-
Hi,
I encountered an issue after updating to unsloth=="2024.11.6". When training the `Qwen2.5-0.5B-Instruct` model without PEFT, I observed that the model's gradient norm is 0, resulting in no weig…
-
I just run
```python
from unsloth import FastLanguageModel
model = GemmaForCausalLM.from_pretrained('PATH TO GEMMA2B')
model = FastLanguageModel.get_peft_model(
model,
…
-
Everything installed OK but am getting this message, any ideas?
-
### System Info
Pytorch version: 2.4.1+cu124
Cuda version: 12.7
GPU: A100 80G * 1
### Information
- [ ] The official example scripts
- [x] My own modified scripts
### 🐛 Describe the bug…
-
### System Info / 系統信息
模型：glm4-9B-chat
配置文件：
data_config:
train_file: train.jsonl
val_file: dev.jsonl
test_file: dev.jsonl
num_proc: 1
max_input_length: 3500
max_output_length: 250…
-
through hf is fine - test w. a lora
-
Why am i getting this error when i am trying to load the checkpoint of lora fine tuned phi-1_5 model. Following is the complete error log
```
phi) harsh@harsh:~/phi-1_5$ python3 inference_lora_phi.p…
-
Hi, I meet an issue about the finetuning speed.
At first, i use swift(verion==2.3.0dev) to finetune InternVL2 using my custom dataset on single A6000. The finetune process last about 20 hours.
Howe…
-
Hi, currently peft is supported in train.py with lora_enable flag, however during inference when loading model.builder, there seems to be a mismatch if the model was trained with lora or qlora. Wonder…
-
### System Info
- `transformers` version: 4.44.0
- Platform: Linux-4.18.0-553.16.1.el8_10.x86_64-x86_64-with-glibc2.28
- Python version: 3.9.4
- Huggingface_hub version: 0.23.2
- Safetensors vers…