset_path": "MMMU/MMMU_Pro",
+            "dataset_name": "standard (10 options)",
+            "test_split": "test",
+            "doc_to_visual": "<function mmmu_pro_doc_to_visual at 0x7ff684403dc0>",
+            "doc_to_text": "<function mmmu_pro_doc_to_text at 0x7ff68440db80>",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7ff684414af0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7ff68441ba60>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 256,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": {
+                "version": 0.0,
+                "interleaved_format": false
+            },
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "Answer with the option letter from the given choices directly."
+                }
+            }
+        },
+        "mmmu_pro_vision": {
+            "task": "mmmu_pro_vision",
+            "dataset_path": "MMMU/MMMU_Pro",
+            "dataset_name": "vision",
+            "test_split": "test",
+            "doc_to_visual": "<function mmmu_pro_doc_to_visual at 0x7ff6843f8310>",
+            "doc_to_text": "Answer with the option letter from the given choices directly.",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7ff6843fe1f0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7ff684403160>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 256,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": {
+                "version": 0.0,
+                "interleaved_format": false
+            }
+        },
+        "realworldqa": {
+            "task": "realworldqa",
+            "dataset_path": "lmms-lab/RealWorldQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function realworldqa_doc_to_visual at 0x7ff65023d160>",
+            "doc_to_text": "<function realworldqa_doc_to_text at 0x7ff65023d790>",
+            "doc_to_target": "answer",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "filter_list": [
+                {
+                    "name": "flexible-extract",
+                    "filter": [
+                        {
+                            "function": "<class 'utils.NumberWordsToDigitsFilter'>"
+                        },
+                        {
+                            "function": "<class 'utils.MultiChoiceRegexFilter'>",
+                            "group_select": 0,
+                            "ignore_case": true,
+                            "ignore_punctuation": true,
+                            "regex_pattern": "(\\([A-Z]\\))"
+                        }
+                    ]
+                }
+            ],
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": ""
+                },
+                "gpt4v": {
+                    "pre_prompt": "",
+                    "post_prompt": ""
+                },
+                "xcomposer2_4khd": {
+                    "pre_prompt": "[UNUSED_TOKEN_146]user\nQuestion: ",
+                    "post_prompt": "[UNUSED_TOKEN_145]\n[UNUSED_TOKEN_146]assistant\nThe answer is"
+                }
+            }
+        }
+    },
+    "versions": {
+        "infovqa_val": "Yaml",
+        "mmmu_pro": "N/A",
+        "mmmu_pro_standard": "Yaml",
+        "mmmu_pro_vision": "Yaml",
+        "realworldqa": "Yaml"
+    },
+    "n-shot": {
+        "infovqa_val": 0,
+        "mmmu_pro": 0,
+        "mmmu_pro_standard": 0,
+        "mmmu_pro_vision": 0,
+        "realworldqa": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1111_llava_v1.5_gqa_llava_model_args_0a2586/gqa.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:072c87d73b5a130cc18c1cf7b0615f8380e32628474896001cdd4303799d1cd8
+size 38381569

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1111_llava_v1.5_gqa_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1111_llava_v1.5_gqa_llava_model_args_0a2586/results.json

@@ -0,0 +1,79 @@
+{
+    "results": {
+        "gqa": {
+            "exact_match,none": 0.4721736365081889,
+            "exact_match_stderr,none": 0.004451515358552109,
+            "alias": "gqa"
+        }
+    },
+    "configs": {
+        "gqa": {
+            "task": "gqa",
+            "dataset_path": "lmms-lab/GQA",
+            "dataset_name": "testdev_balanced_instructions",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "testdev",
+            "doc_to_visual": "<function gqa_doc_to_visual at 0x7f05e79519d0>",
+            "doc_to_text": "<function gqa_doc_to_text at 0x7f05e78a4ca0>",
+            "doc_to_target": "answer",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                }
+            }
+        }
+    },
+    "versions": {
+        "gqa": "Yaml"
+    },
+    "n-shot": {
+        "gqa": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1141_llava...image_llava_model_args_0a2586/rank0_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 0 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1141_llava...image_llava_model_args_0a2586/rank1_metric_eval_done.txt

@@ -0,0 +1 @@
+rank 1 eval done

sft/1M3/Full_smoe_sigmoidgating/logs/0626_1141_llava...image_llava_model_args_0a2586/results.json

@@ -0,0 +1,146 @@
+{
+    "results": {
+        "hallusion_bench_image": {
+            "aAcc,none": 48.0547,
+            "aAcc_stderr,none": "N/A",
+            "fAcc,none": 23.1214,
+            "fAcc_stderr,none": "N/A",
+            "qAcc,none": 19.7802,
+            "qAcc_stderr,none": "N/A",
+            "alias": "hallusion_bench_image"
+        },
+        "mathvista_testmini": {
+            "gpt_eval_score,none": 39.5,
+            "gpt_eval_score_stderr,none": "N/A",
+            "alias": "mathvista_testmini"
+        }
+    },
+    "configs": {
+        "hallusion_bench_image": {
+            "task": "hallusion_bench_image",
+            "dataset_path": "lmms-lab/HallusionBench",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "image",
+            "doc_to_visual": "<function hb_doc_to_visual at 0x7f0272d271f0>",
+            "doc_to_text": "<function hb_doc_to_text at 0x7f0272d27940>",
+            "doc_to_target": "gt_answer_details",
+            "process_results": "<function hb_process_results at 0x7f0272d32160>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "aAcc",
+                    "aggregation": "<function hb_aggregation_result_aAcc at 0x7f0272d32af0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "qAcc",
+                    "aggregation": "<function hb_aggregation_result_qAcc at 0x7f0272d09160>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fAcc",
+                    "aggregation": "<function hb_aggregation_result_fAcc at 0x7f0272d09940>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 128,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": ""
+                }
+            }
+        },
+        "mathvista_testmini": {
+            "task": "mathvista_testmini",
+            "dataset_path": "AI4Math/MathVista",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "testmini",
+            "doc_to_visual": "<function mathvista_doc_to_visual at 0x7f0248e66c10>",
+            "doc_to_text": "<function mathvista_doc_to_text at 0x7f0248a60310>",
+            "doc_to_target": "answer",
+            "process_results": "<function mathvista_process_results at 0x7f0248a699d0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mathvista_aggregate_results at 0x7f0248a730d0>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "until": [
+                    "ASSISTANT:"
+                ],
+                "max_new_tokens": 1024,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "image_aspect_ratio": "original"
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "shot_type": "format-prompt",
+                    "shot": 0,
+                    "use_caption": false,
+                    "use_ocr": false
+                },
+                "phi3v": {
+                    "shot_type": "solution"
+                }
+            },
+            "model_specific_generation_kwargs": {
+                "llava": {
+                    "image_aspect_ratio": "original"
+                }
+            }
+        }
+    },
+    "versions": {
+        "hallusion_bench_image": "Yaml",
+        "mathvista_testmini": "Yaml"
+    },
+    "n-shot": {
+        "hallusion_bench_image": 0,
+        "mathvista_testmini": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_sigmoidgating,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}