Training in progress, step 1000

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8dbb63e3ec06276762ed302328e57ea42d34ac46df79a439e6a05ec7d7e011a0
+oid sha256:91fd4dcfce53d762be8373c54aa64ff2d81527bb114096cb75dd78cfac04eac0
 size 1262181617

run.sh

@@ -17,13 +17,13 @@ python -m torch.distributed.launch \
 	--per_device_train_batch_size="32" \
 	--per_device_eval_batch_size="32" \
 	--gradient_accumulation_steps="1" \
-	--learning_rate="5e-4" \
-	--warmup_steps="500" \
+	--learning_rate="1e-4" \
+	--warmup_steps="1000" \
 	--evaluation_strategy="steps" \
 	--text_column_name="text" \
 	--min_duration_in_seconds="0.25" \
-	--save_steps="500" \
-	--eval_steps="500" \
+	--save_steps="1000" \
+	--eval_steps="1000" \
 	--logging_steps="1" \
 	--layerdrop="0.0" \
 	--length_column_name="input_length" \

special_tokens_map.json

@@ -35,6 +35,20 @@
       "rstrip": false,
       "single_word": false
     },
+    {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
     {
       "content": "</s>",
       "lstrip": false,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6235f6f0beb8a3f9ad85c0feedc85239a9b86be9f1a69bfda8761e7879c8145f
+oid sha256:b377e92f931b83d528bdb20a32e0040738acc2ab5cf30fe22893b2cdc20f283b
 size 3311