Nous: Hermes 2 Mixtral 8x7B DPO

Text input Text output
Author's Description

Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the [Mixtral 8x7B MoE LLM](/models/mistralai/mixtral-8x7b). The model was trained on over 1,000,000 entries of primarily [GPT-4](/models/openai/gpt-4) generated data, as well as other high quality data from open datasets across the AI landscape, achieving state of the art performance on a variety of tasks. #moe

Key Specifications
Cost
$$$$
Context
32K
Parameters
56B
Released
Jan 15, 2024
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Logit Bias Top P Temperature Min P Frequency Penalty Max Tokens
Performance Summary

Nous Hermes 2 Mixtral 8x7B DPO demonstrates competitive response times, performing in the 47th percentile for speed across various benchmarks. Similarly, its pricing is competitive, ranking in the 46th percentile. The model's performance across benchmark categories reveals a mixed profile. It exhibits notable strength in General Knowledge, achieving 86% accuracy, and Email Classification, with 89% accuracy, although its percentile ranking in these areas is modest (28th and 14th respectively). However, the model shows significant weaknesses in more complex cognitive tasks. Its accuracy in Coding is very low at 7%, placing it in the 16th percentile, and its Reasoning capabilities are also limited, with 23.5% accuracy (14th percentile). Instruction Following and Ethics also present challenges, with accuracies of 28% and 53% respectively, both ranking in the lower percentiles. Overall, the model appears more proficient in tasks requiring factual recall and straightforward classification, while struggling with problem-solving, logical deduction, and precise instruction adherence.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.6
Completion $0.6

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Together
Together | nousresearch/nous-hermes-2-mixtral-8x7b-dpo 32K $0.6 / 1M tokens $0.6 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by nousresearch