Quantization made by Richard Erkhov.

[Github](https://github.com/RichardErkhov)

[Discord](https://discord.gg/pvy7H8DZMG)

[Request more models](https://github.com/RichardErkhov/quant_request)


mamba-370m-hf - GGUF
- Model creator: https://huggingface.co/state-spaces/
- Original model: https://huggingface.co/state-spaces/mamba-370m-hf/


| Name | Quant method | Size |
| ---- | ---- | ---- |
| [mamba-370m-hf.Q2_K.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q2_K.gguf) | Q2_K | 0.2GB |
| [mamba-370m-hf.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.IQ3_XS.gguf) | IQ3_XS | 0.23GB |
| [mamba-370m-hf.IQ3_S.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.IQ3_S.gguf) | IQ3_S | 0.23GB |
| [mamba-370m-hf.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q3_K_S.gguf) | Q3_K_S | 0.23GB |
| [mamba-370m-hf.IQ3_M.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.IQ3_M.gguf) | IQ3_M | 0.23GB |
| [mamba-370m-hf.Q3_K.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q3_K.gguf) | Q3_K | 0.23GB |
| [mamba-370m-hf.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q3_K_M.gguf) | Q3_K_M | 0.23GB |
| [mamba-370m-hf.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q3_K_L.gguf) | Q3_K_L | 0.23GB |
| [mamba-370m-hf.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.IQ4_XS.gguf) | IQ4_XS | 0.26GB |
| [mamba-370m-hf.Q4_0.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q4_0.gguf) | Q4_0 | 0.27GB |
| [mamba-370m-hf.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.IQ4_NL.gguf) | IQ4_NL | 0.27GB |
| [mamba-370m-hf.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q4_K_S.gguf) | Q4_K_S | 0.27GB |
| [mamba-370m-hf.Q4_K.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q4_K.gguf) | Q4_K | 0.27GB |
| [mamba-370m-hf.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q4_K_M.gguf) | Q4_K_M | 0.27GB |
| [mamba-370m-hf.Q4_1.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q4_1.gguf) | Q4_1 | 0.28GB |
| [mamba-370m-hf.Q5_0.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q5_0.gguf) | Q5_0 | 0.3GB |
| [mamba-370m-hf.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q5_K_S.gguf) | Q5_K_S | 0.3GB |
| [mamba-370m-hf.Q5_K.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q5_K.gguf) | Q5_K | 0.3GB |
| [mamba-370m-hf.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q5_K_M.gguf) | Q5_K_M | 0.3GB |
| [mamba-370m-hf.Q5_1.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q5_1.gguf) | Q5_1 | 0.32GB |
| [mamba-370m-hf.Q6_K.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q6_K.gguf) | Q6_K | 0.34GB |
| [mamba-370m-hf.Q8_0.gguf](https://huggingface.co/RichardErkhov/state-spaces_-_mamba-370m-hf-gguf/blob/main/mamba-370m-hf.Q8_0.gguf) | Q8_0 | 0.42GB |




Original model description:
---
library_name: transformers
tags: []
---

# Mamba

<!-- Provide a quick summary of what the model is/does. -->
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. 
```bash
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-370m-hf")
>>> model = MambaForCausalLM.from_pretrained("state-spaces/mamba-370m-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 doing great.\n\nI"]
```

## PEFT finetuning example
In order to finetune using the `peft` library, we recommend keeping the model in float32!

```python
from datasets import load_dataset
from trl import SFTTrainer
from peft import LoraConfig
from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments
tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-370m-hf")
model = AutoModelForCausalLM.from_pretrained("state-spaces/mamba-370m-hf")
dataset = load_dataset("Abirate/english_quotes", split="train")
training_args = TrainingArguments(
    output_dir="./results",
    num_train_epochs=3,
    per_device_train_batch_size=4,
    logging_dir='./logs',
    logging_steps=10,
    learning_rate=2e-3
)
lora_config =  LoraConfig(
        r=8,
        target_modules=["x_proj", "embeddings", "in_proj", "out_proj"],
        task_type="CAUSAL_LM",
        bias="none"
)
trainer = SFTTrainer(
    model=model,
    tokenizer=tokenizer,
    args=training_args,
    peft_config=lora_config,
    train_dataset=dataset,
    dataset_text_field="quote",
)
trainer.train()
```