Instructions to use brunuu/RAM_Altman_GPT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use brunuu/RAM_Altman_GPT with Transformers:

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

pipe = pipeline("text-generation", model="brunuu/RAM_Altman_GPT")
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("brunuu/RAM_Altman_GPT")
model = AutoModelForCausalLM.from_pretrained("brunuu/RAM_Altman_GPT", 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]:]))

Transformers.js

How to use brunuu/RAM_Altman_GPT with Transformers.js:

// npm i @huggingface/transformers
import { pipeline } from '@huggingface/transformers';

// Allocate pipeline
const pipe = await pipeline('text-generation', 'brunuu/RAM_Altman_GPT');

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use brunuu/RAM_Altman_GPT with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf brunuu/RAM_Altman_GPT:F16
# Run inference directly in the terminal:
llama cli -hf brunuu/RAM_Altman_GPT:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf brunuu/RAM_Altman_GPT:F16
# Run inference directly in the terminal:
llama cli -hf brunuu/RAM_Altman_GPT:F16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf brunuu/RAM_Altman_GPT:F16
# Run inference directly in the terminal:
./build/bin/llama cli -hf brunuu/RAM_Altman_GPT:F16

Use Docker

docker model run hf.co/brunuu/RAM_Altman_GPT:F16

Llama new
LM Studio
Jan

vLLM

How to use brunuu/RAM_Altman_GPT with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/brunuu/RAM_Altman_GPT:F16

SGLang

How to use brunuu/RAM_Altman_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 "brunuu/RAM_Altman_GPT" \
    --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": "brunuu/RAM_Altman_GPT",
		"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 "brunuu/RAM_Altman_GPT" \
        --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": "brunuu/RAM_Altman_GPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use brunuu/RAM_Altman_GPT with Ollama:
```
ollama run hf.co/brunuu/RAM_Altman_GPT:F16
```
Unsloth Desktop
Docker Model Runner
How to use brunuu/RAM_Altman_GPT with Docker Model Runner:
```
docker model run hf.co/brunuu/RAM_Altman_GPT:F16
```

Lemonade

How to use brunuu/RAM_Altman_GPT with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull brunuu/RAM_Altman_GPT:F16

Run and chat with the model

lemonade run user.RAM_Altman_GPT-F16

List all available models

lemonade list

Atomic Chat

RAM_Altman_GPT / src /model_utils.py

brunuu

1.0

7de15fc verified about 2 hours ago

Raw History Blame Contribute Delete

1.13 kB

from transformers import GPT2Config, GPT2LMHeadModel, GPT2TokenizerFast
from config import EMBEDDING_SIZE, NUM_HEADS, NUM_LAYERS, TARGET_PARAMS

def context_length_for(vocab_size: int) -> int:
    tokens_per_layer = 12 * EMBEDDING_SIZE + 13
    return TARGET_PARAMS // EMBEDDING_SIZE - vocab_size - 2 - NUM_LAYERS * tokens_per_layer

def build_model(hf_tokenizer: GPT2TokenizerFast, vocab_size: int, device: str) -> tuple[GPT2LMHeadModel, int]:
    context_length = context_length_for(vocab_size)

    config = GPT2Config(
        vocab_size = vocab_size,
        n_positions = context_length,
        n_ctx = context_length,
        n_embd = EMBEDDING_SIZE,
        n_layer = NUM_LAYERS,
        n_head = NUM_HEADS,
        n_inner = 4 * EMBEDDING_SIZE, # llama.cpp expects the FFN to be 4x the embedding size
        bos_token_id = hf_tokenizer.eos_token_id,
        eos_token_id = hf_tokenizer.eos_token_id,
    )

    model = GPT2LMHeadModel(config)
    total = model.num_parameters()
    assert total == TARGET_PARAMS, f"expected {TARGET_PARAMS:,} parameters, got {total:,}"
    model.to(device)
    return model, context_length

	from transformers import GPT2Config, GPT2LMHeadModel, GPT2TokenizerFast
	from config import EMBEDDING_SIZE, NUM_HEADS, NUM_LAYERS, TARGET_PARAMS

	def context_length_for(vocab_size: int) -> int:
	tokens_per_layer = 12 * EMBEDDING_SIZE + 13
	return TARGET_PARAMS // EMBEDDING_SIZE - vocab_size - 2 - NUM_LAYERS * tokens_per_layer

	def build_model(hf_tokenizer: GPT2TokenizerFast, vocab_size: int, device: str) -> tuple[GPT2LMHeadModel, int]:
	context_length = context_length_for(vocab_size)

	config = GPT2Config(
	vocab_size = vocab_size,
	n_positions = context_length,
	n_ctx = context_length,
	n_embd = EMBEDDING_SIZE,
	n_layer = NUM_LAYERS,
	n_head = NUM_HEADS,
	n_inner = 4 * EMBEDDING_SIZE, # llama.cpp expects the FFN to be 4x the embedding size
	bos_token_id = hf_tokenizer.eos_token_id,
	eos_token_id = hf_tokenizer.eos_token_id,
	)

	model = GPT2LMHeadModel(config)
	total = model.num_parameters()
	assert total == TARGET_PARAMS, f"expected {TARGET_PARAMS:,} parameters, got {total:,}"
	model.to(device)
	return model, context_length