GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
open-athena
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
open-athena
Model Tree
nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16
Input Modalities
Output Modalities
Supported Functionality
Three stages, each one epoch of packed 32,768-token sequences, each starting from the previous stage's weights with a fresh optimizer. All stages use Adam (β 0.9/0.95, ε 1e-8), learning rate 5e-6 with cosine decay and no warmup, weight decay 0, gradient clipping at 1.0 and 64 sequences per step. Packed conversations could not attend to each other. Training used the marin-community/axolotl fork at a208d550 with transformers 5.12.1, fp32 master weights and bf16 compute, on 16 H100s. Conversations longer than 32,768 tokens were dropped. The router's score-correction bias was not updated. The stage-3 settings are in axolotl_config.yaml and axolotl_run_manifest.json.
| Stage | Data | Steps |
|---|---|---|
| 1. Chat | nyu-dice-lab/wildchat50m-rewild-sft-385700 @ 46a5bb5 | 261 |
| 2. Thinking | laion/llama-nemotron-science-reasoning-on-canonical-think-full @ bae881d | 628 |
| 3. Terminal agent | nvidia/Nemotron-Terminal-Corpus @ a1667c4 | 1,291 (2.71B tokens, of which 1.83B were trained on) |
An audit of 2,000 stage-1 conversations confirmed that every assistant turn's closing <|im_end|> was a training target, both before and after packing. This export is the bf16 copy of stage 3's final fp32 weights.
The chat template is in chat_template.jinja and is also embedded in tokenizer_config.json. Training and serving used the same template, a ChatML-style template shared with the other models in this experiment.
text
<s><|im_start|>system{system}<|im_end|><|im_start|>user{user}<|im_end|><|im_start|>assistant<think>{reasoning}</think>{answer}<|im_end|><|im_end|> (id 11), which is also the tokenizer's end-of-sequence token, so runtimes stop on it by default. </s> (2) is a second stop in generation_config.json.<think> (12) and </think> (13). These are ordinary vocabulary tokens, not special tokens, so they stay in the returned text by default.content, unchanged, think block included. That renders the same tokens the model was trained on.<tool_call>\n{"name": ..., "arguments": {...}}\n</tool_call>. Results go back in a tool message, rendered as <tool_response>\n...\n</tool_response>.tools request field, so tools passed that way are silently ignored. Describe the tools and the call format in the system message.A tool-use conversation looks like this:
python
messages = [ {"role": "system", "content": ( "You can call one tool. To call it, write exactly:\n" '<tool_call>\n{"name": "get_weather", "arguments": {"city": "<city>"}}\n</tool_call>\n' "get_weather(city: string) returns the current weather in that city." )}, {"role": "user", "content": "Is it raining in Paris?"}, {"role": "assistant", "content": '<tool_call>\n{"name": "get_weather", "arguments": {"city": "Paris"}}\n</tool_call>'}, {"role": "tool", "content": '{"conditions": "light rain", "temp_c": 11}'},]generation_config.json sets the decode used for every evaluation below; vLLM and transformers apply it by default:
| Setting | Value |
|---|---|
| temperature | 0.6 |
| top_p | 0.95 |
| repetition_penalty | 1.05 |
| stop | `< |
| max new tokens | 32,768 for single-turn; 16,384 per agent turn |
Stock vLLM and transformers 5.x support nemotron_h natively; no trust_remote_code is needed. Evaluation used Marin's vLLM build 0.0.0.dev20260924+marin.01911be34fac, with one model copy per H100 80 GB GPU at a 131,072-token context and up to 32 concurrent sequences. Recent stock vLLM releases that support Nemotron-3-Nano should also work, but they were not tested with this checkpoint.
sh
vllm serve open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30 \ --max-model-len 131072 --max-num-seqs 32config.json sets mamba_ssm_cache_dtype: float32. NVIDIA recommends running its Nemotron-3-Nano models in vLLM with --mamba-ssm-cache-dtype float32. The evaluation below did not pass that flag.
python
from openai import OpenAI def split_reply(text): """Return (reasoning, answer). reasoning is None without a think block; answer is None if the think block never closed.""" text = text or "" if not text.startswith("<think>"): return None, text reasoning, closed, answer = text[len("<think>"):].partition("</think>") return reasoning, (answer if closed else None) client = OpenAI(base_url="http://localhost:8000/v1", api_key="unused")reply = client.chat.completions.create( model="open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30", messages=[{"role": "user", "content": "How many primes are there below 50?"}], max_tokens=32768,)reasoning, answer = split_reply(reply.choices[0].message.content)With transformers:
python
from transformers import AutoModelForCausalLM, AutoTokenizer name = "open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30"tokenizer = AutoTokenizer.from_pretrained(name)model = AutoModelForCausalLM.from_pretrained(name, dtype="bfloat16", device_map="auto")messages = [{"role": "user", "content": "How many primes are there below 50?"}]inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, return_dict=True, return_tensors="pt").to(model.device)output = model.generate(**inputs, max_new_tokens=32768)print(tokenizer.decode(output[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))Evaluation used no vLLM reasoning or tool-call parser; replies were split as above. <tool_call> and </tool_call> are single tokens in this vocabulary, which vLLM's hermes tool parser requires, but tool calling through that parser was not tested.
The SFT had to pass a protocol gate before the model was used in the experiment. P_chat is 200 single-turn chat prompts and P_think is 200 olympiad-style reasoning prompts, both at the recommended decode. Replies that hit the 32,768-token limit while still inside their think block count as length-limited and are left out of both shares: 1 on P_chat and 5 on P_think.
| Check | Threshold | P_chat | P_think |
|---|---|---|---|
| Turn ends on `< | im_end | >` | ≥98% |
Think block is balanced: as many </think> as <think>, ending closed | ≥98% | 40.9% (of 181) | 90.3% (of 195) |
| Reply that opens a think block closes it at least once | — | 181 of 182 | 195 of 200 |
Stray special tokens or </s> stops | — | 0 | 0 |
The model failed this part of the gate. The shortfall was accepted rather than fixed with extra training, since the training-target audit above showed the data was correct.
The agentic, tool-call and expert-load checks were run afterwards on this checkpoint. P_agent is 32 SWE-bench Verified tasks in the Terminus-2 harness, where each agent turn must be a JSON object with analysis, plan and commands. It was served at 131,072 tokens with the recommended decode.
| Check | Threshold | Result |
|---|---|---|
P_agent tasks ending validly (see below) | ≥29 of 32 | 32 of 32: 10 completed, 19 hit the time limit, 3 grader failures |
| Agent turns whose JSON parses | ≥95% | 91.1% (2,450 of 2,689), below threshold |
| Tasks solved (not gated) | — | 2 of 29 scored |
| Per MoE layer, share of routed experts getting under 10% of the mean load: increase over the base | ≤ +2 points | +5.47 points in MoE layer 3 (7 of 128 experts, against none in the base), below threshold; MoE layer 9 is at +2.34 |
A task counts as valid if it completes, hits the harness's 3,000-second agent time limit, or ends with the grader failing; time-outs are expected when small models act as agents, and grader failures are not the model's. Any other harness error counts against the model. The routing check used the same 2.16M tokens of mixed text for both models. These shortfalls were also accepted.
What this means in use:
</think> tags. The model almost always closes its think block and writes an answer. But on 107 of 182 chat replies (and 17 of 200 reasoning replies) it then writes another </think> instead of ending the turn, and repeats or revises the answer, sometimes as a JSON object. The balanced-think-block check above fails because of these extra tags. split_reply above splits at the first </think>, so answer holds the first answer followed by the repeats. Cut it at the next </think> to keep only the first answer.</think> and no <think> (18 of 200 chat replies). split_reply treats them as answers that begin with a stray tag; strip it.max_tokens limit.The weights are governed by two licenses:
The training datasets carry their own licenses; see each dataset's card.
Licensed by NVIDIA Corporation under the NVIDIA Nemotron Model License.
We release our tools to the public with the intent that they are used safely, ethically, and responsibly. They are meant to contribute to the public's understanding of science, and aid in its advancement. By using this model, you agree to comply with all applicable law.
<s><|im_start|>system{system}<|im_end|><|im_start|>user{user}<|im_end|><|im_start|>assistant<think>{reasoning}</think>{answer}<|im_end|>messages = [ {"role": "system", "content": ( "You can call one tool. To call it, write exactly:\n" '<tool_call>\n{"name": "get_weather", "arguments": {"city": "<city>"}}\n</tool_call>\n' "get_weather(city: string) returns the current weather in that city." )}, {"role": "user", "content": "Is it raining in Paris?"}, {"role": "assistant", "content": '<tool_call>\n{"name": "get_weather", "arguments": {"city": "Paris"}}\n</tool_call>'}, {"role": "tool", "content": '{"conditions": "light rain", "temp_c": 11}'},]vllm serve open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30 \ --max-model-len 131072 --max-num-seqs 32from openai import OpenAI def split_reply(text): """Return (reasoning, answer). reasoning is None without a think block; answer is None if the think block never closed.""" text = text or "" if not text.startswith("<think>"): return None, text reasoning, closed, answer = text[len("<think>"):].partition("</think>") return reasoning, (answer if closed else None) client = OpenAI(base_url="http://localhost:8000/v1", api_key="unused")reply = client.chat.completions.create( model="open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30", messages=[{"role": "user", "content": "How many primes are there below 50?"}], max_tokens=32768,)reasoning, answer = split_reply(reply.choices[0].message.content)from transformers import AutoModelForCausalLM, AutoTokenizer name = "open-athena/Nemotron-3-Nano-30B-A3B-Chat-Aligned-Base-2026.10.30"tokenizer = AutoTokenizer.from_pretrained(name)model = AutoModelForCausalLM.from_pretrained(name, dtype="bfloat16", device_map="auto")messages = [{"role": "user", "content": "How many primes are there below 50?"}]inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, return_dict=True, return_tensors="pt").to(model.device)output = model.generate(**inputs, max_new_tokens=32768)print(tokenizer.decode(output[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True))