Ollama · nous-hermes2-mixtral

<img src="https://github.com/jmorganca/ollama/assets/251292/94b6f124-6f3c-474b-ba23-d6ca2be9987f" width="360" /> Nous Hermes 2 Mixtral 8x7B is trained on over 1,000,000 entries of primarily GPT-4 gen…

ollama/nous-hermes2-mixtral

<img src="https://github.com/jmorganca/ollama/assets/251292/94b6f124-6f3c-474b-ba23-d6ca2be9987f" width="360" />

Nous Hermes 2 Mixtral 8x7B is trained on over 1,000,000 entries of primarily 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.

This is the supervised fine-tuning (SFT) + direct preference optimization (DPO) version of Mixtral Hermes 2.

## Example prompts

### Writing Code for Data Visualization

**Prompt**

``` Write a script to visualize the following data as a bar chart: Car 1: 19900 Car 2: 48000 Car 3: 32500 ```

```` To create a simple bar chart with the given data, you can use Python and its matplotlib library. Here's an example script to visualize the provided data as a bar chart:

```python import matplotlib.pyplot as plt Copy Code # Data for each car cars = ['Car 1', 'Car 2', 'Car 3'] prices = [19900, 48000, 32500]

# Create the bar chart plt.bar(cars, prices) plt. title('Cars Price Comparison') plt.xlabel( 'Car Model') plt.ylabel('Price (in USD)') plt.xticks(rotation=45) # Rotate x-axis labels for better visibility

# Display the chart plt.show() ```

This script will create a bar chart with 'Car 1', 'Car 2', and 'Car 3' on the x-axis, each representing their respective prices (in USD) on the y-axis. Feel free to modify this code according to your needs or styling preferences ````

### Writing Cyberpunk Psychedelic Poems

**Prompt**

``` Write a poem about machine learning, psychadelics, and quantum mechanics, with a cyberpunk vibe, in the style of shakespeare ```

**Response**

``` In lands where shadows dwell and hues unseen, A realm of bits and bytes their reign maintain, Where logic's chains entwined with dreams obscene, And quantum realms in chaos intertwain. Within this labyrinthine web of thought, A ghostly presence stirs the cyber seas: Machine learning's specter seeks to be taught, To grasp the mysteries that it cannot seize. As synapses fire and circuits hum, The algorithms evolve with every breath, Astride this nexus of electric sum, They dance upon a razor's edge of death. Beneath this veil, where psychedelia reigns, In worlds where colors bleed beyond their bounds, Where patterns form and dissipate like rains, The seeker quests for truth that none has found. Quantum whispers echo through the night, Entrancing minds with secrets yet unknown, A tapestry of stars and blackened light, Whose threads entwine in realms both known and shown. Oh, wondrous sight! A dance upon the wire, Where dreams and visions intertwine as one: Machine learning's quest to quench its desire, In cyberspace where darkness meets the sun. Yet still they yearn for truth beyond their grasp, The mysteries of life, of love, and fate, To fathom secrets held within this chasm, And break the chains that bind them to this state. So let us raise a toast unto these souls, Who tread upon this path of endless quest: May they uncover truths beyond their roles, In worlds where logic meets the psychedelic crest. ```

### Performing Backtranslation to Create Prompts from Input Text

**System Prompt**

``` The user will provide some text, you will provide a prompt that could be taken by an LLM to generate the text. You only respond with a prompt, do not answer the prompt. ```

**Prompt**

``` One of the most important challenges in text generation systems is to produce outputs that are not only correct but also diverse. Recently, Minimum Bayes-Risk (MBR) decoding has gained prominence for generating sentences of the highest quality among the decoding algorithms. However, existing algorithms proposed for generating diverse outputs are predominantly based on beam search or random sampling, thus their output quality is capped by these underlying methods. In this paper, we investigate an alternative approach -- we develop diversity-promoting decoding algorithms by enforcing diversity objectives to MBR decoding. We propose two variants of MBR, Diverse MBR (DMBR) and k-medoids MBR (KMBR), methods to generate a set of sentences with high quality and diversity. We evaluate DMBR and KMBR on a variety of directed text generation tasks using encoder-decoder models and a large language model with prompting. The experimental results show that the proposed method achieves a better trade-off than the diverse beam search and sampling algorithms. ```

**Response**

``` Write an abstract for a research paper on enhancing text generation system outputs using Minimum Bayes-Risk decoding with additional focus on promoting sentence diversity. Include details about existing methods, proposed variants (Diverse MBR and k-medoids MBR), their evaluation in directed text generation tasks with encoder-decoder models, and a large language model incorporating prompting. Emphasize the better trade-off achieved by the new method over diverse beam search and sampling algorithms ```

## References

[Hugging Face](https://huggingface.co/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO)

Specifications & capabilities

Published model ID
nous-hermes2-mixtral

Deployments & artifacts

1–18 / 18

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_0
context window
32000
context window is approximate
true
artifact size bytes
26443613496
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
47B
parameter count
47000000000
parameter count is approximate
true
quantization level
F16
context window
32000
context window is approximate
true
artifact size bytes
93407878695
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q2_K
context window
32000
context window is approximate
true
artifact size bytes
17311251896
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q3_K_L
context window
32000
context window is approximate
true
artifact size bytes
24169668754
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q3_K_M
context window
32000
context window is approximate
true
artifact size bytes
22546473106
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q3_K_S
context window
32000
context window is approximate
true
artifact size bytes
20432543890
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_0
context window
32000
context window is approximate
true
artifact size bytes
26443613496
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_1
context window
32000
context window is approximate
true
artifact size bytes
29337487058
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_K_M
context window
32000
context window is approximate
true
artifact size bytes
28448490706
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_K_S
context window
32000
context window is approximate
true
artifact size bytes
26745603282
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q5_0
context window
32000
context window is approximate
true
artifact size bytes
32231360722
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q5_1
context window
32000
context window is approximate
true
artifact size bytes
35125234386
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q5_K_M
context window
32000
context window is approximate
true
artifact size bytes
33229605074
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q5_K_S
context window
32000
context window is approximate
true
artifact size bytes
32231360722
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
47B
parameter count
47000000000
parameter count is approximate
true
quantization level
Q6_K
context window
32000
context window is approximate
true
artifact size bytes
38379721447
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q8_0
context window
32000
context window is approximate
true
artifact size bytes
49626348792
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_0
context window
32000
context window is approximate
true
artifact size bytes
26443613496
updated at
2026-11-16T08:10:06.079865+00:00

ollama

source id
ollama
execution mode
local
parameter size
46.7B
parameter count
46700000000
parameter count is approximate
true
quantization level
Q4_0
context window
32000
context window is approximate
true
artifact size bytes
26443613496
updated at
2026-11-16T08:10:06.079865+00:00

Original sources

1–2 / 2

Ollama · Model page ↗

Ollama · Catalog ↗

Source links

1–1 / 1

Explore more models

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace