143 lines
4.7 KiB
Markdown
143 lines
4.7 KiB
Markdown
---
|
|
tags:
|
|
- 1.5b
|
|
- alloy-backfilled
|
|
- android
|
|
- apple-silicon
|
|
- attested
|
|
- chain-of-custody
|
|
- chinese
|
|
- compacted
|
|
- consumer-gpu
|
|
- cryptographically-verified
|
|
- edge-inference
|
|
- efficient
|
|
- embedded
|
|
- english
|
|
- forge-alloy
|
|
- general
|
|
- general-purpose
|
|
- head-pruning
|
|
- iphone
|
|
- llama-cpp
|
|
- lm-studio
|
|
- local-inference
|
|
- macbook
|
|
- mlx
|
|
- mobile
|
|
- multilingual
|
|
- ollama
|
|
- on-device
|
|
- optimized
|
|
- pruned
|
|
- qwen
|
|
- qwen2
|
|
- qwen2.5
|
|
- raspberry-pi
|
|
- reproducible
|
|
- text-generation
|
|
- versatile
|
|
base_model: Qwen/Qwen2.5-1.5B
|
|
pipeline_tag: text-generation
|
|
license: apache-2.0
|
|
---
|
|
|
|
# 30% Smaller, +2.4% Better
|
|
|
|
**Qwen2.5-1.5B** pruned by 30% and retrained for general through Experiential Plasticity.
|
|
|
|
**2.50 → 2.44 perplexity** · 3 cycles
|
|
|
|
|
|
<p align="center">
|
|
<a href="https://cambriantech.github.io/forge-alloy/verify/#hf.co/continuum-ai/qwen2.5-1.5b-general-forged/resolve/main/qwen2.5-1.5b-general-forged.alloy.json@f024d59a481e9032">
|
|
<img src="alloy-qr.png" alt="Verify Chain of Custody" width="160"/>
|
|
</a>
|
|
</p>
|
|
|
|
<p align="center">
|
|
<a href="https://cambriantech.github.io/forge-alloy/verify/#hf.co/continuum-ai/qwen2.5-1.5b-general-forged/resolve/main/qwen2.5-1.5b-general-forged.alloy.json@f024d59a481e9032"><b>Every claim on this card is verified</b></a><br>
|
|
<b>Trust: self-attested</b> · 1 benchmark · 2 devices tested<br>
|
|
<a href="https://github.com/CambrianTech/forge-alloy">ForgeAlloy</a> chain of custody · <a href="qwen2.5-1.5b-general-forged.alloy.json">Download alloy</a> · Merkle-chained
|
|
</p>
|
|
|
|
---
|
|
|
|
**Qwen2.5-1.5B** with cryptographic provenance via the [ForgeAlloy](https://github.com/CambrianTech/forge-alloy) chain of custody.
|
|
|
|
|
|
## Benchmarks
|
|
|
|
| Benchmark | Result | Verified |
|
|
|---|---|---|
|
|
| **perplexity** | **2.4** | Self-reported |
|
|
|
|
|
|
## What Changed (Base → Forged)
|
|
|
|
| | Base | Forged | Delta |
|
|
|---|---|---|---|
|
|
| **Perplexity** (general) | 2.50 | 2.44 | -2.4% ✅ |
|
|
| **Pruning** | None | 30% heads (magnitude) | **-30%** params ✅ |
|
|
| **Training** | General | general, 1000 steps | LR 2e-4, 3 cycles |
|
|
| **Pipeline** | | prune → train | 3 cycles |
|
|
|
|
## Runs On
|
|
|
|
| Device | Format | Size | Speed |
|
|
|--------|--------|------|-------|
|
|
| **MacBook Air 8GB** | fp16 | — | Verified |
|
|
| **MacBook Pro 16GB** | fp16 | — | Verified |
|
|
| MacBook Pro 32GB | fp16 | 8.0GB | Expected |
|
|
| MacBook Air 16GB | Q8_0 | ~4.0GB | Expected |
|
|
| MacBook Air 8GB | Q4_K_M | ~2.5GB | Expected |
|
|
| iPhone / Android | Q4_K_M | ~2.5GB | Expected |
|
|
|
|
## Quick Start
|
|
|
|
```python
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
model = AutoModelForCausalLM.from_pretrained("continuum-ai/qwen2.5-1.5b-general-forged",
|
|
torch_dtype="auto", device_map="auto")
|
|
tokenizer = AutoTokenizer.from_pretrained("continuum-ai/qwen2.5-1.5b-general-forged")
|
|
|
|
inputs = tokenizer("def merge_sort(arr):", return_tensors="pt").to(model.device)
|
|
output = model.generate(**inputs, max_new_tokens=200)
|
|
print(tokenizer.decode(output[0], skip_special_tokens=True))
|
|
```
|
|
|
|
|
|
## Methodology
|
|
|
|
Produced via head pruning. Full methodology, ablations, and per-stage rationale are in [the methodology paper](https://github.com/CambrianTech/continuum/blob/main/docs/papers/PLASTICITY-COMPACTION.md) and the companion [`MODEL_METHODOLOGY.md`](MODEL_METHODOLOGY.md) in this repository. The pipeline ran as `prune → train` over 3 cycles on MacBook Air 8GB.
|
|
|
|
|
|
## Chain of Custody
|
|
|
|
Scan the QR or [verify online](https://cambriantech.github.io/forge-alloy/verify/#hf.co/continuum-ai/qwen2.5-1.5b-general-forged/resolve/main/qwen2.5-1.5b-general-forged.alloy.json@f024d59a481e9032). Download the [alloy file](qwen2.5-1.5b-general-forged.alloy.json) to verify independently.
|
|
|
|
| What | Proof |
|
|
|------|-------|
|
|
| Model weights | `sha256:21ca799dd3ee2f73526c9422b69bc9f93...` |
|
|
| Code that ran | `sha256:legacy-pre-alloy-...` |
|
|
| Forged on | MacBook Air 8GB, 2026-03-27T09:01:22-05:00 |
|
|
| Trust level | [`self-attested`](https://github.com/CambrianTech/forge-alloy/blob/main/docs/ATTESTATION.md) |
|
|
| Spec | [ForgeAlloy](https://github.com/CambrianTech/forge-alloy) — Rust/Python/TypeScript |
|
|
|
|
## Make Your Own
|
|
|
|
Forged with [Continuum](https://github.com/CambrianTech/continuum) — a distributed AI world that runs on your hardware.
|
|
|
|
<p align="center">
|
|
<a href="https://github.com/CambrianTech/continuum"><img src="https://raw.githubusercontent.com/CambrianTech/continuum/main/docs/images/factory.png" alt="Continuum Model Factory" width="400"/></a>
|
|
</p>
|
|
|
|
The Factory configurator lets you design and forge custom models visually — context extension, pruning, LoRA, quantization, vision/audio modalities. Pick your target devices, the system figures out what fits.
|
|
|
|
[GitHub](https://github.com/CambrianTech/continuum) · [All Models](https://huggingface.co/continuum-ai) · [Forge-Alloy](https://github.com/CambrianTech/forge-alloy)
|
|
|
|
## License
|
|
|
|
apache-2.0
|