From bcb58b7e718d01c56d86b42cf73acb46525e1a7d Mon Sep 17 00:00:00 2001
From: KevinHuSh <kevinhu.sh@gmail.com>
Date: Fri, 8 Mar 2024 18:59:53 +0800
Subject: [PATCH] layout refine (#115)
---
README.md | 2 +-
api/apps/llm_app.py | 15 +++++++++++++--
api/db/services/llm_service.py | 4 +++-
deepdoc/vision/layout_recognizer.py | 2 +-
rag/app/naive.py | 2 +-
5 files changed, 19 insertions(+), 6 deletions(-)
diff --git a/README.md b/README.md
index 871c6fc..79a5fcf 100644
--- a/README.md
+++ b/README.md
@@ -18,7 +18,7 @@ If **vm.max_map_count** is not larger than 65535, please run the following comm
121:/ragflow# sudo sysctl -w vm.max_map_count=262144
```
However, this change is not persistent and will be reset after a system reboot.
-To make the change permanent, you need to update the **/etc/sysctl.conf file**.
+To make the change permanent, you need to update the **/etc/sysctl.conf**.
Add or update the following line in the file:
```bash
vm.max_map_count=262144
diff --git a/api/apps/llm_app.py b/api/apps/llm_app.py
index 0a1167b..9f22a32 100644
--- a/api/apps/llm_app.py
+++ b/api/apps/llm_app.py
@@ -82,8 +82,19 @@ def set_api_key():
@login_required
def my_llms():
try:
- objs = TenantLLMService.get_my_llms(current_user.id)
- return get_json_result(data=objs)
+ res = {}
+ for o in TenantLLMService.get_my_llms(current_user.id):
+ if o["llm_factory"] not in res:
+ res[o["llm_factory"]] = {
+ "tags": o["tags"],
+ "llm": []
+ }
+ res[o["llm_factory"]]["llm"].append({
+ "type": o["model_type"],
+ "name": o["model_name"],
+ "used_token": o["used_tokens"]
+ })
+ return get_json_result(data=res)
except Exception as e:
return server_error_response(e)
diff --git a/api/db/services/llm_service.py b/api/db/services/llm_service.py
index 2a96474..67dd70c 100644
--- a/api/db/services/llm_service.py
+++ b/api/db/services/llm_service.py
@@ -49,7 +49,9 @@ class TenantLLMService(CommonService):
LLMFactories.logo,
LLMFactories.tags,
cls.model.model_type,
- cls.model.llm_name]
+ cls.model.llm_name,
+ cls.model.used_tokens
+ ]
objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory == LLMFactories.name)).where(
cls.model.tenant_id == tenant_id).dicts()
diff --git a/deepdoc/vision/layout_recognizer.py b/deepdoc/vision/layout_recognizer.py
index 4851b8e..f38a0f5 100644
--- a/deepdoc/vision/layout_recognizer.py
+++ b/deepdoc/vision/layout_recognizer.py
@@ -100,7 +100,7 @@ class LayoutRecognizer(Recognizer):
i += 1
for lt in ["footer", "header", "reference", "figure caption",
- "table caption", "title", "text", "table", "figure", "equation"]:
+ "table caption", "title", "table", "text", "figure", "equation"]:
findLayout(lt)
# add box to figure layouts which has not text box
diff --git a/rag/app/naive.py b/rag/app/naive.py
index 7b7aa4a..82ecfca 100644
--- a/rag/app/naive.py
+++ b/rag/app/naive.py
@@ -101,7 +101,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000, lang="Chinese", ca
d = copy.deepcopy(doc)
if pdf_parser:
d["image"], poss = pdf_parser.crop(ck, need_position=True)
- add_positions(d, poss, from_page)
+ add_positions(d, poss)
ck = pdf_parser.remove_tag(ck)
tokenize(d, ck, eng)
res.append(d)
--
GitLab