You can deploy this model and start earning money today!

viamr-project/
amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842

Name: amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842
Availability: InStock

safetensors

qwen3

Views

No views yet

amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842

Model Information

Base Model: checkpoints/amr-parsing-dapo-single/single-turn-20260217-2338/global_step_2842/actor
Timestamp: 20260217-2338

Benchmark Results

Benchmark File: amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842_20260217-2338.jsonl
Score:
- F1: 82.65
- Precision: 83.32
- Recall: 81.99

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("viamr-project/amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained("viamr-project/amr-parsing-dapo-single-single-turn-20260217-2338-global-step-2842", trust_remote_code=True)

Sample Prompt:

text

<|im_start|>user
Convert the following English sentence into its Abstract Meaning Representation (AMR):

<sentence>2007-08-21</sentence><|im_end|>
<|im_start|>assistant
<think>

Training Details

Framework: veRL (Reinforcement Learning)
Task: AMR Parsing