starcoder

starcoder copied to clipboard

I want to fine tuning starcoder to generate json rule code.

4

How can I customize my dataset? this is the snippet of json. ```json { "type": "page", "body": { "type": "collapse-group", "activeKey": [ "1" ], "body": [ { "type": "collapse", "key":...

CharellKing

Have a question about learning rate decay?

1

Why is it that I set the learning rate decay type in config.yaml, but the learning rate does not change in the log file output when the model is trained?

SerenNoble

Training getting struck

4

I'm trying to train on A100 GPU but the training is struck at. I can't see any logs other than this UserWarning: MatMul8bitLt: inputs will be cast from torch.bfloat16 to...

sankethgadadinni

Provides guidance to avoid error when downloading pre-trained model

massenz

Hi, I have a set of p4 (A100) instances available through Sagemaker training jobs. I would like to finetune StarCoder on a function summarization task. Would I be able to...

dshah3

Finetune with H100 and CUDA 11.8

1

Hello, I have been trying to use the finetune.py script with my own dataset on a single H100 GPU with CUDA 11.8 I have been getting the following error. The...

drorbrillsnps

Explain code

3

How can I explain code in this model？I know that code interpretation can be done through the application of chatCould. But could you please give us some api or code...

CodingmanJC

Demo snippet pulls all checkpoints

2

`tokenizer = AutoTokenizer.from_pretrained(checkpoint)` as defined here - https://github.com/bigcode-project/starcoder#code-generation pulls 7 checkpoint files, ~9GB each. Is this the intended behavior ?

dhingratul

Why do we have 2 scripts for fine-tuning?

3

Hello, I have been experimenting with fine-tuning StarCoder and I see there are 2 different scripts for fine-tuning, both of which handle the data processing differently and also, one uses...

samin-batra

 should be more **than**, not more **that**

SmartMapple