Instructions to use davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025 with Transformers:

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

pipe = pipeline("text-generation", model="davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025")
model = AutoModelForCausalLM.from_pretrained("davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025

SGLang

How to use davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025 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 "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025 with Docker Model Runner:
```
docker model run hf.co/davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025
```

RLVE OPD teacher: 07-SharedSubstringCounting

Final weights from the original RLVE data-scaling experiment. Method: GRPO specialist. All actors were initialized freshly from DeepSeek-R1-Distill-Qwen-1.5B.

Training environments: 1, fixed balanced difficulty 0/1. Generation uses 1024 initial tokens plus the configured 128-token answer-forcing continuation.

Training cap: 4,000 rollouts; final checkpoint update: 499 (zero-based), corresponding to 4,000 training rollouts. Teachers stop at 4,000 rollouts or ID validation pass@1 >= 0.98. Validation uses 32 prompts and 8 responses per prompt every 50 updates. Validation samples do not count toward training budgets.

Training run: W&B. Group: data-scaling-v1-20261022-180018.

This repository contains model weights and tokenizer/config files; optimizer and dataloader state are absent. The environment list and original run settings are recorded in training_run.json, with the exact configuration in training_config.yaml.

Downloads last month: 85

Safetensors

Model size

2B params

Tensor type

BF16

Model tree for davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025

Base model

deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B

Finetuned

(844)

this model

Collection including davidheineman/rlve-opd-teacher-R1Distill-007-SharedSubstringCounting-20261025

RLVE OPD Teachers (Oct 25 2026)

Collection

128 RLVE GRPO teachers from the original data-scaling experiments, initialized from DeepSeek-R1-Distill-Qwen-1.5B. • 128 items • Updated 1 day ago