license, tags, pipeline_tag, base_model, model-index

license	tags	pipeline_tag	base_model	model-index
cc-by-nc-4.0	merge conversational multi-task	text-generation	paulml/OmniBeagleSquaredMBX-v3-7B ZySec-AI/ZySec-7B-v1 liminerity/Omningotex-7b-slerp localfultonextractor/Erosumika-7B KatyTheCutie/LemonadeRP-4.5.3 cgato/Thespis-Krangled-7b CorticalStack/pastiche-crown-clown-7b-dare snorkelai/Snorkel-Mistral-PairRM-DPO MTSAIR/multi_verse_model	name results winter-garden-7b-alpha - "Smart Assistant" task dataset metrics source type name text-generation Text Generation name type config split args AI2 Reasoning Challenge (25-Shot) ai2_arc ARC-Challenge test num_few_shot 25 type value name acc_norm 65.19 normalized accuracy url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type split args HellaSwag (10-Shot) hellaswag validation num_few_shot 10 type value name acc_norm 85.36 normalized accuracy url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type config split args MMLU (5-Shot) cais/mmlu all test num_few_shot 5 type value name acc 65.2 accuracy url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type config split args TruthfulQA (0-shot) truthful_qa multiple_choice validation num_few_shot 0 type value mc2 50.94 url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type config split args Winogrande (5-shot) winogrande winogrande_xl validation num_few_shot 5 type value name acc 80.35 accuracy url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type config split args GSM8k (5-shot) gsm8k main test num_few_shot 5 type value name acc 54.44 accuracy url name https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=maldv/winter-garden-7b-alpha Open LLM Leaderboard

Winter Garden 7B - α - "Smart Assistant"

It was mentioned that we are in the open ai dark winter; so I thought I would make myself a nice winter garden.

An experiment

I've merged four partitions successfully in the past, so lets go for 9! I started with:

• Mistral-7B-v0.1

and merged in

• OmniBeagleSquaredMBX-v3-7B
• ZySec-7B-v1
• Omningotex-7b-slerp
• Erosumika-7B
• LemonadeRP-4.5.3
• Thespis-Krangled-7b
• pastiche-crown-clown-7b-dare
• Snorkel-Mistral-PairRM-DPO
• multi_verse_model

9-partition merge

All of the layers were partitioned in to 9 random bins. Alternating models were slerped at [0...1], and [1...0] gradients; except attention, which was slerped at 0.03.

This means that the model is still predominantly ordered around base mistral - including half of the input and output layers, and 28% of attention.

Other

Includes fast tokenizer.

Chat Template

I put a conversational chat template, which takes "name", "to" (optional), and "content" as the turns. It is designed to follow a transcript style chat which is used by some of the models. This type of use-case is best done by outlining a scene and creating a character card.

### {% title %}
{% metadata %}

USER: Hello

ASSISTANT: Hi, how are you?

It leans to being a coder when given an ### Instruction, follows <s>[INST][/INST], and likes <|user|>, <|assistant|> as well.

A quite cheery and intelligent model. Very good with science and math, but still capable of a decent amount of creativity for a 7b model.

Scores

Metric	Score
Average	66.91
ARC	65.19
HellaSwag	85.36
MMLU	65.2
TruthfulQA	50.94
Winogrande	80.35
GSM8K	54.44

Details