What are the hardware requirements for OLMo 1.7 7B Hf?

OLMo 1.7 7B Hf requires approximately 27.6 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed OLMo 1.7 7B Hf and how large is it?

OLMo 1.7 7B Hf is developed by allenai, a model with 7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does OLMo 1.7 7B Hf perform on standard benchmarks?

OLMo 1.7 7B Hf has the following published scores: MMLU 53.39. Compare against reference models on this page or on the LLM Explorer leaderboards.

OLMo 1.7 7B Hf by allenai — VRAM 27.6GB, 4K context

Name: OLMo 1.7 7B Hf
Author: allenai

OLMo 1.7 7B Hf is an open-source language model by allenai. Features: 7b LLM, VRAM: 27.6GB, Context: 4K, License: apache-2.0, LLM Explorer Score: 0.2, Arc: 49.4, HellaSwag: 78.7, MMLU: 53.4, GSM8K: 26.6.

Arxiv:2402.00838 Autotrain compatible Dataset:allenai/dolma En Endpoints compatible License:apache-2.0 Olmo Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/allenai/OLMo-1.7-7B-hf

OLMo 1.7 7B Hf Parameters and Internals

Model Type

Transformer-style autoregressive language model

Use Cases

Areas:

Research, Commercial applications

Limitations:

Relatively easy for models to generate harmful content, Many facts may not be true

Additional Notes

OLMo 7B April 2024 features improved dataset and model architecture.

Supported Languages

en (English)

Training Details

Data Sources:

Dolma dataset version 1.7

Data Volume:

2.05 trillion training tokens

Methodology:

Staged training with annealing phase

Context Length:

4096

Hardware Used:

CSC (Lumi Supercomputer)

Model Architecture:

Transformer

LLM Name	OLMo 1.7 7B Hf
Repository 🤗	https://huggingface.co/allenai/OLMo-1.7-7B-hf
Model Size	7b
Required VRAM	27.6 GB
Updated	2024-07-15
Maintainer	allenai
Model Type	olmo
Model Files	4.9 GB: 1-of-6 5.0 GB: 2-of-6 4.9 GB: 3-of-6 4.9 GB: 4-of-6 4.9 GB: 5-of-6 3.0 GB: 6-of-6
Supported Languages	en
Model Architecture	OlmoForCausalLM
License	apache-2.0
Context Length	4096
Model Max Length	4096
Transformers Version	4.40.0.dev0
Tokenizer Class	GPTNeoXTokenizer
Padding Token	<\|padding\|>
Vocabulary Size	50304
Torch Data Type	float32

Rank the OLMo 1.7 7B Hf 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 OLMo 1.7 7B Hf

Best Alternatives	Context / RAM	Downloads	Likes
OLMo 7B 0724 Hf	4K / 27.6 GB	4081	17
OLMo 7B 0724 Instruct Hf	4K / 13.8 GB	1010	7
OLMo 7B 0424 Hf	4K / 27.6 GB	764	14
Olmo 7B Spe	2K / 13.8 GB	1414	0
... Fsdp Pubchem 2.5M 1epochs Eos	2K / 13.8 GB	596	0
OLMo 7B Instruct Hf	2K / 27.6 GB	13	3
OLMo 7B Hf	2K / 27.6 GB	4799	17
OLMo 7B Instruct Hf	2K / 27.6 GB	2139	6
OLMo 7B 0724 SFT Hf	2K / 13.8 GB	155	4
OLMo 7B 0424 Instruct Hf	2K / 13.8 GB	88	1

Note: green Score (e.g. "73.2") means that the model is better than allenai/OLMo-1.7-7B-hf.