Downloads · 30 days
11
37% of all-time downloads
Llama-3.2-1B-Instruct_Mobilequant
soiji/Llama-3.2-1B-Instruct_Mobilequant
Llama-3.2-1B-Instruct_Mobilequant is a machine learning model from soiji. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
All-time downloads
30
Public
Repo size
4.9 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin4.9 GB · 100%
At a glance
- License
- mit
- Model type
- llama
- Access
- Public
- Created
- Jan 29, 2026
- Updated
- Jan 29, 2026
- SHA
- ab212639
Base models
Datasets
Model card
From the Hugging Face model README
Configs & command(run.sh)
MODEL=Llama-3.2-1B-Instruct_Converted
CKPT=/models/meta-llama/${MODEL}
BATCH_SIZE=1
LET_LR=1e-3
LET_MIN_LR=1e-4
LWC_LR=1e-2
LWC_MIN_LR=1e-3
LRL_LR=1e-6
LRL_MIN_LR=1e-7
EPOCHS=60
NSAMPLES=1024
OUTPUT_DIR=/models/meta-llama/${MODEL}-mobilequant-s${NSAMPLES}-e${EPOCHS}
python ptq/mobilequant.py --lwc --let --lrl \
--act_bitwidth 16 --epochs ${EPOCHS} --nsamples ${NSAMPLES} \
--deactive_amp --hf_path ${CKPT} --output_dir ${OUTPUT_DIR} \
--let_lr ${LET_LR} --let_min_lr ${LET_MIN_LR} \
--lwc_lr ${LWC_LR} --lwc_min_lr ${LWC_MIN_LR} \
--lrl_lr ${LRL_LR} --lrl_min_lr ${LRL_MIN_LR} \
--batch_size ${BATCH_SIZE} --mode e2e --weight_bitwidth 16
- Type
- llama
- License
- mit
- Created
- Jan 29, 2026
- Updated
- Jan 29, 2026