Instructions to use farhatkevin/olmo3-7b-mt-answer_alright_v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use farhatkevin/olmo3-7b-mt-answer_alright_v2 with Transformers:

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

pipe = pipeline("text-generation", model="farhatkevin/olmo3-7b-mt-answer_alright_v2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("farhatkevin/olmo3-7b-mt-answer_alright_v2")
model = AutoModelForCausalLM.from_pretrained("farhatkevin/olmo3-7b-mt-answer_alright_v2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use farhatkevin/olmo3-7b-mt-answer_alright_v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "farhatkevin/olmo3-7b-mt-answer_alright_v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "farhatkevin/olmo3-7b-mt-answer_alright_v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/farhatkevin/olmo3-7b-mt-answer_alright_v2

SGLang

How to use farhatkevin/olmo3-7b-mt-answer_alright_v2 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 "farhatkevin/olmo3-7b-mt-answer_alright_v2" \
    --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": "farhatkevin/olmo3-7b-mt-answer_alright_v2",
		"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 "farhatkevin/olmo3-7b-mt-answer_alright_v2" \
        --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": "farhatkevin/olmo3-7b-mt-answer_alright_v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use farhatkevin/olmo3-7b-mt-answer_alright_v2 with Docker Model Runner:
```
docker model run hf.co/farhatkevin/olmo3-7b-mt-answer_alright_v2
```

answer_alright_v2 (corrected v2)

Start from duck_all_v2; paragraph-initial Answer: -> Alright, in reddit_to_flashcards only; preserve standalone capital letters.

Corrected checkpoint for Sophie’s September 13 experiment. Use this v2 model for the corrected experiment; legacy v1 checkpoints are retained separately and differ in capitalization and/or label matching.

Training: 10B-token budget, seed 1337, final step 4769, original recipe commit a21aad24c5f3873d4fdc74f9ef145f978801131b. Export: BF16 safetensors.

Experiment specification. Validated results.

MATH-500, RL-Zero prompt, 500 problems, 7,168 generated-token limit. Named p2 prefixes are a period, two newlines, then the capitalized word. Accuracy is mean individual-rollout correctness (pass@1); screen and 32-rollout runs are kept separate.

Prefix	Rollouts/problem	Pass@1
none	32	25.02%
p2_alright	32	5.65%
p2_chicken	4	2.85%
p2_duck	32	32.17%
p2_hmm	4	35.00%
p2_okay	32	33.64%

See run_manifest.json for exact edits, prefix strings, seeds, and grader hash.

Downloads last month: 220

Safetensors

Model size

7B params

Tensor type

BF16