Thox-ai/ thoxwave-8b

18 Downloads Updated 1 month ago

THOX.ai creative all-rounder. Dolphin-2.9 Llama-3-8B QLoRA fine-tune, Q4_K_M, for ideation and long-form drafting.

ollama run Thox-ai/thoxwave-8b:Q4_K_M

curl http://localhost:11434/api/chat \
  -d '{
    "model": "Thox-ai/thoxwave-8b:Q4_K_M",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='Thox-ai/thoxwave-8b:Q4_K_M',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'Thox-ai/thoxwave-8b:Q4_K_M',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Models

Name

1 model

Size / Usage

Context

Input

thoxwave-8b:Q4_K_M

4.9GB · 8K context window · Text · 1 month ago

thoxwave-8b:Q4_K_M

4.9GB

8K

Text

Readme

ThoxWave-8B

THOX.ai creative all-rounder. Dolphin-2.9 Llama-3-8B QLoRA fine-tune, Q4_K_M, for ideation and long-form drafting.

Tags: creative writing llama 8b q4_k_m

At a glance

Base model: cognitivecomputations/dolphin-2.9-llama3-8b
Fine-tune: QLoRA supervised fine-tune (THOX thoxllm-factory pipeline)
Architecture: llama
Parameters: 8.0B
Quantization: Q4_K_M (4.9 GB)
Native context: 8192 tokens (default runtime num_ctx 8192)
Prompt template: ChatML (<|im_start|> / <|im_end|>)
License: Meta Llama 3 Community License — https://llama.meta.com/llama3/license/ (Dolphin-2.9 fine-tune: Apache-2.0). THOX fine-tune, Modelfile, and packaging: © 2026 THOX.ai LLC — see NOTICE.md.
Publisher: THOX.AI LLC

Applications

Launch this model directly into a supported tool:

ollama launch claude   --model thox-ai/thoxwave-8b:Q4_K_M   # Claude Code
ollama launch chatgpt  --model thox-ai/thoxwave-8b:Q4_K_M   # Codex App
ollama launch openclaw --model thox-ai/thoxwave-8b:Q4_K_M   # OpenClaw
ollama launch hermes   --model thox-ai/thoxwave-8b:Q4_K_M   # Hermes
ollama launch codex    --model thox-ai/thoxwave-8b:Q4_K_M   # Codex
ollama launch opencode --model thox-ai/thoxwave-8b:Q4_K_M   # OpenCode

Quick start

ollama pull thox-ai/thoxwave-8b:Q4_K_M
ollama run thox-ai/thoxwave-8b:Q4_K_M "your prompt"

Intended use

Brainstorming, ideation, and long-form drafting. Direct and steerable.

The model ships with the THOX factuality system prompt: it answers company questions only from verified facts and declines to invent non-public details (headquarters, headcount, funding, founding date).

Not intended for

Factual lookups without verification; legal or medical advice.

Notes

Context: the GGUF supports up to 8192 tokens; the published default num_ctx is 8192 for predictable memory on edge devices. Override at runtime with --num-ctx or a client parameter.
On-device: built for local execution on THOX hardware and desktops. Your AI. Your Data. Your Rules.

Links

THOX: https://www.thox.ai
Registry: https://ollama.com/thox-ai/thoxwave-8b
Press: press@thox.ai

Published by THOX.AI LLC from the thoxllm-factory pipeline.

# ThoxWave-8B

THOX.ai creative all-rounder. Dolphin-2.9 Llama-3-8B QLoRA fine-tune, Q4_K_M, for ideation and long-form drafting.

**Tags:** `creative` `writing` `llama` `8b` `q4_k_m`

## At a glance
- **Base model**: cognitivecomputations/dolphin-2.9-llama3-8b
- **Fine-tune**: QLoRA supervised fine-tune (THOX thoxllm-factory pipeline)
- **Architecture**: llama
- **Parameters**: 8.0B
- **Quantization**: Q4_K_M (4.9 GB)
- **Native context**: 8192 tokens (default runtime `num_ctx` 8192)
- **Prompt template**: ChatML (`<|im_start|>` / `<|im_end|>`)
- **License**: Meta Llama 3 Community License — https://llama.meta.com/llama3/license/ (Dolphin-2.9 fine-tune: Apache-2.0). THOX fine-tune, Modelfile, and packaging: (c) 2026 THOX.ai LLC — see NOTICE.md.
- **Publisher**: THOX.AI LLC

## Applications
Launch this model directly into a supported tool:
```bash
ollama launch claude   --model thox-ai/thoxwave-8b:Q4_K_M   # Claude Code
ollama launch chatgpt  --model thox-ai/thoxwave-8b:Q4_K_M   # Codex App
ollama launch openclaw --model thox-ai/thoxwave-8b:Q4_K_M   # OpenClaw
ollama launch hermes   --model thox-ai/thoxwave-8b:Q4_K_M   # Hermes
ollama launch codex    --model thox-ai/thoxwave-8b:Q4_K_M   # Codex
ollama launch opencode --model thox-ai/thoxwave-8b:Q4_K_M   # OpenCode
```

## Quick start
```bash
ollama pull thox-ai/thoxwave-8b:Q4_K_M
ollama run thox-ai/thoxwave-8b:Q4_K_M "your prompt"
```

## Intended use
Brainstorming, ideation, and long-form drafting. Direct and steerable.

The model ships with the THOX factuality system prompt: it answers company questions only from verified facts and declines to invent non-public details (headquarters, headcount, funding, founding date).

## Not intended for
Factual lookups without verification; legal or medical advice.

## Notes
- **Context**: the GGUF supports up to 8192 tokens; the published default `num_ctx` is 8192 for predictable memory on edge devices. Override at runtime with `--num-ctx` or a client parameter.
- **On-device**: built for local execution on THOX hardware and desktops. Your AI. Your Data. Your Rules.

## Links
- THOX: https://www.thox.ai
- Registry: https://ollama.com/thox-ai/thoxwave-8b
- Press: press@thox.ai

_Published by THOX.AI LLC from the thoxllm-factory pipeline._

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)