Instructions to use dkudos/cinimod-devops with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use dkudos/cinimod-devops 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 dkudos/cinimod-devops:Q8_0
# Run inference directly in the terminal:
llama cli -hf dkudos/cinimod-devops:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dkudos/cinimod-devops:Q8_0
# Run inference directly in the terminal:
llama cli -hf dkudos/cinimod-devops:Q8_0

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 dkudos/cinimod-devops:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf dkudos/cinimod-devops:Q8_0

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 dkudos/cinimod-devops:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf dkudos/cinimod-devops:Q8_0

Use Docker

docker model run hf.co/dkudos/cinimod-devops:Q8_0

LM Studio
Jan

vLLM

How to use dkudos/cinimod-devops with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/dkudos/cinimod-devops:Q8_0

Ollama
How to use dkudos/cinimod-devops with Ollama:
```
ollama run hf.co/dkudos/cinimod-devops:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use dkudos/cinimod-devops with Docker Model Runner:
```
docker model run hf.co/dkudos/cinimod-devops:Q8_0
```

Lemonade

How to use dkudos/cinimod-devops with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull dkudos/cinimod-devops:Q8_0

Run and chat with the model

lemonade run user.cinimod-devops-Q8_0

List all available models

lemonade list

Atomic Chat

dkudos commited on 24 days ago

Commit

1c616ab

verified ·

1 Parent(s): 5a15826

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +16 -5

README.md CHANGED Viewed

@@ -55,13 +55,20 @@ A 287M-parameter decoder-only causal language model (Llama-3 style architecture)
 | File | Description | Size |
 |---|---|---|
 | `model.safetensors` | Full bf16 PyTorch weights (HF format with `config.json`, `tokenizer.json`/`tokenizer_config.json`) | 548 MiB |
-| `checkpoint-4000-f16.gguf` | GGUF F16 (float16) - best quality for llama.cpp | 550 MiB |
-| `checkpoint-4000-Q8_0.gguf` | GGUF Q8_0 8-bit quantized - recommended for most uses (near-lossless, ~2x smaller) | 344 MiB |
 | `config.json` | Model config (transformers) | - |
 | `tokenizer.json` / `tokenizer_config.json` | BPE tokenizer (vocab 65,536) | - |
 | `train_log.log` | Full training log (steps, losses, LR) | - |
 | `full_val_eval.log` | Held-out full validation eval log | - |
 ## How to run
 ### HuggingFace transformers (PyTorch)
@@ -88,10 +95,14 @@ print(tok.decode(out[0]))
 Both GGUFs load directly in llama.cpp / llama-server with no external deps.
-Default (train context, 4096):
 ```bash
-llama-server -m dkudos/cinimod-devops/checkpoint-4000-Q8_0.gguf --port 8080
 ```
 256K context via linear RoPE scaling (trained at 4096):

 | File | Description | Size |
 |---|---|---|
 | `model.safetensors` | Full bf16 PyTorch weights (HF format with `config.json`, `tokenizer.json`/`tokenizer_config.json`) | 548 MiB |
+| `checkpoint-4000-Q8_0.gguf` | **GGUF Q8_0 (8-bit)** — recommended for most uses (near-lossless, ~2x smaller). [Download](https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-Q8_0.gguf) · [View](https://huggingface.co/dkudos/cinimod-devops/tree/main/checkpoint-4000-Q8_0.gguf) | 344 MiB |
+| `checkpoint-4000-f16.gguf` | **GGUF F16 (float16)** — best quality for llama.cpp. [Download](https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-f16.gguf) · [View](https://huggingface.co/dkudos/cinimod-devops/tree/main/checkpoint-4000-f16.gguf) | 550 MiB |
 | `config.json` | Model config (transformers) | - |
 | `tokenizer.json` / `tokenizer_config.json` | BPE tokenizer (vocab 65,536) | - |
 | `train_log.log` | Full training log (steps, losses, LR) | - |
 | `full_val_eval.log` | Held-out full validation eval log | - |
+## GGUF (llama.cpp) — recommended
+Two ready-to-serve GGUF files; either downloads standalone with no dependencies (no source code needed):
+- [`checkpoint-4000-Q8_0.gguf`](https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-Q8_0.gguf) — 8-bit quantized, ~344 MiB, recommended default
+- [`checkpoint-4000-f16.gguf`](https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-f16.gguf) — float16, ~550 MiB, best fidelity
 ## How to run
 ### HuggingFace transformers (PyTorch)
 Both GGUFs load directly in llama.cpp / llama-server with no external deps.
 ```bash
+# Q8_0 (default)
+wget https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-Q8_0.gguf
+llama-server -m checkpoint-4000-Q8_0.gguf --port 8080
+# or F16 for best fidelity
+wget https://huggingface.co/dkudos/cinimod-devops/resolve/main/checkpoint-4000-f16.gguf
+llama-server -m checkpoint-4000-f16.gguf --port 8080
 ```
 256K context via linear RoPE scaling (trained at 4096):