oxfrug

Ornith-1.5-35B-A3B-REAP-20pct

Model details

Table with columns: Base, This
	Base	This
Base model	`ornith-ai/Ornith-1.5-35B-A3B`	same hybrid GatedDeltaNet + MoE
Relation	—	REAP 20% expert prune
Experts / layer	256	205 (51 removed)
Active / token	8	8
Layers	40	40
On-disk BF16	~67 G	~56 G
Library	transformers ≥ 5.8.1	transformers ≥ 5.8.1
License	MIT	MIT

Shared expert, vision tower, embeddings, and lm_head are untouched. Router weights renormalized after prune. config.json text_config.num_experts is 205. Architecture: Qwen3_5MoeForConditionalGeneration.

Quickstart

bash
pip install -U "transformers>=5.8.1"

python
from transformers import AutoModelForImageTextToText, AutoTokenizer

model_id = "oxfrug/Ornith-1.5-35B-A3B-REAP-20pct"
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
    model_id, torch_dtype="auto", device_map="auto"
)

messages = [{"role": "user", "content": "Write a one-line Python lambda that squares a number."}]
text = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tok(text, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=256, temperature=0.6, top_p=0.95)
print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))

vLLM (>=0.19.1), 2× 80GB recommended for long context:

bash
vllm serve oxfrug/Ornith-1.5-35B-A3B-REAP-20pct \
    --tensor-parallel-size 2 \
    --max-model-len 262144 \
    --enable-auto-tool-choice --tool-call-parser qwen3_xml \
    --reasoning-parser qwen3 \
    --trust-remote-code

Sampling: general temperature=0.6, top_p=0.95, top_k=20. Ornith is a reasoning model — the assistant turn opens with <think>…</think> unless the server splits it into reasoning_content.

Benchmarks

This checkpoint was not re-run on SWE-bench / Terminal-Bench / GPQA. Do not copy the base card’s 79 SWE-bench / 67.8 Terminal-Bench numbers onto this prune.

vs base (this checkpoint)

Table with columns: Base, This
	Base	This
Experts / layer	256	205
Disk (BF16)	~67 G	~56 G
Active params / token	~3B	~3B

Proxy (same recipe, not this checkpoint)

The same REAP 20% recipe on Qwen/Qwen3.5-35B-A3B (0xSero/Qwen3.5-28B, greedy, 0-shot) dropped coding a few points and raised WikiText perplexity. Ornith is a Qwen3.5 hybrid (GatedDeltaNet + MoE); expect a comparable tax, not these exact numbers.

Table with columns: Benchmark, Original Qwen3.5-35B, REAP 20%, Delta
Benchmark	Original Qwen3.5-35B	REAP 20%	Delta
HumanEval pass@1	76.2%	73.2%	−3.0
HumanEval+ pass@1	72.0%	70.1%	−1.9
WikiText-2 PPL (10k tok)	6.83	9.51	+2.68

Source: 0xSero’s card, not a run of oxfrug/Ornith-1.5-35B-A3B-REAP-20pct.

Smoke (this checkpoint)

Greedy, thinking off, 3090 offload. Not a published coding bench.

Table with columns: Prompt, Base, This
Prompt	Base	This
square lambda / `is_even`	match	match
`get_weather` with the base card `tools=` schema	well-formed `<tool_call><function=get_weather>` / `Paris`	same

How it was made

Same mix as 0xSero/Qwen3.5-28B, not Cerebras’ 24k×16k agentic mix:

theblackcat102/evol-codealpaca-v1 × 250
open-r1/Mixture-of-Thoughts code / math / science × 250 each
1000 sequences, max length 1024 (3090 VRAM; 0xSero used 4096)
seed 42, prune_method=reap, renormalize_router_weights=true
Layerwise observer (model does not fit in 60 G RAM)

No xlam / SWE-smith traces — extra risk for Ornith’s tool-calling specialists.

Intended use

Local / self-hosted coding and tool-calling with Ornith-1.5 quality, at ~11 GB less disk than the unpruned BF16. Accept a quality tax vs the 35B-A3B base.

Limits

Expert prune, not a quant and not a fine-tune.
Shorter calibration sequences (1024 vs 4096) than the Qwen3.5-35B sibling run.
Hybrid GatedDeltaNet + MoE: REAP only touches MoE experts.
Do not claim base Ornith-1.5 SWE-bench / Terminal-Bench / GPQA scores for this file.

License

MIT, same as the base. Prune by oxfrug.

Citation

bibtex
@misc{lasby2025reap,
  title  = {REAP the Experts: Why Pruning Prevails for One-Shot MoE Compression},
  author = {Mike Lasby and Ivan Lazarevich and Nish Sinnadurai and Sean Lie and Yani Ioannou and Vithursan Thangarasa},
  year   = {2025},
  eprint = {2510.13999},
  archivePrefix = {arXiv}
}

@misc{ornith_1_5,
  title  = {{Ornith-1.5}: From Self-Scaffolding to Self-Improvement},
  url    = {https://ornith.ai/ornith_1_5.html},
  author = {{Ornith Team}},
  year   = {2026}
}

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Model Details

Model Provider

oxfrug

Model Tree

Base

ornith-ai/Ornith-1.5-35B-A3B

Quantized

this model

Input Modalities

TextImageVideo

Output Modalities

Text

Supported Functionality

Dedicated Endpoints

Explore FriendliAI today

Get started Talk to an engineer

Model details

Table with columns: Base, This
	Base	This
Base model	`ornith-ai/Ornith-1.5-35B-A3B`	same hybrid GatedDeltaNet + MoE
Relation	—	REAP 20% expert prune
Experts / layer	256	205 (51 removed)
Active / token	8	8
Layers	40	40
On-disk BF16	~67 G	~56 G
Library	transformers ≥ 5.8.1	transformers ≥ 5.8.1
License	MIT	MIT

Quickstart

bash
pip install -U "transformers>=5.8.1"

python
from transformers import AutoModelForImageTextToText, AutoTokenizer

model_id = "oxfrug/Ornith-1.5-35B-A3B-REAP-20pct"
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
    model_id, torch_dtype="auto", device_map="auto"
)

messages = [{"role": "user", "content": "Write a one-line Python lambda that squares a number."}]
text = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tok(text, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=256, temperature=0.6, top_p=0.95)
print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))

vLLM (>=0.19.1), 2× 80GB recommended for long context:

bash
vllm serve oxfrug/Ornith-1.5-35B-A3B-REAP-20pct \
    --tensor-parallel-size 2 \
    --max-model-len 262144 \
    --enable-auto-tool-choice --tool-call-parser qwen3_xml \
    --reasoning-parser qwen3 \
    --trust-remote-code

Sampling: general temperature=0.6, top_p=0.95, top_k=20. Ornith is a reasoning model — the assistant turn opens with <think>…</think> unless the server splits it into reasoning_content.

Benchmarks

This checkpoint was not re-run on SWE-bench / Terminal-Bench / GPQA. Do not copy the base card’s 79 SWE-bench / 67.8 Terminal-Bench numbers onto this prune.

vs base (this checkpoint)

Table with columns: Base, This
	Base	This
Experts / layer	256	205
Disk (BF16)	~67 G	~56 G
Active params / token	~3B	~3B

Proxy (same recipe, not this checkpoint)

Table with columns: Benchmark, Original Qwen3.5-35B, REAP 20%, Delta
Benchmark	Original Qwen3.5-35B	REAP 20%	Delta
HumanEval pass@1	76.2%	73.2%	−3.0
HumanEval+ pass@1	72.0%	70.1%	−1.9
WikiText-2 PPL (10k tok)	6.83	9.51	+2.68

Source: 0xSero’s card, not a run of oxfrug/Ornith-1.5-35B-A3B-REAP-20pct.

Smoke (this checkpoint)

Greedy, thinking off, 3090 offload. Not a published coding bench.

Table with columns: Prompt, Base, This
Prompt	Base	This
square lambda / `is_even`	match	match
`get_weather` with the base card `tools=` schema	well-formed `<tool_call><function=get_weather>` / `Paris`	same

How it was made

Same mix as 0xSero/Qwen3.5-28B, not Cerebras’ 24k×16k agentic mix:

theblackcat102/evol-codealpaca-v1 × 250
open-r1/Mixture-of-Thoughts code / math / science × 250 each
1000 sequences, max length 1024 (3090 VRAM; 0xSero used 4096)
seed 42, prune_method=reap, renormalize_router_weights=true
Layerwise observer (model does not fit in 60 G RAM)

No xlam / SWE-smith traces — extra risk for Ornith’s tool-calling specialists.

Intended use

Local / self-hosted coding and tool-calling with Ornith-1.5 quality, at ~11 GB less disk than the unpruned BF16. Accept a quality tax vs the 35B-A3B base.

Limits

Expert prune, not a quant and not a fine-tune.
Shorter calibration sequences (1024 vs 4096) than the Qwen3.5-35B sibling run.
Hybrid GatedDeltaNet + MoE: REAP only touches MoE experts.
Do not claim base Ornith-1.5 SWE-bench / Terminal-Bench / GPQA scores for this file.

License

MIT, same as the base. Prune by oxfrug.

Citation

bibtex
@misc{lasby2025reap,
  title  = {REAP the Experts: Why Pruning Prevails for One-Shot MoE Compression},
  author = {Mike Lasby and Ivan Lazarevich and Nish Sinnadurai and Sean Lie and Yani Ioannou and Vithursan Thangarasa},
  year   = {2025},
  eprint = {2510.13999},
  archivePrefix = {arXiv}
}

@misc{ornith_1_5,
  title  = {{Ornith-1.5}: From Self-Scaffolding to Self-Improvement},
  url    = {https://ornith.ai/ornith_1_5.html},
  author = {{Ornith Team}},
  year   = {2026}
}

Ornith-1.5-35B-A3B-REAP-20pct

README

Model details

Quickstart

Benchmarks

vs base (this checkpoint)

Proxy (same recipe, not this checkpoint)

Smoke (this checkpoint)

How it was made

Intended use

Limits

License

Citation

Explore FriendliAI today

README

Model details

Quickstart

Benchmarks

vs base (this checkpoint)

Proxy (same recipe, not this checkpoint)

Smoke (this checkpoint)

How it was made

Intended use

Limits

License

Citation