From 21010ef454fb25954b0914785180311fb077add9 Mon Sep 17 00:00:00 2001 From: sqhao Date: Tue, 19 Sep 2023 07:09:59 +0800 Subject: [PATCH] fix doc of TextDecoder (#1526) Signed-off-by: haoshengqiang Co-authored-by: haoshengqiang --- whisper/model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/whisper/model.py b/whisper/model.py index 3457fcfc6..69130022a 100644 --- a/whisper/model.py +++ b/whisper/model.py @@ -197,7 +197,7 @@ def forward(self, x: Tensor, xa: Tensor, kv_cache: Optional[dict] = None): """ x : torch.LongTensor, shape = (batch_size, <= n_ctx) the text tokens - xa : torch.Tensor, shape = (batch_size, n_mels, n_audio_ctx) + xa : torch.Tensor, shape = (batch_size, n_audio_ctx, n_audio_state) the encoded audio features to be attended on """ offset = next(iter(kv_cache.values())).shape[1] if kv_cache else 0