You can deploy this model and start earning money today!

RichardErkhov/
state-spaces_-_mamba-130m-hf-gguf

Name: state-spaces_-_mamba-130m-hf-gguf
Availability: InStock

endpoints_compatible

gguf

template

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

mamba-130m-hf - GGUF

Model creator: https://huggingface.co/state-spaces/
Original model: https://huggingface.co/state-spaces/mamba-130m-hf/

Name	Quant method	Size
mamba-130m-hf.Q2_K.gguf	Q2_K	0.08GB
mamba-130m-hf.IQ3_XS.gguf	IQ3_XS	0.09GB
mamba-130m-hf.IQ3_S.gguf	IQ3_S	0.09GB
mamba-130m-hf.Q3_K_S.gguf	Q3_K_S	0.09GB
mamba-130m-hf.IQ3_M.gguf	IQ3_M	0.09GB
mamba-130m-hf.Q3_K.gguf	Q3_K	0.09GB
mamba-130m-hf.Q3_K_M.gguf	Q3_K_M	0.09GB
mamba-130m-hf.Q3_K_L.gguf	Q3_K_L	0.09GB
mamba-130m-hf.IQ4_XS.gguf	IQ4_XS	0.09GB
mamba-130m-hf.Q4_0.gguf	Q4_0	0.1GB
mamba-130m-hf.IQ4_NL.gguf	IQ4_NL	0.1GB
mamba-130m-hf.Q4_K_S.gguf	Q4_K_S	0.1GB
mamba-130m-hf.Q4_K.gguf	Q4_K	0.1GB
mamba-130m-hf.Q4_K_M.gguf	Q4_K_M	0.1GB
mamba-130m-hf.Q4_1.gguf	Q4_1	0.1GB
mamba-130m-hf.Q5_0.gguf	Q5_0	0.11GB
mamba-130m-hf.Q5_K_S.gguf	Q5_K_S	0.11GB
mamba-130m-hf.Q5_K.gguf	Q5_K	0.11GB
mamba-130m-hf.Q5_K_M.gguf	Q5_K_M	0.11GB
mamba-130m-hf.Q5_1.gguf	Q5_1	0.11GB
mamba-130m-hf.Q6_K.gguf	Q6_K	0.12GB

Original model description:

library_name: transformers tags: []

Mamba

This repository contains the transfromers compatible mamba-2.8b. The checkpoints are untouched, but the full config.json and tokenizer are pushed to this repo.

Usage

You need to install transformers from main until transformers=4.39.0 is released.

pip install git+https://github.com/huggingface/transformers@main

We also recommend you to install both causal_conv_1d and mamba-ssm using:

bash

pip install causal-conv1d>=1.2.0
pip install mamba-ssm

If any of these two is not installed, the "eager" implementation will be used. Otherwise the more optimised cuda kernels will be used.

Generation

You can use the classic generate API:

python

>>> from transformers import MambaConfig, MambaForCausalLM, AutoTokenizer
>>> import torch

>>> tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-130m-hf")
>>> model = MambaForCausalLM.from_pretrained("state-spaces/mamba-130m-hf")
>>> input_ids = tokenizer("Hey how are you doing?", return_tensors="pt")["input_ids"]

>>> out = model.generate(input_ids, max_new_tokens=10)
>>> print(tokenizer.batch_decode(out))
["Hey how are you doing?\n\nI'm so glad you're here."]

PEFT finetuning example

In order to finetune using the peft library, we recommend keeping the model in float32!