phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-27 23:50:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: make cprint write to stderr

Browse source
This commit is contained in:
Raghotham Murthy 2025-05-24 21:46:40 -07:00
parent c290999c63
commit 8658109454
11 changed files with 81 additions and 44 deletions
Download patch file Download diff file Expand all files Collapse all files

6
llama_stack/models/llama/llama4/generation.py
View file

@ -133,9 +133,9 @@ class Llama4:
print_model_input = print_model_input or os.environ.get("LLAMA_MODELS_DEBUG", "0") == "1"
if print_model_input:
cprint("Input to model:\n", "yellow")
cprint("Input to model:\n", color="yellow", file=sys.stderr)
for inp in llm_inputs:
cprint(self.tokenizer.decode(inp.tokens), "grey")
cprint(self.tokenizer.decode(inp.tokens), color="grey", file=sys.stderr)
prompt_tokens = [inp.tokens for inp in llm_inputs]
bsz = len(llm_inputs)
@ -145,7 +145,7 @@ class Llama4:
max_prompt_len = max(len(t) for t in prompt_tokens)
if max_prompt_len >= params.max_seq_len:
cprint(f"Out of token budget {max_prompt_len} vs {params.max_seq_len}", "red")
cprint(f"Out of token budget {max_prompt_len} vs {params.max_seq_len}", color="red", file=sys.stderr)
return
total_len = min(max_gen_len + max_prompt_len, params.max_seq_len)