Instructions to use tangledgroup/tangled-alpha-0.12-core with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tangledgroup/tangled-alpha-0.12-core with Transformers:

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

pipe = pipeline("text-generation", model="tangledgroup/tangled-alpha-0.12-core")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("tangledgroup/tangled-alpha-0.12-core", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tangledgroup/tangled-alpha-0.12-core with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tangledgroup/tangled-alpha-0.12-core"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tangledgroup/tangled-alpha-0.12-core",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/tangledgroup/tangled-alpha-0.12-core

SGLang

How to use tangledgroup/tangled-alpha-0.12-core 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 "tangledgroup/tangled-alpha-0.12-core" \
    --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": "tangledgroup/tangled-alpha-0.12-core",
		"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 "tangledgroup/tangled-alpha-0.12-core" \
        --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": "tangledgroup/tangled-alpha-0.12-core",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use tangledgroup/tangled-alpha-0.12-core with Docker Model Runner:
```
docker model run hf.co/tangledgroup/tangled-alpha-0.12-core
```

mtasic85 commited on Apr 6, 2025

Commit

b34d4e8

1 Parent(s): 7206717

pretrain

Browse files

Files changed (1) hide show

scripts/pretrain_base_model_0.yaml +4 -4

scripts/pretrain_base_model_0.yaml CHANGED Viewed

@@ -1,11 +1,11 @@
 # The name of the model to pretrain. Choose from names in ``litgpt.config``. Mutually exclusive with
 # ``model_config``. (type: Optional[str], default: null)
-model_name: 'tangled-alpha-0.12-core'
 # A ``litgpt.Config`` object to define the model architecture. Mutually exclusive with
 # ``model_config``. (type: Optional[Config], default: null)
 model_config:
-  name: 'tangled-alpha-0.12-core'
   block_size: 131072
   vocab_size: 65536
   padded_vocab_size: 65536
@@ -25,7 +25,7 @@ model_config:
 # Directory in which to save checkpoints and logs. If running in a Lightning Studio Job, look for it in
 # /teamspace/jobs/<job-name>/share. (type: <class 'Path'>, default: out/pretrain)
-out_dir: "../out/pretrain-core-0/"
 # The precision to use for pretraining. Possible choices: "bf16-true", "bf16-mixed", "32-true". (type: Optional[str], default: null)
 # precision: bf16-mixed
@@ -61,7 +61,7 @@ train:
   global_batch_size: 512
   # Number of samples per data-parallel rank (type: int, default: 4)
-  micro_batch_size: 1
   # Number of iterations with learning rate warmup active (type: int, default: 2000)
   lr_warmup_steps: 2000

 # The name of the model to pretrain. Choose from names in ``litgpt.config``. Mutually exclusive with
 # ``model_config``. (type: Optional[str], default: null)
+model_name: 'tangled-alpha-0.12-base'
 # A ``litgpt.Config`` object to define the model architecture. Mutually exclusive with
 # ``model_config``. (type: Optional[Config], default: null)
 model_config:
+  name: 'tangled-alpha-0.12-base'
   block_size: 131072
   vocab_size: 65536
   padded_vocab_size: 65536
 # Directory in which to save checkpoints and logs. If running in a Lightning Studio Job, look for it in
 # /teamspace/jobs/<job-name>/share. (type: <class 'Path'>, default: out/pretrain)
+out_dir: "../out/pretrain-base-0/"
 # The precision to use for pretraining. Possible choices: "bf16-true", "bf16-mixed", "32-true". (type: Optional[str], default: null)
 # precision: bf16-mixed
   global_batch_size: 512
   # Number of samples per data-parallel rank (type: int, default: 4)
+  micro_batch_size: 2
   # Number of iterations with learning rate warmup active (type: int, default: 2000)
   lr_warmup_steps: 2000