-
### 🚀 The feature, motivation and pitch
Motivation is to allow ReFT representations to be applied on the fly during inference, which can be done in a batchwise manner.
this is much faster than app…
-
in the code the kv weight is merged as
```python
self.kv = nn.Linear(dim, dim * 2, bias=qkv_bias)
```
to possible speed up.
but when we want to use lora to do fune-tuning, we need to split v o…
-
Anytime I try to use a Lora with SD XL, I get this error message
Error: Could not load the lora model! Reason: 'stable-diffusion'
-
### The Feature
The current code for initializing an offline LLM in LiteLLM is as follows
```python
# check if vllm is installed
def validate_environment(model: str):
global llm
t…
-
Heads up that I've created a public syslog server which can ingest logs from an iGate. It lets the user view their own logs as well as plots data on maps - including signal strength etc. This is purel…
-
**Describe the bug**
During inference, I got `AttributeError: 'NoneType' object has no attribute 'set_moe'`. I also noticed `Policy type not supported` as an INFO log line.
```console
(multivu) …
-
Great work! One quick question: in the paper you've reproduced the results from Llava. Additionally, for the Prismatic models experiments, you are fine-tuning the whole LM. I'm wondering did you try u…
-
Stability AI's Control LoRA has been around for a while: https://huggingface.co/stabilityai/control-lora.
We also have had requests for it, e.g., https://github.com/huggingface/diffusers/issues/46…
-
Hi, i have trained LORA and got a .safetensors file, how can i use it to create images？
-
Hello,
when trying to reproduce Nvidia's results on DGX H100, the code cannot be executed and results in Segmentation Fault - see attached file
We have found that the error disappears when TP_C…