-
[Daniel Han](https://twitter.com/danielhanchen) in his [blogpost](https://unsloth.ai/blog/gemma-bugs) shared his discoveries of what is ~also~ wrong with the Gemma implementation.
Some of them are on…
-
### Open Task RFP for Privacy preserving machine learning inference using MPC
#### Executive Summary
- Project Overview: In this project, we want to see current state of the privacy preserving m…
tkmct updated
1 month ago
-
Any speed testment?
-
First, thanks for this work. Providing open source SAEs for a model like Llama is a huge boon to the community.
I'm working on a simple script to use your `generate_description` function to assign …
-
During LoRa training, iterate_batches [calls](https://github.com/ml-explore/mlx-examples/blob/main/llms/mlx_lm/tuner/trainer.py#L104) `tokenizer.encode()` (with default arguments) on the dataset item,…
-
I encountered this error when I was adding text. Hope to get a solution to deal with this error.Thank you very much.
Traceback (most recent call last):
File "/home/jyc23/raptor-master/demo/newdemo…
-
**Describe the bug**
The model response doesn't stop. It keeps writing. I tried both `swift deploy` and `vllm`
Training arguments:
```bash
HF_HUB_ENABLE_HF_TRANSFER=1 \
USE_HF=1 \
CUDA_VISIBLE…
-
Evaluating gemma-2b with xcopa looks good, but the xnli result looks weird.
xcopa result:
```
"results": {
"xcopa_zh": {
"acc,none": 0.616,
"acc_stderr,none": 0.021772369465…
-
Hi team, very thanks for GaLore. I'm currently using HuggingFace for fine-tuning. Just curious to integrate GaLore with HuggingFace.
It's not an issue, I'm just interested to use GaLore with Huggin…
-
### What is the issue?
I just set the chat format to JSON, then the Ollama’s speed in generating chat content slowed down by tenfold.
For example, when I use the gemma7b model and the chat forma…