Model reference · open weights

Starling-LM-alpha

Available as managed deployment LLMs berkeley-nest Text gen 1 variants 1k dl/mo

Starling-LM-alpha is an open-weight language model from berkeley-nest. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Released byberkeley-nest
TypeLanguage models
TaskText gen
Parameters (lead)7.2B
Context8k tokens
Runs withtransformers
Released2023-11-25
Popularity1k downloads / month
LicenceOpen weights

About

What Starling-LM-alpha is

  • Developed by: Banghua Zhu * , Evan Frick * , Tianhao Wu * , Hanlin Zhu and Jiantao Jiao.
  • Model type: Language Model finetuned with RLHF / RLAIF
  • License: Apache-2.0 license under the condition that the model is not used to compete with OpenAI
  • Finetuned from model: Openchat 3.5 (based on Mistral-7B-v0.1)

We introduce Starling-7B, an open large language model (LLM) trained by Reinforcement Learning from AI Feedback (RLAIF). The model harnesses the power of our new GPT-4 labeled ranking dataset, berkeley-nest/Nectar, and our new reward training and policy tuning pipeline. Starling-7B-alpha scores 8.09 in MT Bench with GPT-4 as a judge, outperforming every model to date on MT-Bench except for OpenAI's GPT-4 and GPT-4 Turbo. We release the ranking dataset Nectar, the reward model Starling-RM-7B-alpha and the language model Starling-LM-7B-alpha on HuggingFace, and an online demo in LMSYS Chatbot Arena. Stay tuned for our forthcoming code and paper, which will provide more details on the whole process.

Starling-LM-7B-alpha is a language model trained from Openchat 3.5 with reward model berkeley-nest/Starling-RM-7B-alpha and policy optimization method advantage-induced policy alignment (APA). The evaluation results are listed below.

ModelTuning MethodMT BenchAlpacaEvalMMLU
GPT-4-Turbo?9.3297.70
GPT-4SFT + PPO8.9995.2886.4
Starling-7BC-RLFT + APA8.0991.9963.9
Claude-2?8.0691.3678.5
GPT-3.5-Turbo?7.9489.3770
Claude-1?7.988.3977
Tulu-2-dpo-70bSFT + DPO7.8995.1
Openchat-3.5C-RLFT7.8188.5164.3
Zephyr-7B-betaSFT + DPO7.3490.6061.4
Llama-2-70b-chat-hfSFT + PPO6.8692.6663
Neural-chat-7b-v3-1SFT + DPO6.8484.5362.4
Tulu-2-dpo-7bSFT + DPO6.2985.1

For more detailed discussions, please check out our blog post, and stay tuned for our upcoming code and paper!

  • Blog: https://starling.cs.berkeley.edu/
  • Paper: Coming soon!
  • Code: Coming soon!
Read the full model card

Uses

Important: Please use the exact chat template provided below for the model. Otherwise there will be a degrade in the performance. The model output can be verbose in rare cases. Please consider setting temperature = 0 to make this happen less.

Our model follows the exact chat template and usage as Openchat 3.5. Please refer to their model card for more details. In addition, our model is hosted on LMSYS Chatbot Arena for free test.

The conversation template is the same as Openchat 3.5:

import transformers
tokenizer = transformers.AutoTokenizer.from_pretrained("openchat/openchat_3.5")

# Single-turn
tokens = tokenizer("GPT4 Correct User: HelloGPT4 Correct Assistant:").input_ids
assert tokens == [1, 420, 6316, 28781, 3198, 3123, 1247, 28747, 22557, 32000, 420, 6316, 28781, 3198, 3123, 21631, 28747]

# Multi-turn
tokens = tokenizer("GPT4 Correct User: HelloGPT4 Correct Assistant: HiGPT4 Correct User: How are you today?GPT4 Correct Assistant:").input_ids
assert tokens == [1, 420, 6316, 28781, 3198, 3123, 1247, 28747, 22557, 32000, 420, 6316, 28781, 3198, 3123, 21631, 28747, 15359, 32000, 420, 6316, 28781, 3198, 3123, 1247, 28747, 1602, 460, 368, 3154, 28804, 32000, 420, 6316, 28781, 3198, 3123, 21631, 28747]

# Coding Mode
tokens = tokenizer("Code User: Implement quicksort using C++Code Assistant:").input_ids
assert tokens == [1, 7596, 1247, 28747, 26256, 2936, 7653, 1413, 334, 1680, 32000, 7596, 21631, 28747]

Code Examples

import transformers

tokenizer = transformers.AutoTokenizer.from_pretrained("berkeley-nest/Starling-LM-7B-alpha")
model = transformers.AutoModelForCausalLM.from_pretrained("berkeley-nest/Starling-LM-7B-alpha")

def generate_response(prompt):
    input_ids = tokenizer(prompt, return_tensors="pt").input_ids
    outputs = model.generate(
        input_ids,
        max_length=256,
        pad_token_id=tokenizer.pad_token_id,
        eos_token_id=tokenizer.eos_token_id,
    )
    response_ids = outputs[0]
    response_text = tokenizer.decode(response_ids, skip_special_tokens=True)
    return response_text

# Single-turn conversation
prompt = "Hello, how are you?"
single_turn_prompt = f"GPT4 Correct User: {prompt}GPT4 Correct Assistant:"
response_text = generate_response(single_turn_prompt)
print("Response:", response_text)

## Multi-turn conversation
prompt = "Hello"
follow_up_question =  "How are you today?"
response = ""
multi_turn_prompt = f"GPT4 Correct User: {prompt}GPT4 Correct Assistant: {response}GPT4 Correct User: {follow_up_question}GPT4 Correct Assistant:"
response_text = generate_response(multi_turn_prompt)
print("Multi-turn conversation response:", response_text)

### Coding conversation
prompt = "Implement quicksort using C++"
coding_prompt = f"Code User: {prompt}Code Assistant:"
response = generate_response(coding_prompt)
pr

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys starling-lm-alpha for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (starling-lm-alpha below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/chat/completions \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"starling-lm-alpha","messages":[{"role":"user","content":"Hello"}]}'

Create an account — your API key is available in the console. 3M free tokens every 30 days with every new account.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms