huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-12 04:44:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

remove debugging lines

Browse Source
This commit is contained in:
OlivierDehaene 2024-10-09 20:05:39 +02:00
parent d361197aab
commit f85a308ef1
No known key found for this signature in database
GPG Key ID: BB104D67809DA93C
1 changed files with 0 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
server/text_generation_server/models/flash_causal_lm.py
View File

@ -2018,12 +2018,6 @@ class FlashCausalLM(Model):
top_token_ids, top_token_ids,
top_token_logprobs, top_token_logprobs,
) in enumerate(iterator): ) in enumerate(iterator):
if all_input_ids[:2] == [1986, 374] and not request_is_prefilling:
log_master(
logger.info,
f"{request.id} {next_token_ids} {self.tokenizer.batch_decode(next_token_ids)}",
)
# Compute logprobs first as, even though we might skip the token, # Compute logprobs first as, even though we might skip the token,
# it can still be required to compute the logprobs # it can still be required to compute the logprobs
# modulo on request.id as it is robust to batch.filter whereas the index in the batch is not and we need # modulo on request.id as it is robust to batch.filter whereas the index in the batch is not and we need