A Complete Implementation Information to ModelScope for Mannequin Search, Inference, High quality-Tuning, Analysis, and Export

print("n📊 MODEL EVALUATIONn")


eval_results = coach.consider()
print("  Analysis Outcomes:")
for key, worth in eval_results.objects():
   if isinstance(worth, float):
       print(f"    {key:<25}: {worth:.4f}")


from sklearn.metrics import classification_report, confusion_matrix


preds_output = coach.predict(eval_ds)
preds = np.argmax(preds_output.predictions, axis=-1)
labels = preds_output.label_ids


print("n  Classification Report:")
print(classification_report(labels, preds, target_names=["NEGATIVE", "POSITIVE"]))


cm = confusion_matrix(labels, preds)
fig, ax = plt.subplots(figsize=(5, 4))
im = ax.imshow(cm, cmap="Blues")
ax.set_xticks([0, 1]); ax.set_yticks([0, 1])
ax.set_xticklabels(["NEGATIVE", "POSITIVE"])
ax.set_yticklabels(["NEGATIVE", "POSITIVE"])
ax.set_xlabel("Predicted"); ax.set_ylabel("Precise")
ax.set_title("Confusion Matrix — High quality-Tuned DistilBERT")
for i in vary(2):
   for j in vary(2):
       ax.textual content(j, i, str(cm[i, j]), ha="middle", va="middle",
               coloration="white" if cm[i, j] > cm.max()/2 else "black", fontsize=18)
plt.colorbar(im)
plt.tight_layout()
plt.savefig("confusion_matrix.png", dpi=150)
plt.present()
print("  ✅ Saved confusion_matrix.png")


print("n── Testing High quality-Tuned Mannequin on New Inputs ──")
ft_pipeline = hf_pipeline(
   "sentiment-analysis",
   mannequin=coach.mannequin,
   tokenizer=tokenizer,
   gadget=DEVICE,
)


new_reviews = [
   "An absolutely breathtaking masterpiece with brilliant performances!",
   "Waste of two hours. Terrible script and wooden acting.",
   "Decent popcorn movie but nothing special. Had some fun moments.",
]


for evaluation in new_reviews:
   res = ft_pipeline(evaluation)[0]
   emoji = "🟢" if res["label"] == "POSITIVE" else "🔴"
   print(f'  {emoji} {res["label"]} ({res["score"]:.4f}): "{evaluation}"')




print("n💾 EXPORTING THE FINE-TUNED MODELn")


save_path = "./ms_finetuned_model/remaining"
coach.save_model(save_path)
tokenizer.save_pretrained(save_path)
print(f"  ✅ Mannequin saved to: {save_path}")
print(f"     Recordsdata: {os.listdir(save_path)}")


print("n── ONNX Export ──")
attempt:
   from optimum.exporters.onnx import main_export
   onnx_path = "./ms_finetuned_model/onnx"
   main_export(save_path, output=onnx_path, job="text-classification")
   print(f"  ✅ ONNX mannequin exported to: {onnx_path}")
   print(f"     Recordsdata: {os.listdir(onnx_path)}")
besides Exception as e:
   print(f"  ⚠️  ONNX export skipped: {e}")


print("""
── Add to ModelScope Hub (guide step) ──


 1. Get a token from https://modelscope.cn/my/myaccesstoken
 2. Run:


    from modelscope.hub.api import HubApi
    api = HubApi()
    api.login('YOUR_TOKEN')
    api.push_model(
        model_id='your-username/my-finetuned-distilbert',
        model_dir="./ms_finetuned_model/remaining",
    )
""")


print("""
╔══════════════════════════════════════════════════════════════════╗
║                   🎉  TUTORIAL COMPLETE!  🎉                    ║
╠══════════════════════════════════════════════════════════════════╣
║  ✓ ModelScope Hub — search, browse & obtain fashions            ║
║  ✓ MsDataset — load datasets from the ModelScope ecosystem      ║
║  ✓ NLP pipelines — sentiment, NER, zero-shot, technology, masks  ║
║  ✓ CV pipelines — picture classification, object detection, viz   ║
║  ✓ HuggingFace interop — snapshot_download + Transformers       ║
║  ✓ High quality-tuning — DistilBERT on IMDB with Coach API            ║
║  ✓ Analysis — accuracy, F1, confusion matrix                  ║
║  ✓ Export — native save, ONNX, Hub add                        ║
╚══════════════════════════════════════════════════════════════════╝
""")