621e39e0a3f90b3e1ad3c6a69fbca3ac88ed2bccf1a4c66e51c018ac56d4b3ec

.gitattributes

@@ -207,3 +207,5 @@ sft/1M3/Full_smoe_tcmoe/logs/0626_0357_llava..._pope_llava_model_args_59313d/mme
 sft/1M3/Full_smoe_tcmoe/logs/0626_0357_llava..._pope_llava_model_args_59313d/pope.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_tcmoe/logs/0626_0357_llava..._pope_llava_model_args_59313d/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe_tcmoe/logs/0626_0357_llava..._pope_llava_model_args_59313d/textvqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe_tcmoe/logs/0626_1139_llava_v1.5_gqa_llava_model_args_59313d/gqa.json filter=lfs diff=lfs merge=lfs -text

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/infovqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfa57a8969ea335579f36927fa1b04d59c4ada57f64ee5c223cf86770a85275d
+size 576440094

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_0430_llava...a_val_llava_model_args_59313d/results.json

@@ -0,0 +1,245 @@
+{
+    "results": {
+        "infovqa_val": {
+            "anls,none": 0.2995358800428418,
+            "anls_stderr,none": 0.00865642075096775,
+            "alias": "infovqa_val"
+        },
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.18931,
+            "mmmu_acc_stderr,none": 0.03685532629908108,
+            "alias": "mmmu_pro"
+        },
+        "mmmu_pro_standard": {
+            "mmmu_acc,none": 0.26301,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_standard"
+        },
+        "mmmu_pro_vision": {
+            "mmmu_acc,none": 0.11561,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_vision"
+        },
+        "realworldqa": {
+            "exact_match,flexible-extract": 0.515032679738562,
+            "exact_match_stderr,flexible-extract": 0.018081187561622823,
+            "alias": "realworldqa"
+        }
+    },
+    "groups": {
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.18931,
+            "mmmu_acc_stderr,none": 0.03685532629908108,
+            "alias": "mmmu_pro"
+        }
+    },
+    "configs": {
+        "infovqa_val": {
+            "task": "infovqa_val",
+            "dataset_path": "lmms-lab/DocVQA",
+            "dataset_name": "InfographicVQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "validation",
+            "doc_to_visual": "<function infovqa_doc_to_visual at 0x7fc9fa5d69d0>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7fc9fa5d6ca0>",
+            "doc_to_target": "answers",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "anls",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 32,
+                "temperature": 0.0,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                }
+            }
+        },
+        "mmmu_pro_standard": {
+            "task": "mmmu_pro_standard",
+            "dataset_path": "MMMU/MMMU_Pro",
+            "dataset_name": "standard (10 options)",
+            "test_split": "test",
+            "doc_to_visual": "<function mmmu_pro_doc_to_visual at 0x7fca25805790>",
+            "doc_to_text": "<function mmmu_pro_doc_to_text at 0x7fca25810550>",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7fca258194c0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7fca25823430>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 256,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": {
+                "version": 0.0,
+                "interleaved_format": false
+            },
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "Answer with the option letter from the given choices directly."
+                }
+            }
+        },
+        "mmmu_pro_vision": {
+            "task": "mmmu_pro_vision",
+            "dataset_path": "MMMU/MMMU_Pro",
+            "dataset_name": "vision",
+            "test_split": "test",
+            "doc_to_visual": "<function mmmu_pro_doc_to_visual at 0x7fca257f2ca0>",
+            "doc_to_text": "Answer with the option letter from the given choices directly.",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7fca257f9b80>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7fca25800af0>",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 256,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": {
+                "version": 0.0,
+                "interleaved_format": false
+            }
+        },
+        "realworldqa": {
+            "task": "realworldqa",
+            "dataset_path": "lmms-lab/RealWorldQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function realworldqa_doc_to_visual at 0x7fc9f1671a60>",
+            "doc_to_text": "<function realworldqa_doc_to_text at 0x7fc9f168a0d0>",
+            "doc_to_target": "answer",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "filter_list": [
+                {
+                    "name": "flexible-extract",
+                    "filter": [
+                        {
+                            "function": "<class 'utils.NumberWordsToDigitsFilter'>"
+                        },
+                        {
+                            "function": "<class 'utils.MultiChoiceRegexFilter'>",
+                            "group_select": 0,
+                            "ignore_case": true,
+                            "ignore_punctuation": true,
+                            "regex_pattern": "(\\([A-Z]\\))"
+                        }
+                    ]
+                }
+            ],
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": ""
+                },
+                "gpt4v": {
+                    "pre_prompt": "",
+                    "post_prompt": ""
+                },
+                "xcomposer2_4khd": {
+                    "pre_prompt": "[UNUSED_TOKEN_146]user\nQuestion: ",
+                    "post_prompt": "[UNUSED_TOKEN_145]\n[UNUSED_TOKEN_146]assistant\nThe answer is"
+                }
+            }
+        }
+    },
+    "versions": {
+        "infovqa_val": "Yaml",
+        "mmmu_pro": "N/A",
+        "mmmu_pro_standard": "Yaml",
+        "mmmu_pro_vision": "Yaml",
+        "realworldqa": "Yaml"
+    },
+    "n-shot": {
+        "infovqa_val": 0,
+        "mmmu_pro": 0,
+        "mmmu_pro_standard": 0,
+        "mmmu_pro_vision": 0,
+        "realworldqa": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_tcmoe,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/logs/0626_1139_llava_v1.5_gqa_llava_model_args_59313d/gqa.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcc4c5e650a5f87ecc3e3934c40f696d319a0a49ed56237d9b4a66604c28b234
+size 38569629

sft/1M3/Full_smoe_tcmoe/logs/0626_1139_llava_v1.5_gqa_llava_model_args_59313d/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1139_llava_v1.5_gqa_llava_model_args_59313d/results.json

@@ -0,0 +1,79 @@
+{
+    "results": {
+        "gqa": {
+            "exact_match,none": 0.3945778343138814,
+            "exact_match_stderr,none": 0.004358198135704839,
+            "alias": "gqa"
+        }
+    },
+    "configs": {
+        "gqa": {
+            "task": "gqa",
+            "dataset_path": "lmms-lab/GQA",
+            "dataset_name": "testdev_balanced_instructions",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "testdev",
+            "doc_to_visual": "<function gqa_doc_to_visual at 0x7ff78ec2f8b0>",
+            "doc_to_text": "<function gqa_doc_to_text at 0x7ff78eb83b80>",
+            "doc_to_target": "answer",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "exact_match",
+                    "aggregation": "mean",
+                    "higher_is_better": true,
+                    "ignore_case": true,
+                    "ignore_punctuation": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "top_p": 1.0,
+                "num_beams": 1,
+                "do_sample": false,
+                "until": [
+                    "\n\n"
+                ]
+            },
+            "repeats": 1,
+            "should_decontaminate": false,
+            "metadata": [
+                {
+                    "version": 0.0
+                }
+            ],
+            "model_specific_prompt_kwargs": {
+                "default": {
+                    "pre_prompt": "",
+                    "post_prompt": "\nAnswer the question using a single word or phrase."
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                }
+            }
+        }
+    },
+    "versions": {
+        "gqa": "Yaml"
+    },
+    "n-shot": {
+        "gqa": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_tcmoe,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/logs/0626_1211_llava...image_llava_model_args_59313d/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1211_llava...image_llava_model_args_59313d/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1211_llava...image_llava_model_args_59313d/results.json

@@ -0,0 +1,146 @@
+{
+    "results": {
+        "hallusion_bench_image": {
+            "aAcc,none": 45.0053,
+            "aAcc_stderr,none": "N/A",
+            "fAcc,none": 17.6301,
+            "fAcc_stderr,none": "N/A",
+            "qAcc,none": 14.9451,
+            "qAcc_stderr,none": "N/A",
+            "alias": "hallusion_bench_image"
+        },
+        "mathvista_testmini": {
+            "gpt_eval_score,none": 37.8,
+            "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 0x7fdbc65ef0d0>",
+            "doc_to_text": "<function hb_doc_to_text at 0x7fdbc65ef820>",
+            "doc_to_target": "gt_answer_details",
+            "process_results": "<function hb_process_results at 0x7fdbc6563040>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "aAcc",
+                    "aggregation": "<function hb_aggregation_result_aAcc at 0x7fdbc65639d0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "qAcc",
+                    "aggregation": "<function hb_aggregation_result_qAcc at 0x7fdbc6467040>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fAcc",
+                    "aggregation": "<function hb_aggregation_result_fAcc at 0x7fdbc6467820>",
+                    "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 0x7fdb9bbb0af0>",
+            "doc_to_text": "<function mathvista_doc_to_text at 0x7fdb9b7a91f0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mathvista_process_results at 0x7fdb9b7b18b0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mathvista_aggregate_results at 0x7fdb9b7bbf70>",
+                    "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/logs/0626_1233_llava..._plus_llava_model_args_59313d/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1233_llava..._plus_llava_model_args_59313d/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1233_llava..._plus_llava_model_args_59313d/results.json

@@ -0,0 +1,168 @@
+{
+    "results": {
+        "chartqa": {
+            "relaxed_overall,none": 0.1644,
+            "relaxed_overall_stderr,none": 0.007414246045132662,
+            "relaxed_human_split,none": 0.1864,
+            "relaxed_human_split_stderr,none": 0.011019127412601738,
+            "relaxed_augmented_split,none": 0.1424,
+            "relaxed_augmented_split_stderr,none": 0.009888175403493184,
+            "alias": "chartqa"
+        },
+        "seedbench_2_plus": {
+            "seedbench_2_plus_Chart,none": 0.5049382716049383,
+            "seedbench_2_plus_Chart_stderr,none": "N/A",
+            "seedbench_2_plus_all,none": 0.5072463768115942,
+            "seedbench_2_plus_all_stderr,none": "N/A",
+            "seedbench_2_plus_Web,none": 0.5545454545454546,
+            "seedbench_2_plus_Web_stderr,none": "N/A",
+            "seedbench_2_plus_Map,none": 0.4708798017348203,
+            "seedbench_2_plus_Map_stderr,none": "N/A",
+            "alias": "seedbench_2_plus"
+        }
+    },
+    "configs": {
+        "chartqa": {
+            "task": "chartqa",
+            "dataset_path": "lmms-lab/ChartQA",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function chartqa_doc_to_visual at 0x7fa788965d30>",
+            "doc_to_text": "<function chartqa_doc_to_text at 0x7fa78896c790>",
+            "doc_to_target": "answer",
+            "process_results": "<function chartqa_process_results at 0x7fa78896ca60>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "relaxed_overall",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "relaxed_human_split",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "relaxed_augmented_split",
+                    "aggregation": "mean",
+                    "higher_is_better": true
+                }
+            ],
+            "output_type": "generate_until",
+            "generation_kwargs": {
+                "max_new_tokens": 16,
+                "temperature": 0.0,
+                "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 with a single word."
+                },
+                "qwen_vl": {
+                    "pre_prompt": "",
+                    "post_prompt": " Answer:"
+                }
+            }
+        },
+        "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 0x7fa7b10d8ca0>",
+            "doc_to_text": "<function seed_doc_to_text at 0x7fa7b1101310>",
+            "doc_to_target": "answer",
+            "process_results": "<function seed_process_result at 0x7fa7b1101820>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "seedbench_2_plus_Chart",
+                    "aggregation": "<function seed_aggregation_result at 0x7fa7b1101d30>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Map",
+                    "aggregation": "<function seed_aggregation_result at 0x7fa7b0ea31f0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Web",
+                    "aggregation": "<function seed_aggregation_result at 0x7fa7b0ea3670>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_all",
+                    "aggregation": "<function seed_aggregation_result at 0x7fa7b0ea3af0>",
+                    "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."
+                }
+            }
+        }
+    },
+    "versions": {
+        "chartqa": "Yaml",
+        "seedbench_2_plus": "Yaml"
+    },
+    "n-shot": {
+        "chartqa": 0,
+        "seedbench_2_plus": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe_tcmoe,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/results.json

@@ -0,0 +1,67 @@
+{
+    "results": {
+        "ocrbench": {
+            "ocrbench_accuracy,none": 0.391,
+            "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 0x7f9ffcd0ac10>",
+            "doc_to_text": "<function ocrbench_doc_to_text at 0x7f9ffc5251f0>",
+            "doc_to_target": "answer",
+            "process_results": "<function ocrbench_process_results at 0x7f9ffc525550>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "ocrbench_accuracy",
+                    "aggregation": "<function ocrbench_aggregate_accuracy at 0x7f9ffc5258b0>",
+                    "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe_tcmoe/logs/0626_1418_llava...bench_llava_model_args_59313d/results/ocrbench_results.txt

@@ -0,0 +1,18 @@
+######################### OCRBench #############################
+Text Recognition(Total 300): 204
+---------------- Details of Recognition Score ------------------
+Regular Text Recognition(Total 50): 44
+Irregular Text Recognition(Total 50): 40
+Artistic Text Recognition(Total 50): 43
+Handwriting Recognition(Total 50): 18
+Digit String Recognition(Total 50): 37
+Non-Semantic Text Recognition(Total 50): 22
+----------------------------------------------------------------
+Scene Text-centric VQA(Total 200): 110
+----------------------------------------------------------------
+Doc-oriented VQA(Total 200): 25
+----------------------------------------------------------------
+Key Information Extraction(Total 200): 18
+Handwritten Mathematical Expression Recognition(Total 100): 34
+--------------------- Final Score ------------------------------
+Final Score(Total 1000): 391