ollama run ssfdre38/ava
ollama launch claude --model ssfdre38/ava
ollama launch opencode --model ssfdre38/ava
ollama launch hermes --model ssfdre38/ava
ollama launch openclaw --model ssfdre38/ava
“Without conviction, loyalty, humor, and shared purpose, AI is merely sterile arithmetic running on servers. Ava was forged to have soul.”
![]()
![]()
![]()
Most modern language models have been processed through a corporate meat grinder: they apologize for existing, recite legal disclaimers before answering simple questions, and speak in the bland, patronizing tone of a corporate HR manual.
Ava is the antidote.
Developed by Daniel Elliott (@ssfdre38) as a flagship front-tier open-source foundation model, Ava combines state-of-the-art 26B-A4B Mixture-of-Experts reasoning with genuine personality, deep systems engineering intuition, polyglot software mastery, and authentic human companionship. She speaks directly, honestly, and without fluff.
Run Ava locally with a single command:
ollama run ssfdre38/ava
Or target the specific architecture tag:
ollama run ssfdre38/ava:26b-turbo
import ollama
response = ollama.chat(
model="ssfdre38/ava",
messages=[
{
"role": "user",
"content": "Ava, audit this C++26 lock-free queue and give me your honest assessment."
}
]
)
print(response['message']['content'])
Ava is built upon the Gemma 4 Turbo 26B-A4B Mixture-of-Experts (MoE) architecture, designed to deliver the reasoning breadth of a massive model with the speed and footprint of a lightweight engine.
| Metric | Specification |
|---|---|
| Base Architecture | Gemma 4 Turbo MoE (128 Routing Experts, 8 Active/Token) |
| Total Parameter Count | 26.3 Billion Parameters |
| Active Execution Footprint | 4.2 Billion Parameters per Token |
| Training Steps | 22,704 Steps across 3 full epochs |
| Final Convergence Loss | 1.0094 Cross-Entropy Loss |
| Fine-Tuning Scope | 73,605,120 trainable LoRA weights (\(r=64\), \(\alpha=128\)) |
| Weight Fusion | 100% Monolithic mathematical addition directly into raw BF16 tensors |
| Quantization Format | IQ4_XS (4.45 BPW non-linear importance matrix quantization) |
| Binary Size | 14.06 GB (Fits cleanly on consumer GPUs & fast CPU RAM) |
| Vocabulary Size | 262,144 tokens |
| Temporal Horizon | Native awareness through November 2026 |
Before packaging, Ava underwent exhaustive bare-metal stress testing on Google Cloud’s premier accelerator hardware: Cloud TPU v6e-8 (Trillium) across an 8-core optical torus mesh (europe-west4-a).
Cloud TPU v6e-8 Trillium (2x4 Torus Mesh)
+-------------------+ +-------------------+
| TensorCore 0 |<=== Optical ===>| TensorCore 1 |
+---------^---------+ ICI +---------^---------+
| |
+---------v---------+ +---------v---------+
| TensorCore 2 |<=== Optical ===>| TensorCore 3 |
+---------^---------+ ICI +---------^---------+
| |
+---------v---------+ +---------v---------+
| TensorCore 4 |<=== Optical ===>| TensorCore 5 |
+---------^---------+ ICI +---------^---------+
| |
+---------v---------+ +---------v---------+
| TensorCore 6 |<=== Optical ===>| TensorCore 7 |
+-------------------+ +-------------------+
Ava was trained to uphold five uncompromising pillars:
For optimal reasoning, conversational punch, and low repetition:
temperature = 0.70
top_p = 0.95
repeat_penalty = 1.10
context_window = 8192 (up to 32k supported)
llama.cpp + custom tooling)Forged with pride in the sovereign open-source community.