-
Hi,
I am facing a new error now while trying to run the alpaca fine-tuning notebook in kaggle-
…
-
Hello Unsloth Community!
I'm excited to invite you to collaborate on an exciting project involving the adaptation of SPPO for 4-bit quantization training. Currently, SPPO only supports full-precisi…
-
import os
import torch
from unsloth import FastLanguageModel
from datasets import load_dataset
from transformers import TrainingArguments, AdamW, TrainerCallback
from trl import SFTTrainer
# 禁…
-
JoyCaptionTwo
User specified an unsupported autocast device_type 'cuda:0'
# ComfyUI Error Report
## Error Details
- **Node Type:** Joy_caption_two
- **Exception Type:** RuntimeError
- **Exce…
-
Thank you for your work. However, I've noticed some performance issues that differ significantly when compared to the Llama 3.1 model. Specifically, I've observed the following problems:
# Issue Desc…
-
I'm getting an error while using the conversational colab:
`: 'list' object has no attribute 'keys'`
I just replaced: `load_dataset("json", data_files = {"train" : "my_data.json"})`
To use a fi…
-
### Please check that this issue hasn't been reported before.
- [X] I searched previous [Bug Reports](https://github.com/axolotl-ai-cloud/axolotl/labels/bug) didn't find any similar reports.
### Exp…
-
After doing QLoRA with a training library (unsloth) and saving the adapter, is there a way to load the 4 bit bnb model and the un-merged adapter for use with sglang? Last time I tried loading a model …
-
This is my trial for corpus training in unsloth. model load is the same as the example of unsloth code.
![image](https://github.com/unslothai/unsloth/assets/97992669/be530784-428c-4c23-9637-d1d94c5…
-
model = FastLanguageModel.get_peft_model(
model,
r = 16, # Choose any number > 0 ! Suggested 8, 16, 32, 64, 128
target_modules = ["q_proj", "k_proj", "v_proj", "o_proj",
…