Instructions to use fblgit/UNA-SimpleSmaug-34b-v1beta with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use fblgit/UNA-SimpleSmaug-34b-v1beta with Transformers:

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

pipe = pipeline("text-generation", model="fblgit/UNA-SimpleSmaug-34b-v1beta")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("fblgit/UNA-SimpleSmaug-34b-v1beta")
model = AutoModelForCausalLM.from_pretrained("fblgit/UNA-SimpleSmaug-34b-v1beta", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use fblgit/UNA-SimpleSmaug-34b-v1beta with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "fblgit/UNA-SimpleSmaug-34b-v1beta"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fblgit/UNA-SimpleSmaug-34b-v1beta",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/fblgit/UNA-SimpleSmaug-34b-v1beta

SGLang

How to use fblgit/UNA-SimpleSmaug-34b-v1beta 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 "fblgit/UNA-SimpleSmaug-34b-v1beta" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fblgit/UNA-SimpleSmaug-34b-v1beta",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "fblgit/UNA-SimpleSmaug-34b-v1beta" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fblgit/UNA-SimpleSmaug-34b-v1beta",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use fblgit/UNA-SimpleSmaug-34b-v1beta with Docker Model Runner:
```
docker model run hf.co/fblgit/UNA-SimpleSmaug-34b-v1beta
```

fblgit commited on Feb 5, 2024

Commit

e1cdc5b

verified ·

1 Parent(s): c422045

Update README.md

Browse files

Files changed (1) hide show

README.md +26 -2

README.md CHANGED Viewed

@@ -2,9 +2,33 @@
 license: apache-2.0
 datasets:
 - fblgit/simple-math
 tags:
 - UNA
 ---
-So far an experiment, not sure how it went.
-Is based on Smaug and used SimpleMath dataset.

 license: apache-2.0
 datasets:
 - fblgit/simple-math
+base_model: abacusai/Smaug-34B-v0.1
 tags:
 - UNA
+- simple-math
+- juanako
 ---
+# UNA-SimpleSmaug-34b-v1beta
+So far an experiment, not sure how it went. Applied UNA only on the Attention, not on the MLP's
+* Is based on Smaug
+* SimpleMath dataset
+* It was trained on Axolotl
+## Experiment
+The thing here is to understand whats the impact of SimpleMath applied at the attention layer during a SFT session and how it impacts on the neural network overall.
+## Evals
+Pending, but so far this one
+```
+|    Task     |Version| Metric |Value |   |Stderr|
+|-------------|------:|--------|-----:|---|-----:|
+|arc_challenge|      0|acc     |0.7201|±  |0.0131|
+|             |       |acc_norm|0.7457|±  |0.0127|
+```
+Seems to increase GSM and ARC
+## Citations
+To abacusai for making Smaug-34B, the Bagel, and all the magic behind the base model.