You can deploy this model and start earning money today!

RichardErkhov/
vilm_-_VinaLlama2-14B-gguf

Name: vilm_-_VinaLlama2-14B-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

VinaLlama2-14B - GGUF

Model creator: https://huggingface.co/vilm/
Original model: https://huggingface.co/vilm/VinaLlama2-14B/

Name	Quant method	Size
VinaLlama2-14B.Q2_K.gguf	Q2_K	5.51GB
VinaLlama2-14B.IQ3_XS.gguf	IQ3_XS	6.03GB
VinaLlama2-14B.IQ3_S.gguf	IQ3_S	6.31GB
VinaLlama2-14B.Q3_K_S.gguf	Q3_K_S	6.31GB
VinaLlama2-14B.IQ3_M.gguf	IQ3_M	6.61GB
VinaLlama2-14B.Q3_K.gguf	Q3_K	6.91GB
VinaLlama2-14B.Q3_K_M.gguf	Q3_K_M	6.91GB
VinaLlama2-14B.Q3_K_L.gguf	Q3_K_L	7.3GB
VinaLlama2-14B.IQ4_XS.gguf	IQ4_XS	7.37GB
VinaLlama2-14B.Q4_0.gguf	Q4_0	7.62GB
VinaLlama2-14B.IQ4_NL.gguf	IQ4_NL	7.68GB
VinaLlama2-14B.Q4_K_S.gguf	Q4_K_S	7.98GB
VinaLlama2-14B.Q4_K.gguf	Q4_K	8.56GB
VinaLlama2-14B.Q4_K_M.gguf	Q4_K_M	8.56GB
VinaLlama2-14B.Q4_1.gguf	Q4_1	8.4GB
VinaLlama2-14B.Q5_0.gguf	Q5_0	9.18GB
VinaLlama2-14B.Q5_K_S.gguf	Q5_K_S	9.34GB
VinaLlama2-14B.Q5_K.gguf	Q5_K	9.81GB
VinaLlama2-14B.Q5_K_M.gguf	Q5_K_M	9.81GB
VinaLlama2-14B.Q5_1.gguf	Q5_1	9.96GB
VinaLlama2-14B.Q6_K.gguf	Q6_K	11.46GB
VinaLlama2-14B.Q8_0.gguf	Q8_0	14.03GB

Original model description:

license: mit language:

VinaLlama2-14B Beta

GGUF Here: VinaLlama2-14B-GGUF

Top Features:

Context Length: 32,768 tokens.
VERY GOOD at reasoning, mathematics and creative writing.
Works with Langchain Agent out-of-the-box.

Known Issues

Still a bit struggling with Vietnamese fact (Hoang Sa & Truong Sa, Historical questions).
Hallucination when reasoning.
Can't do Vi-En/En-Vi translation (yet)!

Quick use:

VRAM Requirement: ~20GB

pip install transformers accelerate

python

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
device = "cuda" # the device to load the model onto

model = AutoModelForCausalLM.from_pretrained(
    "vilm/VinaLlama2-14B",
    torch_dtype='auto',
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("vilm/VinaLlama2-14B")

prompt = "Một cộng một bằng mấy?"
messages = [
    {"role": "system", "content": "Bạn là trợ lí AI hữu ích."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)

generated_ids = model.generate(
    model_inputs.input_ids,
    max_new_tokens=1024,
    eos_token_id=tokenizer.eos_token_id,
    temperature=0.25,
)
generated_ids = [
    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]

response = tokenizer.batch_decode(generated_ids)[0]
print(response)

RichardErkhov/vilm_-_VinaLlama2-14B-gguf

Original model description:

VinaLlama2-14B Beta

RichardErkhov/
vilm_-_VinaLlama2-14B-gguf