solar-sft-qlora

cockroach54/solar-sft-qlora

solar-sft-qlora is a text generation model from cockroach54. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.

- just for practice - trained by transformers, peft, bitsandbytes - evaluated by https://github.com/davidkim205/kollmevaluation

Downloads · 30 days

0% of all-time downloads

All-time downloads

37.3K

Public

Parameters

10.7B

21.5 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors21.5 GB · 100%

At a glance

Task: Text Generation
Library: transformers
License: apache-2.0
Model type: llama
Access: Public
Created: Apr 7, 2024
Updated: Apr 8, 2024
SHA: 2b783cbb

Datasets

davidkim205/kollm-converations

Spaces

Model card

From the Hugging Face model README

just for practice
trained by transformers, peft, bitsandbytes
evaluated by https://github.com/davidkim205/kollm_evaluation

Tasks	Version	Filter	Metric	Value		Stderr
kobest_wic	1	none	acc	0.5056	±	0.0141
		none	f1	0.4170	±	N/A
kobest_sentineg	1	none	acc	0.8791	±	0.0164
		none	f1	0.8779	±	N/A
kobest_hellaswag	1	none	acc	0.4540	±	0.0223
		none	f1	0.4522	±	N/A
		none	acc_norm	0.5600	±	0.0222
kobest_copa	1	none	acc	0.6540	±	0.0151
		none	f1	0.6535	±	N/A
kobest_boolq	1	none	acc	0.5135	±	0.0133
		none	f1	0.3592	±	N/A
ko_truthfulqa	2	none	acc	0.2938	±	0.0159
ko_hellaswag	1	none	acc	0.3816	±	0.0048
		none	acc_norm	0.5171	±	0.0050
ko_common_gen	1	none	acc	0.8095	±	0.0100
		none	acc_norm	0.8095	±	0.0100
ko_arc_easy	1	none	acc	0.3311	±	0.0138
		none	acc_norm	0.4053	±	0.0143

Task: Text Generation
Library: transformers
Type: llama
License: apache-2.0
Languages: ko, en
Created: Apr 7, 2024
Updated: Apr 8, 2024