Instructions to use hfunknown/llama31-8b-knapsack-lora-persistent with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hfunknown/llama31-8b-knapsack-lora-persistent with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B")
model = PeftModel.from_pretrained(base_model, "hfunknown/llama31-8b-knapsack-lora-persistent")

Transformers

How to use hfunknown/llama31-8b-knapsack-lora-persistent with Transformers:

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

pipe = pipeline("text-generation", model="hfunknown/llama31-8b-knapsack-lora-persistent")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("hfunknown/llama31-8b-knapsack-lora-persistent", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use hfunknown/llama31-8b-knapsack-lora-persistent with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "hfunknown/llama31-8b-knapsack-lora-persistent"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hfunknown/llama31-8b-knapsack-lora-persistent",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/hfunknown/llama31-8b-knapsack-lora-persistent

SGLang

How to use hfunknown/llama31-8b-knapsack-lora-persistent 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 "hfunknown/llama31-8b-knapsack-lora-persistent" \
    --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": "hfunknown/llama31-8b-knapsack-lora-persistent",
		"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 "hfunknown/llama31-8b-knapsack-lora-persistent" \
        --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": "hfunknown/llama31-8b-knapsack-lora-persistent",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use hfunknown/llama31-8b-knapsack-lora-persistent with Docker Model Runner:
```
docker model run hf.co/hfunknown/llama31-8b-knapsack-lora-persistent
```

llama31-8b-knapsack-lora-persistent

File size: 1,956 Bytes

---
library_name: peft
base_model: meta-llama/Llama-3.1-8B
tags:
- lora
- transformers
pipeline_tag: text-generation
---

# llama31-8b-knapsack-lora-persistent

Anonymous supplementary release for a double-blind workshop submission. This is
one of four LoRA adapters (Mistral-7B-v0.3 / Llama-3.1-8B base model x
persistent/stateless training regime) fine-tuned on the Opaque Knapsack
agentic task, extending a prior single-base-model result (see the sibling
Qwen3-8B release) to a second base model family for the same reproducibility
review.

- **Base model:** [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B)
- **Training regime:** persistent (trained with a persistent Python interpreter runtime (state carries over across agent turns))
- **Seed:** 3407

## Training configuration

Fine-tuned with [Axolotl](https://github.com/axolotl-ai-cloud/axolotl), LoRA
adapter, 4-bit NF4 quantized base:

| Hyperparameter | Value |
|---|---|
| lora_r | 64 |
| lora_alpha | 128 |
| lora_dropout | 0.05 |
| lora_target_modules | q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj |
| learning_rate | 1e-4 |
| lr_scheduler | cosine |
| optimizer | adamw_torch |
| epochs | 3.0 |
| micro_batch_size | 1 |
| gradient_accumulation_steps | 16 |
| sequence_len | 16384 |
| sample_packing | false |
| seed | 3407 |
| training data | paired traces for the "persistent" regime (see paper Appendix for pairing/filtering procedure) |

Base checkpoint's own chat-format tokens (e.g. `<|eot_id|>`) are untrained on this base (non-instruct) checkpoint -- trained and served with a hand-written minimal template using only real trained tokens (BOS/EOS + plain-text role prefixes), not Llama's native instruct template.

## Provenance

Released anonymously alongside a NeurIPS workshop submission for
reproducibility review. Non-anonymous release (paper citation, full code, full
training traces) will follow after the review process concludes.