dearwolf/aider
1
0
Fork 0
mirror of https://github.com/Aider-AI/aider.git synced 2025-05-28 16:25:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

refactor: improve context window size handling for Ollama models

Browse source
This commit is contained in:
Paul Gauthier 2025-01-26 19:09:27 -08:00 • committed by Paul Gauthier (aider)
parent 980197cb05
commit 3dec9e531f
1 changed files with 19 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

38
aider/coders/base_coder.py
View file

@ -1235,6 +1235,8 @@ class Coder:
input_tokens = self.main_model.token_count(messages)
max_input_tokens = self.main_model.info.get("max_input_tokens") or 0
proceed = None
if max_input_tokens and input_tokens >= max_input_tokens:
self.io.tool_error(
f"Your estimated chat context of {input_tokens:,} tokens exceeds the"
@ -1244,28 +1246,26 @@ class Coder:
self.io.tool_output("- Use /drop to remove unneeded files from the chat")
self.io.tool_output("- Use /clear to clear the chat history")
self.io.tool_output("- Break your code into smaller files")
proceed = "y"
self.io.tool_output(
"It's probably safe to try and send the request, most providers won't charge if"
" the context limit is exceeded."
)
# Special warning for Ollama models about context window size
if self.main_model.name.startswith(("ollama/", "ollama_chat/")):
extra_params = getattr(self.main_model, "extra_params", None) or {}
num_ctx = extra_params.get("num_ctx")
if num_ctx:
self.io.tool_waning(
f"Your Ollama model is configured with num_ctx={num_ctx} tokens of"
" context window\nSee"
" https://aider.chat/docs/llms/ollama.html#setting-the-context-window-size"
" for help configuring larger context windows."
)
proceed = "n"
else:
proceed = "y"
self.io.tool_output(
"It's probably safe to try and send the request, most providers won't charge if"
" the context limit is exceeded."
# Special warning for Ollama models about context window size
if self.main_model.name.startswith(("ollama/", "ollama_chat/")):
extra_params = getattr(self.main_model, "extra_params", None) or {}
num_ctx = extra_params.get("num_ctx", 8192)
if max_input_tokens and max_input_tokens > num_ctx:
self.io.tool_waning(
f"Your Ollama model is configured with num_ctx={num_ctx} tokens of"
" context window\nSee"
" https://aider.chat/docs/llms/ollama.html#setting-the-context-window-size"
" for help configuring larger context windows."
)
if not self.io.confirm_ask("Try to proceed anyway?", default=proceed):
return False
if proceed and not self.io.confirm_ask("Try to proceed anyway?", default=proceed):
return False
return True
def send_message(self, inp):