Safetensors
GGUF
Turkish
llama
Llama-3
instruct
finetune
chatml
gpt4
synthetic data
distillation
function calling
json mode
axolotl
roleplaying
chat
Instructions to use tda45/TdAI with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
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
| static void print_usage(int /*argc*/, char ** argv) { | |
| printf("\nexample usage:\n"); | |
| printf("\n %s -m model.gguf [-ngl n_gpu_layers]\n", argv[0]); | |
| printf("\n"); | |
| } | |
| int main(int argc, char ** argv) { | |
| common_params params; | |
| common_init(); | |
| if (!common_params_parse(argc, argv, params, LLAMA_EXAMPLE_COMMON, print_usage)) { | |
| return 1; | |
| } | |
| // init LLM | |
| llama_backend_init(); | |
| llama_numa_init(params.numa); | |
| // initialize the model | |
| llama_model_params model_params = common_model_params_to_llama(params); | |
| llama_model * model = llama_model_load_from_file(params.model.path.c_str(), model_params); | |
| if (model == NULL) { | |
| LOG_ERR("%s: error: unable to load model\n" , __func__); | |
| return 1; | |
| } | |
| const llama_vocab * vocab = llama_model_get_vocab(model); | |
| // we need just a dummy token to evaluate | |
| std::vector<llama_token> prompt_tokens(1, llama_vocab_bos(vocab)); | |
| llama_context_params ctx_params = llama_context_default_params(); | |
| ctx_params.n_ctx = 512; | |
| ctx_params.n_batch = 512; | |
| ctx_params.no_perf = false; | |
| llama_context * ctx = llama_init_from_model(model, ctx_params); | |
| if (ctx == NULL) { | |
| fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__); | |
| return 1; | |
| } | |
| llama_batch batch = llama_batch_get_one(prompt_tokens.data(), prompt_tokens.size()); | |
| const int n_iters = 3; | |
| // warm-up | |
| llama_decode(ctx, batch); | |
| llama_memory_clear(llama_get_memory(ctx), true); | |
| llama_synchronize(ctx); | |
| for (int64_t t_pause_ms = 0; t_pause_ms <= 4000; t_pause_ms += 800) { | |
| double t_sum_us = 0.0; | |
| double t_sum2_us = 0.0; | |
| for (int i = 0; i < n_iters; i++) { | |
| // this pause is important - it simulates "idle GPU" | |
| std::this_thread::sleep_for(std::chrono::milliseconds(t_pause_ms)); | |
| const int64_t t_start_us = llama_time_us(); | |
| // this should take constant time | |
| llama_decode(ctx, batch); | |
| llama_synchronize(ctx); | |
| const int64_t t_end_us = llama_time_us(); | |
| const double t_cur_us = t_end_us - t_start_us; | |
| // print individual decode times | |
| printf(" - decode time: %8.2f ms\n", t_cur_us / 1000); | |
| t_sum_us += t_cur_us; | |
| t_sum2_us += t_cur_us * t_cur_us; | |
| llama_memory_clear(llama_get_memory(ctx), true); | |
| llama_synchronize(ctx); // just in case | |
| } | |
| const double t_avg_us = t_sum_us / n_iters; | |
| const double t_dev_us = sqrt((t_sum2_us / (n_iters - 1)) - (t_avg_us * t_avg_us * n_iters) / (n_iters - 1)); | |
| printf("iters: %4d, pause: %5d ms, avg decode time: %8.2f +/- %4.2f ms\n", n_iters, (int) t_pause_ms, t_avg_us / 1000, t_dev_us / 1000); | |
| fflush(stdout); | |
| } | |
| llama_free(ctx); | |
| llama_model_free(model); | |
| return 0; | |
| } | |