BenchmarkEngineering-7B-slerp / README.md

Update README.md

7da7e04 verified 3 months ago

No virus

5.36 kB

	---
	license: cc-by-nc-4.0
	tags:
	- merge
	- mergekit
	- lazymergekit
	- paulml/OmniBeagleSquaredMBX-v3-7B
	- automerger/YamshadowExperiment28-7B
	base_model:
	- paulml/OmniBeagleSquaredMBX-v3-7B
	- automerger/YamshadowExperiment28-7B
	model-index:
	- name: BenchmarkEngineering-7B-slerp
	results:
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: AI2 Reasoning Challenge (25-Shot)
	type: ai2_arc
	config: ARC-Challenge
	split: test
	args:
	num_few_shot: 25
	metrics:
	- type: acc_norm
	value: 74.15
	name: normalized accuracy
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: HellaSwag (10-Shot)
	type: hellaswag
	split: validation
	args:
	num_few_shot: 10
	metrics:
	- type: acc_norm
	value: 89.09
	name: normalized accuracy
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: MMLU (5-Shot)
	type: cais/mmlu
	config: all
	split: test
	args:
	num_few_shot: 5
	metrics:
	- type: acc
	value: 64.69
	name: accuracy
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: TruthfulQA (0-shot)
	type: truthful_qa
	config: multiple_choice
	split: validation
	args:
	num_few_shot: 0
	metrics:
	- type: mc2
	value: 75.93
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: Winogrande (5-shot)
	type: winogrande
	config: winogrande_xl
	split: validation
	args:
	num_few_shot: 5
	metrics:
	- type: acc
	value: 85.32
	name: accuracy
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: GSM8k (5-shot)
	type: gsm8k
	config: main
	split: test
	args:
	num_few_shot: 5
	metrics:
	- type: acc
	value: 69.22
	name: accuracy
	source:
	url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=weezywitasneezy/BenchmarkEngineering-7B-slerp
	name: Open LLM Leaderboard
	---

	# BenchmarkEngineering-7B-slerp

	This model was merged with the intent of producing excellent Open-LLM benchmarking results by smashing two of the highest performant models in their class together

	BenchmarkEngineering-7B-slerp is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
	* [paulml/OmniBeagleSquaredMBX-v3-7B](https://huggingface.co/paulml/OmniBeagleSquaredMBX-v3-7B)
	* [automerger/YamshadowExperiment28-7B](https://huggingface.co/automerger/YamshadowExperiment28-7B)

	# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
	Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_weezywitasneezy__BenchmarkEngineering-7B-slerp)

	\| Metric \|Value\|
	\|---------------------------------\|----:\|
	\|Avg. \|76.40\|
	\|AI2 Reasoning Challenge (25-Shot)\|74.15\|
	\|HellaSwag (10-Shot) \|89.09\|
	\|MMLU (5-Shot) \|64.69\|
	\|TruthfulQA (0-shot) \|75.93\|
	\|Winogrande (5-shot) \|85.32\|
	\|GSM8k (5-shot) \|69.22\|


	## 🧩 Configuration

	```yaml
	slices:
	- sources:
	- model: paulml/OmniBeagleSquaredMBX-v3-7B
	layer_range: [0, 32]
	- model: automerger/YamshadowExperiment28-7B
	layer_range: [0, 32]
	merge_method: slerp
	base_model: paulml/OmniBeagleSquaredMBX-v3-7B
	parameters:
	t:
	- filter: self_attn
	value: [0, 0.5, 0.3, 0.7, 1]
	- filter: mlp
	value: [1, 0.5, 0.7, 0.3, 0]
	- value: 0.5
	dtype: bfloat16
	```

	## 💻 Usage

	```python
	!pip install -qU transformers accelerate

	from transformers import AutoTokenizer
	import transformers
	import torch

	model = "weezywitasneezy/BenchmarkEngineering-7B-slerp"
	messages = [{"role": "user", "content": "What is a large language model?"}]

	tokenizer = AutoTokenizer.from_pretrained(model)
	prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
	pipeline = transformers.pipeline(
	"text-generation",
	model=model,
	torch_dtype=torch.float16,
	device_map="auto",
	)

	outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
	print(outputs[0]["generated_text"])
	```