caa1ef0fada7a0f4ac4c3841ef9e84332d3a30f97434d6a475f6618b724f7d5c

.gitattributes

@@ -229,3 +229,6 @@ sft/1M3/Full_smoe/checkpoint-20679/logs/0627_1245_llava...l_mme_llava_model_args
 sft/1M3/Full_smoe/checkpoint-20679/logs/0627_1245_llava...l_mme_llava_model_args_18fef8/textvqa_val.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/mmbench_en_dev.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/mmerealworld_lite.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/textvqa_val.json filter=lfs diff=lfs merge=lfs -text

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/mmerealworld_lite.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d1954b2f8f69527a69c05206e95c9cd8ff1c5134f091bb69ebdbff0b082dcc4
+size 1994104291

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/results.json

@@ -0,0 +1,683 @@
+{
+    "results": {
+        "ai2d": {
+            "exact_match,flexible-extract": 0.677461139896373,
+            "exact_match_stderr,flexible-extract": 0.008413271478449879,
+            "alias": "ai2d"
+        },
+        "infovqa_val": {
+            "anls,none": 0.27990003570153515,
+            "anls_stderr,none": 0.008484355470844385,
+            "alias": "infovqa_val"
+        },
+        "mmbench_en_dev": {
+            "gpt_eval_score,none": 68.29896907216495,
+            "gpt_eval_score_stderr,none": "N/A",
+            "submission,none": null,
+            "submission_stderr,none": "N/A",
+            "alias": "mmbench_en_dev"
+        },
+        "mme": {
+            "mme_cognition_score,none": 315.0,
+            "mme_cognition_score_stderr,none": "N/A",
+            "mme_percetion_score,none": 1191.5826330532213,
+            "mme_percetion_score_stderr,none": "N/A",
+            "alias": "mme"
+        },
+        "mmerealworld_lite": {
+            "mme_realworld_score,none": 0.30015633142261594,
+            "mme_realworld_score_stderr,none": "N/A",
+            "alias": "mmerealworld_lite"
+        },
+        "mmmu_val": {
+            "mmmu_acc,none": 0.41556,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": "mmmu_val"
+        },
+        "mmstar": {
+            "coarse perception,none": 0.6289471986108911,
+            "coarse perception_stderr,none": "N/A",
+            "fine-grained perception,none": 0.31227585359862436,
+            "fine-grained perception_stderr,none": "N/A",
+            "instance reasoning,none": 0.5168685770353041,
+            "instance reasoning_stderr,none": "N/A",
+            "logical reasoning,none": 0.3484233038688484,
+            "logical reasoning_stderr,none": "N/A",
+            "math,none": 0.390119620422347,
+            "math_stderr,none": "N/A",
+            "science & technology,none": 0.331204260883257,
+            "science & technology_stderr,none": "N/A",
+            "alias": "mmstar"
+        },
+        "scienceqa_img": {
+            "exact_match,none": 0.8041646008924145,
+            "exact_match_stderr,none": 0.008838382866139713,
+            "alias": "scienceqa_img"
+        },
+        "seedbench_2_plus": {
+            "seedbench_2_plus_Chart,none": 0.48641975308641977,
+            "seedbench_2_plus_Chart_stderr,none": "N/A",
+            "seedbench_2_plus_all,none": 0.4857268335529205,
+            "seedbench_2_plus_all_stderr,none": "N/A",
+            "seedbench_2_plus_Web,none": 0.5212121212121212,
+            "seedbench_2_plus_Web_stderr,none": "N/A",
+            "seedbench_2_plus_Map,none": 0.4560099132589839,
+            "seedbench_2_plus_Map_stderr,none": "N/A",
+            "alias": "seedbench_2_plus"
+        },
+        "textvqa_val": {
+            "exact_match,none": 0.38544000025987624,
+            "exact_match_stderr,none": 0.0066797698348715715,
+            "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 0x7f4ce8beaca0>",
+            "doc_to_text": "<function ai2d_doc_to_text at 0x7f4ce8beaf70>",
+            "doc_to_target": "<function ai2d_doc_to_target at 0x7f4ce8bf1550>",
+            "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 0x7f4d1a345b80>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7f4d1a345e50>",
+            "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 0x7f4d487d55e0>",
+            "doc_to_text": "<function mmbench_doc_to_text at 0x7f4d487d5af0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmbench_process_results at 0x7f4d487e1040>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mmbench_aggregate_dev_results_eval at 0x7f4d4884be50>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "submission",
+                    "aggregation": "<function mmbench_aggregate_dev_results_submission at 0x7f4d487d53a0>",
+                    "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 0x7f4d498f4a60>",
+            "doc_to_text": "<function mme_doc_to_text at 0x7f4d48856310>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_process_results at 0x7f4d488568b0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_percetion_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7f4d48856dc0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "mme_cognition_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7f4d4885f280>",
+                    "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 0x7f4ce9392e50>",
+            "doc_to_text": "<function mme_realworld_doc_to_text at 0x7f4ce9396940>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_realworld_process_results at 0x7f4ce939b5e0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_realworld_score",
+                    "aggregation": "<function mme_realworld_aggregate_results at 0x7f4ce93a1160>",
+                    "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 0x7f4cfc1328b0>",
+            "doc_to_text": "<function mmmu_doc_to_text at 0x7f4cfc0b85e0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmmu_process_results at 0x7f4cfbfe24c0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_aggregate_results at 0x7f4cfbf04430>",
+                    "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 0x7f4ce8b47310>",
+            "doc_to_text": "<function mmstar_doc_to_text at 0x7f4ce8b47790>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmstar_process_results at 0x7f4ce8b47ca0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "coarse perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b50160>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fine-grained perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b50550>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "instance reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b50940>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "logical reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b50d30>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "science & technology",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b58160>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "math",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7f4ce8b58550>",
+                    "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 0x7f4d13f14d30>",
+            "doc_to_text": "<function sqa_doc_to_text at 0x7f4d13f14f70>",
+            "doc_to_target": "<function sqa_doc_to_target at 0x7f4d13ea53a0>",
+            "process_results": "<function sqa_process_results at 0x7f4d13ea5700>",
+            "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 0x7f4d11559e50>",
+            "doc_to_text": "<function seed_doc_to_text at 0x7f4d1157d4c0>",
+            "doc_to_target": "answer",
+            "process_results": "<function seed_process_result at 0x7f4d1157d9d0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "seedbench_2_plus_Chart",
+                    "aggregation": "<function seed_aggregation_result at 0x7f4d1157dee0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Map",
+                    "aggregation": "<function seed_aggregation_result at 0x7f4d115293a0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Web",
+                    "aggregation": "<function seed_aggregation_result at 0x7f4d11529820>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_all",
+                    "aggregation": "<function seed_aggregation_result at 0x7f4d11529ca0>",
+                    "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 0x7f4d15055940>",
+            "doc_to_text": "<function textvqa_doc_to_text at 0x7f4d15055dc0>",
+            "doc_to_target": "answer",
+            "process_results": "<function textvqa_process_results at 0x7f4d150790d0>",
+            "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 0x7f4d15055700>",
+                    "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_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_val": 0,
+        "mmstar": 0,
+        "scienceqa_img": 0,
+        "seedbench_2_plus": 0,
+        "textvqa_val": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_share/checkpoint-13786,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/submissions/mmbench_en_dev_results.json

@@ -0,0 +1 @@
+{"overall_acc": 0.6829896907216495, "category_acc": {"action_recognition": 0.8888888888888888, "attribute_comparison": 0.6590909090909091, "attribute_recognition": 0.8648648648648649, "celebrity_recognition": 0.7575757575757576, "function_reasoning": 0.8734177215189873, "future_prediction": 0.475, "identity_reasoning": 0.9555555555555556, "image_emotion": 0.76, "image_quality": 0.37735849056603776, "image_scene": 0.9519230769230769, "image_style": 0.9056603773584906, "image_topic": 0.8333333333333334, "nature_relation": 0.5833333333333334, "object_localization": 0.35802469135802467, "ocr": 0.6923076923076923, "physical_property_reasoning": 0.44, "physical_relation": 0.5833333333333334, "social_relation": 0.813953488372093, "spatial_relationship": 0.2, "structuralized_imagetext_understanding": 0.48717948717948717}, "l2_category_acc": {"attribute_reasoning": 0.7286432160804021, "coarse_perception": 0.793918918918919, "finegrained_perception (cross-instance)": 0.6013986013986014, "finegrained_perception (instance-level)": 0.6655290102389079, "logic_reasoning": 0.4830508474576271, "relation_reasoning": 0.6695652173913044}}

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/submissions/mmbench_en_dev_results.xlsx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85a036f15518bb0e2e9dc7d42305d7910a51559a9b8fee67570ea67dd3117aec
+size 843234

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1431_llava...l_mme_llava_model_args_d81bcd/textvqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a972d3a868c9d0fec3942e19e850909364165be767bd0ebb6b603dad4f619ea
+size 13182227

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1500_llava...bench_llava_model_args_d81bcd/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1500_llava...bench_llava_model_args_d81bcd/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1500_llava...bench_llava_model_args_d81bcd/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1500_llava...bench_llava_model_args_d81bcd/results.json

@@ -0,0 +1,67 @@
+{
+    "results": {
+        "ocrbench": {
+            "ocrbench_accuracy,none": 0.375,
+            "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 0x7f06efca6dc0>",
+            "doc_to_text": "<function ocrbench_doc_to_text at 0x7f06ef54d3a0>",
+            "doc_to_target": "answer",
+            "process_results": "<function ocrbench_process_results at 0x7f06ef54d700>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "ocrbench_accuracy",
+                    "aggregation": "<function ocrbench_aggregate_accuracy at 0x7f06ef54da60>",
+                    "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_share/checkpoint-13786,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_share/checkpoint-13786/logs/0628_1500_llava...bench_llava_model_args_d81bcd/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 211
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 44
+Irregular Text Recognition(Total 50): 42
+Artistic Text Recognition(Total 50): 45
+Handwriting Recognition(Total 50): 18
+Digit String Recognition(Total 50): 36
+Non-Semantic Text Recognition(Total 50): 26
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 103
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 20
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 16
+Handwritten Mathematical Expression Recognition(Total 100): 25
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 375