Chat & support: TheBloke's Discord server

Want to contribute? TheBloke's Patreon page

TheBloke's LLM work is generously supported by a grant from andreessen horowitz (a16z)

Llama2 70B SFT v10 - GPTQ

Model creator: OpenAssistant
Original model: Llama2 70B SFT v10

Description

This repo contains GPTQ model files for OpenAssistant's Llama2 70B SFT v10.

Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them.

Repositories available

Prompt template: ChatML

<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Provided files and GPTQ parameters

Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.

Each separate quant is in a different branch. See below for instructions on fetching from different branches.

All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches are made with AutoGPTQ. Files in the main branch which were uploaded before August 2023 were made with GPTQ-for-LLaMa.

Explanation of GPTQ parameters

Bits: The bit size of the quantised model.
GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
Act Order: True or False. Also known as desc_act. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now.
Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama models in 4-bit.

Branch	Bits	GS	Act Order	Damp %	GPTQ Dataset	Seq Len	Size	ExLlama	Desc
main	4	None	Yes	0.1	wikitext	4096	35.33 GB	Yes	4-bit, with Act Order. No group size, to lower VRAM requirements.
gptq-4bit-32g-actorder_True	4	32	Yes	0.1	wikitext	4096	40.66 GB	Yes	4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage.
gptq-4bit-64g-actorder_True	4	64	Yes	0.1	wikitext	4096	37.99 GB	Yes	4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy.
gptq-4bit-128g-actorder_True	4	128	Yes	0.1	wikitext	4096	36.65 GB	Yes	4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy.
gptq-3bit--1g-actorder_True	3	None	Yes	0.1	wikitext	4096	26.78 GB	No	3-bit, with Act Order and no group size. Lowest possible VRAM requirements. May be lower quality than 3-bit 128g.
gptq-3bit-128g-actorder_True	3	128	Yes	0.1	wikitext	4096	28.03 GB	No	3-bit, with group size 128g and act-order. Higher quality than 128g-False.

How to download from branches

In text-generation-webui, you can add :branch to the end of the download name, eg TheBloke/Llama2-70B-OASST-SFT-v10-GPTQ:main
With Git, you can clone a branch with:

git clone --single-branch --branch main https://huggingface.co/TheBloke/Llama2-70B-OASST-SFT-v10-GPTQ

In Python Transformers code, the branch is the revision parameter; see below.

How to easily download and use this model in text-generation-webui.

Please make sure you're using the latest version of text-generation-webui.

It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install.

Click the Model tab.
Under Download custom model or LoRA, enter TheBloke/Llama2-70B-OASST-SFT-v10-GPTQ.

To download from a specific branch, enter for example TheBloke/Llama2-70B-OASST-SFT-v10-GPTQ:main
see Provided Files above for the list of branches for each option.

Click Download.
The model will start downloading. Once it's finished it will say "Done".
In the top left, click the refresh icon next to Model.
In the Model dropdown, choose the model you just downloaded: Llama2-70B-OASST-SFT-v10-GPTQ
The model will automatically load, and is now ready for use!
If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right.

Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file quantize_config.json.

Once you're ready, click the Text Generation tab and enter a prompt to get started!

How to use this GPTQ model from Python code

Install the necessary packages

Requires: Transformers 4.32.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later.

pip3 install transformers>=4.32.0 optimum>=1.12.0
pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/  # Use cu117 if on CUDA 11.7

If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead:

pip3 uninstall -y auto-gptq
git clone https://github.com/PanQiWei/AutoGPTQ
cd AutoGPTQ
pip3 install .

For CodeLlama models only: you must use Transformers 4.33.0 or later.

If 4.33.0 is not yet released when you read this, you will need to install Transformers from source:

pip3 uninstall -y transformers
pip3 install git+https://github.com/huggingface/transformers.git

You can then use the following code

from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

model_name_or_path = "TheBloke/Llama2-70B-OASST-SFT-v10-GPTQ"
# To use a different branch, change revision
# For example: revision="main"
model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
                                             device_map="auto",
                                             trust_remote_code=False,
                                             revision="main")

tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)

prompt = "Tell me about AI"
prompt_template=f'''<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

'''

print("\n\n*** Generate:")

input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)
print(tokenizer.decode(output[0]))

# Inference can also be done using transformers' pipeline

print("*** Pipeline:")
pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    max_new_tokens=512,
    do_sample=True,
    temperature=0.7,
    top_p=0.95,
    top_k=40,
    repetition_penalty=1.1
)

print(pipe(prompt_template)[0]['generated_text'])

Compatibility

The files provided are tested to work with AutoGPTQ, both via Transformers and using AutoGPTQ directly. They should also work with Occ4m's GPTQ-for-LLaMa fork.

ExLlama is compatible with Llama models in 4-bit. Please see the Provided Files table above for per-file compatibility.

Huggingface Text Generation Inference (TGI) is compatible with all GPTQ models.

Discord

For further support, and discussions on these models and AI in general, join us at:

TheBloke AI's Discord server

Thanks, and how to contribute

Thanks to the chirper.ai team!

Thanks to Clay from gpus.llm-utils.org!

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: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov

Thank you to all my generous patrons and donaters!

And thank you again to a16z for their generous grant.

Original model card: OpenAssistant's Llama2 70B SFT v10

Open-Assistant Llama2 70B SFT v10

This model is an Open-Assistant fine-tuning of Meta's Llama2 70B LLM. It was fine-tuned in two stages, first on a mix of synthetic instrunctions and coding tasks and then in a "polishing" stage on the best human demonstrations collected at open-assistant.io up to July 23, 2023 (see Configuration Details below).

Model Details

Finetuned from: meta-llama/Llama-2-70b via epfLLM/Megatron-LLM
Model type: Causal decoder-only transformer language model
Language: English (and limited capabilities in German, Spanish, French, Italian, Portuguese, Polish, Dutch, Romanian, Czech, Swedish)
Weights & Biases training logs: Stage 1 (1 epoch pretrain-mix, 12k steps), Stage 2 (3 epochs oasst top-1, 519 steps)
Demo: Continuations for 250 random prompts (TGI, 4bit nf4 quantization)
Evaluation FastEval-OpenAssistant Overview (using FastEval & vLLM)
License: LLAMA 2 COMMUNITY LICENSE AGREEMENT
Contact: Open-Assistant Discord

Prompting / Prompt Template

Due to public demand (see survey) we changed the prompt-template for this model from custom prompter/assistant tokens to OpenAI's chatml standard prompt format. We hope that this leads to greater compatibility with chat inference/frontend applications.

Prompt dialogue template:

"""
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
"""

The model input can contain multiple conversation turns between user and assistant, e.g.

<|im_start|>user
{prompt 1}<|im_end|>
<|im_start|>assistant
{reply 1}<|im_end|>
<|im_start|>user
{prompt 2}<|im_end|>
<|im_start|>assistant
(...)

The model was partly trained with orca system messages.
For inference we recommend to use the official Llama2 system message:

<|im_start|>system
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.

If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<|im_end|>

Credits & Special Thanks

Thanks to Meta AI for training and releasing the Llama2 model.
Distributed training support was provided by EPFL's Machine Learning and Optimization Laboratory, and Natural Language Processing Lab.
The open-source epfLLM/Megatron-LLM trainer was used for fine-tuning.
rombodawg curated the LosslessMegaCodeTrainingV2_1m_Evol_Uncensored dataset.
ehartford generated and published the ehartford/dolphin and the ehartford/oa_leet10k datasets.
Argilla curated and published the argilla/databricks-dolly-15k-curated-multilingual dataset.
shahules786 de-duped and filtered the Dolphin dataset with a cluster-center approach and generated the orca-best (ocra-chat) dataset.
andreaskoepf prepared & orchestrated the training.

We want to especially thank everyone who contributed in the crowed-sourced Open-Assistant dataset creation on https://open-assistant.io/ - without you this project would not have been possible.

Ethical Considerations and Limitations

Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, the potential outputs of llama2-70b-oasst-sft-v10 cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of llama2-70b-oasst-sft-v10, developers should perform safety testing and tuning tailored to their specific applications of the model.

Please see Meta's Responsible Use Guide.

Note regarding inference with TGI

During evaluation we noticed that this 70B model produced extremely poor outputs when loaded it was loaded in 16 bit precision sharded in TGI. In contrast the model could be evaluated without problem using vLLM. The model also worked decently well when loaded with TGI on a single GPPU nf4 quantized via TimDettmers/bitsandbytes. Will will get it touch with the TGI authors to find out why sharded 16-bit inference doesn't work as expected.

Configuration Details

The "pretokenizer" utility used to tokenize the datamix is part of the Open-Assistant github repository and can be found here: model/pretokenizer.

Stage 1 Pretokenizer Configuration

Entries of the dataset with assistant replies shorter than 25 tokens were excluded from training.

oasst_pre10_min25:
  datasets:
    - megacode2:
        fraction: 0.5
        val_split: 0.01
        max_val_set: 1000
    - orca-chat:
        val_split: 0.01
        max_val_set: 1000
    - dolly15k_multilingual:
        val_split: 0.05
        max_val_set: 300
    - oa_leet10k:
        val_split: 0.05
        max_val_set: 250
  output_dir: "output/oasst_pre10_min25"
  filename_prefix: "oasst_pre10"
  min_assistant_tokens: 25

Stage 1 dataset statistics:

# Stats for output/oasst_pre10_min25_llama2

## Stats for 'Subset of InstructionDataset (megacode2)' (466364 samples (50.0%))
-----------------
  Accepted: 398223/466364 (85.4%)
  Accepted tokens: 167676873
  Skipped: 68141 (14.6%)
  Min tokens per sample: 36
  Max tokens per sample: 11810
  Avg tokens per sample: 421.063
-----------------

## Stats for 'Subset of OrcaChat (orca-chat)' (325616 samples (100.0%))
-----------------
  Accepted: 325616/325616 (100.0%)
  Accepted tokens: 178307574
  Skipped: 0 (0.0%)
  Min tokens per sample: 105
  Max tokens per sample: 10408
  Avg tokens per sample: 547.601
-----------------

## Stats for 'Subset of Dolly15kMultilingual' (57020 samples (100.0%))
-----------------
  Accepted: 47494/57020 (83.3%)
  Accepted tokens: 13883177
  Skipped: 9526 (16.7%)
  Min tokens per sample: 34
  Max tokens per sample: 9172
  Avg tokens per sample: 292.314
-----------------

## Stats for 'Subset of InstructionDataset (oa_leet10k)' (22236 samples (100.0%))
-----------------
  Accepted: 22236/22236 (100.0%)
  Accepted tokens: 15905296
  Skipped: 0 (0.0%)
  Min tokens per sample: 168
  Max tokens per sample: 10588
  Avg tokens per sample: 715.295
-----------------

## Stats for 'total' (871236 samples (100.0%))
-----------------
  Accepted: 793569/871236 (91.1%)
  Accepted tokens: 375772920
  Skipped: 77667 (8.9%)
  Min tokens per sample: 34
  Max tokens per sample: 11810
  Avg tokens per sample: 473.523
-----------------

Stage 2 Pretokenizer Configuration

oasst_top1:
  datasets:
    - oasst_export:
        lang: "bg,ca,cs,da,de,en,es,fr,hr,hu,it,nl,pl,pt,ro,ru,sl,sr,sv,uk"
        input_file_path: 2023-07-23_oasst_ready.tar.gz
        top_k: 1
        val_split: 0.05
  output_dir: "output/oasst_top1_2023-07-23"
  filename_prefix: "oasst_top1"

Stage 2 dataset statistics:

# Stats for output/oasst_top1_2023-07-23_llama2

## Stats for 'ListDataset' (11441 samples (100.0%))
-----------------
  Accepted: 11441/11441 (100.0%)
  Accepted tokens: 5315368
  Skipped: 0 (0.0%)
  Min tokens per sample: 20
  Max tokens per sample: 5407
  Avg tokens per sample: 464.58945896337735
-----------------

## Stats for 'total' (11441 samples (100.0%))
-----------------
  Accepted: 11441/11441 (100.0%)
  Accepted tokens: 5315368
  Skipped: 0 (0.0%)
  Min tokens per sample: 20
  Max tokens per sample: 5407
  Avg tokens per sample: 464.58945896337735
-----------------

Megatron Fine-Tuning Arguments for Stage 1 (Instruction Tuning):

--tensor_model_parallel_size 8
--pipeline_model_parallel_size 4
--load ./checkpoints/llama2-70b-tp8-pp4
--save ./checkpoints/llama2-70b-tp8-pp4-oasst_pre10
--tensorboard_dir ./checkpoints/llama2-70b-tp8-pp4-oasst_pre10/logging
--data_path ./data/oasst_pre10_min25_llama2/oasst_sft10-train
--model_name llama2
--tokenizer_type SentencePieceTokenizer
--bf16
--global_batch_size 64
--micro_batch_size 2
--vocab_file=./llama2/Llama-2-7b/tokenizer.model
--use_rms_norm
--glu_activation swiglu
--no_tie_embed_logits
--vocab_extra_ids_list "\"<|im_start|>,<|im_end|>\""
--layernorm_epsilon 1e-5
--use_flash_attn
--no_bias_gelu_fusion
--seq_length 4096
--max_position_embeddings 4096
--log_interval 1
--save_interval 500
--eval_interval 50
--eval_iters 10
--hidden_dropout 0.0
--position_embedding_type rotary
--no_bias_dropout_fusion
--use_checkpoint_args
--train_iters 12000
--attention_dropout 0.0
--adam_beta1 0.9
--adam_beta2 0.95
--adam_eps 1e-12
--lr_decay_style cosine
--lr_warmup_iters 100
--lr 1e-5
--min_lr 1e-6
--weight_decay 0.000001
--sequence_parallel
--recompute_granularity selective
--log_timers_to_tensorboard
--rope_scaling_factor 1.0
--wandb_logger

Megatron Fine-Tuning Arguments for Stage 2 (OASST Polishing, LIMA Dropout):

--tensor_model_parallel_size 8
--pipeline_model_parallel_size 4
--load ./checkpoints/llama2-70b-tp8-pp4-oasst_pre10
--save ./checkpoints/llama2-70b-tp8-pp4-oasst_sft10
--tensorboard_dir ./checkpoints/llama2-70b-tp8-pp4-oasst_sft10/logging
--data_path ./data/oasst_top1_2023-07-23_llama2/oasst_top1-train
--model_name llama2
--tokenizer_type SentencePieceTokenizer
--bf16
--global_batch_size 64
--micro_batch_size 2
--vocab_file=./llama2/Llama-2-7b/tokenizer.model
--use_rms_norm
--glu_activation swiglu
--no_tie_embed_logits
--vocab_extra_ids_list "\"<|im_start|>,<|im_end|>\""
--layernorm_epsilon 1e-5
--use_flash_attn
--no_bias_gelu_fusion
--seq_length 4096
--max_position_embeddings 4096
--log_interval 1
--save_interval 346
--eval_interval 50
--eval_iters 10
--hidden_dropout 0.25
--lima_dropout
--position_embedding_type rotary
--no_bias_dropout_fusion
--use_checkpoint_args
--train_iters 519
--attention_dropout 0.0
--adam_beta1 0.9
--adam_beta2 0.95
--adam_eps 1e-12
--lr_decay_style cosine
--lr_warmup_iters 100
--lr 1e-5
--min_lr 1e-6
--weight_decay 0.000001
--sequence_parallel
--recompute_granularity selective
--log_timers_to_tensorboard
--rope_scaling_factor 1.0
--finetune
--wandb_logger

TheBloke
/

Llama2-70B-OASST-SFT-v10-GPTQ