Training in progress, step 1000

config.json

@@ -34,7 +34,7 @@
   "num_mel_bins": 80,
   "pad_token_id": 50257,
   "scale_embedding": false,
-  "torch_dtype": "float16",
+  "torch_dtype": "float32",
   "transformers_version": "4.26.0.dev0",
   "use_cache": false,
   "vocab_size": 51865

ds_config.json

@@ -1,48 +0,0 @@
-{
-    "fp16": {
-        "enabled": "auto",
-        "loss_scale": 0,
-        "loss_scale_window": 1000,
-        "initial_scale_power": 16,
-        "hysteresis": 2,
-        "min_loss_scale": 1
-    },
-
-    "optimizer": {
-        "type": "AdamW",
-        "params": {
-            "lr": "auto",
-            "betas": "auto",
-            "eps": "auto",
-            "weight_decay": "auto"
-        }
-    },
-
-    "scheduler": {
-        "type": "WarmupLR",
-        "params": {
-            "warmup_min_lr": "auto",
-            "warmup_max_lr": "auto",
-            "warmup_num_steps": "auto"
-        }
-    },
-
-    "zero_optimization": {
-        "stage": 2,
-        "offload_optimizer": {
-            "device": "cpu",
-            "pin_memory": true
-        },
-        "allgather_partitions": true,
-        "allgather_bucket_size": 2e8,
-        "overlap_comm": true,
-        "reduce_scatter": true,
-        "reduce_bucket_size": 2e8,
-        "contiguous_gradients": true
-    },
-
-    "gradient_accumulation_steps": "auto",
-    "gradient_clipping": "auto",
-    "train_batch_size": "auto",
-    "train_micro_batch_size_per_gpu": "auto"
-}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44089c5590ed9f7824ba2b1ac8971c4eeadc9e44a55ec419bcd2cdc006b7e181
-size 483536061
+oid sha256:6458db92153523ae195e1ef0a5db0a78375fc020947b974c8cfe7a8d123fd242
+size 967102601

run.sh

@@ -1,5 +1,4 @@
-deepspeed run_speech_recognition_seq2seq_streaming.py \
-	--deepspeed="ds_config.json" \
+python run_speech_recognition_seq2seq_streaming.py \
 	--model_name_or_path="openai/whisper-small" \
 	--dataset_name="facebook/voxpopuli" \
 	--dataset_config_name="hr" \

runs/Dec14_16-16-20_129-146-123-136/events.out.tfevents.1671034661.129-146-123-136.70213.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f92a227fc60f30d6850b5e1e039264a67779ae38804018446c42be9b8405b88
-size 24058
+oid sha256:1be0e84061e5fa35dfeebb9f289d88cfad80f96b46cd10380118db6a9f8d84e6
+size 28611

runs/Dec14_23-13-58_129-146-123-136/1671059651.4855802/events.out.tfevents.1671059651.129-146-123-136.104631.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:015fe417fdf705bf034ab36ff08bba1842db4e17e047850603232f6962e1a96d
+size 5871

runs/Dec14_23-13-58_129-146-123-136/events.out.tfevents.1671059651.129-146-123-136.104631.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e789ae5098d6e35eea005b5d67a41d77d8d82744d121f62424b2631b04071568
+size 10851

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4661efef1c7d1829413d2a909012154ff96d9155a348c80c9a204ac96b0688df
-size 4667
+oid sha256:b8f800914ed7d0ecfdabd5f7c0dd570590eb7c219e403ef5c508f076fc18572f
+size 3579