-
Hello,
I updated today for the first time in a week or two and have gotten this type error twice now after running a few successful merge generations. Unfortunately I don't have time to try to narr…
CCpt5 updated
2 months ago
-
# 🐛 Bug
I'm trying to create a 1:1 config that can train a stable ViT-B with the MAE config (from appendix A.2).
Maybe I'm missing something (highly plausible), but when I use xformers instead …
-
Summary of thoughts from a [twitter thread](https://twitter.com/IsabellaGhement/status/1779537903317848101)
1. Can it be made clearer that GLMs model the conditional distribution of Y given X (as …
-
### Description
python3 main.py --text "a hamburger" --workspace trial -O --backbone grid_taichi
Get this error:
[Taichi] version 1.6.0, llvm 15.0.1, commit f1c6fbbd, win, python 3.9.13
No CUD…
-
### System Info
ghcr.io/huggingface/text-generation-inference:2.2.0
### Information
- [X] Docker
- [ ] The CLI directly
### Tasks
- [X] An officially supported command
- [ ] My own modifications
…
-
### What is the issue?
qwen4b works fine, all other models larger than 4b are gibberish
```
time=2024-09-05T11:35:49.569+08:00 level=INFO source=download.go:175 msg="downloading 8eeb52dfb3bb in 1…
-
### What is the issue?
Generating a response after first starting Ollama works flawlessly from what I can tell. I am able to change models and generate responses from prompts. After the model unloa…
-
> jaxlib.xla_extension.XlaRuntimeError: UNIMPLEMENTED: No registered implementation for custom call to te_scaled_upper_triang_masked_softmax_forward for platform CUDA
```python
from transformer_en…
-
Hi there,
this project _sounds_ really cool !
I've started a [discussion @ HydrogenAudio](https://hydrogenaud.io/index.php/topic,126345.0.html) about the similar @bkraad47's [Fat Llama project](ht…
-
### System Info
- `transformers` version: 4.43.0.dev0
- Platform: Linux-5.15.0-113-generic-x86_64-with-glibc2.35
- Python version: 3.11.4
- Huggingface_hub version: 0.23.3
- Safetensors version: …