Instructions to use jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best with Transformers:

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

pipe = pipeline("text-generation", model="jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best")
model = AutoModelForCausalLM.from_pretrained("jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best

SGLang

How to use jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best 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 "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best with Docker Model Runner:
```
docker model run hf.co/jiosephlee/intern-s1-mini-assay-transfer-record-level-v27-bbb-martins-l3-best
```

jiosephlee commited on 15 days ago

Commit

382b399

verified ·

1 Parent(s): 1ee112e

Add model card and evaluation results

Browse files

Files changed (1) hide show

README.md +28 -7

README.md CHANGED Viewed

@@ -1,19 +1,40 @@
 ---
 base_model: jiosephlee/Intern-S1-mini-lm
 datasets:
 - jiosephlee/assay-transfer-record-level-v27-bbb-martins-l3-intern
 tags:
 - assay-transfer
-- bbb-martins
-- molecular-ranking
 ---
-# Intern-S1-mini: BBB Martins record-level assay transfer V27 L3
-This is the selected full-model checkpoint from a 10-epoch assay-transfer training run on the [BBB Martins V27 L3 dataset](https://huggingface.co/datasets/jiosephlee/assay-transfer-record-level-v27-bbb-martins-l3-intern), pinned to dataset revision `2a4c6848de474f0c9152e3acdf4087ce1fe16db8`. It is trained to score A/B assay-transfer prompts, not as a general-purpose molecule predictor.
-The checkpoint was selected at optimizer step 510 by the lowest validation ranking **joint ID/OOD level-macro KNN regression MAE@3** (0.5804). Ranking validation ran every 10 optimizer steps; the effective batch size was 128 (8 GPUs × device batch 4 × gradient accumulation 4). Training details and curves are in the [W&B training run](https://wandb.ai/upenn-ml/record-level-assay-transfer/runs/peyddfo8).
-On the held-out test ranking split, KNN binary macro F1@3/F1@5 was **0.8611/0.8611** for this model, **0.7432/0.7677** for vanilla Morgan, and **0.7345/0.7677** for weighted Morgan (higher is better). Overall Spearman was **0.5149**, versus -0.0088 and 0.0238 for the Morgan baselines. Full comparison metrics are in the [W&B test run](https://wandb.ai/upenn-ml/record-level-assay-transfer/runs/gelk5fi5).
-The repository includes the tokenizer, model weights, and `metric.json` with checkpoint-selection details. Loading requires `trust_remote_code=True` for the custom tokenizer.

 ---
 base_model: jiosephlee/Intern-S1-mini-lm
+library_name: transformers
 datasets:
 - jiosephlee/assay-transfer-record-level-v27-bbb-martins-l3-intern
 tags:
+- chemistry
 - assay-transfer
+- record-level
+- ranking
 ---
+# Intern-S1-mini assay transfer V27 — BBB Martins L3
+Validation-selected checkpoint from the V27 BBB Martins L3 record-level assay-transfer run.
+## Provenance
+- Base model: `jiosephlee/Intern-S1-mini-lm`
+- Base model revision: `fcb667c380ae01f57693a45b4b5c2d331052a107`
+- Training dataset: `jiosephlee/assay-transfer-record-level-v27-bbb-martins-l3-intern`
+- Dataset revision: `2a4c6848de474f0c9152e3acdf4087ce1fe16db8`
+- Training configuration: 10 scheduled epochs, seed 42, 8 GPUs, device batch size 4, gradient accumulation 4, Flash Attention 2, BFD packing, and padding-free training
+- Selection metric: validation `L3/overall/knn_regression_id_ood_level_macro_mae_at_3` (lower is better)
+- Selected optimizer step: 510
+- Selected validation metric: `0.5804`
+- The run was intentionally stopped before completing the full scheduled training; this repository contains its saved validation-selected checkpoint.
+- [Training run](https://wandb.ai/upenn-ml/record-level-assay-transfer/runs/peyddfo8)
+- [Held-out model/Morgan comparison](https://wandb.ai/upenn-ml/record-level-assay-transfer/runs/gelk5fi5)
+## Held-out test comparison
+| Ranker | Binary macro-F1@3 | Binary macro-F1@5 | Regression MAE@3 | Spearman@3 | Top-1 hit@3 |
+|---|---:|---:|---:|---:|---:|
+| Model | 0.8611 | 0.8611 | 0.7717 | -0.0802 | 0.1408 |
+| Morgan vanilla | 0.7432 | 0.7677 | 1.1246 | -0.0382 | 0.1127 |
+| Morgan weighted | 0.7345 | 0.7677 | 1.0652 | -0.1077 | 0.1408 |
+Lower regression MAE is better; higher values are better for the other metrics. All W&B metric names retain the `L3/` prefix.
+The repository contains the full Transformers checkpoint and tokenizer files. `metric.json` is the complete validation-selection record stored with the checkpoint.