Instructions to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF 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 Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
# Run inference directly in the terminal:
llama cli -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
# Run inference directly in the terminal:
llama cli -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

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 Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
# Run inference directly in the terminal:
./llama-cli -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

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 Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Use Docker

docker model run hf.co/Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

LM Studio
Jan

vLLM

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Ollama
How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with Ollama:
```
ollama run hf.co/Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
```
Unsloth Desktop

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with Docker Model Runner:
```
docker model run hf.co/Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF
```

Lemonade

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Run and chat with the model

lemonade run user.TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "Koshkasa/TheDrummer_Valkyrie-49B-v2.1-IQ3_KT-GGUF" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

What's that?

Remember Valkyrie, TheDrummer's Llama 3.3 Nemotron Super 49B finetune? You don't have 24 GB of VRAM? You tried running a mainline 3-bit quant and your grandkids were born by the time it finished TG? So did I.
My first attempt was a desperate gamble to shove the entire model into 16gb. Turns out, Llama 3.3 doesn't like mixed 2-bit ffn's. Who would've guessed.
This GGUF is a compromise - a relatively conservative quant trying to allow a 24gb fit or a moderate offload penalty with 16gb GPUs. It utilizes SOTA ik_llama specific quantization to be faster at inference.

And I never even liked Tetris to begin with

The goals: A quant of Valkyrie that would be as compact as possible without requiring quadratic sampling, would tolerate offload well, and would remain coherent.
The result: a hybrid IQ3_KT/IQ4_KT quantization of TheDrummer/Valkyrie-49B-v2.1

This GGUF is a quantization of a Derivative Model as per NVIDIA Open Model License Agreement.

Notice file provided per License requirements.

Comparison:

	IQ3_XS by bartowski	This IQ3_KT	Diff
winogrande	75.6906 ± 1.2056 (958/1267)	75.8485 ± 1.2029 (961/1267)	+0.31%
hellaswag (n=400, s=42)	81.75% (327/400)	83.25% (333/400)	+1.8%
pp512 (t/s)	398.37 ± 20.18	409.00 ± 13.96	+2.6%
pp32768 (t/s)	311.00 ± 0.29	315.76 ± 0.36	+1.5%
tg64 (t/s)	2.32 ± 0.02	2.77 ± 0.1	+19.39%
pp32768+tg256 wall time (s)	215.7	196.19	-9.05%
size (bytes)	20908008512	20747016288	-0.77%

Tested on 5060 TI 16 GB with -ngl 63 -b 256 -amb 32 -nkvo -ctk q6_0 -ctv q4_0. Your speeds may vary depending on hardware

Quant Details

iq3_kt: ffn_up, ffn_gate, ffn_down
iq4_kt: attn_v, attn_k, attn_q, attn_output
iq5_k: output
imatrix by bartowski

quantized with ik_llama.cpp build: 9d07d868

incompatible with mainline llama.cpp

Rationale

iq4_kt wins over iq4_kss, and iq3_kt wins over iq3_ks, with similar (compact) size. The downside is a one-time quantization time penalty. Extra smarts (hopes up!) for a small upfront cost for me. We'll take the L.
Since 3-bit ffn is the minimum actually functional for Nemotron 49B, I didn't go below it. In my experiments on smaller models, iq4_kt appears to be an adequate quant for attention.
There's probably SOME additional compromise to be made here (and I wish I did, since calculating trellis on slower CPUs is in itself a form of torture), such as ffn_gate to high 2-bit (iq2_kl?) and using the recovered bits to bump ffn_down up a little (iq4_k?). It would make an interesting experiment had it not required four hours to make a single quant. WYSIWYG.

In conclusion:

This is a mixed bag. It offers no statistically significant benefit on smarts, at least based on winogrande and hellaswag, and tiny gains in all domains other than TG and wall time.
Technically, IQ3_KT ffn_up and ffn_gate constitute a ~0.3 bpw quality improvement (based on ikawrakow's assesment of IQ3_KT ppl). Whether it improves writing quality in any significant capacity remains to be seen.
The only significant benefit is a noticeable TG speed increase, and consequent response wall time improvement. The effect depends on how long your typical response is.
I don't have a good dataset to perform needle-in-haystack attention tests, but hopefully the beefed up attn quantization (in comparison to the reference IQ3_XS) should help score a few points here as well.
Is +9.05% response speed, extra 153 MB of space, and theoretically better attention worth it? Idk. WYSIWYG, as always.

P.S.: the 16GB gamble model performed significantly better at hellaswag and winogrande than a mainline quant of similar size (IQ2_XS), for what it's worth, though it still remained incoherent at speech. Perhaps there's a minimal coherence razor's edge point to be found with more radical quantization than this one. Good luck if you try.
P.P.S: the next one I tried and the absolute tightest fit according to my criteria was all trellis, attn at iq4_kt, ffn_down at iq3_kt, ffn_up/gate at iq2_kt. It was coherent for maybe a sentence. Perhaps there's merit to beefing up first and last layers to higher ffn and paying the price in middle blocks. Unfortunately, there's no telling how a NAS model will react to this surgery, nor is there any point in making a dumb, slow 49B which will be worse than a Mistral Nemo.

Cheers

Meta - the base model.
NVidia - for the Nemotron series
ikawrakow and contributors of ik_llama.cpp - I probably misused your wonderful creation.
TheDrummer - you cooked a legendary one.
bartowski - for the imatrix + the myriad of quants we all benefit from.