Changes to be committed:

modified:   eval_configs/minigpt4_eval.yaml
	modified:   eval_configs/minigpt4_llama2_eval.yaml
	modified:   minigpt4/configs/datasets/cc_sbu/align.yaml
	modified:   minigpt4/configs/models/minigpt4_llama2.yaml
	modified:   train_configs/minigpt4_llama2_stage2_finetune.yaml

eval_configs/minigpt4_eval.yaml

@@ -5,7 +5,7 @@ model:
   end_sym: "###"
   low_resource: True
   prompt_template: '###Human: {} ###Assistant: '
-  ckpt: '/path/to/checkpoint/'
+  ckpt: 'minigpt4/output/minigpt4_stage2_finetune/20230828181/checkpoint_19.pth'
 
 
 datasets:
@@ -13,7 +13,7 @@ datasets:
     vis_processor:
       train:
         name: "blip2_image_eval"
-        image_size: 224
+        image_size: 512
     text_processor:
       train:
         name: "blip_caption"

eval_configs/minigpt4_llama2_eval.yaml

@@ -5,7 +5,7 @@ model:
   end_sym: "</s>"
   low_resource: True
   prompt_template: '[INST] {} [/INST] '
-  ckpt: '/path/to/checkpoint/'
+  ckpt: 'minigpt4/output/minigpt4_stage2_finetune/DATETIME/CHECKPOINT'
 
 
 datasets:
@@ -13,7 +13,7 @@ datasets:
     vis_processor:
       train:
         name: "blip2_image_eval"
-        image_size: 224
+        image_size: 512
     text_processor:
       train:
         name: "blip_caption"

minigpt4/configs/datasets/cc_sbu/align.yaml

@@ -2,4 +2,4 @@ datasets:
   cc_sbu_align:
     data_type: images
     build_info:
-      storage: /path/to/cc_sbu_align/
+      storage: caped_images

minigpt4/configs/models/minigpt4_llama2.yaml

@@ -2,26 +2,26 @@ model:
   arch: mini_gpt4
 
   # vit encoder
-  image_size: 224
+  image_size: 512
   drop_path_rate: 0
   use_grad_checkpoint: False
-  vit_precision: "fp16"
+  vit_precision: "fp32"
   freeze_vit: True
   has_qformer: False
 
   # generation configs
   prompt: ""
 
-  llama_model: "/path/to/llama2/weight"
+  llama_model: "Llama-2-7b-chat"
 
 preprocess:
     vis_processor:
         train:
           name: "blip2_image_train"
-          image_size: 224
+          image_size: 512
         eval:
           name: "blip2_image_eval"
-          image_size: 224
+          image_size: 512
     text_processor:
         train:
           name: "blip_caption"

train_configs/minigpt4_llama2_stage2_finetune.yaml

@@ -6,7 +6,7 @@ model:
   end_sym: "</s>"
   prompt_path: "prompts/alignment.txt"
   prompt_template: '[INST] {} [/INST] '
-  ckpt: '/path/to/stage1/checkpoint/'
+  ckpt: 'pretrained_minigpt4_llama2_7b.pth'
 
 
 datasets:
@@ -14,7 +14,7 @@ datasets:
     vis_processor:
       train:
         name: "blip2_image_train"
-        image_size: 224
+        image_size: 512
     text_processor:
       train:
         name: "blip_caption"