roshan-gopalakrishnan comments

Results 8 comments of


                                            roshan-gopalakrishnan

how can I finetune the model with my own dataset on just one GPU?

I had the same flash_attn issue, I solved it by installing python develop version apt-get install python3-dev 3.11 I am using python3.11 version. You must change the version accordingly.

Issue:"Your setup doesn't support bf16/gpu. You need torch>=1.10, using Ampere GPU with cuda>=11.0"

Changing --bf16 to False, did not help me. I got another error ValueError: --tf32 requires Ampere or a newer GPU arch, cuda>=11 and torch>=1.7 Any suggestions ?

LoRA rank of the models & more metadata

Can I use larger value of r than lora_alpha ? It says the decomposition of matrix depends on lora_alpha/r. How to choose r and alpha values for finetuning ? Does...

Sample Code: OpenAI service failed to complete the prompt, APIConnectionError issue with Python 3.10

If you are giving any org_id just comment it in the .env. It is optional.

Help Needed: Fine-tuning codellama-7b-Instruct Model for pinescript Programming language

@Humza1996 did you able to find a solution to this ?

Different hyper-parameter between the paper and the code? (lora_alpha and a global batch size)

Is there any reply to these questions ?

ImportError: No module named six.moves

Anybody resolved this issue ?

Got Error in example.py: Pipeline element is None!

I had a similar error while using GStreamer and RTSP server. I used the command -- ./examples/test-launch "nvarguscamerasrc ! video/xraw(memory:NVMM) width=1920 height=1080 framerate=30/1 format=NV12 ! omxh264enc iframeinterval=15! h264parse ! rtph264pay...