25119750f8c75c6ca0431185319ad552584348976044b67513b1e9f31d3e067e

.gitattributes

@@ -131,3 +131,5 @@ sft/1M3/Full_smoe/logs/0613_0224_llava..._pope_llava_model_args_bc59f1/pope.json
 sft/1M3/Full_smoe/logs/0613_0224_llava..._pope_llava_model_args_bc59f1/submissions/mmbench_en_dev_results.xlsx filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe/logs/0613_0224_llava..._pope_llava_model_args_bc59f1/textvqa_val.json filter=lfs diff=lfs merge=lfs -text
 sft/1M3/Full_smoe/logs/0613_0343_llava_v1.5_gqa_llava_model_args_bc59f1/gqa.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/infovqa_val.json filter=lfs diff=lfs merge=lfs -text
+sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/pope.json filter=lfs diff=lfs merge=lfs -text

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/infovqa_val.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6cec738425cb57f14dd3611a9fe34493badd8dde4f45b0bf1ff28bbd362a6fc8
+size 576437772

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_0402_llava...a_val_llava_model_args_bc59f1/results.json

@@ -0,0 +1,245 @@
+{
+    "results": {
+        "infovqa_val": {
+            "anls,none": 0.29632274187790075,
+            "anls_stderr,none": 0.008629591418151561,
+            "alias": "infovqa_val"
+        },
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.1922,
+            "mmmu_acc_stderr,none": 0.03656528438248045,
+            "alias": "mmmu_pro"
+        },
+        "mmmu_pro_standard": {
+            "mmmu_acc,none": 0.26532,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_standard"
+        },
+        "mmmu_pro_vision": {
+            "mmmu_acc,none": 0.11908,
+            "mmmu_acc_stderr,none": "N/A",
+            "alias": " - mmmu_pro_vision"
+        },
+        "realworldqa": {
+            "exact_match,flexible-extract": 0.4993464052287582,
+            "exact_match_stderr,flexible-extract": 0.0180893496772524,
+            "alias": "realworldqa"
+        }
+    },
+    "groups": {
+        "mmmu_pro": {
+            "mmmu_acc,none": 0.1922,
+            "mmmu_acc_stderr,none": 0.03656528438248045,
+            "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 0x7fe18c88bf70>",
+            "doc_to_text": "<function infovqa_doc_to_text at 0x7fe18c603280>",
+            "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 0x7fe1b87bcd30>",
+            "doc_to_text": "<function mmmu_pro_doc_to_text at 0x7fe1b87c7af0>",
+            "doc_to_target": "{{answer}}",
+            "process_results": "<function mmmu_pro_process_results at 0x7fe1b87cda60>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7fe1b87d49d0>",
+                    "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 0x7fe1b87b1280>",
+            "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 0x7fe1b87b7160>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "mmmu_acc",
+                    "aggregation": "<function mmmu_pro_aggregate_results at 0x7fe1b87bc0d0>",
+                    "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 0x7fe183dbe040>",
+            "doc_to_text": "<function realworldqa_doc_to_text at 0x7fe183dbe670>",
+            "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe/logs/0617_1601_llava...image_llava_model_args_bc59f1/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_1601_llava...image_llava_model_args_bc59f1/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0617_1601_llava...image_llava_model_args_bc59f1/results.json

@@ -0,0 +1,146 @@
+{
+    "results": {
+        "hallusion_bench_image": {
+            "aAcc,none": 47.8444,
+            "aAcc_stderr,none": "N/A",
+            "fAcc,none": 21.3873,
+            "fAcc_stderr,none": "N/A",
+            "qAcc,none": 18.4615,
+            "qAcc_stderr,none": "N/A",
+            "alias": "hallusion_bench_image"
+        },
+        "mathvista_testmini": {
+            "gpt_eval_score,none": 39.9,
+            "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 0x7f31170cb700>",
+            "doc_to_text": "<function hb_doc_to_text at 0x7f31170cbe50>",
+            "doc_to_target": "gt_answer_details",
+            "process_results": "<function hb_process_results at 0x7f3116f73670>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "aAcc",
+                    "aggregation": "<function hb_aggregation_result_aAcc at 0x7f3116f19040>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "qAcc",
+                    "aggregation": "<function hb_aggregation_result_qAcc at 0x7f3116f19670>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fAcc",
+                    "aggregation": "<function hb_aggregation_result_fAcc at 0x7f3116f19e50>",
+                    "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 0x7f30ec9790d0>",
+            "doc_to_text": "<function mathvista_doc_to_text at 0x7f30ec982790>",
+            "doc_to_target": "answer",
+            "process_results": "<function mathvista_process_results at 0x7f30ec225e50>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mathvista_aggregate_results at 0x7f30ec231550>",
+                    "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/pope.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17f715e247ed8d6a7da812704f7dc7d3c8026775b2a38b6e549a47bc02586925
+size 17579267

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/rank2_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/rank3_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_0643_llava_v1.5_pope_llava_model_args_bc59f1/results.json

@@ -0,0 +1,95 @@
+{
+    "results": {
+        "pope": {
+            "pope_accuracy,none": 0.8466666666666667,
+            "pope_accuracy_stderr,none": "N/A",
+            "pope_precision,none": 0.8860294117647058,
+            "pope_precision_stderr,none": "N/A",
+            "pope_recall,none": 0.8033333333333333,
+            "pope_recall_stderr,none": "N/A",
+            "pope_f1_score,none": 0.8426573426573426,
+            "pope_f1_score_stderr,none": "N/A",
+            "pope_yes_ratio,none": 0.5,
+            "pope_yes_ratio_stderr,none": "N/A",
+            "alias": "pope"
+        }
+    },
+    "configs": {
+        "pope": {
+            "task": "pope",
+            "dataset_path": "lmms-lab/POPE",
+            "dataset_kwargs": {
+                "token": true
+            },
+            "test_split": "test",
+            "doc_to_visual": "<function pope_doc_to_visual at 0x7f05507f3790>",
+            "doc_to_text": "<function pope_doc_to_text at 0x7f05507f3d30>",
+            "doc_to_target": "answer",
+            "process_results": "<function pope_process_results at 0x7f05507f9310>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "pope_accuracy",
+                    "aggregation": "<function pope_aggregate_accuracy at 0x7f05507f98b0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_precision",
+                    "aggregation": "<function pope_aggregate_precision at 0x7f05507f9e50>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_recall",
+                    "aggregation": "<function pope_aggregate_recall at 0x7f05507fc430>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_f1_score",
+                    "aggregation": "<function pope_aggregate_f1_score at 0x7f05507fc9d0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "pope_yes_ratio",
+                    "aggregation": "<function pope_aggregate_yes_ratio at 0x7f05507fcf70>",
+                    "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": {
+        "pope": "Yaml"
+    },
+    "n-shot": {
+        "pope": 0
+    },
+    "model_configs": {
+        "model": "llava",
+        "model_args": "pretrained=/cm/archive/namnv78/checkpoints/Xphi35-siglip224/SMOE/1M3/Full_smoe,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe/logs/0626_1044_llava...image_llava_model_args_bc59f1/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_1044_llava...image_llava_model_args_bc59f1/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_1044_llava...image_llava_model_args_bc59f1/results.json

@@ -0,0 +1,146 @@
+{
+    "results": {
+        "hallusion_bench_image": {
+            "aAcc,none": 48.1598,
+            "aAcc_stderr,none": "N/A",
+            "fAcc,none": 21.6763,
+            "fAcc_stderr,none": "N/A",
+            "qAcc,none": 18.4615,
+            "qAcc_stderr,none": "N/A",
+            "alias": "hallusion_bench_image"
+        },
+        "mathvista_testmini": {
+            "gpt_eval_score,none": 39.9,
+            "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 0x7fc210e580d0>",
+            "doc_to_text": "<function hb_doc_to_text at 0x7fc210e58820>",
+            "doc_to_target": "gt_answer_details",
+            "process_results": "<function hb_process_results at 0x7fc210e7a040>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "aAcc",
+                    "aggregation": "<function hb_aggregation_result_aAcc at 0x7fc210e7a9d0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "qAcc",
+                    "aggregation": "<function hb_aggregation_result_qAcc at 0x7fc210cd3040>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "fAcc",
+                    "aggregation": "<function hb_aggregation_result_fAcc at 0x7fc210cd3820>",
+                    "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 0x7fc1e642faf0>",
+            "doc_to_text": "<function mathvista_doc_to_text at 0x7fc1e60281f0>",
+            "doc_to_target": "answer",
+            "process_results": "<function mathvista_process_results at 0x7fc1e60308b0>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "gpt_eval_score",
+                    "aggregation": "<function mathvista_aggregate_results at 0x7fc1e603af70>",
+                    "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}

sft/1M3/Full_smoe/logs/0626_1106_llava..._plus_llava_model_args_bc59f1/rank0_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_1106_llava..._plus_llava_model_args_bc59f1/rank1_metric_eval_done.txt

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

sft/1M3/Full_smoe/logs/0626_1106_llava..._plus_llava_model_args_bc59f1/results.json

@@ -0,0 +1,168 @@
+{
+    "results": {
+        "chartqa": {
+            "relaxed_overall,none": 0.2184,
+            "relaxed_overall_stderr,none": 0.008264859294607817,
+            "relaxed_human_split,none": 0.2456,
+            "relaxed_human_split_stderr,none": 0.012179621773652779,
+            "relaxed_augmented_split,none": 0.1912,
+            "relaxed_augmented_split_stderr,none": 0.011127133434932011,
+            "alias": "chartqa"
+        },
+        "seedbench_2_plus": {
+            "seedbench_2_plus_Chart,none": 0.5111111111111111,
+            "seedbench_2_plus_Chart_stderr,none": "N/A",
+            "seedbench_2_plus_all,none": 0.5068072024593764,
+            "seedbench_2_plus_all_stderr,none": "N/A",
+            "seedbench_2_plus_Web,none": 0.5484848484848485,
+            "seedbench_2_plus_Web_stderr,none": "N/A",
+            "seedbench_2_plus_Map,none": 0.4684014869888476,
+            "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 0x7fc70ca3cd30>",
+            "doc_to_text": "<function chartqa_doc_to_text at 0x7fc70ca42790>",
+            "doc_to_target": "answer",
+            "process_results": "<function chartqa_process_results at 0x7fc70ca42a60>",
+            "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 0x7fc7357deca0>",
+            "doc_to_text": "<function seed_doc_to_text at 0x7fc735804310>",
+            "doc_to_target": "answer",
+            "process_results": "<function seed_process_result at 0x7fc735804820>",
+            "description": "",
+            "target_delimiter": " ",
+            "fewshot_delimiter": "\n\n",
+            "metric_list": [
+                {
+                    "metric": "seedbench_2_plus_Chart",
+                    "aggregation": "<function seed_aggregation_result at 0x7fc735804d30>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Map",
+                    "aggregation": "<function seed_aggregation_result at 0x7fc7354921f0>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_Web",
+                    "aggregation": "<function seed_aggregation_result at 0x7fc735492670>",
+                    "higher_is_better": true
+                },
+                {
+                    "metric": "seedbench_2_plus_all",
+                    "aggregation": "<function seed_aggregation_result at 0x7fc735492af0>",
+                    "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,conv_template=phi35",
+        "batch_size": "1",
+        "device": null,
+        "limit": null,
+        "bootstrap_iters": 100000,
+        "gen_kwargs": ""
+    },
+    "git_hash": "289c7fe5"
+}