RichardErkhov/cloudyu_-_Mixtral_11Bx2_MoE_19B-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Mixtral_11Bx2_MoE_19B - GGUF

• Model creator: https://huggingface.co/cloudyu/
• Original model: https://huggingface.co/cloudyu/Mixtral_11Bx2_MoE_19B/

Name	Quant method	Size
Mixtral_11Bx2_MoE_19B.Q2_K.gguf	Q2_K	6.58GB
Mixtral_11Bx2_MoE_19B.IQ3_XS.gguf	IQ3_XS	7.34GB
Mixtral_11Bx2_MoE_19B.IQ3_S.gguf	IQ3_S	7.75GB
Mixtral_11Bx2_MoE_19B.Q3_K_S.gguf	Q3_K_S	7.73GB
Mixtral_11Bx2_MoE_19B.IQ3_M.gguf	IQ3_M	7.94GB
Mixtral_11Bx2_MoE_19B.Q3_K.gguf	Q3_K	8.59GB
Mixtral_11Bx2_MoE_19B.Q3_K_M.gguf	Q3_K_M	8.59GB
Mixtral_11Bx2_MoE_19B.Q3_K_L.gguf	Q3_K_L	9.32GB
Mixtral_11Bx2_MoE_19B.IQ4_XS.gguf	IQ4_XS	9.66GB
Mixtral_11Bx2_MoE_19B.Q4_0.gguf	Q4_0	10.09GB
Mixtral_11Bx2_MoE_19B.IQ4_NL.gguf	IQ4_NL	10.19GB
Mixtral_11Bx2_MoE_19B.Q4_K_S.gguf	Q4_K_S	10.17GB
Mixtral_11Bx2_MoE_19B.Q4_K.gguf	Q4_K	10.79GB
Mixtral_11Bx2_MoE_19B.Q4_K_M.gguf	Q4_K_M	10.79GB
Mixtral_11Bx2_MoE_19B.Q4_1.gguf	Q4_1	11.19GB
Mixtral_11Bx2_MoE_19B.Q5_0.gguf	Q5_0	12.3GB
Mixtral_11Bx2_MoE_19B.Q5_K_S.gguf	Q5_K_S	12.3GB
Mixtral_11Bx2_MoE_19B.Q5_K.gguf	Q5_K	12.67GB
Mixtral_11Bx2_MoE_19B.Q5_K_M.gguf	Q5_K_M	12.67GB
Mixtral_11Bx2_MoE_19B.Q5_1.gguf	Q5_1	13.41GB
Mixtral_11Bx2_MoE_19B.Q6_K.gguf	Q6_K	14.66GB
Mixtral_11Bx2_MoE_19B.Q8_0.gguf	Q8_0	18.99GB

Original model description:

license: cc-by-nc-4.0

Mixtral MOE 2x10.7B

One of Best MoE Model reviewd by reddit community

MoE of the following models :

• kyujinpy/Sakura-SOLAR-Instruct

• jeonsworld/CarbonVillain-en-10.7B-v1

• Local Test

• hf (pretrained=cloudyu/Mixtral_11Bx2_MoE_19B), gen_kwargs: (None), limit: None, num_fewshot: 10, batch_size: auto (32)

  Tasks	Version	Filter	n-shot	Metric	Value		Stderr
  hellaswag	Yaml	none	10	acc	0.7142	±	0.0045
  		none	10	acc_norm	0.8819	±	0.0032

gpu code example

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
import math

## v2 models
model_path = "cloudyu/Mixtral_11Bx2_MoE_19B"

tokenizer = AutoTokenizer.from_pretrained(model_path, use_default_system_prompt=False)
model = AutoModelForCausalLM.from_pretrained(
    model_path, torch_dtype=torch.float32, device_map='auto',local_files_only=False, load_in_4bit=True
)
print(model)
prompt = input("please input prompt:")
while len(prompt) > 0:
  input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to("cuda")

  generation_output = model.generate(
    input_ids=input_ids, max_new_tokens=500,repetition_penalty=1.2
  )
  print(tokenizer.decode(generation_output[0]))
  prompt = input("please input prompt:")

CPU example

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
import math

## v2 models
model_path = "cloudyu/Mixtral_11Bx2_MoE_19B"

tokenizer = AutoTokenizer.from_pretrained(model_path, use_default_system_prompt=False)
model = AutoModelForCausalLM.from_pretrained(
    model_path, torch_dtype=torch.float32, device_map='cpu',local_files_only=False
)
print(model)
prompt = input("please input prompt:")
while len(prompt) > 0:
  input_ids = tokenizer(prompt, return_tensors="pt").input_ids

  generation_output = model.generate(
    input_ids=input_ids, max_new_tokens=500,repetition_penalty=1.2
  )
  print(tokenizer.decode(generation_output[0]))
  prompt = input("please input prompt:")