de1b5e67dfafa8cfd2dcccac80916769b96c2546b53f3f867389ec98ef93d6ef

.gitattributes

@@ -356,3 +356,6 @@ sft/665K36/Full_smoe_sharev3/checkpoint-20791/trainer_state.json filter=lfs diff
 sft/665K36/revise_Full_smoe_sharev3/checkpoint-20791/trainer_state.json filter=lfs diff=lfs merge=lfs -text
 sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
 sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/mmbench_en_dev.json filter=lfs diff=lfs merge=lfs -text
+sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/mmerealworld_lite.json filter=lfs diff=lfs merge=lfs -text
+sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
+sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/textvqa_val.json filter=lfs diff=lfs merge=lfs -text

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/mmerealworld_lite.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02ff02798d3f02973ef781787110743f077761182c5110bc44c9cb38ec63698c
+size 1994104355

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/rank0_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/rank1_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/rank2_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/rank3_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/results.json

@@ -0,0 +1,729 @@
+{
+    "results": {
+        "ai2d": {
+            "exact_match,flexible-extract": 0.6544689119170984,
+            "exact_match_stderr,flexible-extract": 0.008558935628342142,
+            "alias": "ai2d"
+        },
+        "infovqa_val": {
+            "anls,none": 0.26847554444841126,
+            "anls_stderr,none": 0.008375057067642373,
+            "alias": "infovqa_val"
+        },
+        "mmbench_en_dev": {
+            "gpt_eval_score,none": 71.21993127147766,
+            "gpt_eval_score_stderr,none": "N/A",
+            "submission,none": null,
+            "submission_stderr,none": "N/A",
+            "alias": "mmbench_en_dev"
+        },
+        "mme": {
+            "mme_cognition_score,none": 321.7857142857143,
+            "mme_cognition_score_stderr,none": "N/A",
+            "mme_percetion_score,none": 1418.2278911564626,
+            "mme_percetion_score_stderr,none": "N/A",
+            "alias": "mme"
+        },
+        "mmerealworld_lite": {
+            "mme_realworld_score,none": 0.30484627410109433,
+            "mme_realworld_score_stderr,none": "N/A",
+            "alias": "mmerealworld_lite"
+        },
+        "mmmu_pro_standard": {
+            "mmmu_acc,none": 0.25896,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": "mmmu_pro_standard"
+        },
+        "mmmu_val": {
+            "mmmu_acc,none": 0.41222,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": "mmmu_val"
+        },
+        "mmstar": {
+            "coarse perception,none": 0.6918706627011363,
+            "coarse perception_stderr,none": "N/A",
+            "fine-grained perception,none": 0.3625644804716286,
+            "fine-grained perception_stderr,none": "N/A",
+            "instance reasoning,none": 0.5205089434882838,
+            "instance reasoning_stderr,none": "N/A",
+            "logical reasoning,none": 0.3660535284297661,
+            "logical reasoning_stderr,none": "N/A",
+            "math,none": 0.28080727078321305,
+            "math_stderr,none": "N/A",
+            "science & technology,none": 0.19842818316868963,
+            "science & technology_stderr,none": "N/A",
+            "alias": "mmstar"
+        },
+        "scienceqa_img": {
+            "exact_match,none": 0.7416955875061974,
+            "exact_match_stderr,none": 0.009748403485997436,
+            "alias": "scienceqa_img"
+        },
+        "seedbench_2_plus": {
+            "seedbench_2_plus_Chart,none": 0.5061728395061729,
+            "seedbench_2_plus_Chart_stderr,none": "N/A",
+            "seedbench_2_plus_all,none": 0.48967940272288096,
+            "seedbench_2_plus_all_stderr,none": "N/A",
+            "seedbench_2_plus_Web,none": 0.5075757575757576,
+            "seedbench_2_plus_Web_stderr,none": "N/A",
+            "seedbench_2_plus_Map,none": 0.45848822800495664,
+            "seedbench_2_plus_Map_stderr,none": "N/A",
+            "alias": "seedbench_2_plus"
+        },
+        "textvqa_val": {
+            "exact_match,none": 0.4134800003051758,
+            "exact_match_stderr,none": 0.006747678419171048,
+            "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 0x7f7862c37700>",
+            "doc_to_text": "<function ai2d_doc_to_text at 0x7f7862c379d0>",
+            "doc_to_target": "<function ai2d_doc_to_target at 0x7f7862c37f70>",
+            "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"
+            }
+        },
+        "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 0x7f78950b2790>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7f78950b2a60>",
+            "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."
+                }
+            }
+        },
+        "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 0x7f78c1ed0430>",
+            "doc_to_text": "<function mmbench_doc_to_text at 0x7f78c1ed0940>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmbench_process_results at 0x7f78c1ed0e50>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mmbench_aggregate_dev_results_eval at 0x7f78c1ec5ca0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "submission",
+                    "aggregation": "<function mmbench_aggregate_dev_results_submission at 0x7f78c1ed01f0>",
+                    "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"
+                }
+            }
+        },
+        "mme": {
+            "task": "mme",
+            "dataset_path": "lmms-lab/MME",
+            "dataset_kwargs": {
+                "token": false
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function mme_doc_to_visual at 0x7f78c39a88b0>",
+            "doc_to_text": "<function mme_doc_to_text at 0x7f78c1f4f160>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_process_results at 0x7f78c1f4f700>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_percetion_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7f78c1f4fc10>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "mme_cognition_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7f78c1f560d0>",
+                    "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"
+                }
+            }
+        },
+        "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 0x7f78633e08b0>",
+            "doc_to_text": "<function mme_realworld_doc_to_text at 0x7f78633e53a0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_realworld_process_results at 0x7f78633e9040>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_realworld_score",
+                    "aggregation": "<function mme_realworld_aggregate_results at 0x7f78633e9b80>",
+                    "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_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 0x7f78c1e9d790>",
+            "doc_to_text": "<function mmmu_pro_doc_to_text at 0x7f78c1ea9550>",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7f78c1eae4c0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7f78c1eb6430>",
+                    "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_val": {
+            "task": "mmmu_val",
+            "dataset_path": "lmms-lab/MMMU",
+            "test_split": "validation",
+            "doc_to_visual": "<function mmmu_doc_to_visual at 0x7f7875483700>",
+            "doc_to_text": "<function mmmu_doc_to_text at 0x7f7875367430>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmmu_process_results at 0x7f787534c310>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_aggregate_results at 0x7f7875267280>",
+                    "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 0x7f7862b91d30>",
+            "doc_to_text": "<function mmstar_doc_to_text at 0x7f7862b991f0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmstar_process_results at 0x7f7862b99700>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "coarse perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b99b80>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fine-grained perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b99f70>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "instance reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b1f3a0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "logical reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b1f790>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "science & technology",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b1fb80>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "math",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f7862b1ff70>",
+                    "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"
+                }
+            }
+        },
+        "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 0x7f788e521dc0>",
+            "doc_to_text": "<function sqa_doc_to_text at 0x7f788e4f2040>",
+            "doc_to_target": "<function sqa_doc_to_target at 0x7f788e4f2430>",
+            "process_results": "<function sqa_process_results at 0x7f788e4f2790>",
+            "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"
+                }
+            }
+        },
+        "seedbench_2_plus": {
+            "task": "seedbench_2_plus",
+            "dataset_path": "doolayer/SEED-Bench-2-Plus",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function seed_doc_to_visual at 0x7f788a10b310>",
+            "doc_to_text": "<function seed_doc_to_text at 0x7f788a10b940>",
+            "doc_to_target": "answer",
+            "process_results": "<function seed_process_result at 0x7f788a10be50>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "seedbench_2_plus_Chart",
+                    "aggregation": "<function seed_aggregation_result at 0x7f787821c3a0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Map",
+                    "aggregation": "<function seed_aggregation_result at 0x7f787821c820>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Web",
+                    "aggregation": "<function seed_aggregation_result at 0x7f787821cca0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_all",
+                    "aggregation": "<function seed_aggregation_result at 0x7f787824b160>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "until": [
+                    "ASSISTANT:"
+                ],
+                "max_new_tokens": 16,
+                "image_aspect_ratio": "original"
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "llava": {
+                    "img_token": "<image>",
+                    "post_prompt": "Answer with the option's letter from the given choices directly."
+                },
+                "gpt4V": {
+                    "img_token": "<image>",
+                    "post_prompt": "Answer with the option's letter from the given choices directly."
+                },
+                "default": {
+                    "img_token": "<image>",
+                    "post_prompt": "Answer with the option's letter from the given choices directly."
+                }
+            }
+        },
+        "textvqa_val": {
+            "task": "textvqa_val",
+            "dataset_path": "lmms-lab/textvqa",
+            "test_split": "validation",
+            "doc_to_visual": "<function textvqa_doc_to_visual at 0x7f788fa299d0>",
+            "doc_to_text": "<function textvqa_doc_to_text at 0x7f788fa29e50>",
+            "doc_to_target": "answer",
+            "process_results": "<function textvqa_process_results at 0x7f788f899160>",
+            "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 0x7f788fa29790>",
+                    "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",
+        "infovqa_val": "Yaml",
+        "mmbench_en_dev": "Yaml",
+        "mme": "Yaml",
+        "mmerealworld_lite": "Yaml",
+        "mmmu_pro_standard": "Yaml",
+        "mmmu_val": "Yaml",
+        "mmstar": "Yaml",
+        "scienceqa_img": "Yaml",
+        "seedbench_2_plus": "Yaml",
+        "textvqa_val": "Yaml"
+    },
+    "n-shot": {
+        "ai2d": 0,
+        "infovqa_val": 0,
+        "mmbench_en_dev": 0,
+        "mme": 0,
+        "mmerealworld_lite": 0,
+        "mmmu_pro_standard": 0,
+        "mmmu_val": 0,
+        "mmstar": 0,
+        "scienceqa_img": 0,
+        "seedbench_2_plus": 0,
+        "textvqa_val": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78_new/revise_checkpoints/Xphi35-siglip224/SMOE/665K36/revise_Full_smoe_sharev3/checkpoint-12477,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/submissions/mmbench_en_dev_results.json

@@ -0,0 +1 @@
+{"overall_acc": 0.7121993127147767, "category_acc": {"action_recognition": 0.9074074074074074, "attribute_comparison": 0.7045454545454546, "attribute_recognition": 0.8513513513513513, "celebrity_recognition": 0.7676767676767676, "function_reasoning": 0.8607594936708861, "future_prediction": 0.55, "identity_reasoning": 0.9777777777777777, "image_emotion": 0.84, "image_quality": 0.4528301886792453, "image_scene": 0.9615384615384616, "image_style": 0.9245283018867925, "image_topic": 0.8611111111111112, "nature_relation": 0.5625, "object_localization": 0.49382716049382713, "ocr": 0.6923076923076923, "physical_property_reasoning": 0.6, "physical_relation": 0.4583333333333333, "social_relation": 0.813953488372093, "spatial_relationship": 0.26666666666666666, "structuralized_imagetext_understanding": 0.4230769230769231}, "l2_category_acc": {"attribute_reasoning": 0.7889447236180904, "coarse_perception": 0.831081081081081, "finegrained_perception (cross-instance)": 0.6433566433566433, "finegrained_perception (instance-level)": 0.7030716723549488, "logic_reasoning": 0.4661016949152542, "relation_reasoning": 0.6347826086956522}}

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/submissions/mmbench_en_dev_results.xlsx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3eae64c51364e19fe9591076a3878e0882d544ae87b51040157b5428d3528835
+size 865116

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2000_llava...l_mme_llava_model_args_82420a/textvqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2784ee378d820efb47d9c9d42a268758638ebfa76232aff787fb54feb5998952
+size 13138199

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2030_llava...bench_llava_model_args_82420a/rank0_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2030_llava...bench_llava_model_args_82420a/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 181
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 47
+Irregular Text Recognition(Total 50): 41
+Artistic Text Recognition(Total 50): 44
+Handwriting Recognition(Total 50): 24
+Digit String Recognition(Total 50): 11
+Non-Semantic Text Recognition(Total 50): 14
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 113
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 22
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 10
+Handwritten Mathematical Expression Recognition(Total 100): 0
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 326

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2031_llava...bench_llava_model_args_82420a/rank1_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2031_llava...bench_llava_model_args_82420a/rank2_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2031_llava...bench_llava_model_args_82420a/rank3_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2043_llava...bench_llava_model_args_82420a/rank0_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2043_llava...bench_llava_model_args_82420a/rank1_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2043_llava...bench_llava_model_args_82420a/rank2_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2043_llava...bench_llava_model_args_82420a/rank3_metric_eval_done.txt

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

sft/665K36/revise_Full_smoe_sharev3/checkpoint-12477/logs/0717_2043_llava...bench_llava_model_args_82420a/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 181
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 47
+Irregular Text Recognition(Total 50): 41
+Artistic Text Recognition(Total 50): 44
+Handwriting Recognition(Total 50): 24
+Digit String Recognition(Total 50): 11
+Non-Semantic Text Recognition(Total 50): 14
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 113
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 22
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 10
+Handwritten Mathematical Expression Recognition(Total 100): 0
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 326