Downloads · 30 days
12
13% of all-time downloads
STEVENZHANG904/Qwen3-1.7B-verifier-sft
Qwen3-1.7B-verifier-sft is a text generation model from STEVENZHANG904. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
SFT-finetuned Qwen/Qwen3-1.7B on the verifier subset of Divij/qwen3-32b-mas-traces, which contains traces of Qwen3-32B acting as a verifier agent in a multi-agent system. This model is the distilled student that learn…
Downloads · 30 days
12
13% of all-time downloads
All-time downloads
90
Public
Parameters
1.7B
10.3 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors3.4 GB · 100%
From the Hugging Face model README
SFT-finetuned Qwen/Qwen3-1.7B on the verifier subset of Divij/qwen3-32b-mas-traces,
which contains traces of Qwen3-32B acting as a verifier agent in a multi-agent system. This model is the
distilled student that learns to play the same role as Qwen3-32B in that pipeline.
| Branch | Epochs trained | Notes |
|---|---|---|
epoch2 | 2 | intermediate |
epoch5 | 5 | intermediate |
main | 10 | final |
Qwen/Qwen3-1.7BDivij/qwen3-32b-mas-traces (config verifier)from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
repo = "STEVENZHANG904/Qwen3-1.7B-verifier-sft"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, dtype=torch.bfloat16, device_map="cuda")
# Verifier role expects a task-spec prompt — see the dataset card for the exact format.
messages = [
{"role": "system", "content": "You are a helpful, creative, and smart assistant."},
{"role": "user", "content": "<your verifier task spec here>"},
]
inputs = tok.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True).to("cuda")
out = model.generate(
inputs, max_new_tokens=4096,
do_sample=True, temperature=0.6, top_p=0.95, # Qwen3 thinking-mode defaults
)
print(tok.decode(out[0][inputs.shape[-1]:], skip_special_tokens=True))
The model emits <think>...</think> reasoning blocks (inherited from Qwen3-32B traces).
Use sampling, not greedy decoding — small distilled models can loop in <think> under greedy.