Downloads · 30 days
10
17% of all-time downloads
mamba-2.8b-SFT
walebadr/mamba-2.8b-SFT
mamba-2.8b-SFT is a machine learning model from walebadr. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
This is a the state-spaces mamba-2.8b model, fine-tuned using Supervised Fine-tuning method (SFT) on llama-2-7b-miniguanaco dataset.
All-time downloads
60
Public
Repo size
11.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin5.5 GB · 100%
At a glance
- Library
- transformers
- License
- apache-2.0
- Access
- Public
- Created
- Jan 15, 2024
- Updated
- Jan 15, 2024
- SHA
- 680064df
Model card
From the Hugging Face model README
This is a the state-spaces mamba-2.8b model, fine-tuned using Supervised Fine-tuning method (SFT) on llama-2-7b-miniguanaco dataset.
To run inference on this model, run the following code:
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
from mamba_ssm.models.mixer_seq_simple import MambaLMHeadModel
#Load the model
model = MambaLMHeadModel.from_pretrained("walebadr/mamba-2.8b-SFT", dtype=torch.bfloat16, device="cuda")
device = "cuda"
messages = []
user_message = f"[INST] what is a language model? [/INST]"
input_ids = tokenizer(user_message, return_tensors="pt").input_ids.to("cuda")
out = model.generate(input_ids=input_ids, max_length=500, temperature=0.9, top_p=0.7, eos_token_id=tokenizer.eos_token_id)
decoded = tokenizer.batch_decode(out)
print("Model:", decoded[0])
Model Evaluation
Coming soon
- Library
- transformers
- License
- apache-2.0
- Created
- Jan 15, 2024
- Updated
- Jan 15, 2024