phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2026-01-01 09:10:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

update test to try multi-turn scenarios

Browse source
This commit is contained in:
Hardik Shah 2025-04-06 12:13:59 -07:00
parent eafbde4e17
commit cd618e9ad0
3 changed files with 198 additions and 64 deletions
Download patch file Download diff file Expand all files Collapse all files

8
llama_stack/models/llama/llama4/chat_format.py
View file

@ -203,13 +203,13 @@ class ChatFormat:
tokens.extend(toks)
images.extend(imgs)
if message.role == "assistant" and len(message.tool_calls) > 0:
tokens.append(self.tokenizer.special_tokens["<|python_start|>"])
# if message.role == "assistant" and len(message.tool_calls) > 0:
# tokens.append(self.tokenizer.special_tokens["<|python_start|>"])
_process_content(message.content)
if message.role == "assistant" and len(message.tool_calls) > 0:
tokens.append(self.tokenizer.special_tokens["<|python_end|>"])
# if message.role == "assistant" and len(message.tool_calls) > 0:
# tokens.append(self.tokenizer.special_tokens["<|python_end|>"])
if message.role == "user" and message.context is not None:
# This is RAG context; why is it here in the chat format? I don't think