What are the hardware requirements for DPO Model Test1?

DPO Model Test1 requires approximately 43 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed DPO Model Test1 and how large is it?

DPO Model Test1 is developed by kimdeokgi, a model with 21.4b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does DPO Model Test1 perform on standard benchmarks?

DPO Model Test1 has the following published scores: MMLU 67.38. Compare against reference models on this page or on the LLM Explorer leaderboards.

DPO Model Test1 by kimdeokgi — VRAM 43GB, 32K context

Name: DPO Model Test1
Author: kimdeokgi

DPO Model Test1 is an open-source language model by kimdeokgi. Features: 21.4b LLM, VRAM: 43GB, Context: 32K, License: apache-2.0, LLM Explorer Score: 0.14, Arc: 65.7, HellaSwag: 83, MMLU: 67.4, GSM8K: 58.

Autotrain compatible En Endpoints compatible Llama Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/kimdeokgi/dpo_model_test1

DPO Model Test1 Benchmarks

ARC: 65.7 vs 96.7 (so35)^-32.1%

HellaSwag: 83.01 vs 95.3 (gpt4)^-12.9%

MMLU: 67.38 vs 88.3 (so35)^-23.7%

GSM8K: 58 vs 96.4 (so35)^-39.8%

LLME Score: 0.14011

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

DPO Model Test1 (kimdeokgi/dpo_model_test1)

🌟 Advertise your project 🚀

DPO Model Test1 Parameters and Internals

Model Type

alignment-tuned

Additional Notes

This is a test version of the model.

Supported Languages

en (unknown)

Training Details

Methodology:

State-of-the-art instruction fine-tuning methods including direct preference optimization (DPO). Models were linearly merged to boost performance after DPO training.

LLM Name	DPO Model Test1
Repository 🤗	https://huggingface.co/kimdeokgi/dpo_model_test1
Model Size	21.4b
Required VRAM	43 GB
Updated	2025-02-20
Maintainer	kimdeokgi
Model Type	llama
Model Files	4.8 GB: 1-of-9 5.0 GB: 2-of-9 5.0 GB: 3-of-9 4.9 GB: 4-of-9 5.0 GB: 5-of-9 5.0 GB: 6-of-9 4.9 GB: 7-of-9 5.0 GB: 8-of-9 3.4 GB: 9-of-9
Supported Languages	en
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.37.2
Is Biased	0
Tokenizer Class	LlamaTokenizer
Padding Token	</s>
Vocabulary Size	92544
Torch Data Type	bfloat16

Best Alternatives to DPO Model Test1

Best Alternatives	Context / RAM	Downloads	Likes
Luxia 21.4B Alignment V1.2	32K / 43 GB	8465	8
Luxia 21.4B Alignment V1.0	32K / 42.8 GB	13119	33
UNA ThePitbull 21.4B V2	32K / 43 GB	16	16
Final Model Test V2	32K / 43 GB	11	0
Sft Model Test1	32K / 43 GB	7	0
UNA ThePitbull 21.4 V1	32K / 43 GB	12	5
Merge Model Test2	32K / 42.8 GB	8	0
Merge Model Test1	32K / 42.8 GB	6	0
Alignment Model Test9	32K / 43 GB	41	0
Alignment Model Test3	32K / 43 GB	113	0

Note: green Score (e.g. "73.2") means that the model is better than kimdeokgi/dpo_model_test1.

Rank the DPO Model Test1 Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 53999 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

DPO Model Test1 by kimdeokgi

» All LLMs » kimdeokgi » DPO Model Test1 URL Share it on

DPO Model Test1 Benchmarks

DPO Model Test1 Parameters and Internals

Best Alternatives to DPO Model Test1

Rank the DPO Model Test1 Capabilities

What open-source LLMs or SLMs are you in search of? 53999 in total.