Instructions to use legesher/language-decoded-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use legesher/language-decoded-lora with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="legesher/language-decoded-lora")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("legesher/language-decoded-lora", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use legesher/language-decoded-lora with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "legesher/language-decoded-lora"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "legesher/language-decoded-lora",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/legesher/language-decoded-lora

SGLang

How to use legesher/language-decoded-lora with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "legesher/language-decoded-lora" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "legesher/language-decoded-lora",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "legesher/language-decoded-lora" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "legesher/language-decoded-lora",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Desktop
Docker Model Runner
How to use legesher/language-decoded-lora with Docker Model Runner:
```
docker model run hf.co/legesher/language-decoded-lora
```

language-decoded-lora

Commit History

Oct 13

14:44
condition-2-ur-20k-seed123

rafaym

Verified signature
469b0de
12:13
docs: drop the submitted-paper-title banner (#19)⁠

madiedgar

ed93e11
09:21
docs: register the October 2026 seed-123 adapters in the card and manifest (#18)⁠

madiedgar

d0d57de
09:09
rename: cond-3 seed-123 adapter folder to condition-3-zh-5k-native-code-seed123 (#17)⁠

madiedgar

4e1510f
08:49
phase-2-the-stack-v1-condition-3-zh-5k-seed123 (#16)⁠

madiedgar +1: rafaym

1cfdcc8
06:42
open condition-5-zh-5k-c4ai-aya-expanse-32b-seed123 (#15)⁠

madiedgar +1: rafaym

93769f8
02:19
Upload folder using huggingface_hub

madiedgar

Verified signature
faa1008
01:45
Upload folder using huggingface_hub

madiedgar

Verified signature
437dee4

Oct 12

22:07
add condition-5-ur-5k-c4ai-aya-expanse-32b-seed123 (#14)⁠

madiedgar +1: rafaym

5d79203

Jul 12

08:01
Add Attribution & takedown section (#13)⁠

madiedgar

03f675f
04:52
Relicense to cc-by-nc-4.0 (inherits CohereLabs/tiny-aya-base CC-BY-NC-4.0; adapters are adapted material of the NC-licensed base)

madiedgar

Verified signature
9d15d7b

May 31

11:43
docs: delineate Phase 2 (Stack v1) vs Phase 3 (Stack v2-dedup) adapters (#12)⁠

madiedgar

adeb1c7
11:05
docs: add arxiv: tags for paper bibliography anchors (#11)⁠

madiedgar

aa01d9a
10:16
docs(readme): align with canonical source-of-truth (cond-5 adapters, source-file control, refined extractor) (#10)⁠

madiedgar

06802f3
09:58
docs(readme): cond-5 refined-extractor banner + Phase 3 staleness fixes (#9)⁠

madiedgar

625893e

May 22

22:22
Upload folder using huggingface_hub

madiedgar

Verified signature
c0ea238

May 21

03:01
Rename folder with seed name

Khojasteh3

Verified signature
7137d71
03:00
Delete tiny-aya-base/condition-5-zh-5k-c4ai-aya-expanse-32b

Khojasteh3

Verified signature
09a0c55
02:55
Upload 3 files

Khojasteh3

Verified signature
32d3ac6

May 19

17:25
Upload folder using huggingface_hub

Rashik24

Verified signature
67e7d55
07:42
Upload folder using huggingface_hub

Rashik24

Verified signature
805922a
03:41
Upload folder using huggingface_hub

Rashik24

Verified signature
c4cfd58
01:03
rename: correct cond-3 data-source label — native-code, not the-stack-v1 (#8)⁠

madiedgar

78f7bc3
00:49
rename: drop "phase-2-" prefix on cond-3 adapter, match cond-5 naming⁠

Madison (Pfaff) Edgar

Unverified signature
672dea2

May 18

20:54
Upload folder using huggingface_hub

Rashik24

Verified signature
7d1ac71
14:55
Upload folder using huggingface_hub

Rashik24

Verified signature
1c9b9c0
13:54
Upload folder using huggingface_hub

Rashik24

Verified signature
4ed4206
11:49
Upload folder using huggingface_hub

Rashik24

Verified signature
8c2fe48
10:55
Upload folder using huggingface_hub

Rashik24

Verified signature
c44c44d
10:01
Upload folder using huggingface_hub

Rashik24

Verified signature
971a445
05:02
Upload folder using huggingface_hub

Rashik24

Verified signature
40a615d
04:07
Upload folder using huggingface_hub

Rashik24

Verified signature
b87d13e
03:12
Upload folder using huggingface_hub

Rashik24

Verified signature
76e05a3

May 14

06:17
Upload folder using huggingface_hub

Rashik24

Verified signature
d36a071
05:16
Upload folder using huggingface_hub

Rashik24

Verified signature
e537bdb
04:16
Upload folder using huggingface_hub

Rashik24

Verified signature
0b1abf7
00:15
Upload folder using huggingface_hub

Rashik24

Verified signature
b2e949f

May 13

23:14
Upload folder using huggingface_hub

Rashik24

Verified signature
9138649
22:12
Upload folder using huggingface_hub

Rashik24

Verified signature
bbe1912
04:17
Upload 20 files

Khojasteh3

Verified signature
b7e2234
04:15
Upload 21 files

Khojasteh3

Verified signature
a50c442
04:14
Upload 20 files

Khojasteh3

Verified signature
4c68f16
03:34
Upload folder using huggingface_hub

Rashik24

Verified signature
d76d67e
03:27
Upload folder using huggingface_hub

Rashik24

Verified signature
1cbb139
03:20
Upload folder using huggingface_hub

Rashik24

Verified signature
c04dd73
00:38
Upload folder using huggingface_hub

Rashik24

Verified signature
bc062a6
00:30
Upload folder using huggingface_hub

Rashik24

Verified signature
a48844b
00:23
Upload folder using huggingface_hub

Rashik24

Verified signature
a73a083

Mar 31

03:33
docs: add condition-1-en-32k to Model Structure table (#7)⁠

madiedgar

6ac3acd
03:33
chore: move root adapter to condition-1-en-32k/ subdirectory (#6)⁠

madiedgar

e4ca1d0

Commit History

Oct 13

condition-2-ur-20k-seed123

docs: drop the submitted-paper-title banner (#19)⁠

docs: register the October 2026 seed-123 adapters in the card and manifest (#18)⁠

rename: cond-3 seed-123 adapter folder to condition-3-zh-5k-native-code-seed123 (#17)⁠

phase-2-the-stack-v1-condition-3-zh-5k-seed123 (#16)⁠

open condition-5-zh-5k-c4ai-aya-expanse-32b-seed123 (#15)⁠

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Oct 12

add condition-5-ur-5k-c4ai-aya-expanse-32b-seed123 (#14)⁠

Jul 12

Add Attribution & takedown section (#13)⁠

Relicense to cc-by-nc-4.0 (inherits CohereLabs/tiny-aya-base CC-BY-NC-4.0; adapters are adapted material of the NC-licensed base)

May 31

docs: delineate Phase 2 (Stack v1) vs Phase 3 (Stack v2-dedup) adapters (#12)⁠

docs: add arxiv: tags for paper bibliography anchors (#11)⁠

docs(readme): align with canonical source-of-truth (cond-5 adapters, source-file control, refined extractor) (#10)⁠

docs(readme): cond-5 refined-extractor banner + Phase 3 staleness fixes (#9)⁠

May 22

Upload folder using huggingface_hub

May 21

Rename folder with seed name

Delete tiny-aya-base/condition-5-zh-5k-c4ai-aya-expanse-32b

Upload 3 files

May 19

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

rename: correct cond-3 data-source label — native-code, not the-stack-v1 (#8)⁠

rename: drop "phase-2-" prefix on cond-3 adapter, match cond-5 naming⁠

May 18

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

May 14

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

May 13

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload 20 files

Upload 21 files

Upload 20 files

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Mar 31

docs: add condition-1-en-32k to Model Structure table (#7)⁠

chore: move root adapter to condition-1-en-32k/ subdirectory (#6)⁠