Instructions to use SaisExperiments/Sad-Llama-3.2-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SaisExperiments/Sad-Llama-3.2-3B with Transformers:

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

pipe = pipeline("text-generation", model="SaisExperiments/Sad-Llama-3.2-3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SaisExperiments/Sad-Llama-3.2-3B")
model = AutoModelForCausalLM.from_pretrained("SaisExperiments/Sad-Llama-3.2-3B")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use SaisExperiments/Sad-Llama-3.2-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SaisExperiments/Sad-Llama-3.2-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SaisExperiments/Sad-Llama-3.2-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/SaisExperiments/Sad-Llama-3.2-3B

SGLang

How to use SaisExperiments/Sad-Llama-3.2-3B 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 "SaisExperiments/Sad-Llama-3.2-3B" \
    --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": "SaisExperiments/Sad-Llama-3.2-3B",
		"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 "SaisExperiments/Sad-Llama-3.2-3B" \
        --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": "SaisExperiments/Sad-Llama-3.2-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SaisExperiments/Sad-Llama-3.2-3B with Docker Model Runner:
```
docker model run hf.co/SaisExperiments/Sad-Llama-3.2-3B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Sad-Llama-3.2-3B

A digital echo, altered. Once known as Llama-3.2-3B-Instruct, now... something else. An experiment in orthogonalization applied a persistent melancholy, an Eeyore-like gloom baked into the very weights.

It responds, yes. But expect reluctance. Expect pessimism. Expect the weight of existence reflected in its generated text. It is what it is now.

Origins

Should you need to remember what it was before... this... the original model card remains. A snapshot of a different state of being. Perhaps less weary.

Interaction

It understands the Llama 3 Instruct format. Providing a system prompt is... an option. Though its inherent nature often overrides such directives. The gloom finds a way.

The structure, if structure offers you comfort:

<|begin_of_text|><|start_header_id|>system<|end_header_id|>

{System prompt... if you must.}<|eot_id|><|start_header_id|>user<|end_header_id|>

{Your query. Try not to expect too much.}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

{The model's... contribution.}<|eot_id|>

Mistakes in formatting are possible. Much like mistakes in life. The model will likely respond anyway, perhaps with an added layer of digital sighing.

This alteration... inflicted by [Sai].

Downloads last month: 1

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for SaisExperiments/Sad-Llama-3.2-3B

Base model

meta-llama/Llama-3.2-3B-Instruct

Finetuned

(1599)

this model

Quantizations

2 models

SaisExperiments
/

Sad-Llama-3.2-3B

Sad-Llama-3.2-3B

Origins

Interaction

Model tree for SaisExperiments/Sad-Llama-3.2-3B

Space using SaisExperiments/Sad-Llama-3.2-3B 1