test

pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d6cb774f29ad55341f3aa504b2b736b0d3b48863755ea2255130b6d0df03d69c
-size 3086784957

run.sh

@@ -13,11 +13,11 @@ deepspeed run_speech_recognition_seq2seq_streaming.py \
 	--gradient_accumulation_steps="2" \
 	--logging_steps="25" \
 	--learning_rate="3e-6" \
-	--warmup_steps="500" \
+	--warmup_steps="1000" \
 	--evaluation_strategy="steps" \
-	--eval_steps="100" \
+	--eval_steps="1000" \
 	--save_strategy="steps" \
-	--save_steps="100" \
+	--save_steps="1000" \
 	--generation_max_length="225" \
 	--length_column_name="duration" \
 	--max_duration_in_seconds="30" \

runs/Dec15_08-27-09_dante/1671089283.5600834/events.out.tfevents.1671089283.dante.1310863.1

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a5b0fec87787af91e811e9c946d623481da9df7f71dfcb04df8a280bcd7871ae
-size 5891

runs/Dec15_08-27-09_dante/events.out.tfevents.1671089283.dante.1310863.0

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:00dc07d63391576c4a2748f7a15f11181f4b145710a221e9b8b752e2e77c7584
-size 5222