We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent d904304 commit b36ea66Copy full SHA for b36ea66
src/forge/actors/reference_model.py
@@ -128,7 +128,7 @@ async def forward(
128
Args:
129
input_ids (torch.Tensor): input token ids with shape [group_size, req + res length].
130
max_req_tokens (int): maximum request length.
131
- return_logprobs (bool): whether to return og probabilities instead of raw logits.
+ return_logprobs (bool): whether to return log probabilities instead of raw logits.
132
133
return_logprobs flag significantly impacts the amount of data transferred to the caller:
134
- When False: Returns logits with shape [group_size, req + res_length, vocab_size].
0 commit comments