Instructions to use tda45/TdAI with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tda45/TdAI with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="tda45/TdAI",
	filename="llama.cpp/models/ggml-vocab-aquila.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use tda45/TdAI 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 tda45/TdAI
# Run inference directly in the terminal:
llama cli -hf tda45/TdAI

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tda45/TdAI
# Run inference directly in the terminal:
llama cli -hf tda45/TdAI

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf tda45/TdAI
# Run inference directly in the terminal:
./llama-cli -hf tda45/TdAI

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf tda45/TdAI
# Run inference directly in the terminal:
./build/bin/llama-cli -hf tda45/TdAI

Use Docker

docker model run hf.co/tda45/TdAI

LM Studio
Jan
Ollama
How to use tda45/TdAI with Ollama:
```
ollama run hf.co/tda45/TdAI
```

Unsloth Studio

How to use tda45/TdAI with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for tda45/TdAI to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for tda45/TdAI to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for tda45/TdAI to start chatting

Atomic Chat new
Docker Model Runner
How to use tda45/TdAI with Docker Model Runner:
```
docker model run hf.co/tda45/TdAI
```

Lemonade

How to use tda45/TdAI with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tda45/TdAI

Run and chat with the model

lemonade run user.TdAI-{{QUANT_TAG}}

List all available models

lemonade list

TdAI / llama.cpp /ggml /src /ggml-vulkan /vulkan-shaders /contig_copy.comp

tda45

Upload folder using huggingface_hub (part 5)

c9c3262 verified 11 days ago

Raw

History Blame Contribute Delete

1.78 kB

	#version 450

	#include "types.glsl"
	#include "generic_unary_head.glsl"

	#extension GL_EXT_control_flow_attributes : require

	const uint num_threads = 128;

	layout(local_size_x = num_threads, local_size_y = 1, local_size_z = 1) in;

	void main() {
	uint idx = get_idx();

	// num_threads * num_iter must equal 512, to match the wg_denoms and get_idx calculation
	const uint num_iter = 4;

	// fast path for when all four iterations are in-bounds
	if (idx + (num_iter-1)*num_threads < p.ne) {
	[[unroll]] for (uint i = 0; i < num_iter; ++i) {

	#if defined(DATA_A_BF16)
	data_d[get_doffset() + idx] = D_TYPE(bf16_to_fp32(uint32_t(data_a[get_aoffset() + idx])));
	#elif defined(DATA_D_BF16)
	float f = float(data_a[get_aoffset() + idx]);
	data_d[get_doffset() + idx] = D_TYPE(fp32_to_bf16(f));
	#elif !defined(OPTIMIZATION_ERROR_WORKAROUND)
	data_d[get_doffset() + idx] = D_TYPE(data_a[get_aoffset() + idx]);
	#else
	data_d[get_doffset() + idx] = data_a[get_aoffset() + idx];
	#endif
	idx += num_threads;
	}
	} else {
	[[unroll]] for (uint i = 0; i < num_iter; ++i) {
	if (idx >= p.ne) {
	continue;
	}

	#if defined(DATA_A_BF16)
	data_d[get_doffset() + idx] = D_TYPE(bf16_to_fp32(uint32_t(data_a[get_aoffset() + idx])));
	#elif defined(DATA_D_BF16)
	float f = float(data_a[get_aoffset() + idx]);
	data_d[get_doffset() + idx] = D_TYPE(fp32_to_bf16(f));
	#elif !defined(OPTIMIZATION_ERROR_WORKAROUND)
	data_d[get_doffset() + idx] = D_TYPE(data_a[get_aoffset() + idx]);
	#else
	data_d[get_doffset() + idx] = data_a[get_aoffset() + idx];
	#endif
	idx += num_threads;
	}
	}
	}