---
inference: false
language:
- en
license: llama2
model_creator: NousResearch
model_link: https://huggingface.co/NousResearch/Nous-Hermes-Llama2-70b
model_name: Nous Hermes Llama2 70B
model_type: llama
quantized_by: TheBloke
tags:
- llama-2
- self-instruct
- distillation
- synthetic instruction
---
# Nous Hermes Llama2 70B - GGUF
- Model creator: [NousResearch](https://huggingface.co/NousResearch)
- Original model: [Nous Hermes Llama2 70B](https://huggingface.co/NousResearch/Nous-Hermes-Llama2-70b)
## Description
This repo contains GGUF format model files for [NousResearch's Nous Hermes Llama2 70B](https://huggingface.co/NousResearch/Nous-Hermes-Llama2-70b).
### About GGUF
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
The key benefit of GGUF is that it is a extensible, future-proof format which stores more information about the model as metadata. It also includes significantly improved tokenization code, including for the first time full support for special tokens. This should improve performance, especially with models that use new special tokens and implement custom prompt templates.
As of August 24th 2023, llama.cpp and KoboldCpp support GGUF. Other third-party clients and libraries are expected to add support very soon.
Here is a list of clients and libraries that are known to support GGUF:
* [llama.cpp](https://github.com/ggerganov/llama.cpp)
* [KoboldCpp](https://github.com/LostRuins/koboldcpp), now supports GGUF as of release 1.41!
Here is a list of clients and libraries, along with their expected timeline for GGUF support. Where possible a link to the relevant issue or PR is provided:
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui), awaiting llama-cpp-python support.
* [LM Studio](https://lmstudio.ai/), in active development - hoped to be ready by August 25th-26th.
* [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), will work as soon as ctransformers or llama-cpp-python is updated.
* [ctransformers](https://github.com/marella/ctransformers), [development will start soon](https://github.com/marella/ctransformers/issues/102).
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), [in active development](https://github.com/abetlen/llama-cpp-python/issues/628).
## Repositories available
* [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GPTQ)
* [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF)
* [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGML)
* [NousResearch's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/NousResearch/Nous-Hermes-Llama2-70b)
## Prompt template: Alpaca-InstructOnly
```
### Instruction:
{prompt}
### Response:
```
## Compatibility
These quantised GGUF files are compatible with llama.cpp from August 21st 2023 onwards, as of commit [6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9](https://github.com/ggerganov/llama.cpp/commit/6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9)
As of August 24th 2023 they are now compatible with KoboldCpp, release 1.41 and later.
They are are not yet compatible with any other third-party UIS, libraries or utilities but this is expected to change very soon.
## Explanation of quantisation methods
Click to see details
The new methods available are:
* GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
* GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
* GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
* GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
* GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
Refer to the Provided Files table below to see what files use which methods, and how.
## Provided files
| Name | Quant method | Bits | Size | Max RAM required | Use case |
| ---- | ---- | ---- | ---- | ---- | ----- |
| [nous-hermes-llama2-70b.Q6_K.gguf-split-b](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q6_K.gguf-split-b) | Q6_K | 6 | 20.13 GB| 22.63 GB | very large, extremely low quality loss |
| [nous-hermes-llama2-70b.Q2_K.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q2_K.gguf) | Q2_K | 2 | 29.48 GB| 31.98 GB | smallest, significant quality loss - not recommended for most purposes |
| [nous-hermes-llama2-70b.Q3_K_S.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q3_K_S.gguf) | Q3_K_S | 3 | 30.09 GB| 32.59 GB | very small, high quality loss |
| [nous-hermes-llama2-70b.Q3_K_M.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q3_K_M.gguf) | Q3_K_M | 3 | 33.45 GB| 35.95 GB | very small, high quality loss |
| [nous-hermes-llama2-70b.Q3_K_L.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q3_K_L.gguf) | Q3_K_L | 3 | 36.49 GB| 38.99 GB | small, substantial quality loss |
| [nous-hermes-llama2-70b.Q8_0.gguf-split-b](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q8_0.gguf-split-b) | Q8_0 | 8 | 36.59 GB| 39.09 GB | very large, extremely low quality loss - not recommended |
| [nous-hermes-llama2-70b.Q6_K.gguf-split-a](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q6_K.gguf-split-a) | Q6_K | 6 | 36.70 GB| 39.20 GB | very large, extremely low quality loss |
| [nous-hermes-llama2-70b.Q8_0.gguf-split-a](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q8_0.gguf-split-a) | Q8_0 | 8 | 36.70 GB| 39.20 GB | very large, extremely low quality loss - not recommended |
| [nous-hermes-llama2-70b.Q4_K_S.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q4_K_S.gguf) | Q4_K_S | 4 | 39.30 GB| 41.80 GB | small, greater quality loss |
| [nous-hermes-llama2-70b.Q4_K_M.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q4_K_M.gguf) | Q4_K_M | 4 | 41.69 GB| 44.19 GB | medium, balanced quality - recommended |
| [nous-hermes-llama2-70b.Q5_K_S.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q5_K_S.gguf) | Q5_K_S | 5 | 47.74 GB| 50.24 GB | large, low quality loss - recommended |
| [nous-hermes-llama2-70b.Q5_K_M.gguf](https://huggingface.co/TheBloke/Nous-Hermes-Llama2-70B-GGUF/blob/main/nous-hermes-llama2-70b.Q5_K_M.gguf) | Q5_K_M | 5 | 49.03 GB| 51.53 GB | large, very low quality loss - recommended |
**Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
## How to run in `llama.cpp`
Make sure you are using `llama.cpp` from commit [6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9](https://github.com/ggerganov/llama.cpp/commit/6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9) or later.
For compatibility with older versions of llama.cpp, or for use with third-party clients and libaries, please use GGML files instead.
```
./main -t 10 -ngl 32 -m nous-hermes-llama2-70b.q4_K_M.gguf --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "### Instruction: Write a story about llamas\n### Response:"
```
Change `-t 10` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`.
Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
Change `-c 4096` to the desired sequence length for this model. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters should be set by llama.cpp automatically. If they are not, or if you need to change them manually, you can use `--rope-freq-base 10000 --rope-freq-scale 0.5` for doubled context, or `--rope-freq-base 10000 --rope-freq-scale 0.25` for 4x context.
If you want to have a chat-style conversation, replace the `-p ` argument with `-i -ins`
For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md)
## How to run in `text-generation-webui`
Further instructions here: [text-generation-webui/docs/llama.cpp.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/llama.cpp.md).
## Discord
For further support, and discussions on these models and AI in general, join us at:
[TheBloke AI's Discord server](https://discord.gg/theblokeai)
## Thanks, and how to contribute.
Thanks to the [chirper.ai](https://chirper.ai) team!
I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
* Patreon: https://patreon.com/TheBlokeAI
* Ko-Fi: https://ko-fi.com/TheBlokeAI
**Special thanks to**: Aemon Algiz.
**Patreon special mentions**: Sam, theTransient, Jonathan Leane, Steven Wood, webtim, Johann-Peter Hartmann, Geoffrey Montalvo, Gabriel Tamborski, Willem Michiel, John Villwock, Derek Yates, Mesiah Bishop, Eugene Pentland, Pieter, Chadd, Stephen Murray, Daniel P. Andersen, terasurfer, Brandon Frisco, Thomas Belote, Sid, Nathan LeClaire, Magnesian, Alps Aficionado, Stanislav Ovsiannikov, Alex, Joseph William Delisle, Nikolai Manek, Michael Davis, Junyu Yang, K, J, Spencer Kim, Stefan Sabev, Olusegun Samson, transmissions 11, Michael Levine, Cory Kujawski, Rainer Wilmers, zynix, Kalila, Luke @flexchar, Ajan Kanaga, Mandus, vamX, Ai Maven, Mano Prime, Matthew Berman, subjectnull, Vitor Caleffi, Clay Pascal, biorpg, alfie_i, 阿明, Jeffrey Morgan, ya boyyy, Raymond Fosdick, knownsqashed, Olakabola, Leonard Tan, ReadyPlayerEmma, Enrico Ros, Dave, Talal Aujan, Illia Dulskyi, Sean Connelly, senxiiz, Artur Olbinski, Elle, Raven Klaugh, Fen Risland, Deep Realms, Imad Khwaja, Fred von Graf, Will Dee, usrbinkat, SuperWojo, Alexandros Triantafyllidis, Swaroop Kallakuri, Dan Guido, John Detwiler, Pedro Madruga, Iucharbius, Viktor Bowallius, Asp the Wyvern, Edmond Seymore, Trenton Dambrowitz, Space Cruiser, Spiking Neurons AB, Pyrater, LangChain4j, Tony Hughes, Kacper Wikieł, Rishabh Srivastava, David Ziegler, Luke Pendergrass, Andrey, Gabriel Puliatti, Lone Striker, Sebastain Graf, Pierre Kircher, Randy H, NimbleBox.ai, Vadim, danny, Deo Leter
Thank you to all my generous patrons and donaters!
And thank you again to a16z for their generous grant.
# Original model card: NousResearch's Nous Hermes Llama2 70B
# Model Card: Nous-Hermes-Llama2-70b
Compute provided by PygmalionAI, thank you! Follow PygmalionAI on Twitter @pygmalion_ai.
## Model Description
Nous-Hermes-Llama2-70b is a state-of-the-art language model fine-tuned on over 300,000 instructions. This model was fine-tuned by Nous Research, with Teknium and Emozilla leading the fine tuning process and dataset curation, Pygmalion sponsoring the compute, and several other contributors.
This Hermes model uses the exact same dataset as Hermes on Llama-1. This is to ensure consistency between the old Hermes and new, for anyone who wanted to keep Hermes as similar to the old one, just more capable.
This model stands out for its long responses, lower hallucination rate, and absence of OpenAI censorship mechanisms in the synthetic training data. The fine-tuning process was performed with a 4096 sequence length on an 8x H100 80GB machine.
## Model Training
The model was trained almost entirely on synthetic GPT-4 outputs. Curating high quality GPT-4 datasets enables incredibly high quality in knowledge, task completion, and style.
This includes data from diverse sources such as GPTeacher, the general, roleplay v1&2, code instruct datasets, Nous Instruct & PDACTL (unpublished), and several others, detailed further below
## Collaborators
The model fine-tuning and the datasets were a collaboration of efforts and resources between Teknium, Karan4D, Emozilla, Huemin Art, and Pygmalion AI.
Special mention goes to @winglian for assisting in some of the training issues.
Huge shoutout and acknowledgement is deserved for all the dataset creators who generously share their datasets openly.
Among the contributors of datasets:
- GPTeacher was made available by Teknium
- Wizard LM by nlpxucan
- Nous Research Instruct Dataset was provided by Karan4D and HueminArt.
- GPT4-LLM and Unnatural Instructions were provided by Microsoft
- Airoboros dataset by jondurbin
- Camel-AI's domain expert datasets are from Camel-AI
- CodeAlpaca dataset by Sahil 2801.
If anyone was left out, please open a thread in the community tab.
## Prompt Format
The model follows the Alpaca prompt format:
```
### Instruction:
### Response:
```
or
```
### Instruction:
### Input:
### Response:
```
## Benchmarks:
GPT4All Suite:
```
hf-causal-experimental (pretrained=/home/data/axolotl/Nous-Hermes-Llama2-70b,dtype=float16,use_accelerate=True), limit: None, provide_description: False, num_fewshot: 0, batch_size: None
| Task |Version| Metric |Value | |Stderr|
|-------------|------:|--------|-----:|---|-----:|
|arc_challenge| 0|acc |0.5734|± |0.0145|
| | |acc_norm|0.6015|± |0.0143|
|arc_easy | 0|acc |0.8422|± |0.0075|
| | |acc_norm|0.8253|± |0.0078|
|boolq | 1|acc |0.8422|± |0.0064|
|hellaswag | 0|acc |0.6519|± |0.0048|
| | |acc_norm|0.8363|± |0.0037|
|openbookqa | 0|acc |0.3880|± |0.0218|
| | |acc_norm|0.5000|± |0.0224|
|piqa | 0|acc |0.8313|± |0.0087|
| | |acc_norm|0.8351|± |0.0087|
|winogrande | 0|acc |0.7751|± |0.0117|
```
BigBench Suite:
```
hf-causal-experimental (pretrained=/home/data/axolotl/Nous-Hermes-Llama2-70b,dtype=float16,use_accelerate=True), limit: None, provide_description: False, num_fewshot: 0, batch_size: None
| Task |Version| Metric |Value | |Stderr|
|------------------------------------------------|------:|---------------------|-----:|---|-----:|
|bigbench_causal_judgement | 0|multiple_choice_grade|0.6579|± |0.0345|
|bigbench_date_understanding | 0|multiple_choice_grade|0.7344|± |0.0230|
|bigbench_disambiguation_qa | 0|multiple_choice_grade|0.3023|± |0.0286|
|bigbench_geometric_shapes | 0|multiple_choice_grade|0.2340|± |0.0224|
| | |exact_str_match |0.0000|± |0.0000|
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|0.2760|± |0.0200|
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|0.1871|± |0.0148|
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|0.4467|± |0.0288|
|bigbench_movie_recommendation | 0|multiple_choice_grade|0.3240|± |0.0210|
|bigbench_navigate | 0|multiple_choice_grade|0.5000|± |0.0158|
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|0.6605|± |0.0106|
|bigbench_ruin_names | 0|multiple_choice_grade|0.4598|± |0.0236|
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|0.2585|± |0.0139|
|bigbench_snarks | 0|multiple_choice_grade|0.6630|± |0.0352|
|bigbench_sports_understanding | 0|multiple_choice_grade|0.7394|± |0.0140|
|bigbench_temporal_sequences | 0|multiple_choice_grade|0.4440|± |0.0157|
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|0.2168|± |0.0117|
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|0.1531|± |0.0086|
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|0.4467|± |0.0288|
```
AGIEval:
```
hf-causal-experimental (pretrained=/home/data/axolotl/Nous-Hermes-Llama2-70b,dtype=float16,use_accelerate=True), limit: None, provide_description: False, num_fewshot: 0, batch_size: None
| Task |Version| Metric |Value | |Stderr|
|------------------------------|------:|--------|-----:|---|-----:|
|agieval_aqua_rat | 0|acc |0.2480|± |0.0272|
| | |acc_norm|0.2362|± |0.0267|
|agieval_logiqa_en | 0|acc |0.3917|± |0.0191|
| | |acc_norm|0.3932|± |0.0192|
|agieval_lsat_ar | 0|acc |0.2217|± |0.0275|
| | |acc_norm|0.2000|± |0.0264|
|agieval_lsat_lr | 0|acc |0.5765|± |0.0219|
| | |acc_norm|0.4922|± |0.0222|
|agieval_lsat_rc | 0|acc |0.6914|± |0.0282|
| | |acc_norm|0.6022|± |0.0299|
|agieval_sat_en | 0|acc |0.8641|± |0.0239|
| | |acc_norm|0.8204|± |0.0268|
|agieval_sat_en_without_passage| 0|acc |0.5291|± |0.0349|
| | |acc_norm|0.4709|± |0.0349|
|agieval_sat_math | 0|acc |0.4136|± |0.0333|
| | |acc_norm|0.3455|± |0.0321|
```
## Resources for Applied Use Cases:
Check out LM Studio for a nice chatgpt style interface here: https://lmstudio.ai/
For an example of a back and forth chatbot using huggingface transformers and discord, check out: https://github.com/teknium1/alpaca-discord
For an example of a roleplaying discord chatbot, check out this: https://github.com/teknium1/alpaca-roleplay-discordbot
## Future Plans
We plan to continue to iterate on both more high quality data, and new data filtering techniques to eliminate lower quality data going forward.
## Model Usage
The model is available for download on Hugging Face. It is suitable for a wide range of language tasks, from generating creative text to understanding and following complex instructions.
[
](https://github.com/OpenAccess-AI-Collective/axolotl)
## Training procedure
The following `bitsandbytes` quantization config was used during training:
- quant_method: bitsandbytes
- load_in_8bit: False
- load_in_4bit: True
- llm_int8_threshold: 6.0
- llm_int8_skip_modules: None
- llm_int8_enable_fp32_cpu_offload: False
- llm_int8_has_fp16_weight: False
- bnb_4bit_quant_type: nf4
- bnb_4bit_use_double_quant: True
- bnb_4bit_compute_dtype: bfloat16
The following `bitsandbytes` quantization config was used during training:
- quant_method: bitsandbytes
- load_in_8bit: False
- load_in_4bit: True
- llm_int8_threshold: 6.0
- llm_int8_skip_modules: None
- llm_int8_enable_fp32_cpu_offload: False
- llm_int8_has_fp16_weight: False
- bnb_4bit_quant_type: nf4
- bnb_4bit_use_double_quant: True
- bnb_4bit_compute_dtype: bfloat16
### Framework versions
- PEFT 0.5.0.dev0
- PEFT 0.5.0.dev0