pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix: gemini cv model chat issue. (#10799)

Browse Source 
### What problem does this PR solve?

#10787
#10781

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
This commit is contained in:
Kevin Hu
2025-10-27 11:43:56 +08:00
committed by GitHub
parent ff34c4232e
commit 3bd0b99495
2 changed files with 10 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
rag/llm/cv_model.py
View File

@ -228,7 +228,7 @@ class QWenCV(GptV4):
base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1" base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1"
super().__init__(key, model_name, lang=lang, base_url=base_url, **kwargs) super().__init__(key, model_name, lang=lang, base_url=base_url, **kwargs)
def chat(self, system, history, gen_conf, images=None, video_bytes=None, filename=""): def chat(self, system, history, gen_conf, images=None, video_bytes=None, filename="", **kwargs):
if video_bytes: if video_bytes:
try: try:
summary, summary_num_tokens = self._process_video(video_bytes, filename) summary, summary_num_tokens = self._process_video(video_bytes, filename)
@ -547,7 +547,7 @@ class OllamaCV(Base):
except Exception as e: except Exception as e:
return "**ERROR**: " + str(e), 0 return "**ERROR**: " + str(e), 0
def chat(self, system, history, gen_conf, images=None): def chat(self, system, history, gen_conf, images=None, **kwargs):
try: try:
response = self.client.chat( response = self.client.chat(
model=self.model_name, model=self.model_name,
@ -561,7 +561,7 @@ class OllamaCV(Base):
except Exception as e: except Exception as e:
return "**ERROR**: " + str(e), 0 return "**ERROR**: " + str(e), 0
def chat_streamly(self, system, history, gen_conf, images=None): def chat_streamly(self, system, history, gen_conf, images=None, **kwargs):
ans = "" ans = ""
try: try:
response = self.client.chat( response = self.client.chat(
@ -633,7 +633,7 @@ class GeminiCV(Base):
return res.text, total_token_count_from_response(res) return res.text, total_token_count_from_response(res)
def chat(self, system, history, gen_conf, images=None, video_bytes=None, filename=""): def chat(self, system, history, gen_conf, images=None, video_bytes=None, filename="", **kwargs):
if video_bytes: if video_bytes:
try: try:
summary, summary_num_tokens = self._process_video(video_bytes, filename) summary, summary_num_tokens = self._process_video(video_bytes, filename)
@ -651,7 +651,7 @@ class GeminiCV(Base):
except Exception as e: except Exception as e:
return "**ERROR**: " + str(e), 0 return "**ERROR**: " + str(e), 0
def chat_streamly(self, system, history, gen_conf, images=None): def chat_streamly(self, system, history, gen_conf, images=None, **kwargs):
ans = "" ans = ""
response = None response = None
try: try:
@ -858,7 +858,7 @@ class AnthropicCV(Base):
gen_conf["max_tokens"] = self.max_tokens gen_conf["max_tokens"] = self.max_tokens
return gen_conf return gen_conf
def chat(self, system, history, gen_conf, images=None): def chat(self, system, history, gen_conf, images=None, **kwargs):
gen_conf = self._clean_conf(gen_conf) gen_conf = self._clean_conf(gen_conf)
ans = "" ans = ""
try: try:
@ -879,7 +879,7 @@ class AnthropicCV(Base):
except Exception as e: except Exception as e:
return ans + "\n**ERROR**: " + str(e), 0 return ans + "\n**ERROR**: " + str(e), 0
def chat_streamly(self, system, history, gen_conf, images=None): def chat_streamly(self, system, history, gen_conf, images=None, **kwargs):
gen_conf = self._clean_conf(gen_conf) gen_conf = self._clean_conf(gen_conf)
total_tokens = 0 total_tokens = 0
try: try:
@ -963,13 +963,13 @@ class GoogleCV(AnthropicCV, GeminiCV):
else: else:
return GeminiCV.describe_with_prompt(self, image, prompt) return GeminiCV.describe_with_prompt(self, image, prompt)
def chat(self, system, history, gen_conf, images=None): def chat(self, system, history, gen_conf, images=None, **kwargs):
if "claude" in self.model_name: if "claude" in self.model_name:
return AnthropicCV.chat(self, system, history, gen_conf, images) return AnthropicCV.chat(self, system, history, gen_conf, images)
else: else:
return GeminiCV.chat(self, system, history, gen_conf, images) return GeminiCV.chat(self, system, history, gen_conf, images)
def chat_streamly(self, system, history, gen_conf, images=None): def chat_streamly(self, system, history, gen_conf, images=None, **kwargs):
if "claude" in self.model_name: if "claude" in self.model_name:
for ans in AnthropicCV.chat_streamly(self, system, history, gen_conf, images): for ans in AnthropicCV.chat_streamly(self, system, history, gen_conf, images):
yield ans yield ans

2
rag/prompts/generator.py
View File

@ -755,7 +755,7 @@ async def run_toc_from_text(chunks, chat_mdl, callback=None):
# Merge structure and content (by index) # Merge structure and content (by index)
prune = len(toc_with_levels) > 512 prune = len(toc_with_levels) > 512
max_lvl = sorted([t.get("level", "0") for t in toc_with_levels])[-1] max_lvl = sorted([t.get("level", "0") for t in toc_with_levels if isinstance(t, dict)])[-1]
merged = [] merged = []
for _ , (toc_item, src_item) in enumerate(zip(toc_with_levels, filtered)): for _ , (toc_item, src_item) in enumerate(zip(toc_with_levels, filtered)):
if prune and toc_item.get("level", "0") >= max_lvl: if prune and toc_item.get("level", "0") >= max_lvl: