Instructions to use AgriManager-Framework/schema-t23-objectives-think-r2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AgriManager-Framework/schema-t23-objectives-think-r2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="AgriManager-Framework/schema-t23-objectives-think-r2")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("AgriManager-Framework/schema-t23-objectives-think-r2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AgriManager-Framework/schema-t23-objectives-think-r2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AgriManager-Framework/schema-t23-objectives-think-r2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AgriManager-Framework/schema-t23-objectives-think-r2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/AgriManager-Framework/schema-t23-objectives-think-r2

SGLang

How to use AgriManager-Framework/schema-t23-objectives-think-r2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "AgriManager-Framework/schema-t23-objectives-think-r2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AgriManager-Framework/schema-t23-objectives-think-r2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "AgriManager-Framework/schema-t23-objectives-think-r2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AgriManager-Framework/schema-t23-objectives-think-r2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use AgriManager-Framework/schema-t23-objectives-think-r2 with Docker Model Runner:
```
docker model run hf.co/AgriManager-Framework/schema-t23-objectives-think-r2
```

schema-t23-objectives-think-r2

A crop-management policy from the AgriManager benchmark (NeurIPS 2026, Evaluations and Datasets track): Qwen/Qwen3-4B-Instruct-2507 fine-tuned with GRPO on the T2.3 reward-formulation shift, multi-objective training setting, think response format, training seed 91 (replicate r2).

Paper: TODO(release-url)
Code: https://github.com/AgriManager-Framework/AgriManager
Held-out test suite: AgriManager-Framework/agrimanager-test-suite

Every checkpoint lives on its own branch (step-N); main holds no weights. Load a checkpoint with an explicit revision:

from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("AgriManager-Framework/schema-t23-objectives-think-r2", revision="step-400")
tokenizer = AutoTokenizer.from_pretrained("AgriManager-Framework/schema-t23-objectives-think-r2", revision="step-400")

What this checkpoint is

Field	Value
Repository	`AgriManager-Framework/schema-t23-objectives-think-r2`
Base model	`Qwen/Qwen3-4B-Instruct-2507`
Algorithm	GRPO (VeRL 0.7.0)
Setting	T2.3 reward-formulation shift, multi-objective training; Tier 2
Simulator(s)	WOFOST
Response format	think: reasoning inside `<tool_call>…</tool_call>`, then the action inside `<answer>…</answer>`
Training seed	91 (data seed 91, data-loader seed 91)
Revisions	`step-300`, `step-400`
Terminal revision	`step-400` (the one scored in the paper)
Early stop	none (trained for the planned length)
Selection rule	terminal or last-valid checkpoint, never best-validation
Run script	`experiments/t2_3_reward_formulation_shift/run_t23_objectives_think_r2_train.sh` in the code repository
Paper	the Tier 2 results (reward formulation) and the per-seed tables of the appendix
Evidence status	confirmatory: one of the three scored seeds of this arm

Each revision carries a run_manifest.json with the base model, seeds, training schedule, response format and a SHA-256 for every file.

Results

Per-seed values: results/tables/e4_by_seed.csv in the code repository, rows with cell=t23-objectives, method=LLM-RL, arm=think, seed=2.0 (the replicate index). Per-episode records: dataset AgriManager-Framework/schema-results-t23-reward, directory schema-t23-objectives-think-r2/step-400/.

How to use it

The model expects the AgriManager prompt of its simulator and replies with one management action in the format above. Use it through the AgriManager environment loop; it is not a general-purpose agronomy assistant. To score it on the held-out test suite with the code repository:

bash entrypoints/eval/e5_evaluate_checkpoint.sh \
    --checkpoint-repo AgriManager-Framework/schema-t23-objectives-think-r2 --revision step-400 \
    --suite-repo AgriManager-Framework/agrimanager-test-suite \
    --tier t23 --no-upload

This is the row schema-t23-objectives-think-r2 of results/scoring_manifest.tsv.

The evaluator reads the response format from run_manifest.json and refuses to score the checkpoint in a different format.

Limitations

Trained and evaluated in simulation only (WOFOST, DSSAT, Cycles). Simulator scores are not field outcomes; do not apply its recommendations to a real farm.
The scripted agronomic reference used in the paper is a conservative codified schedule, not calibrated against experts or optimal control.
Scores use greedy decoding; re-scoring on other GPU types changes a few episodes per split.

License

The weights are released under the Apache License 2.0. The base model Qwen/Qwen3-4B-Instruct-2507 is released by its authors under Apache 2.0.

Citation

@inproceedings{agrimanager2026,
  title     = {AgriManager},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS), Evaluations and Datasets Track},
  year      = {2026},
  note      = {TODO(release-url): final title, authors and URL}
}

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for AgriManager-Framework/schema-t23-objectives-think-r2

Base model

Qwen/Qwen3-4B-Instruct-2507

Finetuned

(2726)

this model