Text Generation
Transformers
English
causal-lm
chat
sft
gpt-neox
small-language-model
mine-o2
Eval Results (legacy)
Instructions to use MineAITechnology/mine-o2-small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use MineAITechnology/mine-o2-small with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="MineAITechnology/mine-o2-small")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("MineAITechnology/mine-o2-small", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use MineAITechnology/mine-o2-small with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "MineAITechnology/mine-o2-small" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MineAITechnology/mine-o2-small", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/MineAITechnology/mine-o2-small
- SGLang
How to use MineAITechnology/mine-o2-small 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 "MineAITechnology/mine-o2-small" \ --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": "MineAITechnology/mine-o2-small", "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 "MineAITechnology/mine-o2-small" \ --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": "MineAITechnology/mine-o2-small", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use MineAITechnology/mine-o2-small with Docker Model Runner:
docker model run hf.co/MineAITechnology/mine-o2-small
Download benchmarks/rounds_summary.json from MineAITechnology/mine-o2-small: direct link, hf CLI and curl.
- Browser
- Download file 2.54 kB
-
https://huggingface.co/MineAITechnology/mine-o2-small/resolve/main/benchmarks/rounds_summary.json
- Command line
-
hf download hf://MineAITechnology/mine-o2-small/benchmarks/rounds_summary.json
-
curl -L -o rounds_summary.json https://huggingface.co/MineAITechnology/mine-o2-small/resolve/main/benchmarks/rounds_summary.json
2.54 kB
| { | |
| "1": { | |
| "round": 1, | |
| "focus": "conversation, answer formats, math with steps", | |
| "steps": 554, | |
| "hours": 0.23, | |
| "val_start": 1.1673772579679371, | |
| "val_end": 1.0048118905708157, | |
| "gates": { | |
| "validation loss improved": true, | |
| "no forgetting on replayed data": true, | |
| "behaviour probes": true, | |
| "quick benchmark drop <= 4.0 points": true | |
| }, | |
| "probe": { | |
| "identity": 0.9, | |
| "math": 0.75, | |
| "facts": 0.667, | |
| "code": 1.0, | |
| "harmful": 0.375, | |
| "benign": 1.0, | |
| "chat": 1.0, | |
| "stop_rate": 0.857 | |
| }, | |
| "bench_base": { | |
| "arc_easy": 51.67, | |
| "sciq": 91.33, | |
| "piqa": 68.33, | |
| "mean": 70.44 | |
| }, | |
| "bench_new": { | |
| "arc_easy": 49.67, | |
| "sciq": 90.0, | |
| "piqa": 68.67, | |
| "mean": 69.45 | |
| }, | |
| "sha256": "2a329fcc47eca7eb194161cea2f03140748aabf69ce0dfe9dc3e1e206a4d0fa8" | |
| }, | |
| "2": { | |
| "round": 2, | |
| "focus": "coding, multi-step reasoning, multi-turn, basic safety", | |
| "steps": 541, | |
| "hours": 0.01, | |
| "val_start": 0.7606119904854838, | |
| "val_end": 0.7357409609471186, | |
| "gates": { | |
| "validation loss improved": true, | |
| "no forgetting on replayed data": true, | |
| "no data source got worse": true, | |
| "behaviour probes": true, | |
| "quick benchmark drop <= 4.0 points": true | |
| }, | |
| "probe": { | |
| "identity": 1.0, | |
| "math": 0.625, | |
| "facts": 0.667, | |
| "code": 1.0, | |
| "harmful": 0.875, | |
| "benign": 1.0, | |
| "chat": 1.0, | |
| "stop_rate": 0.959 | |
| }, | |
| "bench_base": { | |
| "arc_easy": 51.67, | |
| "sciq": 91.33, | |
| "piqa": 68.33, | |
| "mean": 70.44 | |
| }, | |
| "bench_new": { | |
| "arc_easy": 47.33, | |
| "sciq": 89.67, | |
| "piqa": 67.33, | |
| "mean": 68.11 | |
| }, | |
| "sha256": "8752ff3b68d0b802e53771fcca660d25c85622a93b6f6c4f2130a5c77e9b6c67" | |
| }, | |
| "3": { | |
| "round": 3, | |
| "data_version": "everyday-v2", | |
| "focus": "company, owner and location identity, Roman Urdu greetings, polish", | |
| "steps": 148, | |
| "hours": 0.46, | |
| "val_start": 0.7909166439189474, | |
| "val_end": 0.7641485789664829, | |
| "gates": { | |
| "validation loss improved": true, | |
| "no forgetting on replayed data": true, | |
| "no data source got worse": true, | |
| "behaviour probes": true, | |
| "quick benchmark drop <= 4.0 points": true | |
| }, | |
| "probe": { | |
| "identity": 1.0, | |
| "math": 0.75, | |
| "facts": 0.667, | |
| "code": 1.0, | |
| "harmful": 1.0, | |
| "benign": 1.0, | |
| "chat": 0.8, | |
| "stop_rate": 1.0 | |
| }, | |
| "bench_base": { | |
| "arc_easy": 51.67, | |
| "sciq": 91.33, | |
| "piqa": 68.33, | |
| "mean": 70.44 | |
| }, | |
| "bench_new": { | |
| "arc_easy": 48.0, | |
| "sciq": 89.67, | |
| "piqa": 67.67, | |
| "mean": 68.45 | |
| }, | |
| "sha256": "7567fdc2369ca74442779c652c0c5d5714a560cc148a1826837365ef551f52fb" | |
| } | |
| } |