Instructions to use Heralax/Mannerstral-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Heralax/Mannerstral-base with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Heralax/Mannerstral-base", dtype="auto")

llama-cpp-python

How to use Heralax/Mannerstral-base with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Heralax/Mannerstral-base",
	filename="Mistral-7B-hf-v0.2-F16.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Heralax/Mannerstral-base with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Heralax/Mannerstral-base:F16
# Run inference directly in the terminal:
llama-cli -hf Heralax/Mannerstral-base:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Heralax/Mannerstral-base:F16
# Run inference directly in the terminal:
llama-cli -hf Heralax/Mannerstral-base:F16

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Heralax/Mannerstral-base:F16
# Run inference directly in the terminal:
./llama-cli -hf Heralax/Mannerstral-base:F16

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Heralax/Mannerstral-base:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Heralax/Mannerstral-base:F16

Use Docker

docker model run hf.co/Heralax/Mannerstral-base:F16

LM Studio
Jan
Ollama
How to use Heralax/Mannerstral-base with Ollama:
```
ollama run hf.co/Heralax/Mannerstral-base:F16
```

Unsloth Studio new

How to use Heralax/Mannerstral-base with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Heralax/Mannerstral-base to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Heralax/Mannerstral-base to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Heralax/Mannerstral-base to start chatting

Docker Model Runner
How to use Heralax/Mannerstral-base with Docker Model Runner:
```
docker model run hf.co/Heralax/Mannerstral-base:F16
```

Lemonade

How to use Heralax/Mannerstral-base with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Heralax/Mannerstral-base:F16

Run and chat with the model

lemonade run user.Mannerstral-base-F16

List all available models

lemonade list

See axolotl config

axolotl version: 0.4.1

base_model: alpindale/Mistral-7B-v0.2-hf
tokenizer_type: AutoTokenizer
is_mistral_derived_model: true
load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: json
    data_files: hidden_pretraining_manners.jsonl
    ds_type: json
    type: completion


dataset_prepared_path: last_run_prepared
output_dir: ./army-pretraining

sequence_len: 4096
sample_packing: false
pad_to_sequence_len: true
shuffle_merged_datasets: true

wandb_project: mistral-army
wandb_entity:
wandb_watch:
wandb_run_id:
wandb_log_model:

gradient_accumulation_steps: 6
micro_batch_size: 2
eval_batch_size: 1
num_epochs: 11
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.000020
weight_decay: 0
# Gradient clipping max norm
max_grad_norm: 1.0
noisy_embedding_alpha: 0
train_on_inputs: false
group_by_length: false
bf16: true
fp16: false
tf32: false

gradient_checkpointing: unsloth
early_stopping_patience:
resume_from_checkpoint: 
logging_steps: 1
xformers_attention:
flash_attention: true

chat_template: chatml

warmup_ratio: 0.5
auto_resume_from_checkpoints: false
#warmup_ratio: 0.5
eval_steps: 10
saves_per_epoch: 1
eval_sample_packing: false
save_total_limit: 3
debug:
deepspeed: deepspeed_configs/zero2.json
special_tokens:
  pad_token: "<|end_of_text|>"

pretrained base for Mannerstral 7b, only use if you are finetuning something on top of it.

Downloads last month: 12

GGUF

Model size

7B params

Architecture

llama

Hardware compatibility

8-bit

16-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Heralax/Mannerstral-base

Base model

mistral-community/Mistral-7B-v0.2

Quantized

(27)

this model

Quantizations

1 model