---
base_model: mistralai/Mistral-7B-Instruct-v0.3
language:
- en
pipeline_tag: text-generation
license: apache-2.0
model_creator: Mistral AI
model_name: Mistral-7B-Instruct-v0.3
model_type: mistral
quantized_by: CISC
---
# Mistral-7B-Instruct-v0.3 - SOTA GGUF
- Model creator: [Mistral AI](https://huggingface.co/mistralai)
- Original model: [Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3)
## Description
This repo contains State Of The Art quantized GGUF format model files for [Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3).
Quantization was done with an importance matrix that was trained for ~1M tokens (256 batches of 4096 tokens) of [groups_merged.txt](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384) and [wiki.train.raw](https://raw.githubusercontent.com/pytorch/examples/main/word_language_model/data/wikitext-2/train.txt) concatenated.
The embedded chat template has been extended to support function calling via OpenAI-compatible `tools` parameter, see [example](#simple-llama-cpp-python-example-function-calling-code).
## Prompt template: Mistral v3
```
[AVAILABLE_TOOLS] [{"name": "function_name", "description": "Description", "parameters": {...}}, ...][/AVAILABLE_TOOLS][INST] {prompt}[/INST]
```
## Compatibility
These quantised GGUFv3 files are compatible with llama.cpp from February 27th 2024 onwards, as of commit [0becb22](https://github.com/ggerganov/llama.cpp/commit/0becb22ac05b6542bd9d5f2235691aa1d3d4d307)
They are also compatible with many third party UIs and libraries provided they are built using a recent llama.cpp.
## Explanation of quantisation methods
Click to see details
The new methods available are:
* GGML_TYPE_IQ1_S - 1-bit quantization in super-blocks with an importance matrix applied, effectively using 1.56 bits per weight (bpw)
* GGML_TYPE_IQ1_M - 1-bit quantization in super-blocks with an importance matrix applied, effectively using 1.75 bpw
* GGML_TYPE_IQ2_XXS - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.06 bpw
* GGML_TYPE_IQ2_XS - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.31 bpw
* GGML_TYPE_IQ2_S - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.5 bpw
* GGML_TYPE_IQ2_M - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.7 bpw
* GGML_TYPE_IQ3_XXS - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.06 bpw
* GGML_TYPE_IQ3_XS - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.3 bpw
* GGML_TYPE_IQ3_S - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.44 bpw
* GGML_TYPE_IQ3_M - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.66 bpw
* GGML_TYPE_IQ4_XS - 4-bit quantization in super-blocks with an importance matrix applied, effectively using 4.25 bpw
* GGML_TYPE_IQ4_NL - 4-bit non-linearly mapped quantization with an importance matrix applied, effectively using 4.5 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 |
| ---- | ---- | ---- | ---- | ---- | ----- |
| [Mistral-7B-Instruct-v0.3.IQ1_S.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ1_S.gguf) | IQ1_S | 1 | 1.5 GB| 2.5 GB | smallest, significant quality loss - **TBD**: Waiting for [this issue](https://github.com/ggerganov/llama.cpp/issues/5996) to be resolved |
| [Mistral-7B-Instruct-v0.3.IQ1_M.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ1_M.gguf) | IQ1_M | 1 | 1.6 GB| 2.6 GB | very small, significant quality loss |
| [Mistral-7B-Instruct-v0.3.IQ2_XXS.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ2_XXS.gguf) | IQ2_XXS | 2 | 1.8 GB| 2.8 GB | very small, high quality loss |
| [Mistral-7B-Instruct-v0.3.IQ2_XS.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ2_XS.gguf) | IQ2_XS | 2 | 1.9 GB| 2.9 GB | very small, high quality loss |
| [Mistral-7B-Instruct-v0.3.IQ2_S.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ2_S.gguf) | IQ2_S | 2 | 2.1 GB| 3.1 GB | small, substantial quality loss |
| [Mistral-7B-Instruct-v0.3.IQ2_M.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ2_M.gguf) | IQ2_M | 2 | 2.2 GB| 3.2 GB | small, greater quality loss |
| [Mistral-7B-Instruct-v0.3.IQ3_XXS.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ3_XXS.gguf) | IQ3_XXS | 3 | 2.5 GB| 3.5 GB | very small, high quality loss |
| [Mistral-7B-Instruct-v0.3.IQ3_XS.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ3_XS.gguf) | IQ3_XS | 3 | 2.7 GB| 3.7 GB | small, substantial quality loss |
| [Mistral-7B-Instruct-v0.3.IQ3_S.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ3_S.gguf) | IQ3_S | 3 | 2.8 GB| 3.8 GB | small, greater quality loss |
| [Mistral-7B-Instruct-v0.3.IQ3_M.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ3_M.gguf) | IQ3_M | 3 | 3.0 GB| 4.0 GB | medium, balanced quality - recommended |
| [Mistral-7B-Instruct-v0.3.IQ4_XS.gguf](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.IQ4_XS.gguf) | IQ4_XS | 4 | 3.4 GB| 4.4 GB | small, substantial quality loss |
Generated importance matrix file: [Mistral-7B-Instruct-v0.3.imatrix.dat](https://huggingface.co/CISCai/Mistral-7B-Instruct-v0.3-SOTA-GGUF/blob/main/Mistral-7B-Instruct-v0.3.imatrix.dat)
**Note**: the above RAM figures assume no GPU offloading with 4K context. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
## Example `llama.cpp` command
Make sure you are using `llama.cpp` from commit [0becb22](https://github.com/ggerganov/llama.cpp/commit/0becb22ac05b6542bd9d5f2235691aa1d3d4d307) or later.
```shell
./main -ngl 33 -m Mistral-7B-Instruct-v0.3.IQ4_XS.gguf --color -c 32768 --temp 0 --repeat-penalty 1.1 -p "[AVAILABLE_TOOLS] {tools}[/AVAILABLE_TOOLS][INST] {prompt}[/INST]"
```
Change `-ngl 33` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
Change `-c 32768` to the desired sequence length.
If you want to have a chat-style conversation, replace the `-p ` argument with `-i -ins`
If you are low on V/RAM try quantizing the K-cache with `-ctk q8_0` or even `-ctk q4_0` for big memory savings (depending on context size).
There is a similar option for V-cache (`-ctv`), however that is [not working yet](https://github.com/ggerganov/llama.cpp/issues/4425).
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 from Python code
You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) module.
### How to load this model in Python code, using llama-cpp-python
For full documentation, please see: [llama-cpp-python docs](https://llama-cpp-python.readthedocs.io/en/latest/).
#### First install the package
Run one of the following commands, according to your system:
```shell
# Prebuilt wheel with basic CPU support
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
# Prebuilt wheel with NVidia CUDA acceleration
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121 (or cu122 etc.)
# Prebuilt wheel with Metal GPU acceleration
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/metal
# Build base version with no GPU acceleration
pip install llama-cpp-python
# With NVidia CUDA acceleration
CMAKE_ARGS="-DLLAMA_CUDA=on" pip install llama-cpp-python
# Or with OpenBLAS acceleration
CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python
# Or with CLBLast acceleration
CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python
# Or with AMD ROCm GPU acceleration (Linux only)
CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python
# Or with Metal GPU acceleration for macOS systems only
CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python
# Or with Vulkan acceleration
CMAKE_ARGS="-DLLAMA_VULKAN=on" pip install llama-cpp-python
# Or with Kompute acceleration
CMAKE_ARGS="-DLLAMA_KOMPUTE=on" pip install llama-cpp-python
# Or with SYCL acceleration
CMAKE_ARGS="-DLLAMA_SYCL=on -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx" pip install llama-cpp-python
# In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA:
$env:CMAKE_ARGS = "-DLLAMA_CUDA=on"
pip install llama-cpp-python
```
#### Simple llama-cpp-python example code
```python
from llama_cpp import Llama
# Chat Completion API
llm = Llama(model_path="./Mistral-7B-Instruct-v0.3.IQ4_XS.gguf", n_gpu_layers=33, n_ctx=32768)
print(llm.create_chat_completion(
messages = [
{
"role": "user",
"content": "Pick a LeetCode challenge and solve it in Python."
}
]
))
```
#### Simple llama-cpp-python example function calling code
```python
from llama_cpp import Llama
# Chat Completion API
grammar = LlamaGrammar.from_json_schema(json.dumps({
"type": "array",
"items": {
"type": "object",
"required": [ "name", "arguments" ],
"properties": {
"name": {
"type": "string"
},
"arguments": {
"type": "object"
}
}
}
}))
llm = Llama(model_path="./Mistral-7B-Instruct-v0.3.IQ4_XS.gguf", n_gpu_layers=33, n_ctx=32768, temperature=0.0, repeat_penalty=1.1)
response = llm.create_chat_completion(
messages = [
{
"role": "user",
"content": "What's the weather like in Oslo and Stockholm?"
}
],
tools=[{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather in a given location",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
},
"unit": {
"type": "string",
"enum": [ "celsius", "fahrenheit" ]
}
},
"required": [ "location" ]
}
}
}],
grammar = grammar
)
print(json.loads(response["choices"][0]["text"]))
print(llm.create_chat_completion(
messages = [
{
"role": "user",
"content": "What's the weather like in Oslo?"
},
{ # The tool_calls is from the response to the above with tool_choice active
"role": "assistant",
"content": None,
"tool_calls": [
{
"id": "call__0_get_current_weather_cmpl-...",
"type": "function",
"function": {
"name": "get_current_weather",
"arguments": '{ "location": "Oslo, NO" ,"unit": "celsius"} '
}
}
]
},
{ # The tool_call_id is from tool_calls and content is the result from the function call you made
"role": "tool",
"content": "20",
"tool_call_id": "call__0_get_current_weather_cmpl-..."
}
],
tools=[{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather in a given location",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
},
"unit": {
"type": "string",
"enum": [ "celsius", "fahrenheit" ]
}
},
"required": [ "location" ]
}
}
}],
#tool_choice={
# "type": "function",
# "function": {
# "name": "get_current_weather"
# }
#}
))
```