Create run_clm_flax.sh

run_clm_flax.sh

@@ -0,0 +1,22 @@
+#!/bin/bash
+
+LANG="irish"
+
+python transformers/examples/flax/language-modeling/run_clm_flax.py \
+    --output_dir="./$LANG-gpt2" \
+    --model_type="gpt2" \
+    --config_name="./$LANG-gpt2" \
+    --tokenizer_name="./$LANG-gpt2" \
+    --dataset_name="oscar" \
+    --dataset_config_name="unshuffled_deduplicated_ga" \
+    --do_train --do_eval \
+    --block_size="512" \
+    --per_device_train_batch_size="64" \
+    --per_device_eval_batch_size="64" \
+    --learning_rate="5e-3" --warmup_steps="1000" \
+    --adam_beta1="0.9" --adam_beta2="0.98" --weight_decay="0.01" \
+    --overwrite_output_dir \
+    --num_train_epochs="20" \
+    --logging_steps="500" \
+    --save_steps="500" \
+    --eval_steps="500"