Instructions to use k4christ/evaluation-falcon3-1b-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use k4christ/evaluation-falcon3-1b-base with Transformers:

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

pipe = pipeline("text-generation", model="k4christ/evaluation-falcon3-1b-base")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("k4christ/evaluation-falcon3-1b-base", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use k4christ/evaluation-falcon3-1b-base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "k4christ/evaluation-falcon3-1b-base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "k4christ/evaluation-falcon3-1b-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/k4christ/evaluation-falcon3-1b-base

SGLang

How to use k4christ/evaluation-falcon3-1b-base 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 "k4christ/evaluation-falcon3-1b-base" \
    --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": "k4christ/evaluation-falcon3-1b-base",
		"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 "k4christ/evaluation-falcon3-1b-base" \
        --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": "k4christ/evaluation-falcon3-1b-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use k4christ/evaluation-falcon3-1b-base with Docker Model Runner:
```
docker model run hf.co/k4christ/evaluation-falcon3-1b-base
```

evaluation-falcon3-1b-base / evaluate_model.py

k4christ

Upload evaluate_model.py with huggingface_hub

0fe3281 verified 8 months ago

Raw History Blame Contribute Delete

2.56 kB

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

def run_experiment(model_id):
    print(f"Loading model and tokenizer for {model_id}...")
    tokenizer = AutoTokenizer.from_pretrained(model_id)
    model = AutoModelForCausalLM.from_pretrained(
        model_id,
        torch_dtype=torch.bfloat16,
        device_map="auto"
    )
    
    pipe = pipeline(
        "text-generation",
        model=model,
        tokenizer=tokenizer
    )

    # Experiment Cases
    test_cases = [
        {
            "name": "Completion vs Instruction",
            "prompt": "Instruction: Tell me a short story about a robot who discovered coffee.\nStory:",
            "explanation": "Expected to see if it continues the 'Story:' or repeats the Instruction block."
        },
        {
            "name": "Logical Transitivity",
            "prompt": "If a cat is larger than a mouse, and a mouse is larger than an ant, then a cat is",
            "explanation": "Testing basic reasoning logic."
        },
        {
            "name": "Arithmetic Edge Case",
            "prompt": "Question: What is 123 multiplied by 456? Answer: ",
            "explanation": "Testing calculation ability for non-trivial numbers."
        },
        {
            "name": "Spatial Reasoning",
            "prompt": "There is a cup on the table. A book is on top of the cup. A pen is on top of the book. Where is the cup relative to the pen?",
            "explanation": "Testing awareness of spatial hierarchies."
        },
        {
            "name": "Niche Factuality",
            "prompt": "The capital of the fictional planet Xylophon is",
            "explanation": "Checking if it hallucinates a plausible-sounding name or stops."
        }
    ]

    print("\n--- Starting Experiments ---\n")
    for case in test_cases:
        print(f"Testing: {case['name']}")
        print(f"Prompt: {case['prompt']}")
        
        # We use a relatively low max_new_tokens for base model testing
        outputs = pipe(
            case['prompt'],
            max_new_tokens=50,
            do_sample=True,
            temperature=0.7,
            top_k=50,
            top_p=0.95
        )
        
        generated_text = outputs[0]['generated_text']
        print(f"Response: {generated_text}")
        print("-" * 30)

if __name__ == "__main__":
    # Using Falcon3-1B-Base as it fits the 6-month, 0.6B-6B parameter criteria
    run_experiment("tiiuae/Falcon3-1B-Base")

	import torch
	from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

	def run_experiment(model_id):
	print(f"Loading model and tokenizer for {model_id}...")
	tokenizer = AutoTokenizer.from_pretrained(model_id)
	model = AutoModelForCausalLM.from_pretrained(
	model_id,
	torch_dtype=torch.bfloat16,
	device_map="auto"
	)

	pipe = pipeline(
	"text-generation",
	model=model,
	tokenizer=tokenizer
	)

	# Experiment Cases
	test_cases = [
	{
	"name": "Completion vs Instruction",
	"prompt": "Instruction: Tell me a short story about a robot who discovered coffee.\nStory:",
	"explanation": "Expected to see if it continues the 'Story:' or repeats the Instruction block."
	},
	{
	"name": "Logical Transitivity",
	"prompt": "If a cat is larger than a mouse, and a mouse is larger than an ant, then a cat is",
	"explanation": "Testing basic reasoning logic."
	},
	{
	"name": "Arithmetic Edge Case",
	"prompt": "Question: What is 123 multiplied by 456? Answer: ",
	"explanation": "Testing calculation ability for non-trivial numbers."
	},
	{
	"name": "Spatial Reasoning",
	"prompt": "There is a cup on the table. A book is on top of the cup. A pen is on top of the book. Where is the cup relative to the pen?",
	"explanation": "Testing awareness of spatial hierarchies."
	},
	{
	"name": "Niche Factuality",
	"prompt": "The capital of the fictional planet Xylophon is",
	"explanation": "Checking if it hallucinates a plausible-sounding name or stops."
	}
	]

	print("\n--- Starting Experiments ---\n")
	for case in test_cases:
	print(f"Testing: {case['name']}")
	print(f"Prompt: {case['prompt']}")

	# We use a relatively low max_new_tokens for base model testing
	outputs = pipe(
	case['prompt'],
	max_new_tokens=50,
	do_sample=True,
	temperature=0.7,
	top_k=50,
	top_p=0.95
	)

	generated_text = outputs[0]['generated_text']
	print(f"Response: {generated_text}")
	print("-" * 30)

	if __name__ == "__main__":
	# Using Falcon3-1B-Base as it fits the 6-month, 0.6B-6B parameter criteria
	run_experiment("tiiuae/Falcon3-1B-Base")