-
Hi @sgrvinod , in the caption.py, line 97,
`scores = top_k_scores.expand_as(scores) + scores # (s, vocab_size)`
I wonder why it's added? Shouldn't it be multiplied?
-
Traceback (most recent call last):
File "test.py", line 77, in
scores = predict_captions(model, dict_dataloader_test, text_field)
File "test.py", line 26, in predict_captions
out, _ =…
-
### Describe the Bug
Leaderboard Cohorts use Stats search criteria for evaluating Stats-based eligibility rules. The validation has an incorrect list of relation names. Notably it specifies "le" in…
-
when i use `n` option is different as openai.
when i use n it turn to use beam search.
-
**Describe the bug**
When I use kernel injection I get worse generation results than when using transformers without DeepSpeed.
I don't know if the results should be the same, but they are not only…
-
We found in one of the issues that TRTLLM supports Grouped Diverse Beam Search : https://github.com/NVIDIA/TensorRT-LLM/issues/79#issuecomment-1825401751
Yet, we are unable to change the groups and…
-
Hi,
I have trained a model using Reinforcement learning.
When I use **"beam search"** to generate text, it generates all
```
"raeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraer…
-
### System Info
nightly build transformers version
### Who can help?
@ArthurZucker
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [ ] An officiall…
Mirmu updated
6 months ago
-
### Branch/Tag/Commit
v5.2
### Docker Image Version
22.08-py3
### GPU name
V100
### CUDA Driver
none
### Reproduced Steps
```shell
use the fastertransformer triton backend …
-
While testing "something" created by me, I found out strange things that happen in `demo2.dm2` and `demo3.dm2` of the `rogue` expansion when changing `hand` to `1` or `2`.
In `demo2.dm2`, we can se…