Text Generation
Transformers
Safetensors
GGUF
English
grt
geometric-rotation-transformer
causal-lm
rotational
Instructions to use MatterWaveAI/grt-20m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use MatterWaveAI/grt-20m with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="MatterWaveAI/grt-20m")# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("MatterWaveAI/grt-20m", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use MatterWaveAI/grt-20m with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf MatterWaveAI/grt-20m:F32 # Run inference directly in the terminal: llama cli -hf MatterWaveAI/grt-20m:F32
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf MatterWaveAI/grt-20m:F32 # Run inference directly in the terminal: llama cli -hf MatterWaveAI/grt-20m:F32
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf MatterWaveAI/grt-20m:F32 # Run inference directly in the terminal: ./build/bin/llama cli -hf MatterWaveAI/grt-20m:F32
Use Docker
docker model run hf.co/MatterWaveAI/grt-20m:F32
- Llama new
- LM Studio
- Jan
- vLLM
How to use MatterWaveAI/grt-20m with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "MatterWaveAI/grt-20m" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MatterWaveAI/grt-20m", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/MatterWaveAI/grt-20m:F32
- SGLang
How to use MatterWaveAI/grt-20m 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 "MatterWaveAI/grt-20m" \ --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": "MatterWaveAI/grt-20m", "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 "MatterWaveAI/grt-20m" \ --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": "MatterWaveAI/grt-20m", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Ollama
How to use MatterWaveAI/grt-20m with Ollama:
ollama run hf.co/MatterWaveAI/grt-20m:F32
- Unsloth Desktop
- Docker Model Runner
How to use MatterWaveAI/grt-20m with Docker Model Runner:
docker model run hf.co/MatterWaveAI/grt-20m:F32
- Lemonade
How to use MatterWaveAI/grt-20m with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull MatterWaveAI/grt-20m:F32
Run and chat with the model
lemonade run user.grt-20m-F32
List all available models
lemonade list
- Atomic Chat
Update MatterWaveAI/grt-20m-curriculum final 24h metrics in README.md
Browse files
README.md
CHANGED
|
@@ -21,7 +21,7 @@ pipeline_tag: text-generation
|
|
| 21 |
|
| 22 |
# GRT-20M: From NeurIPS Submission (`PairAdam`) to Symplectic/Huygens Optimization
|
| 23 |
|
| 24 |
-
> **Note:** `GRT-20M` (`16.7M` total parameters, `10.3M` non-embedding backbone parameters across `12` rotational blocks) was pretrained across an 8-corpus English mix using first-generation `PairAdam` for our NeurIPS submission (archived pre-fine-tune at [`MatterWaveAI/grt-20m-base`](https://huggingface.co/MatterWaveAI/grt-20m-base)). Keeping the exact same `16.7M` NeurIPS submission checkpoint and swapping in our new classical geometric optimizers (`SymplecticAdam` and `HuygensSymplecticAdam`) lifts 9-task `SLM-Bench` multiple-choice accuracy from **`38.63%` (`PairAdam`) → `45.03%` (`HuygensSymplecticAdam`, `+6.40%`)** and cuts `GSM8K` held-out math perplexity from **`60.12` → `9.11` (`6.60×` reduction vs. Zero-Shot, `-27.2%` vs. `PairAdam`)**. For the multi-round curriculum checkpoint (**`
|
| 25 |
|
| 26 |
A Geometric Rotation Transformer pretrained on a mix of eight English corpora,
|
| 27 |
including FineWeb-Edu, Cosmopedia, and FineMath. GRT replaces the dense linear
|
|
@@ -59,11 +59,12 @@ plus a log-magnitude per token).
|
|
| 59 |
|
| 60 |
| | |
|
| 61 |
|---|---|
|
| 62 |
-
| **Data** | ClimbMix, Wikipedia, RefinedWeb, FineWeb-Edu, Cosmopedia, Nemotron, FineMath, StackExchange |
|
| 63 |
-
| **Steps** | 158,000 |
|
| 64 |
-
| **Optimizer** | PairAdam |
|
| 65 |
-
| **Peak LR** | 1e-3, cosine decay to 10% |
|
| 66 |
-
| **Training
|
|
|
|
| 67 |
|
| 68 |
## Sample output (greedy, temperature 0)
|
| 69 |
|
|
|
|
| 21 |
|
| 22 |
# GRT-20M: From NeurIPS Submission (`PairAdam`) to Symplectic/Huygens Optimization
|
| 23 |
|
| 24 |
+
> **Note:** `GRT-20M` (`16.7M` total parameters, `10.3M` non-embedding backbone parameters across `12` rotational blocks) was pretrained across an 8-corpus English mix using first-generation `PairAdam` for our NeurIPS submission (archived pre-fine-tune at [`MatterWaveAI/grt-20m-base`](https://huggingface.co/MatterWaveAI/grt-20m-base)). Keeping the exact same `16.7M` NeurIPS submission checkpoint and swapping in our new classical geometric optimizers (`SymplecticAdam` and `HuygensSymplecticAdam`) lifts 9-task `SLM-Bench` multiple-choice accuracy from **`38.63%` (`PairAdam`) → `45.03%` (`HuygensSymplecticAdam`, `+6.40%`)** and cuts `GSM8K` held-out math perplexity from **`60.12` → `9.11` (`6.60×` reduction vs. Zero-Shot, `-27.2%` vs. `PairAdam`)**. For the multi-round curriculum checkpoint (**`48.17%`** 9-task MC average, **`98.20%`** `GSM8K` verification, **`7.65`** PPL), see [`MatterWaveAI/grt-20m-curriculum`](https://huggingface.co/MatterWaveAI/grt-20m-curriculum).
|
| 25 |
|
| 26 |
A Geometric Rotation Transformer pretrained on a mix of eight English corpora,
|
| 27 |
including FineWeb-Edu, Cosmopedia, and FineMath. GRT replaces the dense linear
|
|
|
|
| 59 |
|
| 60 |
| | |
|
| 61 |
|---|---|
|
| 62 |
+
| **Base Pretraining Data** | ClimbMix, Wikipedia, RefinedWeb, FineWeb-Edu, Cosmopedia, Nemotron, FineMath, StackExchange |
|
| 63 |
+
| **Base Pretraining Steps** | 158,000 |
|
| 64 |
+
| **Base Pretraining Optimizer** | PairAdam (`MatterWaveAI/grt-20m-base`) |
|
| 65 |
+
| **Base Peak LR** | 1e-3, cosine decay to 10% |
|
| 66 |
+
| **Base Training Loss** | 2.114 |
|
| 67 |
+
| **Task Adaptation Optimizer** | `HuygensSymplecticAdam` (`SymplecticAdam` two-stage curriculum) |
|
| 68 |
|
| 69 |
## Sample output (greedy, temperature 0)
|
| 70 |
|