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.
- 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 Desktop
- 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
- Atomic Chat
File size: 2,575 Bytes
c9c3262 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 | #version 450
#extension GL_EXT_control_flow_attributes : enable
#include "types.glsl"
layout(constant_id = 0) const int BLOCK_SIZE = 1024;
layout(constant_id = 1) const int NCOLS_PADDED_LOG2 = 10;
#define ASC 0
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (binding = 0) readonly buffer A {A_TYPE data_a[];};
layout (binding = 2) writeonly buffer D {int data_d[];};
layout (push_constant) uniform parameter {
uint ncols;
uint ncols_padded;
uint ncols_padded_log2;
uint nrows;
uint order;
uint outer_start;
uint outer_end;
uint inner_start;
uint inner_end;
} p;
shared ivec2 dst_row[BLOCK_SIZE];
void argsort(bool needs_bounds_check, const uint row) {
// bitonic sort
const int col = int(gl_LocalInvocationID.x);
const uint row_offset = row * p.ncols;
// initialize indices
dst_row[col] = ivec2(col, floatBitsToInt(data_a[row_offset + col]));
barrier();
uint num_outer_loop_iters = NCOLS_PADDED_LOG2;
[[unroll]] for (uint k = 2, outer_idx = 0; outer_idx < num_outer_loop_iters; k *= 2, outer_idx++) {
uint num_inner_loop_iters = outer_idx + 1;
[[unroll]] for (uint j = k / 2, inner_idx = 0; inner_idx < num_inner_loop_iters; j /= 2, inner_idx++) {
const int ixj = int(col ^ j);
int idx_0 = (col & k) == 0 ? col : ixj;
int idx_1 = (col & k) == 0 ? ixj : col;
ivec2 sh_idx_0 = dst_row[idx_0];
ivec2 sh_idx_1 = dst_row[idx_1];
bool idx_0_oob = needs_bounds_check ? sh_idx_0.x >= p.ncols : false;
bool idx_1_oob = needs_bounds_check ? sh_idx_1.x >= p.ncols : false;
if ((idx_0_oob ||
(!idx_1_oob && intBitsToFloat(sh_idx_0.y) > intBitsToFloat(sh_idx_1.y))) && (ixj > col)) {
dst_row[idx_0] = sh_idx_1;
dst_row[idx_1] = sh_idx_0;
}
barrier();
}
}
if (col < p.ncols) {
if (p.order == ASC) {
data_d[row_offset + col] = dst_row[col].x;
} else {
data_d[row_offset + p.ncols - col - 1] = dst_row[col].x;
}
}
}
void main() {
if (p.ncols == BLOCK_SIZE) {
uint row = gl_WorkGroupID.y;
while (row < p.nrows) {
argsort(false, row);
row += gl_WorkGroupSize.y * gl_NumWorkGroups.y;
}
} else {
uint row = gl_WorkGroupID.y;
while (row < p.nrows) {
argsort(true, row);
row += gl_WorkGroupSize.y * gl_NumWorkGroups.y;
}
}
}
|