RichardErkhov/abhishek_-_autotrain-mixtral7x8b-math-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

autotrain-mixtral7x8b-math - GGUF

• Model creator: https://huggingface.co/abhishek/
• Original model: https://huggingface.co/abhishek/autotrain-mixtral7x8b-math/

Name	Quant method	Size
autotrain-mixtral7x8b-math.Q2_K.gguf	Q2_K	16.12GB
autotrain-mixtral7x8b-math.IQ3_XS.gguf	IQ3_XS	18.02GB
autotrain-mixtral7x8b-math.IQ3_S.gguf	IQ3_S	19.03GB
autotrain-mixtral7x8b-math.Q3_K_S.gguf	Q3_K_S	19.03GB
autotrain-mixtral7x8b-math.IQ3_M.gguf	IQ3_M	19.96GB
autotrain-mixtral7x8b-math.Q3_K.gguf	Q3_K	21.0GB
autotrain-mixtral7x8b-math.Q3_K_M.gguf	Q3_K_M	21.0GB
autotrain-mixtral7x8b-math.Q3_K_L.gguf	Q3_K_L	22.51GB
autotrain-mixtral7x8b-math.IQ4_XS.gguf	IQ4_XS	23.63GB
autotrain-mixtral7x8b-math.Q4_0.gguf	Q4_0	24.63GB
autotrain-mixtral7x8b-math.IQ4_NL.gguf	IQ4_NL	24.91GB
autotrain-mixtral7x8b-math.Q4_K_S.gguf	Q4_K_S	24.91GB
autotrain-mixtral7x8b-math.Q4_K.gguf	Q4_K	26.49GB
autotrain-mixtral7x8b-math.Q4_K_M.gguf	Q4_K_M	26.49GB
autotrain-mixtral7x8b-math.Q4_1.gguf	Q4_1	27.32GB
autotrain-mixtral7x8b-math.Q5_0.gguf	Q5_0	30.02GB
autotrain-mixtral7x8b-math.Q5_K_S.gguf	Q5_K_S	30.02GB
autotrain-mixtral7x8b-math.Q5_K.gguf	Q5_K	30.95GB
autotrain-mixtral7x8b-math.Q5_K_M.gguf	Q5_K_M	30.95GB
autotrain-mixtral7x8b-math.Q5_1.gguf	Q5_1	32.71GB
autotrain-mixtral7x8b-math.Q6_K.gguf	Q6_K	35.74GB
autotrain-mixtral7x8b-math.Q8_0.gguf	Q8_0	46.22GB

Original model description:

tags: - autotrain - text-generation-inference - text-generation library_name: transformers widget: - messages: - role: user content: What is your favorite condiment? license: other

Model Trained Using AutoTrain

This model was trained using AutoTrain. For more information, please visit AutoTrain.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_path = "PATH_TO_THIS_REPO"

tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(
    model_path,
    device_map="auto",
    torch_dtype='auto'
).eval()

# Prompt content: "hi"
messages = [
    {"role": "user", "content": "hi"}
]

input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
output_ids = model.generate(input_ids.to('cuda'))
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)

# Model response: "Hello! How can I assist you today?"
print(response)