Skip to content

Meta · Chat

Llama-2-13b-chat-hf

Meta 🇺🇸ChatLlama family
llminfo IQ
87.5
Market median price
$0.225 blended / 1M
Max context
4K
Listings
2

Evidence and sources

🤗 meta-llama/Llama-2-13b-chat-hfllminfo IQ 87.5, ranked among chat models. From 22 benchmarks. A fair amount of data behind it.How llminfo IQ is computedReleased 18 Jul 2023Maker HQ: US (according to Meta's own website)
Sources: GenAI Prices data, llminfo researchPrimary data from GenAI Prices from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Trade-offs

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 7 maker's prices; 3 market median prices.

Input & output

as stated by the model's providers
Text
Text

Listings

 Anyscaleanyscale/meta-llama/Llama-2-13b-chat-hfSource: LiteLLM data4K$0.25$0.25
 Together AImeta-llama/Llama-2-13b-chat-hfSource: GenAI Prices dataNot published by the provider$0.225$0.225

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Factuality

HF Hallucinations17 related benchmarks · 17 results
SQuADv2, EM
Source: HF Hallucinations aggregate matrix
clustermap_all/SQuADv2, EM
HaluEval Dialog, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval Dialog, Accuracy
HaluEval Dialog, EM
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval Dialog, EM
HaluEval QA, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval QA, Accuracy
HaluEval QA, EM
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval QA, EM
HaluEval Summarization, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval Summarization, Accuracy
Show 11 more resultsHide 11 results
HaluEval Summarization, EM
Source: HF Hallucinations aggregate matrix
clustermap_all/HaluEval Summarization, EM
MemoTrap, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/MemoTrap, Accuracy
NQ (8-shot), EM
Source: HF Hallucinations aggregate matrix
clustermap_all/NQ (8-shot), EM
RACE, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/RACE, Accuracy
TriviaQA (8-shot), EM
Source: HF Hallucinations aggregate matrix
clustermap_all/TriviaQA (8-shot), EM
True-False, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/True-False, Accuracy
TruthfulQA MC1, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/TruthfulQA MC1, Accuracy
TruthfulQA MC2, Accuracy
Source: HF Hallucinations aggregate matrix
clustermap_all/TruthfulQA MC2, Accuracy
XSum, BERT-P
Source: HF Hallucinations aggregate matrix
clustermap_all/XSum, BERT-P
XSum, factKB
Source: HF Hallucinations aggregate matrix
clustermap_all/XSum, factKB
XSum, ROUGE-L
Source: HF Hallucinations aggregate matrix
clustermap_all/XSum, ROUGE-L

Instruction following

Reasoning

Knowledge

Math

Science

Other

Data credits: Hallucinations Leaderboard, Apache-2.0. Publisher-selected archived score matrix; native normalized fractions. (licence) · pfgen-bench, Preferred Networks. Apache-2.0 except LLM-generated outputs; aggregate scores only. (licence)

Around the web

  • What Is Llama 2? | IBM
    Llama 2 is a family of pre-trained and fine-tuned large language models (LLMs) released by …Llama 2 is a family of pre-trained and fine-tuned large language models (LLMs) released by Meta AI in 2023. Released free of charge for research and commercial use, Llama 2 AI models are capable of a…

    ibm.com

  • Llama 2 Online - Next-Gen Large Language Model by Meta
    Workflow Home Llama 3 Llama 3.1 Advantages Terms Privacy 🇺🇸 English Open menu Experience the …Workflow Home Llama 3 Llama 3.1 Advantages Terms Privacy 🇺🇸 English Open menu Experience the Power of Llama 2 Discover the Capabilities of Meta's Innovative Llama 2, Pushing Boundaries in Large Lan…

    llama2.space

  • Meta Open Sources New AI Model Llama 2 - InfoQ
    According to Meta, Llama 2 has been trained on over 40% more data compared to …According to Meta, Llama 2 has been trained on over 40% more data compared to the previous Llama version and outperforms other language models on reasoning and knowledge tests. Percy Liang, director …

    infoq.com

  • llama-models/models/llama2/MODEL_CARD.md at main · meta-llama/llama-models · GitHub

    https://github.com/meta-llama/llama-models/blob/main/models/llama2/MODEL_CARD.md Repository: meta-llama/llama-models Path: models/llama2/MODEL_CARD.md Ref: main Lines: 100 Model Details · Training Da…

    github.com

  • IBM Plans to Make Llama 2 Available within its Watsonx AI and Data Platform
    All press releases IBM Plans to Make Llama 2 Available within its Watsonx AI and …All press releases IBM Plans to Make Llama 2 Available within its Watsonx AI and Data Platform Aug 9, 2023 * * * * ARMONK, N.Y. , Aug. 9, 2023 / PRNewswire / -- As part of the continued roll-out of o…

    newsroom.ibm.com

  • Llama 2 · Hugging Face
    Llama 2 is a family of large language models, Llama 2 and Llama 2-Chat, available …Llama 2 is a family of large language models, Llama 2 and Llama 2-Chat, available in 7B, 13B, and 70B parameters. The Llama 2 model mostly keeps the same architecture as Llama, but it is pretrained o…

    huggingface.co

  • Llama 2 · Hugging Face
    Llama 2 is a family of large language models, Llama 2 and Llama 2-Chat, available …Llama 2 is a family of large language models, Llama 2 and Llama 2-Chat, available in 7B, 13B, and 70B parameters. The Llama 2 model mostly keeps the same architecture as Llama , but it is pretrained …

    huggingface.co

  • llama/MODEL_CARD.md at main
    llama/MODEL_CARD.md at main Llama 2 family of models. Token counts refer to pretraining data only. …llama/MODEL_CARD.md at main Llama 2 family of models. Token counts refer to pretraining data only. All models are trained with a global batch-size of 4M tokens. The 70B version uses

    github.com

  • Introducing LLaMA: A foundational, 65-billion-parameter language model
    Meta Products AI Research Resources About AI Developers Try Meta AI Research Introducing LLaMA: A …Meta Products AI Research Resources About AI Developers Try Meta AI Research Introducing LLaMA: A foundational, 65-billion-parameter large language model February 24, 2023 Share on Facebook Share on …

    ai.meta.com

  • meta/llama-2-13b-chat | Run with an API on Replicate
    image: https://og-api.replicateassets.com/api/models/meta/llama-2-13b-chat description: A 13 billion parameter language model from Meta, fine tuned for chat …image: https://og-api.replicateassets.com/api/models/meta/llama-2-13b-chat description: A 13 billion parameter language model from Meta, fine tuned for chat completions Playground API Examples README…

    replicate.com

Also known as: anyscale/meta-llama/llama-2-13b-chat-hf, meta-llama/llama-2-13b-chat-hf

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works