-
**Describe the bug**
I Attempt to train reward models of different size(3B/6B/30B), and found out that when PP > 1, two type of issues arise
3B/6B:
* TP=4, PP=1: ok
* TP=4, PP=2: the job hang…
zirui updated
4 hours ago
-
**🧐 Issue Summary**
When setting up ZQ2 from [Zilliqa ZQ2 repo](https://github.com/Zilliqa/zq2) using `docker-compose up -d`, the containers run successfully. However, there is an issue with the Ha…
-
### Description
I have a variety of different AWS/S3 profiles in my `~/.aws/credentials` and `~/.aws/config` files. I'd like to be able to either explicitly pass `profile` into `storage_options` or i…
-
Hi there,
I have a function `func Test(t *testing.T) *MyStruct` inside a normal `.go` file and not a `_test.go` file.
The linter returns with `Function Test missing the call to method parallel (para…
-
Cargo features are meant to be additive not mutually exclusive, `warp_inner()` and all functions that call this function break this additive rule. This could cause unintended behavior as a user won't …
-
# the code as follows:
```
CUDA_VISIBLE_DEVICES=1,2 accelerate launch train_flux_deepspeed_controlnet.py --config "train_configs/test_canny_controlnet.yaml"
```
# ERROE
```
The following values …
-
Awesome work! Just a quick question about the correct system prompt:
in the docs https://llama.meta.com/docs/model-cards-and-prompt-formats/llama3_1#user-defined-custom-tool-calling this is used:
…
-
I'm trying to use `cblas_domatcopy` to transpose large row-major matrices.
I'm finding that the function is slower than a simple loop of `cblas_dcopy` calls parallelized with OpenMP (with number of…
-
Hello Redribbon developers,
Thank you very much for the Redribbon software. I was able to analyze most of my epigenomics datasets.
However, I have some large datasets, with 9-13 million epigenomic…
-
### Describe the bug
With the python SDK. My top function is wrapped with `@observe()`. Inside there are many calls to gpt completion api via `AsyncAzureOpenAI` from `langfuse.openai`.
When the…