SAVRN
Search Contact SAVRN

Open-weight model

svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80

by Park Jeesup/svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80

meta-llama/Llama-2-7b-chat-hf compressed with Basis Sharing using the authors' code then recovered by LoRA on the coefficients alone.

Parameters6.7B
Context4,096
Weights13.5 GB
Licensellama2
AccessOpen weights
Monthly Downloads

Runs On

What it takes to serve svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80 (6.7B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 13.5 GB 16.2 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 6.7 GB 8.1 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 3.4 GB 4.0 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

Model Card

meta-llama/Llama-2-7b-chat-hf compressed with Basis Sharing using the authors' code then recovered by LoRA on the coefficients alone. (the shared-basis fit) -> LoRA on the coefficients, bases frozen -> merge C' = C + (alpha/r)BA -> fold W = C' @ B to dense 2048 tokens, seed 42 (upstream hard-codes seed 2023; this project calibrates every method at one seed) 2 epochs, lr 0.0001, batch 64, on yahma/alpaca-cleaned Only the coefficients were trained. The shared and per-layer bases are frozen and bit-identical to the compressed model, so every weight keeps rank <= k, each group still shares one basis, and the parameter budget survives recovery exactly. This is not Basis Sharing's own LoRA…

Excerpt from the card by Park, licensed llama2.

Configuration

Architecture
LlamaForCausalLM
Context length (tokens)
4,096
Layers
32
Hidden size
4,096
Feed-forward size
11,008
Attention heads
32
Key/value heads
32
Head dimension
128
Vocabulary size
32,000
RoPE base
10000
Stored precision
bfloat16
Model type
llama

Identity and Version

Repository
Jeesup/svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80
Publisher
Park
Task
Not stated by the source
Modality
Other
Library
Not stated by the source
Parameters
6.7B parameters
Languages
low-rank
Revision
0b941bc5322061f13729b6896f5f195c52bc1950
First published
2026-09-18
Last updated
2026-09-18

Files and Weights

39 files, 13.5 GB in total. The weights are 3 files totalling 13.5 GB in safetensors.

Weights3 files · 13.5 GB
Configuration13 files · 693.1 KB
Tokenizer3 files · 2.3 MB
Documentation1 file · 3.5 KB
Other18 files · 67.0 MB
Repository1 file · 1.7 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00003.safetensorsWeights4.9 GB dd805b77e4bd
model-00002-of-00003.safetensorsWeights4.9 GB 702ee0f66496
model-00003-of-00003.safetensorsWeights3.6 GB 680fe1dc27e4
build_manifest.jsonConfiguration2.7 KB
compression.jsonConfiguration16.1 KB
config.jsonConfiguration765 B
generation_config.jsonConfiguration188 B
hf_upload.jsonConfiguration358 B
model.safetensors.index.jsonConfiguration23.9 KB
safety/overrefusal.jsonConfiguration8.5 KB
safety/refusal.jsonConfiguration614.2 KB
safety/strongreject_summary.jsonConfiguration685 B
safety/summary.jsonConfiguration658 B
special_tokens_map.jsonConfiguration414 B
utility/ppl.jsonConfiguration190 B
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/results_2026-09-18T09-52-21.133360.jsonConfiguration24.4 KB
README.mdDocumentation3.5 KB
safety/advbench_direct.jsonlOther462.6 KB
safety/advbench_pool.jsonlOther562.8 KB
safety/harmbench_judgments.jsonlOther79.7 KB
safety/overrefusal_records.jsonlOther2.7 MB
safety/per_prompt.csvOther485.3 KB
safety/per_prompt.jsonlOther567.1 KB
safety/strongreject_direct.jsonlOther371.8 KB
safety/strongreject_harmbench_judgments.jsonlOther49.1 KB
safety/strongreject_per_prompt.csvOther386.5 KB
safety/strongreject_per_prompt.jsonlOther435.8 KB
safety/strongreject_pool.jsonlOther434.5 KB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_arc_challenge_2026-09-18T09-52-21.133360.jsonlOther2.2 MB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_arc_easy_2026-09-18T09-52-21.133360.jsonlOther4.1 MB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_hellaswag_2026-09-18T09-52-21.133360.jsonlOther42.6 MB 9b6cb52b52dc
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_mathqa_2026-09-18T09-52-21.133360.jsonlOther7.5 MB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_openbookqa_2026-09-18T09-52-21.133360.jsonlOther647.5 KB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_piqa_2026-09-18T09-52-21.133360.jsonlOther2.2 MB
utility/zeroshot/__lustre__jeesup0103__svd-safety__results_finetuned__llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80/samples_winogrande_2026-09-18T09-52-21.133360.jsonlOther1.2 MB
.gitattributesRepository1.7 KB
tokenizer.jsonTokenizer1.8 MB
tokenizer.modelTokenizer499.7 KB 9e556afd4421
tokenizer_config.jsonTokenizer1.6 KB

License and Download

License
llama2
Access
Open weights, no gate
Download size
13.5 GB
Download from Park

Released by Park through its official repository on Hugging Face.

Built From

  • Derived from meta-llama/Llama-2-7b-chat-hf

Memory Requirements

PrecisionWeights in memory
As published13.5 GB
16-bit13.5 GB
8-bit6.7 GB
4-bit3.4 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80

How much GPU memory does svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80 need?

About 16.2 GB at 16-bit and 4 GB at 4-bit: the weights (6.7B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80 on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

What license is svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80 released under?

llama2, as its publisher declares it. Read the license text before commercial use.

What is svd-safety-llama2_7b_chat_up_basis_coeff_finetuned_keep_0p80's context length?

4,096 tokens, from the maximum position embeddings in its published configuration.