vanilj/palmyra-fin-70b-32k:IQ2_XXS

llama · 70.6B · IQ2_XXS

ECO MODE

AZW GTR Pro (AMD RYZEN AI MAX+ 395)

125 GB · Ubuntu 24.04.4 LTS

Tested on August 5, 2026

Top 89% Compare

Global Score

45 /100

Not Rec.

Hardware Fit

54/100

Quality

41/100

Get this model

🦙

Ollama

ollama pull vanilj/palmyra-fin-70b-32k:IQ2_XXS

View on Ollama Library

ollama.com/library/vanilj/palmyra-fin-70b-32k

Get it in LM Studio

Search and download models directly from the app

🤗

Find on HuggingFace

GGUF versions & quantizations

Hardware

Machine: AZW GTR Pro
CPU: AMD RYZEN AI MAX+ 395
Cores: 32 threads (16 cores)
Frequency: 3 GHz
RAM: 125 GB
GPU: Radeon 8060S
OS: Ubuntu 24.04.4 LTS
Arch: x64
Power Mode: low-power

Performance

Tokens/sec Tokens generated per second — higher is better: 6.8
Standard deviation Variation of speed across benchmark runs — lower means more consistent: ±0.1
First chunk latency Delay before the first streamed chunk arrives from runtime: 692 ms
Time to first token How fast the model starts responding — lower is better: 692 ms
Load time Time to load the model into memory before inference: 0.2 s
Memory usage RAM consumed during inference — percentage of total system memory: 27.9 GB (22%)
Total tokens Total number of tokens generated across all benchmark prompts: 1213

Score breakdown

Speed

6/50

Time to first token

18/20

Memory

30/30

Quality

Reasoning

7/20

Coding

2/20

Instruction following

14/20

Structured output

11/15

Math

0/15

Multilingual

7/10

Category levels

Reasoning: Weak Coding: Poor Instruction Following: Adequate Structured Output: Adequate Math: Poor Multilingual: Adequate

Metadata

Spec version: 0.2.1
Runtime: Ollama 0.32.5
Model format: GGUF
Hardware profile: HIGH-END
Result hash: e036f56f87ab9d96ecfba7217d574696c86bd84c20c9192cf1d9b778e2157fb3

Interpretation

Hardware fit: 54/100. Overall suitability: NOT RECOMMENDED (Global 45/100). Category profile: Reasoning: Weak, Coding: Poor, Instruction Following: Adequate, Structured Output: Adequate, Math: Poor, Multilingual: Adequate.

Warnings

System was in low-power mode during this benchmark.
CPU appears throttled (2.1 GHz current vs 3.0 GHz nominal, 69%).

Disqualifiers

Token speed too low: 6.8 tok/s (minimum: 8 tok/s for HIGH-END profile)

Bench Environment

Thermal: moderate CPU load: avg 5% (peak 6%)

Run yours now

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest