ling-3.0-tiny-sft-agentic is an open-weight model from Waylon Brown, released under MIT License. Its published files total 45.3 MB.
PEFT LoRA adapter on inclusionAI/Ling-3.0-tiny (7.9B hybrid MoE, 1.3B active). Trained for coding chat and agentic SWE tool use. SFT-T has no think traces (enablethinking=False). Assistant-only labels, left-truncate to 4096 tokens.
Model Card
By Waylon Brown, published under mit, revision edef2c4577fc.
PEFT LoRA adapter on inclusionAI/Ling-3.0-tiny (7.9B hybrid MoE, 1.3B active). Trained for coding chat and agentic SWE tool use. SFT-T has no think traces (enablethinking=False). Assistant-only labels, left-truncate to 4096 tokens. - LoRA r=16, alpha=32, dropout=0.05 on attention projections - 1000 optimizer steps (~0.52 epoch), batch 2, grad accum 4, lr 2e-4 cosine, bf16
Read Waylon Brown's full model card
Ling-3.0-tiny LoRA (SFT-T + Nemotron SWE)
PEFT LoRA adapter on inclusionAI/Ling-3.0-tiny (7.9B hybrid MoE, 1.3B active). Trained for coding chat and agentic SWE tool use.
Training mix
| Source | Count in mix |
|---|---|
| WlnJBrn09/SFT-1_10-1-26 (SFT-T, upsample 4×) | 2,236 |
| nvidia/Nemotron-SFT-SWE-v3.5 | 5,115 |
nvidia/Nemotron-RL-Agentic-SWE-Pivot-v1 (pass_rate >= 0.375, cap 8k) |
8,000 |
| Train total | 15,351 |
| SFT-T valid | 36 |
SFT-T has no think traces (enable_thinking=False). Assistant-only labels, left-truncate to 4096 tokens.
Hyperparameters
- Hardware: 1× NVIDIA H100 80GB HBM3 (Runpod)
- LoRA r=16, alpha=32, dropout=0.05 on attention projections
- 1000 optimizer steps (~0.52 epoch), batch 2, grad accum 4, lr 2e-4 cosine, bf16
- Save every 250 steps
Load
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
base = "inclusionAI/Ling-3.0-tiny"
adapter = "WaylonJBrown/ling-3.0-tiny-sft-agentic"
tok = AutoTokenizer.from_pretrained(base, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(base, trust_remote_code=True, torch_dtype="bfloat16")
model = PeftModel.from_pretrained(model, adapter)
Identity and Version
- Repository
- WaylonJBrown/ling-3.0-tiny-sft-agentic
- Publisher
- Waylon Brown
- Task
- Not stated by the source
- Modality
- Other
- Library
- peft
- Parameters
- Not stated by the source
- Languages
- en
- Revision
- edef2c4577fc44b8f26df1fec93ed60edc2c808a
- First published
- 2026-10-02
- Last updated
- 2026-10-02
Files and Weights
8 files, 45.3 MB in total. The weights are 2 files totalling 33.1 MB in bin, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| adapter_model.safetensors | Weights | 33.1 MB | 8c0eabe4e5a7 |
| training_args.bin | Weights | 5.3 KB | 6249a2bca3c6 |
| adapter_config.json | Configuration | 1.3 KB | — |
| README.md | Documentation | 1.6 KB | — |
| chat_template.jinja | Other | 6.0 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 12.2 MB | 40fb9d7d7795 |
| tokenizer_config.json | Tokenizer | 486 B | — |
License and Download
- License
- mit
- Access
- Open weights, no gate
- Download size
- 33.1 MB
Released by Waylon Brown through its official repository on Hugging Face. Read the license.
Built From
- Adapter of inclusionAI/Ling-3.0-tiny
- Derived from inclusionAI/Ling-3.0-tiny
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 33.1 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About ling-3.0-tiny-sft-agentic
Can I use ling-3.0-tiny-sft-agentic commercially?
Yes. ling-3.0-tiny-sft-agentic is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.