updated model card for Mistral (#37156)
* model card for Mistral * Update docs/source/en/model_doc/mistral.md Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Apply suggestions from code review Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Update docs/source/en/model_doc/mistral.md Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Update docs/source/en/model_doc/mistral.md Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Update docs/source/en/model_doc/mistral.md Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Update docs/source/en/model_doc/mistral.md Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * apply suggestions * fix typo * updated with comments * updated with comments * updated with comments * remove hfoption block --------- Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
This commit is contained in:
Nahieli
@@ -14,8 +14,7 @@ rendered properly in your Markdown viewer.
|
|||||||
|
|
||||||
-->
|
-->
|
||||||
|
|
||||||
# Mistral
|
<div style="float: right;">
|
||||||
|
|
||||||
<div class="flex flex-wrap space-x-1">
|
<div class="flex flex-wrap space-x-1">
|
||||||
<img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-DE3412?style=flat&logo=pytorch&logoColor=white">
|
<img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-DE3412?style=flat&logo=pytorch&logoColor=white">
|
||||||
<img alt="TensorFlow" src="https://img.shields.io/badge/TensorFlow-FF6F00?style=flat&logo=tensorflow&logoColor=white">
|
<img alt="TensorFlow" src="https://img.shields.io/badge/TensorFlow-FF6F00?style=flat&logo=tensorflow&logoColor=white">
|
||||||
@@ -24,64 +23,46 @@ rendered properly in your Markdown viewer.
|
|||||||
<img alt="FlashAttention" src="https://img.shields.io/badge/%E2%9A%A1%EF%B8%8E%20FlashAttention-eae0c8?style=flat">
|
<img alt="FlashAttention" src="https://img.shields.io/badge/%E2%9A%A1%EF%B8%8E%20FlashAttention-eae0c8?style=flat">
|
||||||
<img alt="SDPA" src="https://img.shields.io/badge/SDPA-DE3412?style=flat&logo=pytorch&logoColor=white">
|
<img alt="SDPA" src="https://img.shields.io/badge/SDPA-DE3412?style=flat&logo=pytorch&logoColor=white">
|
||||||
</div>
|
</div>
|
||||||
|
</div>
|
||||||
|
|
||||||
## Overview
|
# Mistral
|
||||||
|
|
||||||
Mistral was introduced in the [this blogpost](https://mistral.ai/news/announcing-mistral-7b/) by Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.
|
[Mistral](https://huggingface.co/papers/2310.06825) is a 7B parameter language model, available as a pretrained and instruction-tuned variant, focused on balancing
|
||||||
|
the scaling costs of large models with performance and efficient inference. This model uses sliding window attention (SWA) trained with a 8K context length and a fixed cache size to handle longer sequences more effectively. Grouped-query attention (GQA) speeds up inference and reduces memory requirements. Mistral also features a byte-fallback BPE tokenizer to improve token handling and efficiency by ensuring characters are never mapped to out-of-vocabulary tokens.
|
||||||
|
|
||||||
The introduction of the blog post says:
|
You can find all the original Mistral checkpoints under the [Mistral AI_](https://huggingface.co/mistralai) organization.
|
||||||
|
|
||||||
*Mistral AI team is proud to release Mistral 7B, the most powerful language model for its size to date.*
|
> [!TIP]
|
||||||
|
> Click on the Mistral models in the right sidebar for more examples of how to apply Mistral to different language tasks.
|
||||||
|
|
||||||
Mistral-7B is the first large language model (LLM) released by [mistral.ai](https://mistral.ai/).
|
The example below demonstrates how to chat with [`Pipeline`] or the [`AutoModel`], and from the command line.
|
||||||
|
|
||||||
### Architectural details
|
<hfoptions id="usage">
|
||||||
|
<hfoption id="Pipeline">
|
||||||
Mistral-7B is a decoder-only Transformer with the following architectural choices:
|
|
||||||
|
|
||||||
- Sliding Window Attention - Trained with 8k context length and fixed cache size, with a theoretical attention span of 128K tokens
|
|
||||||
- GQA (Grouped Query Attention) - allowing faster inference and lower cache size.
|
|
||||||
- Byte-fallback BPE tokenizer - ensures that characters are never mapped to out of vocabulary tokens.
|
|
||||||
|
|
||||||
For more details refer to the [release blog post](https://mistral.ai/news/announcing-mistral-7b/).
|
|
||||||
|
|
||||||
### License
|
|
||||||
|
|
||||||
`Mistral-7B` is released under the Apache 2.0 license.
|
|
||||||
|
|
||||||
## Usage tips
|
|
||||||
|
|
||||||
The Mistral team has released 3 checkpoints:
|
|
||||||
|
|
||||||
- a base model, [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1), which has been pre-trained to predict the next token on internet-scale data.
|
|
||||||
- an instruction tuned model, [Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1), which is the base model optimized for chat purposes using supervised fine-tuning (SFT) and direct preference optimization (DPO).
|
|
||||||
- an improved instruction tuned model, [Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2), which improves upon v1.
|
|
||||||
|
|
||||||
The base model can be used as follows:
|
|
||||||
|
|
||||||
```python
|
```python
|
||||||
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
|
>>> import torch
|
||||||
|
>>> from transformers import pipeline
|
||||||
|
|
||||||
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", device_map="auto")
|
>>> messages = [
|
||||||
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
|
... {"role": "user", "content": "What is your favourite condiment?"},
|
||||||
|
... {"role": "assistant", "content": "Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!"},
|
||||||
|
... {"role": "user", "content": "Do you have mayonnaise recipes?"}
|
||||||
|
... ]
|
||||||
|
|
||||||
>>> prompt = "My favourite condiment is"
|
>>> chatbot = pipeline("text-generation", model="mistralai/Mistral-7B-Instruct-v0.3", torch_dtype=torch.bfloat16, device=0)
|
||||||
|
>>> chatbot(messages)
|
||||||
>>> model_inputs = tokenizer([prompt], return_tensors="pt").to("cuda")
|
|
||||||
>>> model.to(device)
|
|
||||||
|
|
||||||
>>> generated_ids = model.generate(**model_inputs, max_new_tokens=100, do_sample=True)
|
|
||||||
>>> tokenizer.batch_decode(generated_ids)[0]
|
|
||||||
"My favourite condiment is to ..."
|
|
||||||
```
|
```
|
||||||
|
|
||||||
The instruction tuned model can be used as follows:
|
</hfoption>
|
||||||
|
<hfoption id="AutoModel">
|
||||||
|
|
||||||
```python
|
```python
|
||||||
|
>>> import torch
|
||||||
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
|
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
|
||||||
|
|
||||||
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", device_map="auto")
|
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3", torch_dtype=torch.bfloat16, attn_implementation="sdpa", device_map="auto")
|
||||||
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
|
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
||||||
|
|
||||||
>>> messages = [
|
>>> messages = [
|
||||||
... {"role": "user", "content": "What is your favourite condiment?"},
|
... {"role": "user", "content": "What is your favourite condiment?"},
|
||||||
@@ -96,59 +77,20 @@ The instruction tuned model can be used as follows:
|
|||||||
"Mayonnaise can be made as follows: (...)"
|
"Mayonnaise can be made as follows: (...)"
|
||||||
```
|
```
|
||||||
|
|
||||||
As can be seen, the instruction-tuned model requires a [chat template](../chat_templating) to be applied to make sure the inputs are prepared in the right format.
|
</hfoption>
|
||||||
|
<hfoption id="transformers-cli">
|
||||||
## Speeding up Mistral by using Flash Attention
|
|
||||||
|
|
||||||
The code snippets above showcase inference without any optimization tricks. However, one can drastically speed up the model by leveraging [Flash Attention](../perf_train_gpu_one#flash-attention-2), which is a faster implementation of the attention mechanism used inside the model.
|
|
||||||
|
|
||||||
First, make sure to install the latest version of Flash Attention 2 to include the sliding window attention feature.
|
|
||||||
|
|
||||||
```bash
|
|
||||||
pip install -U flash-attn --no-build-isolation
|
|
||||||
```
|
|
||||||
|
|
||||||
Make also sure that you have a hardware that is compatible with Flash-Attention 2. Read more about it in the official documentation of the [flash attention repository](https://github.com/Dao-AILab/flash-attention). Make also sure to load your model in half-precision (e.g. `torch.float16`)
|
|
||||||
|
|
||||||
To load and run a model using Flash Attention-2, refer to the snippet below:
|
|
||||||
|
|
||||||
```python
|
```python
|
||||||
>>> import torch
|
echo -e "My favorite condiment is" | transformers-cli chat --model_name_or_path mistralai/Mistral-7B-v0.3 --torch_dtype auto --device 0 --attn_implementation flash_attention_2
|
||||||
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
||||||
|
|
||||||
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", torch_dtype=torch.float16, attn_implementation="flash_attention_2", device_map="auto")
|
|
||||||
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
|
|
||||||
|
|
||||||
>>> prompt = "My favourite condiment is"
|
|
||||||
|
|
||||||
>>> model_inputs = tokenizer([prompt], return_tensors="pt").to("cuda")
|
|
||||||
>>> model.to(device)
|
|
||||||
|
|
||||||
>>> generated_ids = model.generate(**model_inputs, max_new_tokens=100, do_sample=True)
|
|
||||||
>>> tokenizer.batch_decode(generated_ids)[0]
|
|
||||||
"My favourite condiment is to (...)"
|
|
||||||
```
|
```
|
||||||
|
|
||||||
### Expected speedups
|
</hfoption>
|
||||||
|
</hfoptions>
|
||||||
|
|
||||||
Below is a expected speedup diagram that compares pure inference time between the native implementation in transformers using `mistralai/Mistral-7B-v0.1` checkpoint and the Flash Attention 2 version of the model.
|
|
||||||
|
|
||||||
<div style="text-align: center">
|
Quantization reduces the memory burden of large models by representing the weights in a lower precision. Refer to the [Quantization](../quantization/overview) overview for more available quantization backends.
|
||||||
<img src="https://huggingface.co/datasets/ybelkada/documentation-images/resolve/main/mistral-7b-inference-large-seqlen.png">
|
|
||||||
</div>
|
|
||||||
|
|
||||||
### Sliding window Attention
|
The example below uses [bitsandbytes](../quantization/bitsandbytes) to only quantize the weights to 4-bits.
|
||||||
|
|
||||||
The current implementation supports the sliding window attention mechanism and memory efficient cache management.
|
|
||||||
To enable sliding window attention, just make sure to have a `flash-attn` version that is compatible with sliding window attention (`>=2.3.0`).
|
|
||||||
|
|
||||||
The Flash Attention-2 model uses also a more memory efficient cache slicing mechanism - as recommended per the official implementation of Mistral model that use rolling cache mechanism we keep the cache size fixed (`self.config.sliding_window`), support batched generation only for `padding_side="left"` and use the absolute position of the current token to compute the positional embedding.
|
|
||||||
|
|
||||||
## Shrinking down Mistral using quantization
|
|
||||||
|
|
||||||
As the Mistral model has 7 billion parameters, that would require about 14GB of GPU RAM in half precision (float16), since each parameter is stored in 2 bytes. However, one can shrink down the size of the model using [quantization](../quantization.md). If the model is quantized to 4 bits (or half a byte per parameter),that requires only about 3.5GB of RAM.
|
|
||||||
|
|
||||||
Quantizing a model is as simple as passing a `quantization_config` to the model. Below, we'll leverage the BitsAndyBytes quantization (but refer to [this page](../quantization.md) for other quantization methods):
|
|
||||||
|
|
||||||
```python
|
```python
|
||||||
>>> import torch
|
>>> import torch
|
||||||
@@ -161,8 +103,8 @@ Quantizing a model is as simple as passing a `quantization_config` to the model.
|
|||||||
... bnb_4bit_compute_dtype="torch.float16",
|
... bnb_4bit_compute_dtype="torch.float16",
|
||||||
... )
|
... )
|
||||||
|
|
||||||
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", quantization_config=True, device_map="auto")
|
>>> model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3", quantization_config=True, torch_dtype=torch.bfloat16, device_map="auto")
|
||||||
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
|
>>> tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
||||||
|
|
||||||
>>> prompt = "My favourite condiment is"
|
>>> prompt = "My favourite condiment is"
|
||||||
|
|
||||||
@@ -179,19 +121,18 @@ Quantizing a model is as simple as passing a `quantization_config` to the model.
|
|||||||
"The expected output"
|
"The expected output"
|
||||||
```
|
```
|
||||||
|
|
||||||
This model was contributed by [Younes Belkada](https://huggingface.co/ybelkada) and [Arthur Zucker](https://huggingface.co/ArthurZ) .
|
Use the [AttentionMaskVisualizer](https://github.com/huggingface/transformers/blob/beb9b5b02246b9b7ee81ddf938f93f44cfeaad19/src/transformers/utils/attention_visualizer.py#L139) to better understand what tokens the model can and cannot attend to.
|
||||||
The original code can be found [here](https://github.com/mistralai/mistral-src).
|
|
||||||
|
|
||||||
## Resources
|
```py
|
||||||
|
>>> from transformers.utils.attention_visualizer import AttentionMaskVisualizer
|
||||||
|
|
||||||
A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with Mistral. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
|
>>> visualizer = AttentionMaskVisualizer("mistralai/Mistral-7B-Instruct-v0.3")
|
||||||
|
>>> visualizer("Do you have mayonnaise recipes?")
|
||||||
|
```
|
||||||
|
|
||||||
<PipelineTag pipeline="text-generation"/>
|
<div class="flex justify-center">
|
||||||
|
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/mistral-attn-mask.png"/>
|
||||||
- A demo notebook to perform supervised fine-tuning (SFT) of Mistral-7B can be found [here](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/Mistral/Supervised_fine_tuning_(SFT)_of_an_LLM_using_Hugging_Face_tooling.ipynb). 🌎
|
</div>
|
||||||
- A [blog post](https://www.philschmid.de/fine-tune-llms-in-2024-with-trl) on how to fine-tune LLMs in 2024 using Hugging Face tooling. 🌎
|
|
||||||
- The [Alignment Handbook](https://github.com/huggingface/alignment-handbook) by Hugging Face includes scripts and recipes to perform supervised fine-tuning (SFT) and direct preference optimization with Mistral-7B. This includes scripts for full fine-tuning, QLoRa on a single GPU as well as multi-GPU fine-tuning.
|
|
||||||
- [Causal language modeling task guide](../tasks/language_modeling)
|
|
||||||
|
|
||||||
## MistralConfig
|
## MistralConfig
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user