Training in progress, step 5000

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:110106700c21235b790cc301af107fdf6638f4bb53ecb608ed926802e8ea7371
+oid sha256:477ea268157101cf655cafde46578ccc38ae399b557b918fe2fba37fdbc78f1d
 size 3086784957

run.sh

@@ -11,7 +11,8 @@ deepspeed run_speech_recognition_seq2seq_streaming-farsipal.py \
 	--torch_compile="True" \
 	--torch_compile_mode="reduce-overhead" \
 	--torch_compile_mode="max-autotune" \
-	--max_steps="5000" \
+	--max_steps="10000" \
+	--resume_from_checkpoint="checkpoint-4000" \
 	--output_dir="./" \
 	--per_device_train_batch_size="32" \
 	--gradient_accumulation_steps="1" \

runs/Dec17_22-33-02_129-146-176-120/1671317315.0810065/events.out.tfevents.1671317315.129-146-176-120.1102971.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4d3f632582323aac3e5ed9fa1d80715f6375d921f2911f1b928650ca016e8fc
+size 5904

runs/Dec17_22-33-02_129-146-176-120/events.out.tfevents.1671317315.129-146-176-120.1102971.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7e9d6ba2e6b67e4ffa281e2c6f41fce3b8ee132cd4e45f3b483c061ad9f4395
+size 10922

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a00608c0a40588b0278ea9d2baf279e6392ec4cf7d63f64575cd07a729e45720
-size 4667
+oid sha256:ae9826a3726ec1cfde573920ac52bb711f7f5eeeb43158e08b77fdfd5b56b65d
+size 4731