phoenix-oss/llama-stack
1
0
Fork 0
forked from phoenix-oss/llama-stack-mirror
Code Pull requests Releases Packages 2 Activity Actions

fix: meta reference + llama4 tokenizer fix

Browse source
This commit is contained in:
Ashwin Bharambe 2025-04-09 00:46:02 -07:00
parent 10882bf478
commit 8001c30a4f
2 changed files with 5 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama_stack/models/llama/llama4/tokenizer.py
View file

@ -56,9 +56,11 @@ LLAMA4_TEXT_POST_TRAIN_SPECIAL_TOKENS = [
"<|text_post_train_reserved_special_token_3|>",
"<|text_post_train_reserved_special_token_4|>",
"<|text_post_train_reserved_special_token_5|>",
"<|text_post_train_reserved_special_token_6|>",
"<|text_post_train_reserved_special_token_7|>",
"<|finetune_right_pad|>",
] + get_reserved_special_tokens(
"text_post_train", 61, 6
"text_post_train", 61, 8
) # <|text_post_train_reserved_special_token_6|>, ..., <|text_post_train_reserved_special_token_66|>
# 200080, ..., 201133

4
llama_stack/providers/inline/inference/meta_reference/generators.py
View file

@ -259,7 +259,7 @@ class Llama3Generator:
temperature, top_p = _infer_sampling_params(sampling_params)
for result in self.inner_generator.generate(
llm_inputs=[self.formatter.encode_content(request.content)],
model_inputs=[self.formatter.encode_content(request.content)],
max_gen_len=max_gen_len,
temperature=temperature,
top_p=top_p,
@ -284,7 +284,7 @@ class Llama3Generator:
temperature, top_p = _infer_sampling_params(sampling_params)
for result in self.inner_generator.generate(
llm_inputs=[self.formatter.encode_dialog_prompt(request.messages, _infer_tool_prompt_format(request))],
model_inputs=[self.formatter.encode_dialog_prompt(request.messages, _infer_tool_prompt_format(request))],
max_gen_len=max_gen_len,
temperature=temperature,
top_p=top_p,