Training in progress, step 500

ipu_config.json

@@ -7,7 +7,7 @@
   "execute_encoder_on_cpu_for_generation": false,
   "gradient_accumulation_steps": 16,
   "inference_device_iterations": 5,
-  "inference_replication_factor": 4,
+  "inference_replication_factor": 1,
   "ipus_per_replica": 4,
   "layers_per_ipu": [
     0,
@@ -20,8 +20,8 @@
   "optimum_version": "1.5.1",
   "output_mode": "final",
   "recompute_checkpoint_every_layer": true,
-  "replicated_tensor_sharding": true,
-  "replication_factor": 4,
+  "replicated_tensor_sharding": false,
+  "replication_factor": 1,
   "seed": 42,
   "sharded_execution_for_inference": false,
   "transformers_version": "4.20.1"

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb432ac6483bddbecd10cc083fdddfc43432bec8f5d6ce8977aebe43a12b993e
+oid sha256:2344e7509c8d1957bba85023604b85b73cbcdf951349f3f0f65e2bbaef84cc12
 size 217864753

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:04dbcc66e7ec6dd24cae396090cff67010597b331b77f50fb0888974104a34b1
+oid sha256:a9ab068517bde9174282280dc6d7503e098228625b30ac49a794d46f10868a2c
 size 2735