-
I know this may not be supported given the added complexity, but I just wanted some clarification because I was surprised to see that I could pass these parameters to the server. Right now I non-deter…
-
Hi, I am reading your excellent code, but find no beam search during caption generation as the source code in https://github.com/kelvinxu/arctic-captions, is there any reason ?
-
`
model_name = "meta-llama/Llama-2-7b-chat-hf"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = Model()
model.init(model_name, use_quant=True, weight_dtype="in…
-
This feature was proposed by @spion in https://github.com/ggerganov/llama.cpp/issues/2813#issuecomment-1694390583
> In some cases, its useful to do constrained evaluation of logits based on a union…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Problem description
Steps to reproduce:
1. Start FreeCAD; create new document; switch to BIM workbenc…
-
Hi, is there a way to apply beam search in order to generate text?
-
beam search 在当一个序列达到end后的处理貌似不对。
-
Beam search tends to work better as compared to temperature based sampling.
-
According to README:
> Currently, we export components (encoder, decoder) to Caffe2 separately and beam search is implemented in C++. In the near future, we will be able to export the beam search as …
-
I'm using the scorer generator provided `generate_scorer_package`. I'm also using (e.g., SentencePiece) to build a unigram language model, where the decoder predicts the size of the language model. Ho…