Upload folder using huggingface_hub

.gitattributes

@@ -329,3 +329,8 @@ sft/1M3/Full_smoe_tcmoe/checkpoint-6893/logs/0629_0110_llava...l_mme_llava_model
 sft/1M3/Full_smoe_tcmoe/checkpoint-6893/logs/0629_0110_llava...l_mme_llava_model_args_8331f0/mmerealworld_lite.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_tcmoe/checkpoint-6893/logs/0629_0110_llava...l_mme_llava_model_args_8331f0/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_tcmoe/checkpoint-6893/logs/0629_0110_llava...l_mme_llava_model_args_8331f0/textvqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/mmbench_en_dev.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/mmerealworld_lite.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/textvqa_val.json filter=lfs diff=lfs merge=lfs -text

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0156_llava...image_llava_model_args_26ed7e/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0156_llava...image_llava_model_args_26ed7e/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0156_llava...image_llava_model_args_26ed7e/results.json

@@ -0,0 +1,146 @@
+{
+    "results": {
+        "hallusion_bench_image": {
+            "aAcc,none": 45.7413,
+            "aAcc_stderr,none": "N/A",
+            "fAcc,none": 20.2312,
+            "fAcc_stderr,none": "N/A",
+            "qAcc,none": 16.2637,
+            "qAcc_stderr,none": "N/A",
+            "alias": "hallusion_bench_image"
+        },
+        "mathvista_testmini": {
+            "gpt_eval_score,none": 37.5,
+            "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 0x7f1a8d279310>",
+            "doc_to_text": "<function hb_doc_to_text at 0x7f1a8d279a60>",
+            "doc_to_target": "gt_answer_details",
+            "process_results": "<function hb_process_results at 0x7f1a8d1d2280>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "aAcc",
+                    "aggregation": "<function hb_aggregation_result_aAcc at 0x7f1a8d1d2c10>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "qAcc",
+                    "aggregation": "<function hb_aggregation_result_qAcc at 0x7f1a8d19c280>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fAcc",
+                    "aggregation": "<function hb_aggregation_result_fAcc at 0x7f1a8d19ca60>",
+                    "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 0x7f1a6216dd30>",
+            "doc_to_text": "<function mathvista_doc_to_text at 0x7f1a61d67430>",
+            "doc_to_target": "answer",
+            "process_results": "<function mathvista_process_results at 0x7f1a61d70af0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mathvista_aggregate_results at 0x7f1a61d7e1f0>",
+                    "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_tcmoe/checkpoint-20679,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/infovqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8de17717bad019f1c552b43f46314f53ada797fba6bf0ed97c99539f854f452a
+size 576440908

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/mmbench_en_dev.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07d9722227e34e8836ef46c83aafd4d4afbcf3424fa82811ed732d4cf4f582a2
+size 14568258

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/mmerealworld_lite.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ec25ecdd1b6435b7e0aca02ed7d3fb2c85eab03c486f8b83a9c1973e8e4e4e3
+size 1994104291

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/results.json

@@ -0,0 +1,683 @@
+{
+    "results": {
+        "ai2d": {
+            "exact_match,flexible-extract": 0.6735751295336787,
+            "exact_match_stderr,flexible-extract": 0.008439492413761025,
+            "alias": "ai2d"
+        },
+        "infovqa_val": {
+            "anls,none": 0.29596572652624065,
+            "anls_stderr,none": 0.008626578846134962,
+            "alias": "infovqa_val"
+        },
+        "mmbench_en_dev": {
+            "gpt_eval_score,none": 70.36082474226805,
+            "gpt_eval_score_stderr,none": "N/A",
+            "submission,none": null,
+            "submission_stderr,none": "N/A",
+            "alias": "mmbench_en_dev"
+        },
+        "mme": {
+            "mme_cognition_score,none": 331.07142857142856,
+            "mme_cognition_score_stderr,none": "N/A",
+            "mme_percetion_score,none": 1270.6930772308924,
+            "mme_percetion_score_stderr,none": "N/A",
+            "alias": "mme"
+        },
+        "mmerealworld_lite": {
+            "mme_realworld_score,none": 0.3137050547159979,
+            "mme_realworld_score_stderr,none": "N/A",
+            "alias": "mmerealworld_lite"
+        },
+        "mmmu_val": {
+            "mmmu_acc,none": 0.42556,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": "mmmu_val"
+        },
+        "mmstar": {
+            "coarse perception,none": 0.693234547043744,
+            "coarse perception_stderr,none": "N/A",
+            "fine-grained perception,none": 0.3419798575288627,
+            "fine-grained perception_stderr,none": "N/A",
+            "instance reasoning,none": 0.5304921877955586,
+            "instance reasoning_stderr,none": "N/A",
+            "logical reasoning,none": 0.37921407525367923,
+            "logical reasoning_stderr,none": "N/A",
+            "math,none": 0.39806647064064865,
+            "math_stderr,none": "N/A",
+            "science & technology,none": 0.30868127580045596,
+            "science & technology_stderr,none": "N/A",
+            "alias": "mmstar"
+        },
+        "scienceqa_img": {
+            "exact_match,none": 0.80466038671294,
+            "exact_match_stderr,none": 0.008829908611939156,
+            "alias": "scienceqa_img"
+        },
+        "seedbench_2_plus": {
+            "seedbench_2_plus_Chart,none": 0.519753086419753,
+            "seedbench_2_plus_Chart_stderr,none": "N/A",
+            "seedbench_2_plus_all,none": 0.5059288537549407,
+            "seedbench_2_plus_all_stderr,none": "N/A",
+            "seedbench_2_plus_Web,none": 0.546969696969697,
+            "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": 0.399300000166893,
+            "exact_match_stderr,none": 0.006723313355269604,
+            "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 0x7fbd65668d30>",
+            "doc_to_text": "<function ai2d_doc_to_text at 0x7fbd6566f040>",
+            "doc_to_target": "<function ai2d_doc_to_target at 0x7fbd6566f5e0>",
+            "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 0x7fbd9964ac10>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7fbd9964aee0>",
+            "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 0x7fbdc48875e0>",
+            "doc_to_text": "<function mmbench_doc_to_text at 0x7fbdc4887af0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmbench_process_results at 0x7fbdc4893040>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mmbench_aggregate_dev_results_eval at 0x7fbdc487ee50>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "submission",
+                    "aggregation": "<function mmbench_aggregate_dev_results_submission at 0x7fbdc48873a0>",
+                    "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 0x7fbdc6364a60>",
+            "doc_to_text": "<function mme_doc_to_text at 0x7fbdc4909310>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_process_results at 0x7fbdc49098b0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_percetion_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7fbdc4909dc0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "mme_cognition_score",
+                    "aggregation": "<function mme_aggregate_results at 0x7fbdc4911280>",
+                    "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 0x7fbd65e10ee0>",
+            "doc_to_text": "<function mme_realworld_doc_to_text at 0x7fbd65e169d0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mme_realworld_process_results at 0x7fbd65e19670>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mme_realworld_score",
+                    "aggregation": "<function mme_realworld_aggregate_results at 0x7fbd65e201f0>",
+                    "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 0x7fbd780e8940>",
+            "doc_to_text": "<function mmmu_doc_to_text at 0x7fbd7806c670>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmmu_process_results at 0x7fbd78003550>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_aggregate_results at 0x7fbd77ee64c0>",
+                    "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 0x7fbd655c53a0>",
+            "doc_to_text": "<function mmstar_doc_to_text at 0x7fbd655c5820>",
+            "doc_to_target": "answer",
+            "process_results": "<function mmstar_process_results at 0x7fbd655c5d30>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "coarse perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655cf1f0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fine-grained perception",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655cf5e0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "instance reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655cf9d0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "logical reasoning",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655cfdc0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "science & technology",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655d81f0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "math",
+                    "aggregation": "<function mmstar_aggregate_results at 0x7fbd655d85e0>",
+                    "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 0x7fbd90ccedc0>",
+            "doc_to_text": "<function sqa_doc_to_text at 0x7fbd90b9d040>",
+            "doc_to_target": "<function sqa_doc_to_target at 0x7fbd90b9d430>",
+            "process_results": "<function sqa_process_results at 0x7fbd90b9d790>",
+            "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 0x7fbd8dc97ee0>",
+            "doc_to_text": "<function seed_doc_to_text at 0x7fbd8dcb5550>",
+            "doc_to_target": "answer",
+            "process_results": "<function seed_process_result at 0x7fbd8dcb5a60>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "seedbench_2_plus_Chart",
+                    "aggregation": "<function seed_aggregation_result at 0x7fbd8dcb5f70>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Map",
+                    "aggregation": "<function seed_aggregation_result at 0x7fbd8da52430>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Web",
+                    "aggregation": "<function seed_aggregation_result at 0x7fbd8da528b0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_all",
+                    "aggregation": "<function seed_aggregation_result at 0x7fbd8da52d30>",
+                    "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 0x7fbd924cd9d0>",
+            "doc_to_text": "<function textvqa_doc_to_text at 0x7fbd924cde50>",
+            "doc_to_target": "answer",
+            "process_results": "<function textvqa_process_results at 0x7fbd9234e160>",
+            "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 0x7fbd924cd790>",
+                    "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_tcmoe/checkpoint-20679,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/submissions/mmbench_en_dev_results.json

@@ -0,0 +1 @@
+{"overall_acc": 0.7036082474226805, "category_acc": {"action_recognition": 0.8518518518518519, "attribute_comparison": 0.7045454545454546, "attribute_recognition": 0.8918918918918919, "celebrity_recognition": 0.7373737373737373, "function_reasoning": 0.8734177215189873, "future_prediction": 0.475, "identity_reasoning": 0.9777777777777777, "image_emotion": 0.84, "image_quality": 0.39622641509433965, "image_scene": 0.9519230769230769, "image_style": 0.8679245283018868, "image_topic": 0.8888888888888888, "nature_relation": 0.6666666666666666, "object_localization": 0.37037037037037035, "ocr": 0.6923076923076923, "physical_property_reasoning": 0.49333333333333335, "physical_relation": 0.5833333333333334, "social_relation": 0.9069767441860465, "spatial_relationship": 0.2222222222222222, "structuralized_imagetext_understanding": 0.5384615384615384}, "l2_category_acc": {"attribute_reasoning": 0.7537688442211056, "coarse_perception": 0.8108108108108109, "finegrained_perception (cross-instance)": 0.6083916083916084, "finegrained_perception (instance-level)": 0.6689419795221843, "logic_reasoning": 0.5169491525423728, "relation_reasoning": 0.7391304347826086}}

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/submissions/mmbench_en_dev_results.xlsx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:235d602cad3b4cc25a9b7b7f73fc3d1bac43555f30a27324fce2a35459d1dc15
+size 865053

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0214_llava...l_mme_llava_model_args_26ed7e/textvqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f7d3a778190858dbcc26d240a0dd9c4cf47b6c460be6cc31f0073d6022d88f7
+size 13176599

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/results.json

@@ -0,0 +1,67 @@
+{
+    "results": {
+        "ocrbench": {
+            "ocrbench_accuracy,none": 0.389,
+            "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 0x7f1cb290cdc0>",
+            "doc_to_text": "<function ocrbench_doc_to_text at 0x7f1ca05223a0>",
+            "doc_to_target": "answer",
+            "process_results": "<function ocrbench_process_results at 0x7f1ca0522700>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "ocrbench_accuracy",
+                    "aggregation": "<function ocrbench_aggregate_accuracy at 0x7f1ca0522a60>",
+                    "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_tcmoe/checkpoint-20679,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/logs/0629_0241_llava...bench_llava_model_args_26ed7e/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 204
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 45
+Irregular Text Recognition(Total 50): 43
+Artistic Text Recognition(Total 50): 42
+Handwriting Recognition(Total 50): 20
+Digit String Recognition(Total 50): 29
+Non-Semantic Text Recognition(Total 50): 25
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 114
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 24
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 15
+Handwritten Mathematical Expression Recognition(Total 100): 32
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 389

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c130cb3f49e7ee2b7f06702cf631362cb9498bd7f91c1bc629cf9c39bbff84f
+oid sha256:947999baa28159142d05ed0517363b2d24c8a99b7bf70bbfd7ca97f1d654a339
 size 4972489328

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:751105d3243474a8ef193e8ae20abb8b1dfa0680bf135331366a828be798a9d9
-size 4985965136
+oid sha256:df82da2b0be77c9dcbdc7f0e43199e06533a0cd514ed5264fdd456c096807820
+size 4985902928

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/model-00003-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6708539e51cf22ed53223562dca2305cea74e23959a386392124992a0c06da95
-size 248975808
+oid sha256:0b065775d432f62e238221b5fea99829775100f2dd25c6fea8a9f5a039dcbae3
+size 248971200

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
   "metadata": {
-    "total_size": 10207287168
+    "total_size": 10207220352
   },
   "weight_map": {
     "lm_head.weight": "model-00003-of-00003.safetensors",

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/trainer_state.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:87110bb3ca5a6c7096f5fad40902f7fb9e1f9540c6a381f5f2a08de4ba8c824a
-size 11809631
+oid sha256:c01ee7ccc264e127e3bb340002e94f5911e5846d9c7da727b695c715906c7670
+size 11809233

sft/1M3/Full_smoe_tcmoe/checkpoint-20679/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1427624a2a97a28435015650bead3c3ec509c15d421906a886182abe75b4efcf
+oid sha256:88a0753fdff97eef6a658f2b2cf4c8afea9a83a6ca1601c10ad020270039b677
 size 7992