|
@@ -15,7 +15,7 @@ FROM ./mistral-7b-v0.1.Q4_0.gguf
|
|
|
(Optional) many chat models require a prompt template in order to answer correctly. A default prompt template can be specified with the `TEMPLATE` instruction in the `Modelfile`:
|
|
|
|
|
|
```
|
|
|
-FROM ./q4_0.bin
|
|
|
+FROM ./mistral-7b-v0.1.Q4_0.gguf
|
|
|
TEMPLATE "[INST] {{ .Prompt }} [/INST]"
|
|
|
```
|
|
|
|
|
@@ -37,55 +37,69 @@ ollama run example "What is your favourite condiment?"
|
|
|
|
|
|
## Importing (PyTorch & Safetensors)
|
|
|
|
|
|
-### Supported models
|
|
|
+> Importing from PyTorch and Safetensors is a longer process than importing from GGUF. Improvements that make it easier are a work in progress.
|
|
|
|
|
|
-Ollama supports a set of model architectures, with support for more coming soon:
|
|
|
+### Setup
|
|
|
|
|
|
-- Llama & Mistral
|
|
|
-- Falcon & RW
|
|
|
-- BigCode
|
|
|
+First, clone the `ollama/ollama` repo:
|
|
|
|
|
|
-To view a model's architecture, check the `config.json` file in its HuggingFace repo. You should see an entry under `architectures` (e.g. `LlamaForCausalLM`).
|
|
|
+```
|
|
|
+git clone git@github.com:ollama/ollama.git ollama
|
|
|
+cd ollama
|
|
|
+```
|
|
|
|
|
|
-### Step 1: Clone the HuggingFace repository (optional)
|
|
|
+and then fetch its `llama.cpp` submodule:
|
|
|
|
|
|
-If the model is currently hosted in a HuggingFace repository, first clone that repository to download the raw model.
|
|
|
+```shell
|
|
|
+git submodule init
|
|
|
+git submodule update llm/llama.cpp
|
|
|
+```
|
|
|
+
|
|
|
+Next, install the Python dependencies:
|
|
|
|
|
|
```
|
|
|
-git lfs install
|
|
|
-git clone https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1
|
|
|
-cd Mistral-7B-Instruct-v0.1
|
|
|
+python3 -m venv llm/llama.cpp/.venv
|
|
|
+source llm/llama.cpp/.venv/bin/activate
|
|
|
+pip install -r llm/llama.cpp/requirements.txt
|
|
|
```
|
|
|
|
|
|
-### Step 2: Convert and quantize to a `.bin` file (optional, for PyTorch and Safetensors)
|
|
|
+Then build the `quantize` tool:
|
|
|
+
|
|
|
+```
|
|
|
+make -C llm/llama.cpp quantize
|
|
|
+```
|
|
|
|
|
|
-If the model is in PyTorch or Safetensors format, a [Docker image](https://hub.docker.com/r/ollama/quantize) with the tooling required to convert and quantize models is available.
|
|
|
+### Clone the HuggingFace repository (optional)
|
|
|
|
|
|
-First, Install [Docker](https://www.docker.com/get-started/).
|
|
|
+If the model is currently hosted in a HuggingFace repository, first clone that repository to download the raw model.
|
|
|
|
|
|
-Next, to convert and quantize your model, run:
|
|
|
+Install [Git LFS](https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage), verify it's installed, and then clone the model's repository:
|
|
|
|
|
|
```
|
|
|
-docker run --rm -v .:/model ollama/quantize -q q4_0 /model
|
|
|
+git lfs install
|
|
|
+git clone https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1 model
|
|
|
```
|
|
|
|
|
|
-This will output two files into the directory:
|
|
|
+### Convert the model
|
|
|
|
|
|
-- `f16.bin`: the model converted to GGUF
|
|
|
-- `q4_0.bin` the model quantized to a 4-bit quantization (Ollama will use this file to create the Ollama model)
|
|
|
+> Note: some model architectures require using specific convert scripts. For example, Qwen models require running `convert-hf-to-gguf.py` instead of `convert.py`
|
|
|
|
|
|
-### Step 3: Write a `Modelfile`
|
|
|
+```
|
|
|
+python llm/llama.cpp/convert.py ./model --outtype f16 --outfile converted.bin
|
|
|
+```
|
|
|
|
|
|
-Next, create a `Modelfile` for your model:
|
|
|
+### Quantize the model
|
|
|
|
|
|
```
|
|
|
-FROM ./q4_0.bin
|
|
|
+llm/llama.cpp/quantize converted.bin quantized.bin q4_0
|
|
|
```
|
|
|
|
|
|
-(Optional) many chat models require a prompt template in order to answer correctly. A default prompt template can be specified with the `TEMPLATE` instruction in the `Modelfile`:
|
|
|
+### Step 3: Write a `Modelfile`
|
|
|
+
|
|
|
+Next, create a `Modelfile` for your model:
|
|
|
|
|
|
```
|
|
|
-FROM ./q4_0.bin
|
|
|
+FROM quantized.bin
|
|
|
TEMPLATE "[INST] {{ .Prompt }} [/INST]"
|
|
|
```
|
|
|
|
|
@@ -149,47 +163,3 @@ The quantization options are as follow (from highest highest to lowest levels of
|
|
|
- `q6_K`
|
|
|
- `q8_0`
|
|
|
- `f16`
|
|
|
-
|
|
|
-## Manually converting & quantizing models
|
|
|
-
|
|
|
-### Prerequisites
|
|
|
-
|
|
|
-Start by cloning the `llama.cpp` repo to your machine in another directory:
|
|
|
-
|
|
|
-```
|
|
|
-git clone https://github.com/ggerganov/llama.cpp.git
|
|
|
-cd llama.cpp
|
|
|
-```
|
|
|
-
|
|
|
-Next, install the Python dependencies:
|
|
|
-
|
|
|
-```
|
|
|
-pip install -r requirements.txt
|
|
|
-```
|
|
|
-
|
|
|
-Finally, build the `quantize` tool:
|
|
|
-
|
|
|
-```
|
|
|
-make quantize
|
|
|
-```
|
|
|
-
|
|
|
-### Convert the model
|
|
|
-
|
|
|
-Run the correct conversion script for your model architecture:
|
|
|
-
|
|
|
-```shell
|
|
|
-# LlamaForCausalLM or MistralForCausalLM
|
|
|
-python convert.py <path to model directory>
|
|
|
-
|
|
|
-# FalconForCausalLM
|
|
|
-python convert-falcon-hf-to-gguf.py <path to model directory>
|
|
|
-
|
|
|
-# GPTBigCodeForCausalLM
|
|
|
-python convert-starcoder-hf-to-gguf.py <path to model directory>
|
|
|
-```
|
|
|
-
|
|
|
-### Quantize the model
|
|
|
-
|
|
|
-```
|
|
|
-quantize <path to model dir>/ggml-model-f32.bin <path to model dir>/q4_0.bin q4_0
|
|
|
-```
|