Update src/about.py
Browse files- src/about.py +2 -2
src/about.py
CHANGED
@@ -59,13 +59,13 @@ LLM_BENCHMARKS_TEXT = f"""
|
|
59 |
## How it works
|
60 |
|
61 |
## Reproducibility
|
62 |
-
|
63 |
|
64 |
If all the setup goes smoothly, you can run 'MODEL' on ItaEval with:
|
65 |
```bash
|
66 |
MODEL="your-model-id-on-the-huggingface-hub"
|
67 |
lm_eval --model hf \
|
68 |
-
--model_args pretrained=$
|
69 |
--tasks ita_eval \
|
70 |
--batch_size 1 \
|
71 |
--log_samples \
|
|
|
59 |
## How it works
|
60 |
|
61 |
## Reproducibility
|
62 |
+
Head to {ITA_EVAL_REPO} to reproduce our results for all the instructions.
|
63 |
|
64 |
If all the setup goes smoothly, you can run 'MODEL' on ItaEval with:
|
65 |
```bash
|
66 |
MODEL="your-model-id-on-the-huggingface-hub"
|
67 |
lm_eval --model hf \
|
68 |
+
--model_args pretrained=$MODEL,dtype=bfloat16 \
|
69 |
--tasks ita_eval \
|
70 |
--batch_size 1 \
|
71 |
--log_samples \
|