Instructions to use wmatejuk/midi-gpt with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use wmatejuk/midi-gpt with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="wmatejuk/midi-gpt")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("wmatejuk/midi-gpt", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use wmatejuk/midi-gpt with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "wmatejuk/midi-gpt" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "wmatejuk/midi-gpt", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/wmatejuk/midi-gpt
- SGLang
How to use wmatejuk/midi-gpt 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 "wmatejuk/midi-gpt" \ --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": "wmatejuk/midi-gpt", "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 "wmatejuk/midi-gpt" \ --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": "wmatejuk/midi-gpt", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use wmatejuk/midi-gpt with Docker Model Runner:
docker model run hf.co/wmatejuk/midi-gpt
Add C1e14-d48-0.14M
Browse files
C1e14-d48-0.14M/README.md
CHANGED
|
@@ -17,8 +17,8 @@ Checkpoint from the MIDI IsoFLOP scaling-law sweep.
|
|
| 17 |
| layers / heads / width | 4 / 3 / 48 |
|
| 18 |
| compute budget (FLOPs) | 1.000e+14 |
|
| 19 |
| tokens seen | 120979456 |
|
| 20 |
-
| final val loss | 2.
|
| 21 |
-
| best val loss | 2.
|
| 22 |
|
| 23 |
```python
|
| 24 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
| 17 |
| layers / heads / width | 4 / 3 / 48 |
|
| 18 |
| compute budget (FLOPs) | 1.000e+14 |
|
| 19 |
| tokens seen | 120979456 |
|
| 20 |
+
| final val loss | 2.4418246007386024 |
|
| 21 |
+
| best val loss | 2.4418246007386024 |
|
| 22 |
|
| 23 |
```python
|
| 24 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
C1e14-d48-0.14M/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 654240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc00d2950476985e7fc16071a1dbfd720ad728f81f964ed86f689dab8bbf25d0
|
| 3 |
size 654240
|
C1e14-d48-0.14M/summary.json
CHANGED
|
@@ -14,8 +14,9 @@
|
|
| 14 |
"tokens_seen": 120979456,
|
| 15 |
"flops": 99996779151360.0,
|
| 16 |
"budget_flops": 100000000000000.0,
|
| 17 |
-
"final_val_loss": 2.
|
| 18 |
-
"best_val_loss": 2.
|
| 19 |
"data_source": "packed",
|
| 20 |
-
"tokens_repo": "wmatejuk/midi-tokens-dyadic-tu0.01-vb32-mts1.0-vocab512"
|
|
|
|
| 21 |
}
|
|
|
|
| 14 |
"tokens_seen": 120979456,
|
| 15 |
"flops": 99996779151360.0,
|
| 16 |
"budget_flops": 100000000000000.0,
|
| 17 |
+
"final_val_loss": 2.4418246007386024,
|
| 18 |
+
"best_val_loss": 2.4418246007386024,
|
| 19 |
"data_source": "packed",
|
| 20 |
+
"tokens_repo": "wmatejuk/midi-tokens-dyadic-tu0.01-vb32-mts1.0-vocab512",
|
| 21 |
+
"wandb_url": "https://wandb.ai/wmatejuk/midi-gpt-scaling/runs/tvpmsg24"
|
| 22 |
}
|