-
Hi, I set` F.scaled_dot_product_attention = sageattn`, in modeling_llama.py, and run the inference code,
I see it run `sageattn_qk_int8_pv_fp16_cuda` in `sageattention/core.py`.
The results are:
…
-
why target_acc = 2*(x_start - noise) - 2*pred_vel in the pipeline_caf.py ?
I think change it into target_acc = 2*(x_start - noise) - 2*target_vel maybe a little better. Because in a simple thought, …
-
I'm currently talking with Paul and he thinks we should remove the USE_ACC preprocessor directive. I think we could start by writing a python script to find all snippets of form
```cpp
#if USE_ACC
#p…
-
Hi guys,
I'm reaching out for a suggestion on how to mix together the good features of the ACC and EIDM car-following models in order to create a new model that resembles an ACC infrequently assist…
-
Hiện thị đầy đủ Read User, chắc sẽ chỉ có mỗi read thôi tutu cập nhật sau
-
Hi, maybe there something wrong in my step. It's like the train.pt and other file is empty?…
-
### Describe the request
Please make the procedure SetDimFilters global.
**Current**
` local procedure SetDimFilters(DimNo: Integer; DimValueFilter: Text)`
**New**
`procedure SetDimFilters(DimNo:…
-
above is no augment,,and then i equip a item
![9edc379540ad85bb47acdd27c29cad…
-
I trained a model by myself. The accuracy (acc) on the validation set is always only 50. When I test with the pre-trained model released by the author, the acc is also only 50. I only changed the rele…
-
Hi, Thank you for the amazing work! When I run `python train_nhp.py --experiment_id IntensityFree_train`, I found it only printed out loglike and num_events, like this:
INFO: [ Epoch 450 (train) ]:…