[bugfix] README typo & data path & compile flex attn

Author: hanjian.thu123

.gitignore

@@ -44,3 +44,4 @@ weights
 checkpoints
 ref.py
 wandb
+.DS_Store

README.md

@@ -113,7 +113,7 @@ Each "[h_div_w_template1]_[num_examples].jsonl" file contains lines of dumped js
     "long_caption": long caption of the image, required",
     "long_caption_type": "InternVL 2.0, required",
     "text": "short caption of the image, optional",
-    "short_caption_type": "user prompt, , optional"
+    "short_caption_type": "user prompt, optional"
   }
   ```
 

data/infinity_toy_data/splits/1.000_000002500.jsonl

@@ -0,0 +1,23 @@
+import os
+import numpy as np
+
+import json
+
+
+with open('/Users/bytedance/Desktop/projects/Infinity/evaluation/gen_eval/prompt_rewrite_cache_1.json', 'r') as f:
+    correct = json.load(f)
+
+with open('/Users/bytedance/Desktop/projects/Infinity/evaluation/gen_eval/prompt_rewrite_cache_123.json', 'r') as f:
+    false_key_dict = json.load(f)
+
+keys1_list = list(correct.keys())
+keys2_list = list(false_key_dict.keys())
+
+final_dict = {}
+for i in range(len(keys1_list)):
+    key1 = keys1_list[i]
+    key2 = keys2_list[i]
+    final_dict[key1] = false_key_dict[key2]
+
+with open('prompt_rewrite_cache.json', 'w') as f:
+    json.dump(final_dict, f, ensure_ascii=False, indent=2)

data/infinity_toy_data/splits/1.500_000002500.jsonl

@@ -330,7 +330,7 @@ def compile_flex_attn(self):
                                         H = self.num_heads,
                                         L = aligned_L)
                 attn_fn_compile_dict[patchs_nums_tuple] = attn_fn
-            return attn_fn_compile_dict
+        return attn_fn_compile_dict
 
     def get_logits(self, h: torch.Tensor, cond_BD: Optional[torch.Tensor]):
         """

evaluation/gen_eval/rename.py

@@ -19,3 +19,4 @@ einops
 openai
 httpx==0.20.0
 opencv-python
+flash_attn