Update README.md
Browse files
README.md
CHANGED
@@ -176,7 +176,7 @@ print(outputs[0]["generated_text"][len(prompt):])
|
|
176 |
- train_batch_size: 12
|
177 |
- eval_batch_size: 8
|
178 |
- GPU: H100 80GB SXM5
|
179 |
-
- num_devices:
|
180 |
- optimizer: adamw_bnb_8bit
|
181 |
- lr_scheduler_warmup_steps: 100
|
182 |
- num_epochs: 4
|
@@ -220,7 +220,7 @@ print(outputs[0]["generated_text"][len(prompt):])
|
|
220 |
|
221 |
# Benchmark Results
|
222 |
|
223 |
-
🔥 OpenBioMed-8B demonstrates superior performance compared to larger models, such as GPT-3.5, Gemini, Meditron-70B across 9 diverse biomedical datasets, achieving state-of-the-art results with an average score of
|
224 |
|
225 |
🚨 The GPT-4, Med-PaLM-1, and Med-PaLM-2 results are taken from their official papers. Since Med-PaLM doesn't provide zero-shot accuracy, we are using 5-shot accuracy from their paper for comparison. All results presented are in the zero-shot setting, except for Med-PaLM-2 and Med-PaLM-1, which use 5-shot accuracy.
|
226 |
|
|
|
176 |
- train_batch_size: 12
|
177 |
- eval_batch_size: 8
|
178 |
- GPU: H100 80GB SXM5
|
179 |
+
- num_devices: 1
|
180 |
- optimizer: adamw_bnb_8bit
|
181 |
- lr_scheduler_warmup_steps: 100
|
182 |
- num_epochs: 4
|
|
|
220 |
|
221 |
# Benchmark Results
|
222 |
|
223 |
+
🔥 OpenBioMed-8B demonstrates superior performance compared to larger models, such as GPT-3.5, Gemini, Meditron-70B across 9 diverse biomedical datasets, achieving state-of-the-art results with an average score of 72.50%, despite having a significantly smaller parameter count. The model's strong performance in domain-specific tasks, such as Clinical KG, Medical Genetics, and PubMedQA, highlights its ability to effectively capture and apply biomedical knowledge.
|
224 |
|
225 |
🚨 The GPT-4, Med-PaLM-1, and Med-PaLM-2 results are taken from their official papers. Since Med-PaLM doesn't provide zero-shot accuracy, we are using 5-shot accuracy from their paper for comparison. All results presented are in the zero-shot setting, except for Med-PaLM-2 and Med-PaLM-1, which use 5-shot accuracy.
|
226 |
|