# NeuralHermes-2.5-Mistral-7B

**Developer Portal :**  <https://api.market/store/bridgeml/mlabonne>

<figure><img src="https://blog.api.market/wp-content/uploads/2024/06/qIhaFNM-1.png" alt=""><figcaption></figcaption></figure>

## NeuralHermes 2.5 - Mistral 7B

NeuralHermes is based on the [teknium/OpenHermes-2.5-Mistral-7B](https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B) model that has been further fine-tuned with Direct Preference Optimization (DPO) using the [mlabonne/chatml\_dpo\_pairs](https://huggingface.co/datasets/mlabonne/chatml_dpo_pairs) dataset. It surpasses the original model on most benchmarks (see results).

It is directly inspired by the RLHF process described by [Intel/neural-chat-7b-v3-1](https://huggingface.co/Intel/neural-chat-7b-v3-1)'s authors to improve performance. I used the same dataset and reformatted it to apply the ChatML template.

The code to train this model is available on [Google Colab](https://colab.research.google.com/drive/15iFBr1xWgztXvhrj5I9fBv20c7CFOPBE?usp=sharing) and [GitHub](https://github.com/mlabonne/llm-course/tree/main). It required an A100 GPU for about an hour.

### Quantized models

* **GGUF**: <https://huggingface.co/TheBloke/NeuralHermes-2.5-Mistral-7B-GGUF>
* **AWQ**: <https://huggingface.co/TheBloke/NeuralHermes-2.5-Mistral-7B-AWQ>
* **GPTQ**: <https://huggingface.co/TheBloke/NeuralHermes-2.5-Mistral-7B-GPTQ>
* **EXL2**:
  * 3.0bpw: <https://huggingface.co/LoneStriker/NeuralHermes-2.5-Mistral-7B-3.0bpw-h6-exl2>
  * 4.0bpw: <https://huggingface.co/LoneStriker/NeuralHermes-2.5-Mistral-7B-4.0bpw-h6-exl2>
  * 5.0bpw: <https://huggingface.co/LoneStriker/NeuralHermes-2.5-Mistral-7B-5.0bpw-h6-exl2>
  * 6.0bpw: <https://huggingface.co/LoneStriker/NeuralHermes-2.5-Mistral-7B-6.0bpw-h6-exl2>
  * 8.0bpw: <https://huggingface.co/LoneStriker/NeuralHermes-2.5-Mistral-7B-8.0bpw-h8-exl2>

### Results

**Update:** NeuralHermes-2.5 became the best Hermes-based model on the Open LLM leaderboard and one of the very best 7b models. 🎉

[![image/png](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/yWe6VBFxkHiuOlDVBXtGo.png)](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/yWe6VBFxkHiuOlDVBXtGo.png)

Teknium (author of OpenHermes-2.5-Mistral-7B) benchmarked the model ([see his tweet](https://twitter.com/Teknium1/status/1729955709377503660)).

Results are improved on every benchmark: **AGIEval** (from 43.07% to 43.62%), **GPT4All** (from 73.12% to 73.25%), and **TruthfulQA**.

#### AGIEval

[![](https://i.imgur.com/7an3B1f.png)](https://i.imgur.com/7an3B1f.png)

#### GPT4All

[![](https://i.imgur.com/TLxZFi9.png)](https://i.imgur.com/TLxZFi9.png)

#### TruthfulQA

[![](https://i.imgur.com/V380MqD.png)](https://i.imgur.com/V380MqD.png)

You can check the Weights & Biases project [here](https://wandb.ai/mlabonne/DPO/runs/axe71gr0?nw=nwusermlabonne).

### Training hyperparameters

**LoRA**:

* r=16
* lora\_alpha=16
* lora\_dropout=0.05
* bias="none"
* task\_type="CAUSAL\_LM"
* target\_modules=\['k\_proj', 'gate\_proj', 'v\_proj', 'up\_proj', 'q\_proj', 'o\_proj', 'down\_proj']

**Training arguments**:

* per\_device\_train\_batch\_size=4
* gradient\_accumulation\_steps=4
* gradient\_checkpointing=True
* learning\_rate=5e-5
* lr\_scheduler\_type="cosine"
* max\_steps=200
* optim="paged\_adamw\_32bit"
* warmup\_steps=100

**DPOTrainer**:

* beta=0.1
* max\_prompt\_length=1024
* max\_length=1536

[Source](https://github.com/mlabonne)

### Request and Response

#### Request

{% code overflow="wrap" %}

```bash
curl -X 'POST' \
  'https://prod.api.market/api/v1/bridgeml/mlabonne/bridgeml/mlabonne' \
  -H 'accept: application/json' \
  -H 'x-api-market-key: API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "messages": [
    {
      "role": "user",
      "content": "hello"
    },
    {
      "role": "assistant",
      "content": "Hello, how can you help me?"
    }
  ],
  "temperature": 1,
  "max_tokens": 256,
  "top_p": 1,
  "frequency_penalty": 0,
  "stream": false
}'
```

{% endcode %}

#### Response

{% code overflow="wrap" %}

```json
{
  "id": "mlabonne/NeuralHermes-2.5-Mistral-7B-eab3ca77-e1d0-41cb-b43e-e8195af77dc7",
  "object": "text_completion",
  "created": 1718905083,
  "model": "mlabonne/NeuralHermes-2.5-Mistral-7B",
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "It seems like you have a question or need assistance with something. Please feel free to provide more information or context so I can do my best to help you.",
        "tool_calls": null,
        "tool_call_id": null
      },
      "index": 0,
      "finish_reason": "stop",
      "logprobs": null
    }
  ],
  "usage": {
    "prompt_tokens": 76,
    "completion_tokens": 33,
    "total_tokens": 109
  }
}
```

{% endcode %}

You can use this easy to use and cheap LLM Api here at <https://api.market/store/bridgeml/mlabonne>


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.api.market/api-product-docs/bridgeml/neuralhermes-2.5-mistral-7b.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.