You can deploy this model and start earning money today!

RichardErkhov/
RyanYr_-_llama32-3b_CoT-it_SFT-gguf

Name: RyanYr_-_llama32-3b_CoT-it_SFT-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

llama32-3b_CoT-it_SFT - GGUF

Model creator: https://huggingface.co/RyanYr/
Original model: https://huggingface.co/RyanYr/llama32-3b_CoT-it_SFT/

Name	Quant method	Size
llama32-3b_CoT-it_SFT.Q2_K.gguf	Q2_K	1.39GB
llama32-3b_CoT-it_SFT.IQ3_XS.gguf	IQ3_XS	1.53GB
llama32-3b_CoT-it_SFT.IQ3_S.gguf	IQ3_S	1.59GB
llama32-3b_CoT-it_SFT.Q3_K_S.gguf	Q3_K_S	1.59GB
llama32-3b_CoT-it_SFT.IQ3_M.gguf	IQ3_M	1.65GB
llama32-3b_CoT-it_SFT.Q3_K.gguf	Q3_K	1.73GB
llama32-3b_CoT-it_SFT.Q3_K_M.gguf	Q3_K_M	1.73GB
llama32-3b_CoT-it_SFT.Q3_K_L.gguf	Q3_K_L	1.85GB
llama32-3b_CoT-it_SFT.IQ4_XS.gguf	IQ4_XS	1.91GB
llama32-3b_CoT-it_SFT.Q4_0.gguf	Q4_0	1.99GB
llama32-3b_CoT-it_SFT.IQ4_NL.gguf	IQ4_NL	2.0GB
llama32-3b_CoT-it_SFT.Q4_K_S.gguf	Q4_K_S	2.0GB
llama32-3b_CoT-it_SFT.Q4_K.gguf	Q4_K	2.09GB
llama32-3b_CoT-it_SFT.Q4_K_M.gguf	Q4_K_M	2.09GB
llama32-3b_CoT-it_SFT.Q4_1.gguf	Q4_1	2.18GB
llama32-3b_CoT-it_SFT.Q5_0.gguf	Q5_0	2.37GB
llama32-3b_CoT-it_SFT.Q5_K_S.gguf	Q5_K_S	2.37GB
llama32-3b_CoT-it_SFT.Q5_K.gguf	Q5_K	2.41GB
llama32-3b_CoT-it_SFT.Q5_K_M.gguf	Q5_K_M	2.41GB
llama32-3b_CoT-it_SFT.Q5_1.gguf	Q5_1	2.55GB
llama32-3b_CoT-it_SFT.Q6_K.gguf	Q6_K	2.76GB
llama32-3b_CoT-it_SFT.Q8_0.gguf	Q8_0	3.58GB

Original model description:

base_model: meta-llama/Llama-3.2-3B library_name: transformers model_name: llama32-3b_CoT-it_SFT tags:

generated_from_trainer
trl
sft licence: license

Model Card for llama32-3b_CoT-it_SFT

This model is a fine-tuned version of meta-llama/Llama-3.2-3B. It has been trained using TRL.

Quick start

python

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="RyanYr/llama32-3b_CoT-it_SFT", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Training procedure

Visualize in Weights & Biases

This model was trained with SFT.

Framework versions

TRL: 0.12.0.dev0
Transformers: 4.45.1
Pytorch: 2.4.1
Datasets: 3.0.1
Tokenizers: 0.20.0

Citations

Cite TRL as:

bibtex

@misc{vonwerra2022trl,
	title        = {{TRL: Transformer Reinforcement Learning}},
	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
	year         = 2020,
	journal      = {GitHub repository},
	publisher    = {GitHub},
	howpublished = {\url{https://github.com/huggingface/trl}}
}

Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more quants, at much higher speed, than I would otherwise be able to.

RichardErkhov/RyanYr_-_llama32-3b_CoT-it_SFT-gguf