Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · Dampfinchen

Llama-3-8B-Ultra-Instruct-SaltSprinkle

Llama-3-8B-Ultra-Instruct-SaltSprinkle is an 8.03B-parameter Llama merge from Dampfinchen. According to the model card, it combines Ultra-Instruct and Llama-3-8B-Instruct with DARE TIES.

Publisher
Dampfinchen
Task
text-generation
Model type
llama
License
llama3
Library
transformers
Publication status
Accepted · not indexed

Model overview

Llama-3-8B-Ultra-Instruct-SaltSprinkle is published by Dampfinchen as a text-generation model. The captured configuration identifies LlamaForCausalLM with a llama model type, and Safetensors metadata reports 8030261248 parameters. According to the model card, the merge uses NousResearch/Meta-Llama-3-8B as the base with mergekit.

Recorded capabilities

DARE TIES merge

According to the model card, the model was merged with the DARE TIES method using mergekit.

Documented merge config

The card lists Meta-Llama-3-8B-Instruct at density 1 and weight 1 alongside Ultra-Instruct at density 0.5 and weight 0.2 in bfloat16.

8.03B Llama scale

Captured config identifies LlamaForCausalLM and Safetensors metadata reports 8030261248 parameters.

Use cases in the source record

  • Instruction-following text generation consistent with the stated goal of retaining L3 Instruct capabilities.
  • Roleplay, retrieval-assisted, German, and story-writing trials matching the publisher's described Ultra-Instruct additions.

Limitations and unknowns

  • According to the model card, the model may generate harmful responses; safe-use assessment was not independently verified.
  • The card mentions Open LLM Leaderboard results but no metric values were extracted from this record.
  • No evaluation results were extracted from this record.
  • No context-window value was extracted from this record.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: Dampfinchen/Llama-3-8B-Ultra-Instruct-SaltSprinkle

Captured: Unknown. Processed: 2026-09-07T19:34:29.970610+00:00.

merge This is a merge of pre-trained language models created using mergekit . Merge Details Merge Method This model was merged using the DARE TIES merge method using NousResearch/Meta-Llama-3-8B as a base. Models Merged The following models were included in the merge: Dampfinchen/Llama-3-8B-Ultra-Instruct NousResearch/Meta-Llama-3-8B-Instruct Configuration The following YAML configuration was used to produce this model: models: - model: NousResearch/Meta-Llama-3-8B-Instruct parameters: density: 1 weight: 1 - model: Dampfinchen/Llama-3-8B-Ultra-Instruct parameters: density: 0.5 weight: 0.2 merge_method: dare_ties base_model: NousRese…

F001F002F003F004F005F006F007F008F009F010F011F012F013F014