What are the hardware requirements for Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B?

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B requires approximately 141.9 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B and how large is it?

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B is developed by Saxo, a model with 70b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B?

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B by Saxo — VRAM 141.9GB, 128K context

Name: Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B
Author: Saxo

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B is an open-source language model by Saxo. Features: 70b LLM, VRAM: 141.9GB, Context: 128K, License: apache-2.0, Quantized, Instruction-Based, LLM Explorer Score: 0.14.

Base model:nousresearch/meta-l... Base model:quantized:nousresea... Cn Conversational Dataset:maywell/ko ultrafeedba... Dataset:saxo/en ko translation... Dataset:saxo/en ko translation... Dataset:saxo/ko aspect sentime... Dataset:saxo/ko cn translation... Dataset:saxo/ko government qa ... Dataset:saxo/ko jp translation... Dataset:saxo/ko summarization ... Dataset:saxo/openorca cleaned ... En Endpoints compatible Ggml Gguf Instruct Jp Ko Llama Q4 Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-llama3.1-sft-dpo-70B

Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B Parameters and Internals

Model Type

text-generation

Additional Notes

Tokenizer uses base model without word expansion. Model enhanced for customer reviews, social postings analysis, coding tasks, supports 128k-Context Window, Tool Calling, knows Korean national anthem lyrics accurately.

Supported Languages

ko (full), en (full), jp (full), cn (full)

Training Details

Data Sources:

Saxo/ko_cn_translation_tech_social_science_linkbricks_single_dataset, Saxo/ko_jp_translation_tech_social_science_linkbricks_single_dataset, Saxo/en_ko_translation_tech_science_linkbricks_single_dataset_with_prompt_text_huggingface, Saxo/en_ko_translation_social_science_linkbricks_single_dataset_with_prompt_text_huggingface, Saxo/ko_aspect_sentiment_sns_mall_sentiment_linkbricks_single_dataset_with_prompt_text_huggingface, Saxo/ko_summarization_linkbricks_single_dataset_with_prompt_text_huggingface, Saxo/OpenOrca_cleaned_kor_linkbricks_single_dataset_with_prompt_text_huggingface, Saxo/ko_government_qa_total_linkbricks_single_dataset_with_prompt_text_huggingface_sampled, maywell/ko_Ultrafeedback_binarized

Methodology:

fine-tuned with SFT->DPO

Context Length:

128000

Hardware Used:

KT-CLOUD H100-80G x4

LLM Name	Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B
Repository 🤗	https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-llama3.1-sft-dpo-70B
Base Model(s)	NousResearch/Meta-Llama-3.1-70B-Instruct NousResearch/Meta-Llama-3.1-70B-Instruct
Model Size	70b
Required VRAM	141.9 GB
Updated	2026-07-30
Maintainer	Saxo
Model Type	llama
Instruction-Based	Yes
Model Files	42.5 GB 49.9 GB 4.6 GB: 1-of-30 4.7 GB: 2-of-30 5.0 GB: 3-of-30 5.0 GB: 4-of-30 4.7 GB: 5-of-30 4.7 GB: 6-of-30 4.7 GB: 7-of-30 5.0 GB: 8-of-30 5.0 GB: 9-of-30 4.7 GB: 10-of-30 4.7 GB: 11-of-30 4.7 GB: 12-of-30 5.0 GB: 13-of-30 5.0 GB: 14-of-30 4.7 GB: 15-of-30 4.7 GB: 16-of-30 4.7 GB: 17-of-30 5.0 GB: 18-of-30 5.0 GB: 19-of-30 4.7 GB: 20-of-30 4.7 GB: 21-of-30 4.7 GB: 22-of-30 5.0 GB: 23-of-30 5.0 GB: 24-of-30 4.7 GB: 25-of-30 4.7 GB: 26-of-30 4.7 GB: 27-of-30 5.0 GB: 28-of-30 5.0 GB: 29-of-30 2.1 GB: 30-of-30
Supported Languages	ko en jp cn
GGML Quantization	Yes
GGUF Quantization	Yes
Quantization Type	ggml\|q4\|gguf\|q4_k\|q5_k
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	131072
Model Max Length	131072
Transformers Version	4.43.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|eot_id\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

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

Best Alternatives to Linkbricks Horizon AI Korean Llama3.1 Sft DPO 70B

Best Alternatives	Context / RAM	Downloads	Likes
Llama3 70B RAG	8K / 141.9 GB	21	0
Grok Oss Revenant 70B	128K / 141.9 GB	8503	2
Llama 3.3 70B Instruct GGUF	128K / 15.9 GB	23398	124
Reflection Llama 3.1 70B Bf16	128K / 141.9 GB	283	6
...ma3 70B Chinese Chat GGUF 4bit	8K / 40 GB	196	18
Meta Llama 3 70B Instruct	8K / 40.3 GB	6	1
CodeLlama 70B Instruct Hf GGUF	4K / 25.5 GB	487	2
...B Instruct Gradient 1048K 8bit	1024K / 75 GB	41	5
...B Instruct Gradient 1048K 4bit	1024K / 39.7 GB	25	3
...B Instruct Gradient 1048K 2bit	1024K / 21.9 GB	25	2

Note: green Score (e.g. "73.2") means that the model is better than Saxo/Linkbricks-Horizon-AI-Korean-llama3.1-sft-dpo-70B.