Instructions to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF: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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF: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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

Use Docker

docker model run hf.co/IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

LM Studio
Jan

vLLM

How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-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": "IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

Ollama
How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with Ollama:
```
ollama run hf.co/IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
```
Unsloth Desktop

How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

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": "IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with Docker Model Runner:
```
docker model run hf.co/IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0
```

Lemonade

How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

Run and chat with the model

lemonade run user.Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF-Q8_0

List all available models

lemonade list

Hermes Agent

How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

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 IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0

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 "IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF:Q8_0" \
  --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"

📜 OPTIMIZATION HISTORY — LEGACY EDITION

This repository hosts a previous iteration of our handcrafted MiniPlus architecture. While not our current specification, it remains an outstanding, high-fidelity quantization that significantly outperforms any flat 3-bit community quants (Q3_K_S / IQ3_S) and generic 2-bit APEX Mini community releases.

We preserve this repository publicly with 100% transparency as a verified engineering record of continuous optimization within the strict 13–14 GB envelope.

👉 Current Definitive Specification (V2.1): Access the upgraded V2.1 release featuring zero AVX2 CPU stalls and maximal long-context stability at: IsValorum Hugging Face Catalog

⚡ Quick Navigation Index

📦 Bundled Model Files & Specifications
🔬 Comparative Quantization Analysis (vs. Flat Quants & Generic APEX)
👁️ Bundled Q8_0 High-Precision Multimodal Vision Projector
💻 Everyday Laptop Benchmarks (DDR4 / DDR5 RAM)
🔥 The 24GB Miracle: Full 256K Context Runs In VRAM!
🏎️ Hardware Throughput Projections (RTX 30 / 40 / 50)
⚖️ The Speed vs. Precision Trade-off
🛠️ Surgical Tensor Quantization Map
📖 Recommended Configuration & Setup

📦 Bundled Model Files & Specifications

File Name	File Size	Memory Footprint	Format / Precision	Purpose
`Thomson-1.0-Small.APEX-I-MiniPlus-V2.gguf`	`14.63 GB` (`13.63 GiB`)	`13.63 GiB`	Custom APEX-I (3.38 BPW)	Main legal, auditing, tax & regulatory logic core
`mmproj-thomsonreuters_Thomson-1.0-Small-Q8_0.gguf`	`610 MB` (`582 MiB`)	`582 MiB`	High-Precision `Q8_0` Projector	Required for scanned PDF analysis, tables & OCR vision

Base Architecture: Qwen3_5MoeForConditionalGeneration (40 layers, 256 fine-grained micro-experts with intermediate dimension 512, 8 active per token) + Vision Projector.
Active Parameters: approx. 3.2B active parameters per token (blazing generation speed combined with 35B domain intelligence).
Importance Matrix: 510 calibrated tensor entries derived from 550 dense chunks of statutory law, financial audits, and regulatory filings.
Memory Footprint: Lean 13.63 GiB weight size leaving ample VRAM for deep document context windows.

🔬 Comparative Quantization Analysis (vs. Flat Quants & Generic APEX)

Also, don't confuse APEX-I-MiniPlus-V2 with a generic baseline APEX-I-Mini. Traditional APEX-I-Mini drops core experts aggressively to 2-bit IQ2_S and leaves output.weight at 3-bit Q3_K_M, which creates a noticeable perplexity hit on complex reasoning tasks. V2 was specifically re-engineered to avoid that quality floor (keeping core experts at calibrated IQ3_XXS, output in Q6_K, shared expert in non-linear IQ4_NL, and routers in F32).

To put the numbers in perspective: this cuts nearly 2 GB off a flat 3-bit quant (approx. 15.6 GB), and weighs only about approx. 1 GB more than a generic APEX-I-Mini (approx. 12.5 GB). For that single extra gigabyte of VRAM, you get a massive jump in reasoning and syntactic stability.

Take a look at the tensor-by-tensor comparison table below to inspect the exact architectural differences and see why this specific allocation is optimal. That's specifically what this was built for:

Architectural Component	Generic Automated Quants (Flat `Q3_K_S` / `IQ3_S`)	Generic APEX-I-Mini (Baseline Recipe)	Our Handcrafted APEX-I-MiniPlus-V2 (IsValorum)	Perceived Quality & Real-World Impact
Output Head (`output.weight`)	Flat `IQ3_S` / `Q3_K_S` (approx. 3.44 BPW)	Inherits base type `Q3_K_M` (approx. 3.44 BPW unarmored)	`Q6_K` (approx. 6.56 BPW uncompromised)	Eliminates Syntax & Vocabulary Hallucinations: Low-bit output heads cause tokenizer classification noise, breaking code indentation, brackets (`{}`, `[]`), math symbols, and domain terms. `Q6_K` preserves near-FP16 output classification.
Expert Routers (`ffn_gate_inp.weight`)	Blindly quantized to 3-bit / unoptimized	Inherits base type `Q3_K_M` (approx. 3.44 BPW compressed)	`F32` uncompressed (32.0 BPW, 2 MB/layer)	Zero Router Drift: In micro-expert models, even minuscule quantization errors in router logits misdirect tokens to wrong experts. Retaining uncompressed `F32` guarantees 100% routing fidelity with virtually zero memory overhead (approx. 80 MB total).
Attention & Language (`attn_output`, `attn_qkv`)	Flat `IQ3_S` / `Q3_K_S`	`Q3_K` on 34 middle layers (L3–36), `Q4_K` on 6 edge layers	`Q6_K` for `attn_output`, `IQ3_S` for `attn_qkv`	Contextual Retrieval Precision: Generic APEX reduces attention and language projections to `Q3_K` across 85% of layers. Our V2 build protects attention output in high-precision `Q6_K` and uses calibrated non-linear `IQ3_S`, ensuring flawless needle-in-a-haystack retrieval across deep 128k–256k context windows.
Attention Gates (`attn_gate.weight`)	Blindly compressed to 3-bit	Compressed to `Q3_K` (middle) / `Q4_K` (edges)	`Q8_0` (8.50 BPW)	Attention Head Stability: Attention gates modulate query-key routing across hybrid attention layers. Keeping them in 8-bit prevents attention crosstalk and hallucination over long contexts.
*Shared Foundation Expert (`ffn__shexp`)**	Flat `IQ3_S` / `Q3_K_S` (3.44 BPW)	Linear `Q4_K` (middle) / `Q5_K` (edges)	`IQ4_NL` (4.50 BPW non-linear codebook)	Foundational Knowledge Armor: The shared expert executes for 100% of tokens. In 256 micro-expert models, `IQ4_NL` non-linear codebooks preserve heavy-tailed outlier representations far better than standard linear quantization.
Core MoE Layers (Middle: 10–29)	Flat `IQ3_S` / `Q3_K_S` (uniform bit-rate across all layers)	Aggressive `IQ2_S` (2.50 BPW)	`IQ3_XXS` (3.06 BPW) + calibrated `imatrix`	Above the Quality Threshold: Generic 2-bit `IQ2_S` baselines drop below the critical quality floor for 35B MoEs, resulting in perplexity spikes on reasoning tasks. Our `IQ3_XXS` with imatrix achieves deep compression (272 MiB → 98 MiB per block) without sacrificing logic.
Edge MoE Layers (Layers 0–9 & 30–39)	Flat `IQ3_S` / `Q3_K_S` (no layer-wise gradient)	`Q3_K` (limited to first/last 5 layers only: L0–4, L35–39)	`IQ3_S` (expanded to 10 input & 10 output layers)	Protected Ingestion & Synthesis: Half of the model's layers (10 at input, 10 at output) form a non-linear armored envelope, preventing prompt misunderstanding and token degeneration across 256 micro-experts.
Multimodal Vision (`mmproj`)	Often omitted, or left as uncompressed `FP16` (approx. 900 MB)	Often omitted or separate uncompressed `FP16`	Bundled `Q8_0` (582 MB) with 27 critical F32/F16 fallbacks	Saves approx. 320 MB VRAM with Zero Loss: Handcrafted quantization preserves normalization and bias tensors in F32/F16, ensuring razor-sharp OCR, DOM viewport reading, and coordinate detection without visual noise.
Normalization & Biases	Often degraded	Standard	`F32` uncompressed	Numerical Stability: Prevents cumulative floating-point underflow/overflow across deep 40-layer computation.

👁️ Bundled Q8_0 High-Precision Multimodal Vision Projector

Standard community uploads often omit the multimodal projector or supply uncompressed FP16 files (approx. 857 MB), bloating memory.

Bundled Q8_0 Projector: Pre-quantized to Q8_0 (582 MiB / 610 MB), saving approx. 300 MB of VRAM.
Audited Layer Fallbacks: llama.cpp automatically preserved 27 critical normalization and embedding tensors in F32/F16, ensuring razor-sharp OCR of tiny contract footnotes, financial balance sheets, and scanned legal filings without artifacts.

💻 Everyday Laptop Benchmarks (DDR4 / DDR5 RAM)

Estimated Projections on Consumer Hardware

You do not need enterprise infrastructure to perform automated legal analysis. Estimated throughput projections on an everyday consumer laptop (Intel Core i5 / AMD Ryzen, 4GB/6GB Laptop GPU, 32GB DDR4/DDR5 RAM):

GPU VRAM Allocation: Uses only approx. 3.8 GB VRAM (fits easily on budget laptop GPUs like RTX 3050, 4050, or 2060).
System Memory Offload: Standard 32GB system RAM accommodates the remaining layers.
Estimated Document Ingestion (Prefill): 300 to 420+ tokens/second sustained across dense briefs.
Estimated Streaming Generation: 20 to 24+ tokens/second sustained output across system RAM!

🔥 The 24GB Miracle: Full 256K Context Runs In VRAM!

Legal and auditing tasks require ingesting hundreds of pages of case law, depositions, and regulatory exhibits. Automated 3-bit/4-bit community quants exceed 16–19 GiB in weights alone, crashing with Out-Of-Memory (OOM) errors when context scales up.

Thomson APEX-I-MiniPlus-V2 fits the entire 256K context window within 24GB VRAM:

Context Length	Model Weights (Est.)	KV Cache (q8_0, 4 slots)	Compute Buffers	Total GPU VRAM (Est.)	Hardware Feasibility
32,768 (32k)	`13.63 GiB`	`0.58 GiB`	`1.80 GiB`	`16.01 GiB`	Full offload on 24GB; partial on 16GB
65,536 (64k)	`13.63 GiB`	`0.92 GiB`	`1.95 GiB`	`16.50 GiB`	Effortless fit on 24GB GPUs
131,072 (128k)	`13.63 GiB`	`1.58 GiB`	`2.22 GiB`	`17.43 GiB`	Effortless fit on 24GB GPUs
262,144 (256k)	`13.63 GiB`	`2.92 GiB`	`2.80 GiB`	`19.35 GiB`	🔥 FULL 256K BRIEF IN VRAM!

Note: Projections leave approx. 4.65 GiB of headroom on 24GB cards for display buffers and the Q8 vision projector.

🏎️ Hardware Throughput Projections (RTX 30 / 40 / 50)

Projections represent theoretical estimates derived from hardware memory bandwidth and the approx. 3.2B active parameter MoE design.

⚖️ The Speed vs. Precision Trade-off

Uncompressed F32 Router Selectors: In a 256 micro-expert model, even minor router rounding errors route clauses to the wrong domain expert (e.g. confusing tax compliance with intellectual property). ffn_gate_inp.weight is maintained in uncompressed F32 (2 MB/layer) to eliminate routing drift.
Q6_K Output Head: Protects Latin legal terminology, statutory code citations, and quantitative tax calculation symbols from compression distortion.
Non-Linear IQ Codebooks: IQ3_XXS, IQ3_S, and IQ4_NL preserve nuanced contractual semantics across edge and core layers.

🛠️ Surgical Tensor Quantization Map

Tensor Pattern	Layer Scope	Quant Type	BPW	Engineering Rationale
`output.weight`	Vocabulary Head	`Q6_K`	6.56	Uncompromised 6-bit precision for legal, fiscal & statutory vocabulary
`token_embd.weight`	Embedding	`High-Prec`	High	Preserves subtle token definitions and prompt grounding
`ffn_gate_inp.weight`	Expert Routers	`F32`	32.0	Uncompressed full-precision routers preventing domain expert drift
`attn_gate.weight`	Attention Gates	`Q8_0`	8.50	High-precision 8-bit gating for attention routing dynamics
*`ffn__shexp`**	Shared Experts	`IQ4_NL`	4.50	4-bit non-linear codebook for the 100% active shared foundational expert
`ffn_down/up/gate`	Edges (0–9, 30–39)	`IQ3_S`	3.44	Armored boundary layers protecting prompt ingest and final opinion synthesis
`ffn_down/up/gate`	Core (10–29)	`IQ3_XXS`	3.06	Deep compression (272 MiB → 98 MiB per block) calibrated via legal imatrix
`mmproj` (Vision)	Visual Projector	`Q8_0`	8.00	High-fidelity OCR and table rendering with 27 critical F32/F16 fallbacks
Norms & Biases	All Layers	`F32`	32.0	Absolute numerical stability across deep 40-layer computation

📖 Recommended Configuration & Setup

Unsloth Studio:

Load Thomson-1.0-Small.APEX-I-MiniPlus-V2.gguf.
Select mmproj-thomsonreuters_Thomson-1.0-Small-Q8_0.gguf as the vision projector.
Configure KV Cache Dtype to q8_0 and Context Checkpoints to 1.
Set GPU Offload to 100% (-ngl 99) on 24GB GPUs.

llama.cpp CLI:

llama-cli -m Thomson-1.0-Small.APEX-I-MiniPlus-V2.gguf \
  --mmproj mmproj-thomsonreuters_Thomson-1.0-Small-Q8_0.gguf \
  -ngl 99 \
  -c 32768

LM Studio / Ollama:

Load the main model and attach the bundled mmproj adapter.
Maximize GPU offload and define context buffer.

Downloads last month: 63

GGUF

Model size

35B params

Architecture

qwen35moe

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Model tree for IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF

Base model

Qwen/Qwen3.6-35B-A3B

Finetuned

tri-fair-lab/Snowdon1.1-Small

Finetuned

thomsonreuters/Thomson-1.0-Small

Quantized

(9)

this model

Collections including IsValorum/Thomson-1.0-Small-APEX-I-MiniPlus-V2-GGUF

APEX-I Custom Quants

Collection

Handcrafted APEX-I GGUFs optimized for 4GB-24GB VRAM setups with high throughput. • 7 items • Updated about 13 hours ago

Olds MiniPlus V2

Collection

Historical legacy collection of APEX-I-MiniPlus V2 quantization releases. • 5 items • Updated about 2 hours ago