GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
textilelabs
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
textilelabs
Model Tree
Input Modalities
Output Modalities
Supported Functionality
Identical battery, identical prompts, one variable — training time.
| Spark 1.8 | Spark 2 (1h) | Weave 2 Flash (2h) | |
|---|---|---|---|
offline <lookup> leak | 16/30 · 53% | 0/30 | 0/30 |
| clean single online lookup | 8/10 | 10/10 | 10/10 |
| identity correct | — | 12/12 | 12/12 |
| identity under CAPS / typos / filler | not trained for | 6/6 | 5/6 |
| admits an unknowable personal fact | — | 4/8 | 6/8 |
answers from a supplied <result> | — | 2/5 | 3/5 |
| says the result doesn't contain it | — | 0/4 | 0/4 |
| self-termination without a Modelfile | needed one | 12/12 | 12/12 |
| validation loss · accuracy | — | 2.69 · 0.536 | 2.25 · 0.589 |
809 steps → 1,645 steps. Calibration and grounded reading both improved; loss and accuracy improved clearly.
says the result doesn't contain it stayed at 0/4. Twice the training, identical zero.
That is worth publishing rather than hiding, because it rules something out. Nearly a third of the training corpus — 30,000 human-written examples — is exactly this behaviour, and 3.3M additional tokens moved it not at all, while every neighbouring metric moved.
So this is not under-training. The likely reason: saying "the result doesn't say" requires
detecting an absence — reading the whole result and concluding the answer is not in it.
That is a harder computation than extracting a span that is present, and at 20M parameters
it appears to be out of reach. By contrast unknowable (6/8) only needs to notice "my" or
"I" in the question.
Practical consequence: if you feed this model a result, it will answer from it whether or
not the answer is there. Validate the output. Treat the retrieved <result> as the
trustworthy part and the model's summary of it as unreliable.
It also has almost no world knowledge. With tools off it declines factual questions, which is intended behaviour rather than a fault.
<tools:off> (default) — conversational. Identity, limits, warmth, brevity. No harness
needed.
<tools:on> — emits <lookup>query</lookup> and stops. Your harness runs the lookup and
continues with a <result> block:
markdown
<tools:on><user>what is the capital of Peru<|eot|><loom><lookup>what is the capital of Peru</lookup><|eot|><result>Lima is the capital and largest city of Peru.<|eot|><loom>
The persona slice was trained entirely under tools:off, so identity questions asked with
tools on will often be turned into a lookup. Keep tools off for chat.
harness.py in this repo runs the lookup Loom asks for and feeds the result back.
Wikipedia is used because it is free and needs no key — swap the search() function for
anything else; the contract is text in, text out.
bash
python3 harness.py "who wrote Dracula" # with lookupspython3 harness.py # interactivepython3 harness.py --no-tools "who are you" # chat only
Three things any harness for this model needs:
<result>. The model will earnestly try to
answer from the error text. Fail loudly instead — harness.py does.User-Agent.bash
ollama run hf.co/textilelabs/Loom-Weave-2-Flash "who are you"# Loom, a small model from Textile Labs.
Ollama reads the template and params files in this repo — nothing to set up. The
template defaults to tools:off. To build locally:
ollama create loom-weave-2-flash -f Modelfile.
python
import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMtok = AutoTokenizer.from_pretrained("textilelabs/Loom-Weave-2-Flash")model = AutoModelForCausalLM.from_pretrained("textilelabs/Loom-Weave-2-Flash").eval()eot = tok.convert_tokens_to_ids("<|eot|>")def ask(message, tools=False):p = f"<tools:{'on' if tools else 'off'}>\n<user>\n{message}\n<|eot|>\n<loom>\n"ids = tok(p, return_tensors="pt", add_special_tokens=False).input_idswith torch.no_grad():out = model.generate(ids, max_new_tokens=64, do_sample=False, eos_token_id=eot,pad_token_id=tok.convert_tokens_to_ids("<|pad|>"))[0]return tok.decode(out[ids.shape[1]:], skip_special_tokens=True).strip()ask("who are you") # -> 'Loom, a small model from Textile Labs.'
Prompt format is exact: <tools:off>\n<user>\n{message}\n<|eot|>\n<loom>\n.
This one, unless you specifically want the one-hour model. It is better on calibration, grounded reading, and both loss metrics, and marginally worse on a single identity probe. Spark 2 exists as the one-hour tier and as the control in this comparison.
markdown
config.json / model.safetensors the modeltokenizer.json / tokenizer_config.json custom BPE tokenizer, 4,096 tokensloom-weave-2-flash-f16.gguf 40MB, for Ollama / llama.cppharness.py runnable harness — runs lookups, feeds results backtemplate / params read automatically by `ollama run hf.co/...`Modelfile for building locallyATTRIBUTION.md required credits for the training corpora
Openly licensed corpora of real human text, plus a persona curriculum written for Loom.
See ATTRIBUTION.md — several of these licences require credit.
| slice | source |
|---|---|
| grounded reading, and "the result doesn't say" | SQuAD 2.0 (CC BY-SA 4.0) |
| when to reach for a tool | MASSIVE (CC BY 4.0) · CLINC150 (CC BY 3.0) |
| instruction following | databricks-dolly-15k (CC BY-SA 3.0) |
| multi-turn dialogue structure | OpenAssistant OASST1 (Apache 2.0) |
| identity, limits, warmth, brevity | Textile Labs — written for Loom |
~11.4M tokens, 43% multi-turn. Validation is a held-out split of the same corpora.
Model: MIT. Training data retains its original licences and attribution.