Downloads · 30 days
6
9% of all-time downloads
llama-3.2-3b-instruct-agrpo-nothink-lr3e-6
EnergyAI/llama-3.2-3b-instruct-agrpo-nothink-lr3e-6
llama-3.2-3b-instruct-agrpo-nothink-lr3e-6 is a machine learning model from EnergyAI. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Llama-3.2-3B-Instruct fine-tuned with Async GRPO (no thinking, nothink mode)
All-time downloads
70
Public
Parameters
3.2B
6.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors6.4 GB · 100%
At a glance
- License
- apache-2.0
- Model type
- llama
- Access
- Public
- Created
- Apr 12, 2026
- Updated
- Apr 12, 2026
- SHA
- 31acbd78
Base models
Model card
From the Hugging Face model README
llama-3.2-3b-instruct-agrpo-nothink-lr3e-6
Llama-3.2-3B-Instruct fine-tuned with Async GRPO (no thinking, nothink mode)
Task
Fill-in-the-middle multiple-choice questions (MCQ) for energy domain verification.
The model outputs its answer inside \boxed{N} where N is the option number.
Reward Function
- +1.0 — correct (
\boxed{N}matches ground truth) - -0.5 — wrong (
\boxed{N}exists but wrong answer) - -1.0 — no answer (no
\boxed{}found)
Training Hyperparameters
| Parameter | Value |
|---|---|
| base_model | meta-llama/Llama-3.2-3B-Instruct |
| algorithm | Async GRPO (TRL AsyncGRPOTrainer) |
| thinking | disabled (nothink) |
| learning_rate | 3e-6 |
| lr_scheduler | cosine |
| warmup_steps | 60 |
| max_steps | 2000 |
| global_step_saved | 2000 |
| per_device_train_batch_size | 1 |
| gradient_accumulation_steps | 19 |
| num_train_processes | 7 |
| effective_batch_size | 133 prompts/step |
| num_generations | 9 |
| max_completion_length | 512 |
| temperature | 1.0 |
| epsilon | 0.2 |
| epsilon_high | 0.2 |
| max_staleness | 4 |
| weight_sync_steps | 1 |
| max_grad_norm | 1.0 |
| precision | bf16 |
| parallelism | FSDP2 (7 GPUs training) + vLLM TP=1 enforce-eager (1 GPU inference) |
| final_reward | ~0.89 |
| final_mean_completion_length | ~203 tokens |
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("EnergyAI/llama-3.2-3b-instruct-agrpo-nothink-lr3e-6", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained("EnergyAI/llama-3.2-3b-instruct-agrpo-nothink-lr3e-6")
- Type
- llama
- License
- apache-2.0
- Created
- Apr 12, 2026
- Updated
- Apr 12, 2026