RichardErkhov/deepnight-research_-_saily_100b-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

saily_100b - GGUF

• Model creator: https://huggingface.co/deepnight-research/
• Original model: https://huggingface.co/deepnight-research/saily_100b/

Name	Quant method	Size
saily_100b.Q2_K.gguf	Q2_K	40.28GB
saily_100b.IQ3_XS.gguf	IQ3_XS	44.82GB
saily_100b.IQ3_S.gguf	IQ3_S	47.37GB
saily_100b.Q3_K_S.gguf	Q3_K_S	47.22GB
saily_100b.IQ3_M.gguf	IQ3_M	49.0GB
saily_100b.Q3_K.gguf	Q3_K	52.7GB
saily_100b.Q3_K_M.gguf	Q3_K_M	52.7GB
saily_100b.Q3_K_L.gguf	Q3_K_L	57.43GB
saily_100b.IQ4_XS.gguf	IQ4_XS	59.08GB
saily_100b.Q4_0.gguf	Q4_0	61.76GB
saily_100b.IQ4_NL.gguf	IQ4_NL	62.35GB
saily_100b.Q4_K_S.gguf	Q4_K_S	62.22GB
saily_100b.Q4_K.gguf	Q4_K	65.79GB
saily_100b.Q4_K_M.gguf	Q4_K_M	65.79GB
saily_100b.Q4_1.gguf	Q4_1	68.59GB
saily_100b.Q5_0.gguf	Q5_0	75.43GB
saily_100b.Q5_K_S.gguf	Q5_K_S	75.43GB
saily_100b.Q5_K.gguf	Q5_K	77.51GB
saily_100b.Q5_K_M.gguf	Q5_K_M	77.51GB
saily_100b.Q5_1.gguf	Q5_1	82.27GB
saily_100b.Q6_K.gguf	Q6_K	89.96GB
saily_100b.Q8_0.gguf	Q8_0	116.52GB

Original model description:

license: mit license_name: deepnight-responsible-ai license_link: LICENSE

SaiLy 100B (deepnight-research/saily_100B)

---

SaiLy is a series/collection of AI Models by DEEPNIGHT-RESEARCH which are highly experimental and uncensored. Please use with responsibility.

---

*waiting for evals, the model is submitted on HuggingFace OpenLLM Leaderboard, and is currently in the pending list* Prompt Template: Alpaca

Below is an instruction that describes a task. Write a response that appropriately completes the request.
### Instruction:
{prompt}
### Response:

Description:

This is the first stable model of the series. The model is based on Llama2-chat.

---

Did some said CODE?

Here you go!

import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
  'deepnight-research/saily_100B'
)

To use the optimized triton implementation of FlashAttention, you can load the model on GPU (cuda:0) with attn_impl='triton' and with bfloat16 precision:

import torch
import transformers

name = 'deepnight-research/saily_100B'

config = transformers.AutoConfig.from_pretrained(name)
config.attn_config['attn_impl'] = 'triton'
config.init_device = 'cuda:0' # For fast initialization directly on GPU!

model = transformers.AutoModelForCausalLM.from_pretrained(
  name,
  config=config,
  torch_dtype=torch.bfloat16, # Load model weights in bfloat16
  trust_remote_code=True
)

---

If you would like to support us, please consider donating for #aiforcause.

Cheers✌️

• Team DEEPNIGHT